T1. Altered CREB Binding to Activity-dependent Genes in Serine Racemase Deficient Mice, a Mouse Model of Schizophrenia

Darrick Balu*, Joseph Coyle

Harvard University, McLean Hospital, Belmont, Massachusetts

Background: cAMP-response-element-binding protein (CREB) is a ubiquitously expressed transcription factor in the brain that regulates neuroplasticity by modulating gene expression. There are numerous signaling pathways by which information is transmitted from the cell membrane to the nucleus that in turn affects CREB binding to DNA. The influx of calcium through N-methyl-D-aspartate receptors (NMDARs) is a well-defined mechanism that leads to the increased expression of CREB-dependent genes, including brain derived neurotrophic factor (BDNF), microRNA-132 (miR132), and activity-regulated cytoskeleton-associated protein (Arc). We have previously demonstrated that serine racemase knockout (SR-/-) mice, which exhibit NMDAR hypofunction, also have reduced mRNA and protein levels of the aforementioned CREB-dependent genes in the hippocampus. In addition, these molecules are reduced in schizophrenia.

Methods: Wild-type (WT; n=5-6) and SR-/- (n=5) mice were killed and their hippocampi were flash frozen on dry ice. Tissue samples were dissociated and then fixed with 1.5% paraformaldehyde. Chromatin was digested using the SimpleChIP Plus Enzymatic Chromatin IP kit (Cell Signaling Technologies). The chromatin was incubated with a rabbit anti-CREB antibody (1:50; Cell Signaling Technologies) overnight at 4°C on a rotisserie. The crosslinks were reversed and the DNA was purified for analysis using qPCR (Sybr Green). To determine the levels of CREB bound at each gene of interest, PCR primers were directed near the CRE sequence of each promoter. All of the primers were validated in previous publications. Melting temperature analysis was performed after every PCR to ensure accuracy.

Results: We first validated the ChIP kit for brain tissue using a positive control antibody and primers to ensure proper chromatin shearing. SR-/- mice have approximately 50% less CREB binding at the miR-132 gene than WT mice. Using BDNF exon-specific primers, SR-/- mice have significantly less CREB binding at the BDNF I promoter, but not at the BDNF IV promoter. Finally, SR-/- mice have 80% less CREB binding to the synaptic activity response element (SARE) region of the Arc gene.

Conclusions: These data demonstrate that in vivo, NMDAR hypofunction caused by the selective removal of the NMDAR co-agonist, D-serine, leads to altered CREB binding on known activity-dependent genes. The ChIP results suggest that reduced CREB binding contributes to the lower levels of BDNF, miR-132, and Arc that we previously observed in SR-/- mice. Future studies will determine whether D-serine administration, which reverses these abnormalities at the RNA and protein level, also normalizes the epigenetic perturbations.

Keywords: CREB, Arc, BDNF, miR132.

Disclosure: JTC served as a consultant for EnVivo and Abbvie in the last 2 years. A patent owned by MGH for the use of D-serine as a treatment for serious mental illness could yield royalties for Dr. Coyle.

T2. Wnt Signaling, Neurodevelopmental, and Behavioral Phenotypes in a Dixdc1 Knock-out Mouse Model of Psychiatric Illness

Benjamin Cheyette*, Robert Stanley, Pierre-Marie Martin

University of California at San Francisco, San Francisco, California

Background: DIXDC1 encodes a scaffold protein enriched in neurons that physically interacts with the DISC1 protein implicated in major mental illness, as well as with the Dishevelled protein central to Wnt/beta-catenin intercellular signaling. The Wnt/beta-catenin signaling pathway is broadly involved in neural development and for this reason has been considered a pontential pathogenic neurodevelopmental pathway in psychiatric illness. Moreover, GSK3, the central kinase in the Wnt/beta-catenin pathway, is a pharmacological target of lithium at therapeutic doses. Recent human genetic studies have buttressed interest in this pathway by showing that loss-of-function mutations in the CHD8 gene, which up-regulate Wnt/beta-catenin pathway target genes, are a susceptibility factor in the autism spectrum disorders.

Methods: We have generated a Dixdc1 knock-out (KO) mouse line using traditional (homologous recombination) gene-targeting techniques. We have analyzed and compared behavior in littermates homozygous for the wild type Dixdc1 locus, heterozygous, or homozygous for the Dixdc1 KO allele. We have tested the response of behavioral differences in these animals to psychopharmacological agents, including lithium. We have also analyzed neurodevelopmental phenotypes in the brains of these animals and in cultured forebrain (hippocampal) neurons. Separately, we have been probing for SNVs at the DIXDC1 locus in human psychiatric patients (both in our own samples and in publicly-available sequence databases) and are testing these for functional differences in assays of Wnt/beta-catenin signaling and in assays based on phenotypic findings in this knock-out mouse model.

Results: Mice homozygous for loss-of-function at the Dixdc1 locus are viable, fertile, and grossly normal in development. In extended (unpublished) behavioral assays we have found that loss-of-function at the Dixdc1 locus leads to a specific dose-sensitive behavioral deficit: reduced motivation (Forced Swim and Tail Suspension tests). This deficit is responsive to correction by acute administration of either lithium or a specific GSK3-inhibitor. We have concurrently found that forebrain pyramidal neurons from Dixdc1 knock-out animals have reduced dendritic spines and glutamatergic synapses - tests of the response of this synapse phenotype to lithium and to pharmacological GSK3 inhibition are ongoing at the time of this abstract. Unlike some other Wnt signal scaffold proteins that we have studied previously, loss of Dixdc1 does not lead to more generally reduced dendrite complexity as quantified by Sholl analysis, nor to changes in inhibitory synaptic markers. Excitatory synapse phenotypes in the mouse KO can be rescued by expression of a wild type human DIXDC1 cDNA.

Conclusions: Our data support a role for the Dixdc1 protein in the generation and maintenance of excitatory synapses in the forebrain. Our data further suggest that this occurs downstream of activity in the Wnt/beta-catenin pathway within developing neurons. Behavioral consequences of Dixdc1 inactivation in mice are relatively specific for motivation and can be reversed by lithium administration or by pharmacologically-specific GSK3 inhibition. This supports a role for this protein and for its associated pathway in major affective disorders and their pharmacological treatment. Using this animal model, we are testing for differences in the signaling and neurodevelopmental activities of DIXDC1 SNVs found in psychiatric patients. Our studies provide independent support for the contribution of abnormal Wnt/beta-catenin signaling pathway activation during neural development to psychiatric pathogenesis.

Keywords: Wnt signaling, Excitatory Synapses, Motivation, Lithium.

Disclosure: Nothing to Disclose.

T3. Toward an Understanding of Eating Disorders: Are PVH PACAP Neurons an Interface Between Stress Response and Feeding Behavior?

Rachel Ross*, Michael Krashes, Bhavik Shah, Bradford Lowell

Harvard Medical School/Massachussets General Hospital, Boston, Massachusetts

Background: Pituitary adenylate cyclase activating polypeptide (PACAP) is a neuromodulator implicated in human anxiety and feeding behavior. Notably, high expression of PACAP is found in the paraventricular nucleus of the hypothalamus (PVH), a region known to be involved metabolism and coordinating sympathetic activation. Furthermore, mouse knock out models of PACAP, and of single minded-1 (sim-1), a protein that is required for development of the PVH, display behaviors that point to a potential role in both anxiety and eating disorders. However, the PACAP-expressing neurons of the PVH have yet to be examined in this regard. We are harnessing genetic tools that focus on PACAP neurocircuitry in the PVH to investigate whether PVH PACAP neurons, and PACAP itself, serve as an interface for the integration of stress and feeding related signals.

Methods: Taking advantage of cre-lox technology we have devised two approaches to investigate the importance of PACAP neurons in the PVH. First, we generated knock-in mice expressing cre-recombinase under the control of the PACAP promotor. The PACAP-cre mice facilitate real time manipulation of PACAP expressing neurons, through neuroanatomically directed viral injections conferring expression of chemogenetic tools. These neurons are then stimulated or inhibited by intraparenchymal injection of the otherwise inert molecule, clozapine-N-oxide (CNO). The same animals injected with saline serve as controls, and data is analyzed by paired t-test. In our second approach, we created lox-PACAP mice that possess loxP sites flanking the 2nd exon of the PACAP gene, allowing for deletion of functional PACAP in the presence of cre-recombinase. The PACAP-lox mice were bred with Sim1-cre mice to induce the conditional knock-out of PACAP within the PVH in order to discern the specific effect of PACAP in the circuit. Mice were subjected to assays for energy homeostasis and anxiety, and compared to littermate controls using ANOVA.

Results: Acute chemogenetic stimulation of paraventricular hypothalamic PACAP neurons caused an 2.5x increase in food intake during the light cycle, when animals do not normally eat, and a trend toward increased anxiety like behavior, evidenced by less time spent in the open arms of the elevated plus maze. Conditional knock out animals lacking PACAP in Sim1-expressing neurons show a small, but significant, increase in body weight (10%), as well as a trend to decrease anxiety-like behavior.

Conclusions: Stimulation of PVH PACAP neurons leads to hyperphagia and increased anxiety, while knock out of PACAP from the same neurons leads to weight gain and decreased anxiety-like behavior. The non-reciprocal difference in behavioral effects of stimulating synaptic transmission versus effects of genetic knock out of PACAP in the same neurons points to the complex role neuromodulators play in neuronal communication. This circuitry may be important in understanding the genesis of eating disorders, as well as phenotypic differences in feeding behavior seen in anxiety and mood disorders.

Keywords: eating disorders, energy homeostasis, PACAP, PVH.

Disclosure: Nothing to Disclose.

T4. Stem Cell Therapy as a Candidate Treatment Approach for Neural Plasticity Change in Refractory Depression

Toshikazu Saito*, Wataru Ukai, Yoshiyasu Kigawa, Takao Ishi, Kengo Furuse, Hanako Tsujino, Masaya Tayama, Eri Hashimoto

Sapporo Medical University School of Medicine, Sapporo, Japan

Background: Repeated exposure to environmental adversity during development can cause treatment resistant, refractory depression. Although genetic research has revealed that the vulnerability to depression is partially heritable, there is also strong support for the pathogenic role of early environmental adversity in the development and expression of this disorder. Prenatal alcohol exposure has been identified as a potential risk factor for refractory depression and a high rate of comorbid depression has been reported in fetal alcohol spectrum disorder (FASD) patients. Stem cell-based regenerative therapy is considered a promising cellular therapeutic approach for some patients with incurable brain diseases. Previously, we found that intravenous neural stem cell (NSC) treatment reversed impaired memory/cognitive function and social interaction behaviors in rats tested using a fetal alcohol exposure-induced psychiatric disorder model.

Methods: In this study, we tested this approach to treat the refractory type of depression model rats caused by the combined stress exposures to prenatal ethanol and adolescent corticosterone treatments.

Results: We showed that aspects of depressive-like behavior resulting from a forced swim test were reversed by the combined treatment of an antidepressant and the intravenous administration of fetal rat brain derived NSCs with atelocollagen which was used for reducing immune rejection and for potentiate effective migration of administrated cells into the brain. In addition, we showed that alterations of PV-containing GABAergic interneurons and synaptic density protein levels resulting from the combined treatment of antidepressant and NSCs in this model.

Conclusions: Our results established a new role for PV-containing GABAergic interneurons associated with treatment resistant, refractory depression-like behavior in rats and suggested possible new therapeutic options of stem cell therapies for treating refractory types of psychiatric disorders.

Keywords: depression, neural stem cell, regenerative therapy, fetal alcohol syndrome.

Disclosure: Suntory.

T5. Novel Role of Microtubules in the Dysregulation of Reward Learning By Cocaine

Erin Calipari*, Michael Cahil, Diane Damez-Werno, Deena Walker, Joseph Landry, Yasmin Hurd, Eric Nestler

Icahn School of Medicine at Mount Sinai, New York, New York

Background: Reward learning for natural reinforcers is robust, and cue presentation following periods of abstinence elicits seeking behavior for the previously paired reward. While reward learning for drugs is similar to natural rewards, abstinence results in seeking behavior that increases (“incubates”) with withdrawal. Incubation does not occur for natural rewards and highlights how drugs of abuse adapt normal learning mechanisms to exert potent control over motivated behaviors. Reward learning has been attributed to the formation of new synaptic connections on medium spiny neurons, and the mechanism by which this process occurs, and how it is dysregulated in addiction, is not well understood. Recently it was determined that microtubules play an integral role in dendritic spine structure/stability suggesting that they may have an important role in this process; however, their function in vivo as well as their role in learning has not previously been examined. Utilizing self-administration paired with viral approaches, we dissected the specific mechanisms driving microtubule signaling in spines and determined how these pathways converge to drive addictive behaviors.

Methods: Animals were allowed to self-administer cocaine (0.8mg/kg/inj) or saline on a fixed-ratio one schedule of reinforcement for 10 consecutive days. Following a 24hr or 30 day withdrawal period, drug seeking was assessed and tissue was processed for neurochemical analysis. Western blot hybridization on the synaptic fraction was used to determine levels of the dendritic spine/microtubule associated proteins (MAPs): EB3, β-tubulin, Src kinase, and the NR2B subunit of the NMDA receptor. Using confocal imagining and dendritic spine analysis, alterations in spine density and morphology in the NAc following cocaine exposure were determined. Further, the role of microtubule polymerization in the rewarding/reinforcing effects of cocaine was confirmed with behavioral analysis, whereby the microtubule inhibitor podophyllotoxin was infused directly into the NAc before daily conditioned place preference (CPP) sessions. Finally, viral-mediated gene transfer in self-administering animals was used to over-express targets (EB3 and Srcin1-a negative regulator of Src kinase) via HSV vectors to determine their effects on the reinforcing efficacy of cocaine.

Results: Here we show that following cocaine self-administration and 24hr withdrawal Src phosphorylation at the active site (Y416) was increased, while microtubules and MAPs were unchanged. The initial increases in Src phosphorylation at 24hrs likely mediate the increased cocaine reinforcement observed at that time point, as overexpressing Srcin1 (which reduces Src activity) reduced the motivation to administer self-administer cocaine. Following longer withdrawal periods, Src phosphorylation returned to control levels, and levels of beta-tubulin and EB3 were elevated. Increases in EB3, a MAP that regulates microtubule polymerization, were positively correlated with cocaine intake during self-administration. Additionally, the increases in MAPs were associated with a cocaine intake-dependent increase in levels of NR2B, a traditional marker of dendritic spine stability. These changes likely regulate cocaine reward and reinforcement, as evidence by the ability of podophyllotoxin to reduce CPP for cocaine.

Conclusions: Here we show that following cocaine self-administration and withdrawal microtubule-associated signaling is increased and associated with cocaine intake and drug seeking behavior. We hypothesize that, cocaine self-administration leads to the formation of immature dendritic spines, and that during abstinence/withdrawal microtubule entry into spines enhances EB3-dependent signaling cascades leading to long-term stabilization of spines and enhanced cue-reward associations that incubate and drive addiction-related behaviors. Here we show that inhibiting microtubule polymerization reduces cocaine reward, suggesting that microtubule inhibitors, which are already FDA approved for a number of other conditions, may serve as efficacious pharmacotherapies to reduce drug seeking following prolonged withdrawal periods, and thus prevent relapse.

Keywords: Self-Administration, Medium Spiny Neuron, Dendritic Spine, Cocaine.

Disclosure: Nothing to Disclose.

T6. Translational Research Supporting the Relevance of PTRPG to the Etiology of Schizophrenia

Arnaud Cressant, Dolores Malaspina,* Jing Kong, Jacques Caliber, Jean-Marie Launay, Francoise Lazarini, Moses Chao, Sylvie Granon, Shiela Harroch

New York University School of Medicine, New York, New York

Background: PTRPG, a phosphatase of receptor protein tyrosine, is expressed in the CNS throughout vertebrate development and adulthood and plays important roles in adult synaptic plasticity and cognitive function. The gene had previously been linked to affective disorders and this translational study considered its relevance to schizophrenia.

Methods: A de novo loss of function point mutation was discovered in PTRPG for a sporadic schizophrenia case compared to the parents using next generation sequencing in 11 parent-child trios derived from the Jerusalem Perinatal Schizophrenia Study as a nested sample. A leading mouse genetics group examined the phenotype of PTPRg-/- mice were compared to wild type mice.

Results: Significant differences between PTPRg-/- and wild type mice were demonstrated in the mutants persistence in swimming in the forced swim test, longer latency to escape anxiogenic areas in the light/dark test, and decreased time exploring novel objects. These effects are consistent with perseverative behaviors and decreased emotional processing. These mice also demonstrated robustly elevated prefrontal and amygdala levels of 5-HT, DA and NA.

Conclusions: Findings that the knock out mice demonstrated reduced responsiveness to external social stimuli, motor deficits, enhanced sensory processing for acoustic stimuli and reduced performance with cued fear conditioning the robust evidence linking the mouse phenotype to the clinical picture of schizophrenia. However this translational collaboration bolsters support for the conclusion that this phosphatase gene is specifically relevant to behavior and that it is a susceptibility gene for schizophrenia more comprehensively than either bench-top or bedside approaches could individually accomplish.

Keywords: phosphatase, receptor tyrosine kinase, genetics, schizophrenia.

Disclosure: Nothing to Disclose.

T7. Epigenetic Dysregulation of MEF2C in Schizophrenia

Amanda Mitchell*, Venu Pothula, Erica Shen, W.E. Bunney Jr., Andree Lessard, Schahram Akbarian

Icahn School of Medicine at Mount Sinai, New York, New York

Background: Only about 2% of the human genome codes for protein and the remaining 98% harbor the majority of sequence variants and polymorphisms implicated in schizophrenia and other major psychiatric disease. Here, we use next generation sequencing to profile the epigenomic landscape of prefrontal cortex (PFC) neurons from 17 subjects with schizophrenia (SCZ) and matched controls, focusing on trimethyl-histone H3-lysine 4 (H3K4me3), a mark sharply regulated at transcription start sites and a subject of enhancers and other additional cis-regulatory elements associated with gene expression.

Methods: Neuronal (NeuN+) nuclei were collected from PFC and processed by ChIP-seq. ChIP-seq data was analyzed using bowtie, MACS, homer, and RSAT motifs. Motif analysis predicted variants of the MADS box transcription enhancer factor 2, polypeptide C, (MEF2C) motif from the differential peaks. MEF2C is an enhancer thought to function as a transcription factor important in cell fate and identify.

Results: Altogether we identified 1331 up- and 637 down peaks in our SCZ cohort with a significant enrichment for sequences that matched variants of the MEF2C transcription factor motif. This included a risk-associated regulatory sequence of the MEF2C gene itself (rs16867576). SCZ subjects with increased H3K4me3 at the MEF2C promoter showed a significant decrease in MEF2C mRNA. Disease-associated H3K4me3 changes were enriched for gene categories related to cell-cell adhesion, synaptic membrane and postsynaptic signaling.

Conclusions: These results point to changes in the histone methylation landscape of PFC neurons in SCZ, affecting neuronal signaling and communication. Further work is necessary to explore whether epigenetic dysregulation of MEF2C could play a key role for PFC dysfunction in SCZ.

Keywords: epigenetics, transcription factor, PFC, schizophrenia.

Disclosure: Nothing to Disclose.

T8. Calsyntenin-3: Molecular Architecture and Interaction with Neurexin 1alpha

Zhuoyang Lu, Yun Wang, Fang Chen, Huimin Tong, Sekhar Reddy, Lin Luo, Suchithra Seshadrinathan, Lei Zhang, Luis Marcelo Holthauzen, Ann Marie Craig, Gang Ren, Gabrielle Rudenko*

University of Texas Medical Branch, Galveston, Texas

Background: Calsyntenin 3 (Cstn3 or Clstn3) is a recently identified synaptic organizer that promotes synapse development. Cstn3 localizes to the postsynaptic membrane and triggers presynaptic differentiation of inhibitory and excitatory synapses. Cstn3 knockout mice show decreased inhibitory and excitatory synaptic densities and deficits in synaptic transmission. Importantly, calsyntenin members play a role in memory and learning. Cstn3 was recently shown to interact with neurexin 1alpha in cell-based assays (1). Neurexins are synaptic organizers implicated in diseases such as autism spectrum disorder and schizophrenia, and they play an essential role in synapse maturation and synaptic transmission. Alpha-neurexins are tethered predominantly to the presynaptic membrane, and their extracellular domains interact with many partners in the synaptic cleft, including neuroligins, LRRTM family members, alpha-dystroglycan, latrophilin, cerebellins, neurexophilin and the GABAA-receptor. The interaction between Cstn3 and n1alpha is contentious however (2) and it is not clear if these molecules bind directly to form a trans-synaptic complex promoting synaptic differentiation.

Methods: We used a combination of biochemical, biophysical and structural methods to gain insight into the structure and function of Cstn3 and its interaction with n1alpha. We used biochemical and cell-based assays to probe the molecular state of Cstn3. We used a combination of electron microscopy and tomography to reveal the architecture of purified Cstn3 ectodomains. Finally, we tested the direct interaction between Cstn3 and n1alpha using solid phase binding, surface plasmon resonance and cell-based assays.

Results: We show that the Cstn3 ectodomain forms monomers as well as tetramers. While the Cstn3 monomer is flexible, adopting more open or more closed conformations, the Cstn3 tetramer is more fixed and resembles an unopened flower. We show further that the extracellular domains of Cstn3 and n1alpha interact directly and that both Cstn3 monomers and tetramers bind n1alpha with nanomolar affinity. Furthermore, Cstn3 uses a fundamentally different mechanism to bind n1alpha compared to other neurexin partners such as the synaptic organizer neuroligin.

Conclusions: Our data suggest how Cstn3 as a synaptic organizer on the postsynaptic membrane, particularly in the tetrameric form, may assemble radially symmetric trans-synaptic bridges with the presynaptic synaptic organizer n1alpha to recruit and spatially organize proteins into networks essential for synaptic function. These trans-synaptic complexes involve proteins that are implicated in neuropsychiatric disorders and regulate crucial functional neuronal circuits. By studying the molecular interactions between molecules that guide inhibitory and/or excitatory synapse formation we hope in future to identify new therapeutic targets that could be leveraged for the treatment of neuropsychiatric disorders. (1) Pettem KL, et al. (2013) The Specific α-Neurexin Interactor Calsyntenin-3 Promotes Excitatory and Inhibitory Synapse Development. Neuron, 80(1):113-128. (2) Um JW, et al. (2014) Calsyntenins function as synaptogenic adhesion molecules in concert with neurexins. Cell Rep, 6(6):1096-109.

Keywords: synaptic organizer, calsyntenin, neurexin, structure-function.

Disclosure: Nothing to Disclose.

T9. Molecular Mechanisms of Opiate-induced Plasticity

David Dietz*, Gabrielle Schroeder, Kevin Braunscheidel, Clarisse Panganiban, Amy Gancarz

State University of New York at Buffalo, Buffalo, New York

Background: Repeated exposure to drugs of abuse causes morphologic changes to neurons within the mesolimbic dopamine circuitry. There is a great deal of evidence demonstrating that chronic psychostimulant exposure increases the number dendritic spines on medium spiny neurons in the nucleus accumbens. In contrast, exposure to opioids such as morphine and heroin, which similarly induce behavioral sensitization, leads to a decrease in dendritic spine density. However, there is minimal data regarding the molecular events that regulate this opiate-induced plasticity.

Methods: To examine the molecular mechanisms of dendritic spine plasticity following behavioral sensitization to a morphine regimen (10 mg/kg daily) or heroin self-administration (0.075 mg/kg/infusion), tissue punches from several brain regions were collected and analyzed for mRNA, protein and chromatin regulation.

Results: Following opiate-induced behavioral sensitization and heroin self-administration, there is a decrease in the expression of the actin bundling protein drebrin. This decrease results from an increase in hdac2 expression and a concordant decrease in H3 acetylation at the transcriptional start site of the drebrin promoter.

Conclusions: These results implicate epigenetic control of cytoskeletal protein regulation as a mechanism for opiate-induced morphologic changes. Ongoing studies are examining in more depth the critical functional role of hdac2 and drebrin in the regulation of structural, cellular and behavioral plasticity in response to repeated opiate exposure.

Keywords: opiates, self-administration, dendritic spines, sensitization.

Disclosure: Nothing to Disclose.

T10. A New Roadmap for Brain Development: Regionally Specialized Astrocytes in Neural Circuit Formation and Function

Anna Victoria Molofsky*, Kevin Kelley, Hui Hsin Tsai, Sergio Baranzini, Eric Ullian, David Rowitch

University of California San Francisco, Langley Porter Psychiatric Institute, San Francisco, California

Background: Neural circuits form over a prolonged period of developmental time, during which the brain expands dramatically in size. How is positional information maintained throughout development, ensuring the accuracy and specificity of neural circuit formation? Astrocytes, a type of glial cell, are implicated in a growing number of neurodevelopmental diseases, including autism-associated syndromes (Rett and Fragile X), depression, and neurodegenerative disease. While astrocytes are known to support neurons by performing housekeeping functions in the mature brain, their normal role during development is unclear. A major obstacle has been understanding whether astrocytes are heterogeneous, and if so, what might distinguish astrocyte subtypes at a molecular or functional level. We discovered that astrocytes in different regions of the central nervous system (CNS) are different “from birth” as a result of embryonic patterning. Here we show that developing region-restricted astrocytes are molecularly and functionally specialized. Using the mouse spinal cord sensorimotor circuit as a model, I demonstrate that astrocytes in the 'sensory' and 'motor' portions of the circuits are molecularly distinct, and encode multiple positional cues such as the guidance molecule Semaphorin3a (Sema3a.) Deletion of this gene from astrocytes led to defects in axon guidance and neuronal polarity, followed by neuronal dysfunction, altered excitatory/inhibitory synaptic balance, and neuronal death. Thus, this novel form of astrocyte functional heterogeneity is essential for normal neural circuit development. Further investigating this critical function of astrocytes could shed light on aberrant neural circuit development in diseases like autism, schizophrenia, and many others.

Methods: To test whether astrocytes in different regions of the developing mouse CNS are molecularly and/or functionally heterogeneous, I focused on the spinal cord sensorimotor circuit – the simplest monosynaptic circuit in the CNS, which is functionally divided into dorsal (sensory) and ventral (motor) areas. Using the transgenic mouse line Aldh1l1-eGFP (Gensat), I isolated astrocytes and non-astrocytes from dorsal and ventral spinal cord at postnatal day 7 by flow cytometry. Using affymetrix microarray gene chips (Mouse Gene 1.0) were used for gene expression profiling, we identified differential expression of numerous axon guidance and extracellular matrix molecules, including the guidance cue semaphorin3a (Sema3a), which is expressed in ventral radial glia and ventral astrocytes, but not dorsal astrocytes from mid embryogenesis through early postnatal development (E14.5-P14). Further experiments focused on the role of Sema3a in sensorimotor circuit development. In vivo approaches include conditional deletion of Sema3a from developing astrocytes using Aldh1l1-cre and the human GFAP cre (hGFAPcre) transgenic lines, both of which were astrocyte-specific in the spinal cord. In vitro work utilized astrocyte-neuronal cocultures. Dorsal and ventral spinal cords at P0-P1 were cultured at high density in the presence of serum to generate astrocyte monolayers, which retained differential expression of Sema3a. Sensory neurons from mouse E13.5 dorsal root ganglia, or rat embryonic motorneurons were isolated and cocultured with astrocytes. In some experiments, recombinant Sema3a was added to motor neurons in culture.

Results: Ventral and dorsal astrocytes differentially expressed numerous genes, many of which were extracellular matrix genes or guidance cues, including Epha5 (Eph receptor A5), reelin, and Sema3a. Sema3a was ventral-astrocyte encoded from mid-embryogenesis until at least postnatal day 15. Sema3a deletion in radial glia and immature astrocytes via Aldh1l1-cre led to abnormal ventral extension of DRG sensory axons and perinatal lethality. Loss of astrocyte-encoded Sema3a led to motor neuron misorientation with respect to the ventral root, as evidenced by abnormal positioning of axon initial segments. Sema3a deletion from astrocytes using the hGFAPcre led to a less penetrant phenotype: these mice that survived into adulthood. hGFAPcreSema3a fl/fl mice had motor neuron misorientation and abnormal dendritogenesis. Later in development they had altered excitatory/inhibitory balance, including decreased vglut1 and increased VGAT puncta on ventral motor neurons, which was confirmed electrophysiologically (decreased excitatory post synaptic currents and increased inhibitory postsynaptic currents.) Motor neurons were also hyperexcitable after loss of astrocyte-encoded Sema3a. By adulthood, there was a two-fold loss of large alpha-motor neurons. To confirm that there were direct effects of astrocyte-encoded Sema3a on motor neurons, we performed a series of coculture experiments. Dorsoventral differences in region restricted astrocytes were maintained in culture. Furthermore, Sema3a dependent differences in axon outgrowth were also evident when dorsal or ventral astrocytes were cocultured with DRG neurons. Ventral spinal cord astrocytes also repelled motor neuron axons in a Sema3a dependent manner. Recombinant Sema3a had a dose-dependent effect on motor neuron survival that was correlated with its effects on neuronal polarization.

Conclusions: Strikingly, many well-known guidance cues are encoded and maintained by astrocytes well into postnatal life. I identified a panel of genes differentially expressed by dorsal and ventral astrocytes, and tested the functional roles of one of these astrocyte-encoded genes: the guidance molecule Semaphorin 3a. I found that this astrocyte encoded positional cue played key roles in neural circuit refinement, including guiding axons, orienting neurons, promoting dendritogenesis, maintaining excitatory/inhibitory synaptic balance, and ultimately promoting neuronal survival. These data demonstrate that developing astrocytes are a positional scaffold that maintains guidance cues during the key developmental period of synaptogenesis and neural circuit refinement. Thus, astrocytes play a key role in normal circuit formation. This suggests that defects in astrocyte positional encoding could potentially contribute to a wide array of neurodevelopmental diseases, including autism and schizophrenia, diseases of synaptogenesis for which neuroanatomical substrates are not well understood.

Keywords: glia, astrocytes, neural circuits, autism.

Disclosure: Nothing to Disclose.

T11. Role of Hippocampal ΔFosB in Spatial Learning and Cocaine Responses

Andrew Eagle, Paula Gajewski, A.J. Robison*

Michigan State University, East Lansing, Michigan

Background: The hippocampus is necessary for learned associations between the environmental context and salient stimuli, including drugs of abuse. In addition, stimulation of the hippocampus causes enhanced mesolimbic dopamine function, suggesting that the hippocampus may play a role in both drug acquisition and in drug reward. However, it is unknown how chronic exposure to drugs alters hippocampal gene expression and function, and whether such alterations may underlie the aberrant valuation of drugs and the strong associations of drugs with environmental cues that comprise the addiction phenotype. Although the transcription factor ΔFosB is induced in the hippocampus by chronic drug exposure, and its expression in other brain regions regulates drug reward, the role of ΔFosB in hippocampal function remains unknown. Therefore, we explored the induction and function of ΔFosB in the hippocampus. Using viral-mediated overexpression or inhibition of ΔFosB in the dorsal hippocampus, we sought to determine the role of ΔFosB in general learning, association of cocaine with environment, and hippocampal cell morphology and physiology.

Methods: Male, C57Bl6/J mice from Jackson Labs were used in this study and all experiments were performed in accordance with Michigan State University IACUC-approved protocols. Behavioral assays were performed 21-23 d post-surgery in AAV studies and 1-3 d post-surgery in HSV studies. Novel object recognition was performed over three days (1 hr habituation day 1, 30 min acquisition day 2, 5 min testing day (3) in a square white polyvinylchloride foam box (38 cm x 38 cm x 35 cm). Contextual fear conditioning was performed in an operant chamber (Coulbourn Instruments) using 3 mild electric footshocks (0.8 mA, 1 sec) and the following day freezing behavior was video-scored by a blind, independent observer. Temporally dissociated passive avoidance was assessed over 5 days in a light-dark passive avoidance chamber (Coulbourn Instruments). Footshock (0.8 mA, 2 sec) was delivered 5 min after entry into the dark chamber. Conditioned place preference was performed in custom 3-chambered boxes (San Diego Instruments), with two days of 5 mg/kg cocaine pairing. ΔFosB immunohistochemistry was visualized using DAB staining (SC-48 antibody, Santa Cruz), and ΔFosB Western blotting used the 5G4 antibody (Cell Signaling). Chromatin immunoprecipitations were performed from nuclear fractions with micrococcal nuclease digestion using antibodies against various histone marks (AbCam, Millipore). Viral vectors were purchased from the Viral Gene Transfer Core at MIT (HSV) or from the UNC Gene Therapy Center Vector Core (AAV) and injected into the hippocampus by stereotaxic surgery (10° angle; -2.2 mm anteroposterior (AP),±2.5 mm mediolateral (ML) and 0.6 uL of purified high-titer virus was separately infused (0.3 uL/infusion) over 3 min periods at 2 sites: -2.1 mm dorsoventral (DV) and at -1.9 mm DV). Dendrites were visualized by confocal microscopy and analyzed using Neuron Studio software.

Results: We find that ΔFosB is induced in the hippocampus, but not the nucleus accumbens, by exposure to novel environments and spatial learning, and is induced in both brain regions by chronic cocaine exposure. We also demonstrate that this induction of ΔFosB by spatial learning is specific to the CA1 region of the hippocampus. In addition, we find changes in multiple covalent modifications of histones associated with the FosB gene in animals exposed to cocaine or spatial learning, indicating that induction of ΔFosB in the hippocampus may be regulated by epigenetic changes in chromatin structure. We found that inhibition of ΔFosB transcriptional activity using either HSV- or AAV-mediated expression of ΔJunD in the dorsal hippocampus abrogated spatial learning in the novel object recognition, fear conditioning, and temporally displaced passive avoidance tasks. Importantly, we found that ΔJunD expression in the dorsal hippocampus also prevented conditioned place preference to low doses of cocaine. We are currently examining whether overexpression of ΔFosB in the dorsal hippocampus affects spatial learning and cocaine responses using the same behavioral paradigms. We show that overexpression of ΔFosB can control dendritic spine formation in the CA1 but not the CA3 region of the hippocampus, correlating with our expression data. Finally, we examine whether overexpression of ΔFosB affects the synaptic function of hippocampal neurons using a variety of electrophysiological assays.

Conclusions: ΔFosB is a unique candidate factor for spatial memory associations because, unlike other immediate early genes, it possesses remarkable stability (weeks in a living brain). Here, we demonstrate that ΔFosB is induced in the mouse hippocampus specifically by spatial learning and by cocaine exposure. Because known gene targets of ΔFosB are key players in hippocampal learning and synaptic function (i.e., CaMKII), we hypothesize that ΔFosB regulates hippocampal synaptic function. Indeed, we find that ΔFosB overexpression increases dendritic spines specifically in the CA1 region, and regulates hippocampal synaptic properties. We suggest that these ΔFosB-mediated changes in hippocampal synapses are responsible for the critical role we demonstrate for hippocampal ΔFosB in spatial learning and cocaine responses. Overall, these studies suggest that hippocampal ΔFosB may be important for the associations of drug and environment that underlie location-induced drug craving and relapse, and thus that hippocampal ΔFosB and its downstream targets may represent novel therapeutic inroads for the treatment or prevention of drug addiction.

Keywords: ΔFosB, hippocampus, cocaine, learning.

Disclosure: Nothing to Disclose.

T12. The Role of NR2B in CA1 Pyramidal Spine Morphology Following Morphine Conditioned Place Preference

Amanda Fakira*, George Portugal, Ream Al-Hasani, Sam Golden, Scott Russo, Michael Bruchas, Dave Sulzer, Jose Moron-Concepcion

Columbia College of Physicians and Surgeons, New York, New York

Background: The hippocampus plays a critical role in the development of opiate preference and cravings triggered by environmental cues. Increasing studies demonstrate that long term memories are maintained by structural and functional plasticity in post-synaptic dendritic spines of which there are three types; thin (65%), mushroom (25%) and stubby (10%). Studies have demonstrated that morphine decreases spine density on hippocampal neurons. However, these studies used Golgi staining and 2D counts manual counts which cannot differentiate spine type. New techniques using high resolution 3D reconstruction of fluorescently labeled dendritic segments and automated spine counting to examine the types of spines have been developed. Dendritic spine morphology, controlled by actin polymerization, is regulated by NR2B-mediated Ca2+influx and activation of the Rho signaling cascade. By determining the types of spines affected by morphine and mechanisms that underlie this structural plasticity we can highlight new therapeutic targets in the treatment of opiate relapse.

Methods: For these studies, a morphine conditioned place preference paradigm in which subcutaneously (s.c.) morphine (5, 8, 10 and 15 mg/kg) is paired with visual cues (horizontal vs vertical stripes) over 4 consecutive days. After preference testing, mice undergo extinction training for 5 days followed by morphine induced reinstatement. Following preference and extinction tests mice are sacrificed for either confocal imaging or western blotting. Subcellular fractionation and western blot analysis of NMDA receptor subunits, NR1 NR2B and NR2B, and Rho family signaling proteins, Rho and Rac1, in hippocampal lysates are performed. For confocal imaging intra-hippocampal injections of AAV-CamKII-eYFP are performed three weeks prior to preference test after which coronal brain sections are used to collect high resolution images of CA1 pyramidal neuron dendritic segments. Raw confocal images are deconvolved using Velocity software and then automatic spine counts are performed using NeuroStudio software. All experimental protocols in animal studies were approved by the Institutional Animal Care and Use Committee at Columbia University.

Results: We observed that following Mor CPP NR2B, NR1 and NR2A NMDA receptor subunits are increased in post-synaptic density fractions of the hippocampus. Additionally, these mice have decreases in the total number of spines on CA1 pyramidal neurons which is driven by a decrease in thin spines since mushroom spines were unaffected. Moreover, initial studies indicate that total Rho (A and B) expression is increased in the hippocampus following Mor CPP and that Rac1 expression is unaffected. Further analysis reveals that only post-synaptic expression of NR2B remains increased following extinction training. Finally, we are able to prevent morphine induced reinstatement by infusing NR2B antagonist, Ifenprodil (2μg/ml), intra-hippocampally, prior to acute morphine challenge.

Conclusions: NR2B has been shown to be a critical mediator of structural plasticity. Our studies indicate that NR2B-Rho signaling mechanisms may be mediating opiate induced decreases in thin spines. Our data also demonstrate that elevated NR2B synaptic expression is essential for morphine induced reinstatement. Future studies are designed to reveal if NR2B-mediated Ca2+and Rho signaling is involved in decreases in thin spines and whether this decrease persists following extinction training. Altogether, our data highlight new mechanisms underlying context-dependent drug seeking behavior and may lead to development new therapeutic approaches to preventing relapse.

Keywords: morphine, spines, hippocampus, Rho.

Disclosure: Nothing to Disclose.

T13. A Neural Circuit Mechanism for Differentiating Positive and Negative Associative Memories

Praneeth Namburi, Anna Beyeler, Suzuko Yorozu, Romy Wichmann, Stephanie Holden, Kim Mertens, Sarah Halbert, Ada Felix-Ortiz, Jesse Gray, Ian Wickersham, Kay Tye*

Massachusetts Institute of Technology, Cambridge, Massachusetts

Background: The ability to differentiate environmental stimuli that predict positive or negative outcomes is critical for survival, and perturbations of emotional processing can be manifested in many psychiatric disease states. Synaptic plasticity in the amygdala has been shown to be critical for the acquisition of associative memories, both positive and negative. While there is evidence that different populations of neurons in the amygdala may encode fearful or rewarding associations, the identifying features of these populations and the circuit and synaptic mechanisms of differentiating positive and negative emotional valence has remained an enigma.

Methods: We trained animals on fear or reward conditioning tasks before evaluating synaptic strength by performing whole-cell patch-clamp recordings in basolateral amygdala (BLA) neurons that were labeled with retrogradely-travelling beads injected into the nucleus accumbens (NAc) or centromedial amygdala (CeM). We then used rabies viral vectors to retrogradely express ChR2 in specific projections within the BLA. We also filled cells with biocytin during patch-clamp recordings to allow for post-hoc morphological reconstruction. Finally, we dissociated retrogradely labelled BLA neurons projecting to the NAc from those that projected to the CeM and performed RNA-Seq.

Results: Here we show that neurons in the basolateral amygdala complex (BLA) projecting to the nucleus accumbens (NAc) or the centromedial nucleus of the amygdala (CeM) undergo opposing synaptic changes following fear or reward conditioning. We also show that photostimulation of BLA cell bodies projecting to the NAc is positively reinforcing while photostimulation of BLA neurons projecting to the CeM causes aversion. Because we could not detect defining characteristics of these functionally-distinct neuronal populations based on electrophysiological firing properties, nor morphology, we performed RNA sequencing to characterize the transcriptome of these populations. RNA-Seq results provided a list of candidate genes that could contribute to differential recruitment of NAc and CeM projectors during fear or reward conditioning. One of the candidate genes differentially expressed in BLA neurons projecting to the NAc and CeM, was the neurotensin-1 receptor. We validated that neurotensin had opposing modulatory effects on glutamatergic inputs to NAc and CeM projectors.

Conclusions: Our findings demonstrate opposing functional roles for amygdala neurons depending on projection target. Even more importantly, these results provide a mechanistic explanation, on both a synaptic and circuit level, for how positive and negative associations can be rapidly formed, represented and expressed within the amygdala.

Keywords: amygdala, fear, reward, plasticity.

Disclosure: Nothing to Disclose.

T14. Chondroitin Sulfate Proteoglycan Abnormalities in the Hippocampus of Subjects with Schizophrenia

Harry Pantazopoulos*, Caroline Sawyer, Stephan Heckers, Sabina Berretta, Matej Markota

Harvard Medical School, Belmont, Massachusetts

Background: Increasing evidence points to the involvement of chondroitin sulfate proteoglycans (CSPGs) in schizophrenia (SZ). CSPGs are involved in processes that shape neural circuits, including migration of neurons, guidance of axons, neuronal and glial cell maturation, and regulation of plasticity and firing properties, all relevant to SZ. During late postnatal stages, coinciding with the age of onset of SZ, CSPGs contribute to the assemblage of perineuronal nets (PNNs), specialized extracellular matrix aggregates. Once formed, PNNs stabilize neural connections and contribute to the acquisition and maintenance of mature neuronal properties. PNNs envelop fast-firing interneurons, including neurons expressing parvalbumin (PVB). We have previously reported decreases of PNNs in the amygdala, entorhinal cortex and prefrontal cortex of SZ, regions in which numbers of PVB neurons are not altered, indicating that PNN decreases do not reflect loss of these neurons. In addition, we reported marked increases of CSPG expressing glial cells. In order to examine how widespread these abnormalities are, we focused on a region strongly implicated in SZ, the hippocampus (HP).

Methods: We tested the hypothesis that CSPG abnormalities are present in the HP of SZ subjects. Wisteria floribunda agglutinin (WFA) lectin, a broad spectrum CSPG histochemical marker commonly used to study PNNs, was used to label CSPGs in the HP of control (n=19), SZ (n=14), and bipolar disorder (BD) (n=14) subjects. Numerical densities (Dn) and total numbers (N) of WFA positive (WFA+) PNNs and glial cells were counted using computer-assisted light microscopy; group differences were tested using stepwise linear regression models. Statistical correlations between PNN measures and previously published PVB measures from the same subjects were used to examine the relationship of PNNs with PVB neurons.

Results: In subjects with SZ, decreases of WFA+PNNs were detected in the whole HP (N, p=0.0002, 51.1% decrease; Dn, p=0.0003, 51.40% decrease), and were significant in all hippocampal subregions with the exception of stratum lacunosum moleculare. N and Dn of WFA+PNNs were significantly correlated with N and Dn of PVB neurons reported previously in these subjects. In comparison, WFA+glial cells were markedly increased in the whole HP of SZ subjects (N, p=0.0001, 1266.9% increase; Dn, p=0.0001, 1205.3% increase). The majority of WFA+glial cells were concentrated in the dentate gyrus, an established site of adult neurogenesis, although increases were present all subregions with the exception of stratum lacunosum moleculare. In addition, numbers of WFA+glia were negatively correlated with age, suggestive of a decrease of adult neurogenesis with age reported in this region. In subjects with BD, N and Dn of WFA+PNNs and glia were normal in all HP regions.

Conclusions: WFA+PNN and glial abnormalities in the HP are selective for SZ. In comparison to PVB neuron decreases reported in the HP of these SZ subjects, PNN decreases may reflect failure to form or stabilize PNNs, or loss of PVB neurons. Alternatively, PVB neuron maturation may be disrupted, resulting in impaired PNN formation and decreased PVB. Ongoing investigations are examining these possibilities. PNN decreases in absence of neuronal loss in other regions, and genetic associations of CSPG and matrix metalloproteases, support the idea of a primary PNN abnormality. If so, decreases of PNNs in the HP may result in decreased inhibition, affecting interconnected brain regions, as suggested by rodent work. Lack of WFA+PNN decreases in BD subjects which PVB decreases were reported may reflect loss of PVB neurons that do not have PNNs, possibly from oxidative stress. Evidence that PNNs protect PVB neurons from oxidative stress supports this possibility. Alternatively, decreased PVB may occur in neurons that properly form and maintain PNNs. Increases of WFA+glia in SZ in the dentate gyrus raises the possibility that they are associated with disregulated cell maturation.

Keywords: perineuronal nets, bipolar disorder, parvalbumin, neurodevelopment.

Disclosure: Nothing to Disclose.

T15. Adolescent Intermittent Ethanol and DNA Methylation Mechanisms in Amygdala: A Role in Anxiety-like and Alcohol-drinking Behaviors

Subhash Pandey*, Amul Sakharkar, David Gavin, Huaibo Zhang, Ying Chen, Dennis Grayson

University of Illinois at Chicago, Chicago, Illinois

Background: Binge alcohol drinking during adolescence causes molecular changes in the developing brain resulting in substance abuse and alcoholism in adulthood. DNA methylation is an important epigenetic mechanism that regulates neuronal gene expression during brain development. We investigated the effects of adolescent intermittent ethanol (AIE) exposure on DNA methylation and demethylation mechanisms in the amygdala and its role in anxiety-like and alcohol-drinking behaviors in adulthood.

Methods: In order to model a binge pattern of drinking we administered eight intraperitoneal injections of ethanol (2 g/kg) or n-saline to adolescent rats during post-natal days (PND) 28-41 with a 2-day on/off paradigm. We studied the effects of AIE and adolescent intermittent n-saline (AIS) exposure on DNA methylation/demethylation mechanisms and neuropeptide Y (NPY) expression in the amygdala and the resultant anxiety-like behaviors during adolescence and adulthood (PND 92). We also examined the effects of 5-azacytidine, a DNA methytransferases (DNMT) inhibitor, on anxiety-like and alcohol-drinking behaviors in adulthood.

Results: AIE-exposed animals displayed anxiety-like behaviors 24 hrs after the last ethanol injection (ethanol withdrawal), which persisted into adulthood. Concomitantly, DNA methyltransferase (DNMT) activity, DNMT3b mRNA and DNA demethylating factors, i.e. GADD45a, b and g were found to be altered in the amygdala of AIE-exposed rats during adolescence and some of these changes persisted into adulthood. To understand the down-stream molecular mechanisms by which DNA methylation may be regulating the anxiety-like and alcohol-drinking behaviors, we examined the DNA methylation specific to the NPY gene promoter and NPY protein levels in the amygdala. NPY protein levels were down-regulated in the central and medial amygdaloid structures of AIE-exposed rats at 24 hrs, with a persistent decrease in adulthood. Reciprocally, DNA methylation at the NPY gene promoter was increased in the amygdala of AIE adult rats, which is consistent with the increase in DNMT3b expression and DNMT activity and decrease in the GADD45g levels. To test if DNMT inhibition could reverse AIE-induced anxiety-like and alcohol-drinking behaviors and DNA hypermethylation of the NPY gene, we treated AIS and AIE adult rats with 5’-azacytidine. Treatment with 5’-azacytidine attenuated AIE exposure-induced anxiety-like behaviors and alcohol intake at adulthood and also inhibited the DNA hypermethylation of the NPY gene promoter in the amygdala.

Conclusions: These results suggest that AIE altered the DNA methylation/demethylation pathways in the amygdala during adolescence with long-lasting up regulation of DNMT activity persistent in adulthood causing anxiety-like and alcohol-drinking behaviors most likely via decreased NPY gene expression. These results also raise the possibility of DNMT inhibitors as a promising therapeutic option to treat alcohol-abuse and co-morbid anxiety disorders (supported by NADIA grant from NIH-NIAAA to SCP).

Keywords: DNA methylation, Adolescent binge drinking, Anxiety, Amygdala.

Disclosure: Nothing to Disclose.

T16. Ketamine Produces Structural Plasticity in Dopaminergic Neurons via Co-Activation of MEK-ERK and Akt-mTOR Pathways: Translation from Mouse Primary Culture to Human Ipsc-Derived Neurons

Ginetta Collo, Laura Cavalleri, Federica Bono, Mark Millan, Cristian Chiamulera, Pierfranco Spano, Emilio Merlo Pich*

F. Hoffmann-La Roche, Basel, Switzerland

Background: Addictive drugs produce structural changes in the dopaminergic system, changes that may act as pathological substrate for drug-related chronic symptoms, including craving and relapse. Ketamine, a psychotrope anaesthetic drug with addictive properties, is known to increases dopamine (DA) release, but its capacity to induce structural plasticity is unknown. To explore its occurrence and the key biological mechanisms eventually involved we used two in vitro models: primary cultures of mesencephalic dopaminergic neurons obtained from the mouse embryo (Collo et al. 2009) and dopaminergic neurons differentiated from human IPSC (Kricks et al. 2011). In these models neurons organize functionally and morphologically in local networks and intracellular pathways can be effectively studied, offering a paradigm for assessing disease-relevant targets and effects of pharmacologic intervention.

Methods: Primary neuronal cultures of mesencephalon containing dopaminergic neurons and obtained from E17 mouse embryos as described (Collo et al. 2008, 2012, 2013). Both wild-type and D3KO mice were used. We also derived iPSC from human volunteers and differentiated them into dopaminergic neurons following the method of Kricks et al. (2011) with modification. Cells with dopaminergic phenotype were cultured for 80 days and profiled for co-expression of TH and DAT using morphological and biochemical methods. Ketamine was added to the incubation media at various concentrations. Inhibitor of kinase, dopaminergic antagonists and other compounds were used to pharmacologically probe various pathways. Structural changes in neurons in cultures were studied using confocal, fluorescence and transmission microscopy associated with computer assisted morphometry. Western blood, RNA assessment and dopamine HPLC measurements were performed according to published methods. A group of wild-type and D3KO mice was challenge with ketamine and sacrificed at different time points to study the activation of mTORC1 in the ventral mesencephalon in vivo.

Results: In a series of studies we showed that ketamine (0.01-1 uM) dose-dependently increased dendritic arborisation and soma size of dopaminergic neurons obtained from the mouse embryo or differentiated from human iPS cells. Similar effects were observed using the NMDA antagonist Ro 25-6981 (0.01-1 uM). Structural neuroplasticity induced by ketamine did not occur following pre-treatment with the AMPA antagonists NBQX and GYKI 52466 (0.01-10 uM). These data indicate the presence of functional glutamatargic input on dopaminergic neurons. Accordingly, morphological assessments of both the mouse and human cultures identified interconnected DAergic, glutamatergic and GABA neurons, supporting the existence of complex networks in vitro. The involvement of Akt-mTORC pathway was based on the evidence of ketamine-induced increase of p70S6 kinase phosphorylation that was blocked by LY294002, a PI3 kinase inhibitor, and by rapamycin, a mTORC1 inhibitor. The effects of ketamine were also blocked by the DA D3 antagonist SB277011-A and were absent in primary cultures of D3-KO mice, effects confirmed in vivo. Immunoneutralization of BDNF also blocked the effects of ketamine, suggesting the necessity of the co-activation of the TrkB-MEK-ERK and D3-Akt-mTOR pathways to induce structural plasticity in dopaminergic neurons.

Conclusions: In dopaminergic neurons Ketamine induced a dose-dependent increase of dendritic arborization and soma size in both mouse primary culture and differentiated human iPSC. Both cells respond similarly to ketamine, showing activation of the mTOR and ERK intracellular prototypical pathways for cell growth and survival. Ketamine effect was blocked by D3 antagonists and was absent in D£-KO mice preparations, an evidence collected also in vivo, suggesting a autocrine role for DA released by ketamine via activation of the D3 autoreceptors. Ketamine effects wee also blocked by AMPA antagonists, indicating a cross-talk between glutamateric and dopaminergic neurons present in the cultures. Since we recently showed that cocaine and nicotine produced structural plasticity both in vitro, in mouse dopaminergic neuron primary cultures, and in vivo, after repeated in utero exposures (Collo et al. 2012, 2013), we conclude that D3 antagonists or partial D3 agonists can offer a potential therapeutic and disease-modifying action for some forms of addiction, in particular those related with early life exposure.

Keywords: dendrites, dopamine D3 receptors, AMPA antagonist, BDNF.

Disclosure: Nothing to Disclose.

T17. Lymphocytes Adoptively Transferred from Chronically Stressed Mice Confer Rapid Antidepressant Effects to Naive Mice

Miles Herkenham*, Michael L. Lehmann, Rebecca A. Brachman

National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland

Background: Mood and cognition are generally thought of as being controlled almost exclusively by the CNS. However, peripheral systems, i.e., the endocrine and immune systems, influence behavior. Psychological stress and stress hormones alter the status of the immune system, and in turn, immune system perturbations affect CNS function and structure via humoral and cellular pathways conveying signals to the brain. The bidirectional neuroimmune pathways serve to achieve internal homeostasis. Disequilibrium, if sustained or severe, results in disease. Scant attention has been paid to the role played by the adaptive immune system in responding to and affecting mood states. It is known that chronic psychological stress affects lymphocyte numbers and function via sympathetic and steroid hormone effector pathways. However, very little is known about whether stress-modulated lymphocytes in turn influence affective behavior and CNS function. Studies have shown that lymphocyte subsets adoptively transferred into or selectively deleted from mice can affect cognition, mood, and adult hippocampal neurogenesis. However, there is no information about whether the psychological status of the donor animals plays a role in exerting these effects on the recipient. To address that question, we designed a simple experiment in which lymphocytes from chronically defeat-stressed or non-stressed mice were transferred to healthy, naïve Rag2-/- mice, which lack mature lymphocytes. We found that Rag2-/- mice repopulated with cells from socially defeated (SD) mice showed reduced anxious and depressive-like behaviors and increased hippocampal neurogenesis, a correlate of antidepressant efficacy, compared to Rag2-/- mice receiving either no cells or cells from home-cage (HC) control mice. We interpret the data to mean that psychological stress primes the adaptive immune system to confer stress resiliency to its host.

Methods: Animals. Donor and control mice were adult male C57BL/6 mice (Taconic). Recipient mice were adult male Rag2-/- mice on a C57BL/6 background (Taconic). Social Defeat (SD). Repeated SD was used to induce alterations in behavioral affect in the donor mice. An experimental intruder mouse was co-housed for 14 days in the home cage of a dominant aggressor CD-1 male mouse. A partition separated the pair. Each day, the partition was removed for 5 min to allow agonistic encounters between the pair. Home-cage (HC) control mice were kept in group housing conditions. Adoptive transfer. After 14 days of SD or HC, cervical, axillary, inguinal, and mesenteric lymph nodes of donor mice were removed, isolated and injected retro-orbitally at a concentration of 10–20 million cells per host. Recipient Rag2-/- mice received cells from either SD (SD->Rag) or HC (HC->Rag) mice or saline (Naive). Behavioral analysis. Beginning 10 days after the transfer, the host Rag2-/- mice were tested in one behavioral test per day using automated video-based tracking of behavior in the Light/dark (L/D) box, the Open field test (OFT), the Social interaction (SI) test, the Urine scent marking (USM), and the Tail suspension test (TST). Neurogenesis analysis by BrdU immunohistochemistry. One day after behavioral testing, BrdU was administered i.p. at 200 mg/kg. Three h later, mice were deeply anesthetized and perfused with paraformaldehyde. Sections cut through the hippocampus were immunostained for BrdU. Stained cells in the dentate gyrus subgranular zone were counted from every 12th slide-mounted section. Protein determinations. Serum was analyzed according to manufacturer’s instructions by MS-Q-Plex Mouse Cytokine screen (Quansys) measuring IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-10, IL-12p70, IL-17, MCP-1, IFNγ, TNFα, MIP-1α, GMCSF, and RANTES.

Results: Two weeks after adoptive transfer to lymphopenic mice, lymphocytes have migrated to the spleen and lymph nodes, proliferated, and reached a plateau. At this stage, the recipient Rag2-/- mice were behaviorally phenotyped. In the L/D box, SD->Rag mice made more transitions and spent more time in the light compartment than either the saline-injected or HC->Rag mice, suggesting an anxiolytic effect of SD cells. In the OFT, the SD->Rag mice relative to naïve Rag2-/- mice spent more time in the center. In the SI test, SD->Rag mice showed more interaction with the CD-1 mouse than did the other groups, suggesting a greater sociability in these mice. In the USM test, SD->Rag mice showed enhanced scent marking to female urine, suggesting increased hedonic drive compared to naïve and HC->Rag mice. In the TST, where immobility is a measure of depressive-like states, the SD->Rag mice showed more time mobile than the naïve Rag mice. Thus in all tests, the SD->Rag group showed reductions in anxiety- and depressive-like behaviors relative to other groups. Hippocampal neurogenesis is decreased by psychological stressors and increased by antidepressant interventions. BrdU+cell counts in the dentate gyrus showed significantly more (about 50%) newborn cells in the SD->Rag group than the HC->Rag group. Cytokine profiles in the blood of donor and recipient mice showed opposite stress-induced and cell-induced changes. Chronic defeat stress affected the blood cytokine levels in the donor mice by increasing TNFα, IL-1β, IL-2, IL-3, IL-6, IL-17, and IFNγ in the SD relative to HC groups. Interestingly, in the Rag2-/- recipient mice, the levels of TNFα, L-1β, IL-2, IL-3, and IL-17 were reduced in SD->Rag compared to HC->Rag mice, and thus the direction of the difference was opposite to that of the donor mice. Naïve Rag2-/- mice had levels that fell between the HC->Rag and SD>Rag levels for some cytokines.

Conclusions: We showed that lymphocytes can be modified in vivo by psychological manipulations to exert antidepressant effects on their host. The view that psychopathology entails a destructive spiral in the immune system and brain overlooks the immune system’s proactive agenda—to achieve homeostasis in conditions of disequilibrium. The data presented here show that lymphocytes of the adaptive immune system are programmed by psychosocial stress to confer antidepressant effects. They also increased new cell proliferation in the hippocampal dentate gyrus. A clue to how stress-programmed lymphocytes work to reset homeostasis is provided by the profile of cytokine levels in the blood of the donors and recipients. As expected, chronic stress elevated inflammatory cytokines in the blood of the SD donor mice relative to HC control mice, but that profile was reversed in the blood of SD->Rag relative to HC->Rag recipient mice, consistent with its likely role in mediating the “restorative” effects on behavior and neurogenesis. Future work will explore the cellular and humoral pathways by which the changes access the brain to influence its structure and function.

Keywords: neuroimmune, lymphocytes, antidepressant, neurogenesis.

Disclosure: Nothing to Disclose.

T18. Effects of Extended Access Cocaine Self-administration on Inhibitory Neurotransmission in the Nucleus Accumbens

Anthony Purgianto*, Julia Miao, Mike Milovanovic, Marina Wolf

The Chicago Medical School at Rosalind Franklin University, North Chicago, Illinois

Background: Excitatory synaptic transmission in the nucleus accumbens (NAc) undergoes time-dependent changes during withdrawal from extended access cocaine self-administration in concert with the incubation of cue-induced cocaine craving. The goal of this project is to determine if incubation of cocaine craving is also accompanied by plasticity of inhibitory transmission in the NAc. While prior studies have found changes in both presynaptic and postsynaptic aspects of inhibitory transmission after different cocaine regimens, very little is known about whether these measures are affected during incubation. Our goal is to investigate whether inhibitory GABAergic transmission is altered in the NAc during incubation.

Methods: Three types of studies are underway to compare rats (male Sprague-Dawley) after prolonged withdrawal (>45 days) from extended access cocaine or saline self-administration (6h/day x 10 days). The first is an immunocytochemical study of parvalbumin expression, which provides an index of the activity of GABAergic parvalbumin positive (PV+) interneurons. The second study aims to measure GABAA receptor subunit surface expression at several different withdrawal time points (2, 25, and 48 days) using biotinylation and immunoblotting. Our third study uses electrophysiological local field potential (LFP) recordings to obtain a measure of overall inhibitory tone in the NAc. We are currently analyzing LFP in the NAc after stimulation of medial prefrontal cortex (mPFC).

Results: From our immunocytochemical study, we found no differences in the number of PV+cells in the NAc, indicating there was no loss or de novo generation of PV+interneurons. However, we did find an increase in PV immunoreactivity in the NAc of the cocaine group compared to the saline group. When we separated our data into 3 rostral to caudal levels, we found that the increase in PV immunoreactivity was specific to middle- and caudal NAc. Our immunoblotting study indicated a significant decrease in surface expression of the GABAA α2 subunit after 48 days of withdrawal from cocaine. No changes were observed for α1 on any withdrawal day or for α4 on withdrawal day 2. Our electrophysiological study indicated loss of potentiation after 5Hz and 10Hz stimulation in the cocaine group. No alteration in the input/output (I/O) curve was observed.

Conclusions: Our first two studies described above suggest potential pre- and post-synaptic alterations in inhibitory neurotransmission. An increase in PV+interneuron activity after prolonged withdrawal is suggested by our immunocytochemical results. This would result in increased GABA tone, which seems to be accompanied by a reduction in GABAA α2 receptor expression. Using LFP to analyze the mPFC-NAc circuit, we found no alteration in the I/O curve, indicating no change in the basal network strength. However, our results after train stimulations seem to indicate either an increase of inhibitory tone or a reduction in excitatory tone. To determine whether this is due to changes in inhibitory GABA transmission, we are repeating these studies after intra-NAc injection of picrotoxin, a GABAA receptor antagonist. We are also investigating changes in the basolateral amygdala-NAc circuit. Overall, these studies are the first to explore changes in inhibitory neurotransmission in the NAc that may contribute to the withdrawal-dependent incubation of cocaine craving.

Keywords: Cocaine, Self-Administration, Animal Model, Nucleus Accumbens.

Disclosure: Nothing to Disclose.

T19. The Schizophrenia and Autism Spectrum Disorder Gene TCF4 Regulates Cortical Structure and Neuronal Physiology

Matthew Rannals, Stephanie Cerceo-Page, Andrew Jaffe, Morganne Campbell, Ryan Gallo, BaDoi Phan, Thomas Hyde, Joel Kleinman, Daniel Weinberger, Brady Maher*

Lieber Institute for Brain Development, Baltimore, Maryland

Background: Genome-wide association studies (GWAS) have identified a number of loci associated with increase risk for SZ and several of these risk variants are located within introns of Transcription Factor 4 (TCF4; E2-2, ITF2). In addition, autosomal dominant mutations in TCF4 result in Pitt Hopkins Syndrome (PTHS), a rare neurodevelopmental disorder characterized by a spectrum of symptoms including hyperventilation, seizures, autistic behaviors, mental retardation, and brain malformations. Currently, the molecular mechanisms and underlying pathophysiology responsible for these two disorders are not understood. Our goal is to determine the function of TCF4 during cortical development and to understand the molecular mechanism of risk that is associated with genetic variants of TCF4. TCF4 is a ubiquitous basic helix-loop-helix (bHLH) protein that binds to E-box DNA sequences (CANNTG) and regulates transcription. It has 41 exons of which 21 are alternative 5’ exons and potentially code for 18 N-terminally distinct protein isoforms named TCF4A – TCF4R. TCF4 can form homodimers or heterodimers with other bHLH proteins and appears to be regulated by activity through its interaction with the Ca2+binding protein calmodulin.

Methods: To test the function of TCF4 in the developing neocortex we altered its expression by transfecting layer 2/3 pyramidal cells in the rat medial prefrontal cortex by in utero electroporation (IUE). We knockdown TCF4 expression using two shRNA constructs that target independent sequences within the TCF4 transcript and over-expressed human TCF4 with recombinant TCF4 constructs. Functional analysis was performed using whole-cell electrophysiology and confocal imaging in acute brain slices. To identify candidate genes regulated by TCF4 we combined IUE with translating ribosome affinity purification (TRAP).

Results: Embryonic knockdown of TCF4 decreased intrinsic excitability and resulted in the ectopic appearance of spike-frequency adaptation (SFA, p<0.002; TCF4 shRNA n=24, Con shRNA n=32). Decreased intrinsic excitability resulted from a significantly increased afterhyperpolarization (AHP) potential following a train of action potentials (p<0.0001; TCF4 shRNA n=29, Con shRNA n=26) and increased slow AHP currents following 100ms voltage-steps (p<0.05; TCF4 shRNA n=13, Con shRNA n=14). To identify candidate genes responsible for these phenotypes we performed TRAP from IUE transfected rat brains. Two candidate ion channel genes, KCNQ1 and SCN10a were significantly up regulated in neurons expressing TCF4 shRNA (KCNQ1 p<0.004; SCN10a p<0.01; TCF4 shRNA n=3, Con shRNA n=3). Remarkably, both KCNQ1 and SCN10a are known to be associated with SFA in a variety neuronal cell types. In addition, over expression of TCF4B dramatically alters cortical development and structure. in vivo terminal differentiation was significantly increased in neuroprogenitor cells expressing TCF4B (p<0.003; TCF4B n=5, GFP control n=6) and neuronal migration was increased such that greater than 30% of the transfected neurons had reached the intermediate zone or cortical plate compared to less than 10% of the cells transfected with GFP alone (p<0.0001; TCF4B n=8, GFP n=8). TCF4B overexpression also resulted in postnatal formation of abnormal cortical columns. Formation of abnormal columns was rescued by co-expressing TCF4B+calmodulin (CaM) and these columns appear to be formed in an activity-dependent manner because a mutant CaM (CaM1,2,3,4) that cannot interact with calcium was ineffective at rescue.

Conclusions: Our results suggest the dosage of TCF4 is critical to proper cortical development and neuronal physiology. By combining IUE with TRAP we demonstrate a novel approach to identifying underlying molecular mechanisms associated with genetic manipulations. Using this unique protocol we identify two candidate genes whose expression appears to be regulated by TCF4 and thus may be dysregulated in Schizophrenia and PTHS. In addition, we show that TCF4 in an activity-dependent manner is critical to cortical development by regulating the number and positioning of neurons found in cortical circuits.

Keywords: schizophrenia, autism, electrophysiology, neurodevelopment.

Disclosure: Nothing to Disclose.

T20. Sonic Hedgehog Signaling Disruption in Ellis-van Creveld Dwarfism Confers Protection Against Bipolar Affective Disorder

Edward Ginns*, Marzena Galdzicka, Robert Elston, Yeunjoo Song, Steven Paul, Janice Egeland

University of Massachusetts Medical School, Shrewsbury, Massachusetts

Background: Decades of longitudinal research on bipolar affective disorder (BPAD) revealed co-segregation of high numbers of Ellis-van Creveld dwarfism (EvC) and Bipolar I (BPI) cases in several large Amish families descending from the same pioneer. Despite more than forty years of research documenting the high prevalence of both disorders in these families, no EvC individual has ever been reported with BPI. Although compelling evidence supports a significant genetic susceptibility to develop BPAD, identification of the genetic variants and an associated underlying molecular mechanism or pathophysiology have remained elusive. To date, most genetic studies of affective disorders have been limited to identifying genetic variants that increase the risk of disease. By contrast, we have previously provided evidence that, in addition to rare susceptibility alleles, there may be rare alleles that reduce the risk of developing BPAD. We previously reported strong evidence for a BPAD locus with protective minor alleles on chromosome 4p at D4S2949 and suggestive evidence for a locus on chromosome 4q at D4S397. The EVC and hedgehog interacting protein (Hhip) genes were subsequently cloned and located within 5 million bases of our chromosome 4p16 and 4q putative loci for BPAD, respectively. The proximity of the EvC gene to our previously reported chromosome 4p16 BPAD locus, coupled with detailed clinical observations and statistical confirmation that EvC and BPI do not occur in the same Amish individuals, led us to postulate that the molecular mechanism underlying EvC is protective against BPI. Since homozygous Amish EVC mutations causing EvC dwarfism do so by disrupting sonic hedgehog (Shh) signaling, our data implicate Shh signaling in the underlying pathophysiology of BPAD. We now report clinical and statistical evidence that disruption of sonic hedgehog signaling in EvC confers protection from BPI, and perhaps more generally against major affective disorder.

Methods: EvC families were ascertained systematically by Amish church districts. EvC subjects were examined and diagnosed at home or at the Johns Hopkins Moore Clinic. Deceased were certified by death records. A pedigree trace was conducted for every EvC sibship. The resulting progenitor charts tested if all Amish EvC cases descended from a common progenitor. Under written informed consent, genomic DNA was obtained from peripheral blood samples, immortalized lymphoblastoid cell lines and/or skin fibroblasts. Analysis for the Amish EVC gene intron13 (IVS13+5G>T) mutation was performed using Sanger sequencing and the MassARRAY MALDI-TOF (Sequenom/Agena Biosciences) platforms. Two types of clinical data (psychiatric interviews and medical records) were processed for independent and “blinded” assessment by a five member Amish Study Psychiatric Board using Research Diagnostic Criteria (RDC) and Diagnostic Criteria from DSM-III-IV (DSM) to make categorical diagnoses. Only one primary hypothesis was tested, that of association between BPI/no disease status and EvC/no disease status. Two types of association analysis were performed on all 358 individuals in the BPAD Master Pedigree for whom both EvC genotype and BPAD diagnosis were available: a Fisher’s exact test, which ignores all relationships, and large sample tests based on a logistic regression that allowed for the familial dependencies.

Results: The EvC sample comprised 67 EvC families with 156 dwarfs and 33 surviving to adulthood. Among these 33 individuals with EvC were a number in the 40-70 years of age categories (67% over age 40 years; 52% over age 50 years) and past the window of risk for onset of BPI disorder. Progenitor traces on confirmed BPI cases yielded the same pioneer, Koenig, identified previously as the progenitor for EvC and giving early evidence for significant co-segregation of these two illnesses. Although the bipolar pedigree comprises 42 BPI sub-pedigrees, the EvC families were mainly included in four of them. Another relevant co-segregation is the clustering of BPAD Amish male suicides in this same extended family. Apart from the documented co-segregation of EvC and BPI, unbiased case ascertainment over time never revealed subjects with co-morbidity. Analyses performed to test our primary hypothesis, i.e. that of association between absence of BPI and EvC (P=.029, Fisher’s exact test, two-sided, confirmed by permutation test), supported our hypothesis that EvC confers protection (i.e mental health wellness) from BPI, as well as suggesting a more general protection against the spectrum of affective disorders in these families.

Conclusions: Our study of the co-segregation of EvC and BPAD in the Old Order Amish implicates Shh signaling in the pathophysiology of BPAD. Elucidating how disrupted Shh signaling protects against BPI could uncover mutations in the Shh pathway that cause or increase risk for this and related mood disorders. Our understanding of the molecular basis of Shh signaling and reactivation occurring in a wide range of cancers has led to identification of Shh signaling antagonists that are already in human clinical studies. Redirection of drugs targeting Shh signaling that are in clinical development for other medical conditions could lead to better treatments for affective disorders in the near future.

Keywords: BPAD, Sonic Hedgehog, Mood Disorders, Ellis-van Creveld.

Disclosure: Nothing to Disclose.

T21. Involvement of the Ventral Hippocampus (vHipp) Ascending Pathway to the Medial Prefrontal Cortex (mPFC) in the Antidepressant-like Effect of Ketamine in Rats

Flavia Carreno*, Amiksha Shah, Alan Frazer, Daniel Lodge

University of Texas Health Science Center at San Antonio, San Antonio, Texas

Background: The mechanisms and brain regions contributing to the antidepressant effects of ketamine are still unclear. Our initial studies demonstrated that the transient inactivation of the vHipp with 2% lidocaine at the time of ketamine administration (10 mg/kg), completely blocked the sustained (1 week), but not the acute (30 min) antidepressant-like response to ketamine, measured by the forced swim test (FST). A significant limbic projection of the vHipp is to the mPFC. We examined in rodents the role of the ascending projections from the vHipp to the mPFC in the response to ketamine using the FST. In addition, the role of the neurotrophin receptor TrkB, which is known to be required for ketamine’s antidepressant-like response, was studied.

Methods: Different approaches were used to test the hypothesis that activation of the vHipp-mPFC pathway is involved in the antidepressant-like response to ketamine. In one, an asymmetric disconnection procedure was utilized whereby 2% lidocaine (0.5ul) was injected via guide cannulas to the vHipp on one side of the brain and into the mPFC on the other. This transiently disconnects these two regions at the time of ketamine administration. Other approaches were to evaluate the effects of either optogenetic stimulation or DREADD activation of the vHipp-mPFC pathway. For the optogentic experiments, AAV2-hsyn-ChR2-EYFP (or control virus) was injected into the vHipp and bilateral fiber optic cannulas implanted into the mPFC. ChR2 was activated by laser light and the behavioral effect in the FST measured. For the DREADD (Gq) approach, vHipp-mPFC activation was achieved by combining a trans-synaptic (WGE)-Cre virus (in the vHipp) with a virus expressing a floxed Gq-protein (DIO-hM3D in the mPFC). The involvement of BDNF/TrkB was studied in two separate cohorts of rats implanted with bilateral cannulas into the vHipp for injection of the tyrosine kinase inhibitor, K252a, or a TrkB/Fc chimera, which scavenges BDNF.

Results: The sustained effect of ketamine in the FST was completely blunted following the asymmetric disconnection procedure. Consistent with the effect of ketamine, optogenetic stimulation of vHipp-mPFC pathway significantly increased climbing in the FST; however, in contrast to the effect of ketamine, a decrease in immobility was not observed but there was a significant decrease in swimming behavior. To test if the decrease in swimming was due to a feedback pathway from the mPFC to the dorsal raphe nucleus (DRN), we examined optogenetic stimulation following administration of the GABA-A receptor antagonist, bicuculine, directly into the DRN. Pretreatment with bicuculine abolished the decrease in swimming observed when the mPFC was stimulated and now a significant decrease in immobility was revealed. Activation of specific neurons in the mPFC that receive inputs from the vHipp with the designer drug, clozapine-n-oxide (0.5mg/kg i.p.), recapitulated the antidepressant-like response to ketamine in the FST. We also examined whether a transient increase in TrkB receptor phosphorylation in the vHipp contributed to the sustained antidepressant response to ketamine. Similar to the effects of administration of lidocaine into the v-Hipp, blockade of TrkB receptor phosphorylation with K252a into the vHipp at the time of ketamine administration prevented its antidepressant-like effects measured 1 week following its administration. This result led to the hypothesis that ketamine induces plastic changes in vHipp afferents, e.g., an increase in ionotropic glutamate receptor subunit GluR1, that underlie its sustained antidepressant response. To test this, we administered a TrkB/Fc chimera, which scavenges BDNF, into the vHipp at the time of ketamine administration and examined the expression of GluR1 in the mPFC 24h later. Ketamine produced an increase in GluR1 in mPFC synaptosomes that was blocked in rats administered the TrkB/Fc chimera. These data show that initial blockade of TrkB signaling by the time of ketamine injection prevented the sustained (1 week) antidepressant response to ketamine as well as a marker of plasticity in the mPFC.

Conclusions: In conclusion, these data demonstrate a role for vHipp-mPFC plasticity in the sustained antidepressant-like effects of ketamine.

Keywords: ketamine, hippocampus-mPFC pathway, optogenetics, DREADD.

Disclosure: Nothing to Disclose.

T22. Oligodendrocyte Morphometry and Expression of Myelin Related mRNA in Ventral Prefrontal White Matter in Major Depressive Disorder

Grazyna Rajkowska*, Gouri Mahajan, Monica Sathyanesan, Abiye Iyo, Mohadetheh Moulana, Patrick Kyle, William Woolverton, Javier Miguel-Hidalgo, Craig Stockmeier, Samuel Sathyanesan

University of Mississippi Medical Center, Jackson, Mississippi

Background: Involvement of ventral prefrontal (vPFC) white matter in the pathophysiology of major depressive disorder (MDD) has been revealed in several diffusion tensor imaging studies. However, the cellular and molecular pathology underlying prefrontal white matter abnormalities in MDD and the possible influence of antidepressant medications have not been examined. The present study investigated oligodendrocyte morphometry and myelin-related mRNA expression in the vPFC white matter in MDD. Oligodendrocyte morphometric parameters were also studied in the analogous region of the vPFC white matter of male rhesus monkeys treated chronically with the antidepressant fluoxetine.

Methods: Triplicate coronal frozen sections (20 μm) of the vPFC deep or gyral white matter were collected for immunohistochemistry from 21 subjects diagnosed in the last month of life with MDD and 17 age-matched non-psychiatric control subjects. An antidepressant medication was detected in the blood of five of the depressed subjects. There were no significant differences between cohorts in age, gender, postmortem interval (PMI), tissue pH or time in freezer. The density and volume (size) of CNPase-immunoreactive oligodendrocytes was estimated in the deep and gyral white matter using a 3-dimensional cell counting method implemented with StereoInvestigator (Microbrightfield) software. Similar analyses were performed in the analogous region of the 6 male monkeys (Macaca mulatta) treated daily for 9 months with oral fluoxetine (3 mg/kg) and in 6 control monkeys receiving the vehicle. Peak fluoxetine levels in blood, seen at two hours after administration, approximated clinically relevant levels. In 11-13 pairs of MDD and matched control subjects, expression of oligodendrocyte/myelin-related mRNA for CNPase, PLP1, MBP, MOG, MOBP, Olig1 and Olig2 was measured by quantitative real time PCR in homogenates of combined deep and gyral tissue taken from 50 μm sections of vPFC white matter.

Results: There was no significant difference between the two human cohorts for either the density or the size of oligodendrocyte cell bodies in the deep vPFC white matter. In contrast, the size of oligodendrocyte cell bodies was significantly decreased in the gyral vPFC white matter in depression (ANCOVA F(1, 33)=13.283, p=0.001). Death by suicide had no significant effect on these parameters in depressed subjects. Pearson correlation analyses revealed no significant association between smaller sizes in deep or gyral white matter and age, PMI, tissue pH, time in freezer or duration of depression. There were no significant differences between fluoxetine-treated and control monkeys in the density or size of oligodendrocytes in vPFC white matter. There was as a significant 34 percent decrease in the expression of PLP1 mRNA in depressed subjects vs. controls (t=2.049, df=23, p=0.05). In contrast, there was a significant increase in the expression of mRNA for CNPase (109%, t=3.209; df=24, p=0.004), MOG (51%, t=2.452, df=19, p=0.024) and Olig1 (47%, t=2.108, df=22, p=0.047) in depression. There were no significant correlations between the expression of mRNA for PLP1, CNPase, MOG or Olig1 and age, PMI, tissue pH, time in freezer or duration of depression. Death by suicide had no significant effect on these parameters in depressed subjects. There was no significant effect of depression on the expression of mRNA for MBP, MOBP and Olig2.

Conclusions: This is the first observation in postmortem tissue from depressed subjects that examined the density and size of oligodendrocytes immunolabeled specifically for this glial cell type. Our results suggest that subtle changes in oligodendrocyte morphology in the gyral vPFC white matter may be related to altered axonal integrity as revealed by imaging studies. The proximity of gyral white matter to the overlying vPFC gray matter suggests a relationship between smaller soma sizes of white matter oligodendrocytes and previously noted reductions in neuronal and glial cell density in vPFC gray matter. Neuroimaging studies specifically targeting gyral vPFC white matter have not been performed in MDD. Our study in monkey tissue suggests that chronic antidepressant treatment does not appear to affect soma size of oligodendrocytes. As noted in animal models of myelin disturbances, changes in gene expression for four myelin proteins (i.e., PLP1, CNPase, MOG and OLIG1) in the vPFC white matter suggest a possible mechanism for the degeneration of cortical axons and dysfunctional maturation of oligodendrocytes in depression. Supported by P30 GM103328.

Keywords: postmortem, depression, white matter, morphometry.

Disclosure: Nothing to Disclose.

T23. Altered Astrocyte-Microglia Communication as Potential Immune/Inflammatory Changes in Major Mental Illness

Shin-ichi Kano*, Brian Lo, Akira Sawa

Johns Hopkins University, Baltimore, Maryland

Background: Clinical studies have reported that patients with major mental illness such as schizophrenia and bipolar disorder have mild inflammation or immune abnormalities. The underlying biological mechanisms, however, are not fully understood. Although microglia and astrocytes have been extensively studied in brain inflammatory changes, less is known about their mutual regulations. Here we have focused on astrocyte-microglia interaction and explored its alterations in immune/inflammatory responses by studying the role of glutathione-S-transferase theta 2 (GSTT2), which has been implicated in major mental illness.

Methods: Mouse models are used to address the role of GSTT2 in immune/inflammatory responses. Primary mouse glial cell culture was employed to address immune/inflammatory responses to defined sets of immune stimuli such as LPS and polyI:C. mRNA expression of inflammatory mediators was measured by quantitative RT-PCR. Systemic LPS-induced neuroinflammation was assessed in mice with astrocyte-specific modulation of GSTT2 expression levels.

Results: We found that GSTT2 could facilitate the production of proinflammatory cytokines such as TNF-α and IL-6 in primary mouse glial cells against various immune stimuli. Cell type-specific analysis indicated that GSTT2 controlled astrocytic communication with microglia via soluble factors (e.g., CCL2, CXCL10, IL-34) in immune/inflammatory responses. in vivo neuroinflammation models showed that the modification of astrocytic expression of GSTT2 altered microglial activation in the brain. Further studies on the underlying mechanisms and in vivo phenotypes are in progress.

Conclusions: Our data suggest that GSTT2 may be involved in altered astrocyte-microglia communication under immune/inflammatory conditions relevant for major mental illness. This study provides further insight into the biological mechanisms underlying immune/inflammatory changes in major mental illness.

Keywords: Astrocytes, Microglia, Immune, GSTT2.

Disclosure: Nothing to Disclose.

T24. Knockdown of mu-Opioid Receptors in Rat Ventral Tegmental Area Prevents Social Stress-induced Cross-sensitization and BDNF Expression While Altering Intracellular AKT Phosphorylation in GABA neurons

Ella Nikulina*, Caitlin Johnston, Amy Lasek, Ronald Hammer, Jr

University of Arizona, Phoenix, Arizona

Background: We showed previously that exposure to intermittent social defeat stress up-regulates mu-opioid receptor (MOR) expression in the ventral tegmental area (VTA) and induces prolonged cross-sensitization to amphetamine, together with a long-lasting increase of VTA brain-derived neurotrophic factor (BDNF) expression. In the VTA, MORs are localized in GABA neurons, and their activation inhibits GABA release to indirectly stimulate VTA dopamine (DA) neurons. Stimulation of MORs activates the PI3K-AKT signaling cascade that is involved in stress- and psychostimulant-induced behaviors. However, the specific intracellular systems by which VTA MORs alter social stress-induced behavior and cellular function are unknown. The purpose of the current study was to determine the effects of lentivirus-mediated knockdown of MOR in the VTA on amphetamine cross-sensitization, VTA BDNF expression and cellular co-localization of AKT with GABA or DA in the VTA.

Methods: We used lentivirus-mediated gene transfer and RNA interference to induce persistent VTA MOR knockdown. Short hairpin MOR or control scrambled virus constructs were infused bilaterally into the VTA of adult Sprague Dawley rats. Intermittent social defeat stress consisted of four brief confrontations between the experimental intruder rat and an aggressive resident rat over the course of 10 days. Control rats were handled according to the same schedule. Amphetamine (1.0 mg/kg, ip) challenge was performed ten days after the last stress or handling procedure. Animals were euthanized and fresh-frozen brains were removed and sectioned for MOR autoradiography using [3H]DAMGO; labeling was quantified using X-ray film co-exposed with radiostandards, and ImageJ. VTA BDNF expression and AKT labeling were examined in different groups of rats treated similarly, and perfused ten days after the last episode of stress or handling during the time period in which cross-sensitization is known to be present in defeated rats. Brains were removed and processed by immunohistochemistry for BDNF or phospho-AKT and either glutamic acid decarboxylase (GAD 65/67) or tyrosine hydroxylase (TH) to characterize labeling in GABA or dopamine neurons, respectively.

Results: Knockdown of MORs significantly reduced MOR binding in the VTA, thereby confirming the efficacy of MOR knockdown. VTA MOR knockdown prevented stress-induced cross-sensitization to amphetamine. Following intermittent social defeat stress, significantly more phospho-AKT labeling was found in VTA GABA cells than in DA cells. In rats with intact VTA MORs, intermittent social stress significantly increased AKT phosphorylation in GABA cells by more than three times that seen in handled rats. By contrast, VTA MOR knockdown prevented intermittent social stress-induced AKT phosphorylation in GABA cells. VTA MOR knockdown and a history of social stress produced significantly fewer phospho-AKT-labeled GABAergic VTA cells. In addition, MOR knockdown prevented social stress-induced increase of VTA BDNF labeling.

Conclusions: These results provide evidence that intermittent social defeat stress-induced MOR upregulation in VTA GABA cells is necessary for amphetamine cross-sensitization. Subsequent phosphorylation of AKT downstream of VTA MORs may alter VTA GABA transmission in a phospho-AKT-dependent manner, possibly by facilitating the insertion GABA-A receptors into the cell membrane. Such MOR-dependent, phospho-AKT-mediated reduction of local VTA GABA transmission would disinhibit DA neurons, and could explain both amphetamine cross-sensitization and the MOR-dependent stress-induced increases of VTA BDNF. These data reveal that cross-sensitization following social defeat stress is accompanied by increased phospho-AKT in GABA cells and BDNF in DA neurons, all of which are prevented by knockdown of VTA MORs, suggesting that MOR-AKT signaling may serve as a potential target for the intervention of stress-induced drug-sensitization.

Keywords: social stress, cross-sensitization, GABA, mu-opioid receptors.

Disclosure: Nothing to Disclose.

T25. Brain-enriched Sorting Nexin Family Proteins Regulate Spine Morphogenesis and Are Associated with Risk for Schizophrenia

Takanobu Nakazawa*, Ryota Hashimoto, Asami Tanimura, Kazutaka Ohi, Hidenaga Yamamori, Yuka Yasuda, Satomi Umeda-Yano, Yuji Kiyama, Kohtarou Konno, Takafumi Inoue, Shusuke Numata, Tohru Ohnuma, Nakao Iwata, Norio Ozaki, Hitoshi Hashimoto, Masahiko Watanabe, Toshiya Manabe, Tadashi Yamamoto, Masatoshi Takeda, Masanobu Kano

Osaka University, Osaka, Japan

Background: Accumulating evidence suggests that altered function and morphology of synapses in central neurons underlie pathophysiology of schizophrenia. We previously identified novel brain-enriched sorting nexin proteins, ARHGAP32 and ARHGAP33, both of which have Rho GTPase-activating (RhoGAP) activity. These proteins constitute a subfamily in the large RhoGAP family of proteins, because, in addition to the RhoGAP domain, they unusually share their PX domain which is involved in surface protein trafficking. Genetic manipulation of ARHGAP32 and ARHGAP33 resulted in altered spine morphology, suggesting the possible involvement of these genes in schizophrenia.

Methods: We investigated the possible association between schizophrenia and the ARHGAP32 and ARHGAP33 genes using samples from schizophrenia patients. We further examined the molecular roles of these genes in schizophrenia using genetically-engineered mice.

Results: We found single nucleotide polymorphisms associated with risk for schizophrenia at both the ARHGAP32 and ARHGAP33 gene loci. The allele of the ARHGAP32 gene, which was overrepresented in schizophrenia patients, was associated with higher scores of schizotypal personality traits in mentally healthy subjects. Likewise, the affected allele of the ARHGAP33 gene was associated with reduced brain volume in several regions of schizophrenia patients. Furthermore, we found that ARHGAP33 expression was decreased in peripheral lymphocytes of schizophrenia patients and that ARHGAP33 knockout mice showed schizophrenia-related behavioral abnormalities.

Conclusions: From these results, ARHGAP32/ARHGAP33-mediated regulation of spine morphogenesis may be a new molecular pathophysiology of schizophrenia.

Keywords: schizophrenia, sorting nexin, spine morphogenesis, Rho family GTPase.

Disclosure: Nothing to Disclose.

T26. Raphe Neuroligin 2/Serotonin Transporter Protein Complex Regulates Serotonin Signaling

Ran Ye*, Meagan Quinlan, Hideki Iwamoto, Hsiao-Huei Wu, David Airey, Noah Green, Christopher Jetter, Douglas McMahon, Jeremy Veenstra-VanderWeele, Pat Levitt, Randy Blakely

Vanderbilt University Medical Center, Nashville, Tennessee

Background: Compromised serotonin (5-HT, 5-hydroxytryptamine) signaling has been implicated in multiple neuropsychiatric and neurodevelopmental disorders including depression, obsessive-compulsive disorder (OCD), anxiety and autism. 5-HT neurons localized in midbrain dorsal raphe nuclei (DRN) project extensively throughout the forebrain to modulate a number of complex behaviors. Although basic aspects of these projections and their signaling are understood, the molecular networks that support 5-HT signaling continue to be elaborated. In an effort to identify novel genes that regulate 5-HT signaling, we reported a 5-HT gene network based on correlational analyses of midbrain transcriptomes of BXD recombinant inbred mice (Ye et al, Genes Brain and Behavior, 2014). Here we report the intersection of these efforts with results of a complementary strategy that involves a proteomic analysis of 5-HT transporter (SERT) protein complexes. Remarkably, both unbiased approaches converge on the synaptic cell adhesion protein neuroligin 2 (NLGN2) a gene previously implicated in multiple psychiatric disorders.

Methods: For SERT proteomic analysis, midbrain protein lysates were prepared from C57BL/6J and SERT KO mice followed by SERT co-immunoprecipitation. Bead eluants were separated by SDS-PAGE, digested, and subjected to MudPIT-MS/MS to obtain peptide sequence information of possible SERT-interacting proteins. Chromogenic and fluorescence-based In situ hybridization as well as immunohistochemistry approaches were used to elucidate raphe localization. Frozen brain tissues were extracted for HPLC-based neurochemistry measurements. Western blotting approaches were used to quantify protein levels. To assess the impact of genotype on behavior, 8-12 week old male mice were subjected to tail suspension, forced swim, DOI-induced head twitch, or Tube Test evaluations. For whole-cell recordings, WT or NLGN2 KO mice were crossed with Tph2:eYFP reporter line to label 5-HT neurons. To examine 5-HT neuron firing rates, coronal sections containing the RN were subjected to multi-electrode array recordings. Neuronal identify was verified through assessment of basal firing rates suppression after exposure to the 5HT1A receptor agonist 8-OH-DPAT.

Results: Using in situ hybridization and immunohistochemistry methods, we validated NLGN2 gene and protein expression in RN 5-HT neurons, with particular enrichment evident in the ventromedial DRN. Co-immunoprecipitation studies revealed that SERT-NLGN2 protein complexes can be recovered from the midbrain, but not the hippocampus, indicating that SERT-NLGN2 associations likely derive from somatodendritic sites of co-expression, a hypothesis supported by immunofluorescence co-localization studies. Biochemical studies reveal that SERT-NLGN2 interaction is not indirectly dependent upon the NLGN2 presynaptic binding partner, Neurexin, consistent with a somatodendritic site of interaction. Interestingly, we discovered a requirement of NLGN2 to maintain normal levels of SERT in both midbrain and hippocampus. Electrophysiological recordings obtained from DRN 5-HT neurons in NLGN2 KO mice revealed reduced firing rates and diminished spontaneous IPSCs, consistent with disrupted GABAergic inputs. Finally, NLGN2 KO mice displayed altered performances on multiple neurobehavioral paradigms, including tasks sensitive to antidepressants (tail suspension test, forced swim test), tasks that monitor 5-HT receptor sensitivity (DOI-induced head twitch), and social interactions sensitive to SERT mutations (Tube Test).

Conclusions: Our efforts have identified a novel, raphe-specific SERT-NLGN2 protein complex and establish an essential role of NLGN2 in control of raphe excitability as monitored in vitro. NLGN2 is a cell adhesion protein enriched at GABAergic synapses and an essential component of postsynaptic GABAA receptor protein complex. DRN 5-HT neurons receive extensive inputs from local GABAergic interneurons located ventrolaterally to 5HT neurons in the DRN. Our results indicated that these inhibitory inputs are intimately and physically connected with the somatodendritic 5-HT re-uptake machinery. Furthermore, our characterization of NLGN2 KO mice suggests important functional roles of of NLGN2/SERT associations in the control of GABAergic input, with alterations leading to disrupted 5-HT neuron physiology and behavior. Together our findings suggest that a disruption of 5-HT signaling may underlie the etiology of neuropsychiatric disorders associated with rare NLGN2 variants and suggest an unsuspected link between SERT-dependent autoreceptor regulation of raphe neurons and GABAergic inhibitory control mechanisms.

Keywords: serotonin, GABA, neuroligin, raphe.

Disclosure: Nothing to Disclose.

T27. ATP Regulation of Glutamatergic Transmission Following Cocaine Self-administration

Haley Andersen, Luyi Zhou, Pavel Ortinski*

University of South Carolina, Columbia, South Carolina

Background: Widespread release of adenosine triphosphate (ATP) from glial and neuronal cells regulates neuronal excitability in normal brain and in pathological conditions. ATP and its breakdown product, adenosine, have been proposed to play important roles in drug addiction, including in animal models of cocaine self-administration. In this study we investigated the interaction between ATP/adenosine receptor systems with glutamatergic signaling in the nucleus accumbens (NAc) shell, an area implicated in several cocaine-associated behaviors.

Methods: Acute brain slices, containing NAc shell were prepared and whole-cell patch-clamp recordings obtained from visually identified medium spiny neurons.

Results: We demonstrate that exogenously applied ATP (100 μM) powerfully inhibits evoked excitatory postsynaptic currents (eEPSCs) in the NAc shell. The ATP-mediated block affects both AMPA and NMDA receptor components of eEPSCS and is fully and quickly reversible. A broad spectrum P2-receptor antagonist, PPADS, does not prevent ATP-mediated inhibition of eEPSCs whereas adenosine receptor antagonist, DPCPX, has only a weak effect. ATP inhibition of eEPSCs does not appear to recruit pre-synaptic mechanisms as measured by the paired pulse ratio and the frequency of spontaneous synaptic currents.

Conclusions: These results suggest that ATP modulation of excitatory neurotransmission in the NAc shell may rely on post-synaptic mechanisms that do not involve adenosine or P2-type receptors. The consequences of short access (2 hr/day for 14 days) cocaine self-administration training on ATP regulation of glutamatergic neurotransmission in the NAc are explored.

Keywords: adenosine triphosphate (ATP), ampa receptor, nmda receptor, cocaine.

Disclosure: Nothing to Disclose.

T28. Proteomic Analysis of the PSD-95 Interactome in Postmortem Brain

Robert McCullumsmith*, Adam Funk

University of Cincinnati, Cincinatti, Ohio

Background: Ionotropic glutamate receptors are concentrated in the postsynaptic density (PSD), a subcellular organelle consisting of densely packed structural, scaffolding, signaling, and receptor proteins. Receptors in the PSD transduce external stimuli onto intracellular signaling cascades. Dynamic changes are well documented in the trafficking of glutamate receptors and signaling molecules at postsynaptic sites. Through its PSD-95/Discs large/Zona occludens 1 (PDZ) domain, PSD-95 interacts with many postsynaptic proteins that regulate synaptic function. PSD-95 is the most abundant scaffolding protein in the PSD, with well characterized protein-protein interactions that modulate the trafficking of glutamate receptors, making it an ideal target for proteomic studies in the human brain.

Methods: We developed a modified immunoisolation-LCMS/MS protocol to assess the PSD-95 interactome. Magnetic dynabeads conjugated with anti-PSD-95 antibody were used to isolate PSD-95 protein complexes. The complexes were eluted from the beads and processed by 1D gel-electrophoresis and analyzed by Western blot, or were reduced, alkylated, digested with modified trypsin, extracted and resuspended in 0.1% formic acid for LCMS/MS analysis on an ABSciex 5600+mass spectrometer. Data independent acquisition (DIA) was used to identify peptides, and the data were analyzed using Protalizer proteomic software (Birmingham AL).

Results: We explored the extent of protein-protein interactions affiliated with PSD-95 in the frontal cortex. First, we verified the specificity of this immunoisolation from postmortem frontal cortex with Western blot analysis. We found significant enrichment of PSD-95, capturing all available PSD-95 from 1.4 mg of postmortem brain homogenate. PSD-95 signal was undetectable in an isotype specific, negative control immunoisolation. Using Western blot analyses, our PSD-95 immunoisolations were free of non-PSD proteins such as synaptophysin, but included PSD components such as Calcium/Calmodulin Kinase II alpha (CaMKIIα). Further verification using LCMS/MS showed co-isolation of over 700 proteins, including key postsynaptic components: PSD-93, SAP97, SAP102, SynGAP, GluA1, GluA2, GluA3, GluN1, GluN2A, GluN2B, Homer1, CaMKII, AKAP1, Shank1, and Shank3. IPA analyses indicates hundreds of primary connections between proteins identified in our data set, with hubs centered on PSD-95 (DLG4), CACNA1E, SRC, CTNNB1, GRIN1, STAT3, and MECP2. PSD-95 interactome data from the ACC, STG, and hippocampus will also be presented.

Conclusions: Our approach shows promise as a scalable method to isolate high-yield targets from postmortem brain, permitting assessment of the PSD-95 interactome in studies of psychiatric, developmental, and neurodegenerative diseases with synaptic pathology. This method represents an important advance in the study of postmortem protein-protein interactions for two reasons: (1) Traditional PSD isolations require substantial amounts of starting material (300-500 mg). Here we show robust isolation that is quantifiable with as little as 1.4 mg of tissue, (2) This method is highly specific, targeting PSD-95 protein-protein interactions of PSD related molecules. This allows for a high-level of reproducibility between preparations and subjects, making it ideal for quantitative analyses. Regional proteomic and pathway analyses of the specific interactomes will enable researchers to identify abnormalities in the protein connectome, leading to the development of new hypotheses based on mechanisms of physiologic and pathologic synaptic plasticity. Additionally, it allows the field to move beyond stratified approaches of one or a few targets and into a systems/protein-protein interaction pathway analysis, in alignment with the broader goals of the Human BRAIN Initiative.

Keywords: PSD-95, mass spectrometry, interactome, pathway analyses.

Disclosure: Nothing to Disclose.

T29. Differential Modulation of Cocaine-related Behaviors Consequent to Knockdown of Serotonin (5-HT) 5-HT2C Receptor (5-HT2CR) in the Nucleus Accumbens Shell (NAcSh) Vs. Ventral Tegmental Area (VTA)

Sarah E. Swinford-Jackson*, Noelle C. Anastasio, Sonja J. Stutz, Robert G. Fox, Kathryn A. Cunningham

University of Texas Medical Branch of Galveston, Galveston, Texas

Background: The mesoaccumbens pathway is known to mediate the hyperlocomotive and reinforcing effects of cocaine as well as cue-evoked cocaine-seeking behavior. The serotonin 5-HT2C receptor (5-HT2CR) localized to this pathway is poised to regulate dopamine neurotransmission and cocaine-related behaviors. Acute intra-NAcSh administration of a 5-HT2CR agonist augments the expression of behaviors evoked by cocaine, including cocaine-evoked hyperactivity. Conversely, acute intra-VTA administration of a 5-HT2CR agonist attenuates cocaine-evoked hyperactivity and cocaine-taking. Here, we employ a virally-mediated genetic knockdown strategy to test the hypothesis that knockdown of the 5-HT2CR in the NAcSh vs. VTA will differentially modulate cocaine-related behaviors, including cocaine-evoked hyperactivity, cocaine self-administration and cue reactivity assessed 24 hr after cocaine self-administration.

Methods: Male Sprague-Dawley rats were evaluated for basal motor activity and cocaine-evoked hyperactivity (10 mg/kg; i.p.) following intra-NAcSh or intra-VTA infusion of a 5-HT2CR shRNA-eGFP adeno-associated virus (AAV; 5-HT2CR knockdown) or a non-silencing control shRNA-eGFP AAV (control). Rats were then trained in daily three-hour sessions to self-administer cocaine (0.25 mg/kg/infusion; FR1-5); the cocaine dose-response (0.05, 0.125, 0.25 and 0.75 mg/kg/infusion; FR5) curve was then established in NAcSh or VTA 5-HT2CR knockdown and control rats to interrogate differences in sensitivity to the reinforcing properties of cocaine. Cocaine self-administration was re-stabilized and cue reactivity in NAcSh or VTA 5-HT2CR knockdown and control rats was measured 24 hr later by lever presses for the discrete cues previously associated with cocaine intake during a 1 hr cue test session.

Results: Rats with a genetic knockdown of the 5-HT2CR in the NAcSh, but not the VTA, displayed enhanced cocaine-evoked hyperactivity, but no difference in basal locomotion relative to control rats. Cocaine infusions received were lower at 0.05 mg/kg/infusion and 0.125 mg/kg/infusion of cocaine after loss of 5-HT2CR in the NAcSh, but not the VTA, which suggests that the dose-response curve for cocaine was shifted downward. Cue reactivity 24 hr after cocaine self-administration was higher following knockdown of the 5-HT2CR in the NAcSh, but not the VTA, relative to control rats.

Conclusions: These data suggest that diminished 5-HT2CR tone in the NAcSh, but not the VTA, confers an enhanced sensitivity to the hyperlocomotive and reinforcing properties of cocaine as well as cue reactivity. Our genetic approach captures the long-term consequences of shifts in 5-HT2CR functional capacity and contrasts some of the previous data obtained by intracranial pharmacology, possibly due to compensatory homeostatic mechanisms in response to the chronic loss of 5-HT2CR control over the mesoaccumbens pathway. Overall, a neuroanatomically distinct role for the 5-HT2CR as a critical, multifaceted regulator of cocaine hyperactivity, cocaine-taking and cue reactivity is indicated. Future studies will explore brain region-dependent 5-HT2CR-mediated shifts in dopamine transmission in relationship to cocaine-related behaviors.

Keywords: Serotonin 5-HT2C Receptor, Cocaine, Nucleus Accumbens Shell, Ventral Tegmenal Area.

Disclosure: Dr. Cunningham is a consultant for Arena Pharmaceuticals and an editor of Neuropsychopharmacology Reviews for which she receives compensation from the American College of Neuropsychopharmacology.

T30. Contribution of Dorsal Hippocampal Src Family Tyrosine Kinases and NMDA Receptors to Cocaine-memory Reconsolidation in Rats

Audrey Wells, Xiaohu Xie, Kelley Harmon, Amy Arguello, Kati Healey, Rita Fuchs*

Washington State University, Pullman, Washington

Background: The ability of cocaine-paired environmental stimuli to promote relapse in humans and reinstatement of cocaine-seeking behavior in rats requires the retrieval and utilization of context-response-cocaine associative memories. These memories become destabilized when retrieved and must be reincorporated into long-term memory storage via a protein synthesis-dependent memory reconsolidation process for subsequent availability. Identification of the molecular underpinnings of cocaine-memory reconsolidation within critical processing sites, like the dorsal hippocampus (DH), may inform the development of new treatments that mitigate the impact of pathological cocaine memories on relapse vulnerability. The present study utilized the rat extinction-reinstatement paradigm to test the hypothesis that the activation of Src family tyrosine kinases (SFK) in the DH is necessary for the cocaine memory reconsolidation and subsequent drug context-induced cocaine-seeking behavior.

Methods: Rats received bilateral microinfusions of the SFK inhibitor, PP2 (62.5ng/0.5μl/hemisphere) or vehicle (VEH) into the DH following re-exposure to a cocaine-associated or an unpaired context (i.e., cocaine-memory reactivation and no memory reactivation, respectively). We assessed alterations in the phosphorylation state of recognized SFK targets, including NR2a and NR2b NMDA and GluR2 AMPA receptor subunits, at the putative time of memory re-stabilization. In addition, we tested drug context-induced cocaine-seeking behavior 72 h later in a separate cohort of rats. In a follow-up experiment, a third cohort of rats received bilateral microinfusions of the NR2a subunit-containing NMDA receptor antagonist, PEAQX (2.5μg/0.5μl/hemisphere), or VEH into the DH following re-exposure to the cocaine-associated context, and drug context-induced cocaine-seeking behavior was assessed 72 h later.

Results: NR2a NMDA, NR2b NMDA, or GluR2 AMPA receptor subunit phosphorylation did not change following exposure to the cocaine-associated context, relative to exposure to the home cage (i.e., no memory reactivation). However, PP2 selectively reduced NR2a phosphorylation at the time of putative memory re-stabilization. PP2 also attenuated cocaine-seeking behavior at test, relative to VEH, in a memory reactivation-dependent fashion. Furthermore, this effect was recapitulated using PEAQX.

Conclusions: The findings suggest that SFKs in the DH may maintain NR2a-subunit containing NMDA receptor function that is necessary for memory reconsolidation. Thus, SFKs in the DH critically contribute to the long-term stability of maladaptive cocaine-related memories that underlie contextual stimulus control over cocaine-seeking behavior.

Keywords: cocaine, memory reconsolidation, relapse, glutamate.

Disclosure: Nothing to Disclose.

T31. Elucidating Serotonin's Contribution to the Addicted Brain: Molecular and Behavioral Effects of Cocaine Without Serotonin-Reuptake Inhibition in the Sert M172 Mouse Model

Linda D. Simmler*, Michael H. Levin, Alexander G. Nackenoff, Paul J. Gresch, Randy D. Blakely

Vanderbilt University Medical Center, Nashville, Tennessee

Background: Cocaine dependence remains a world-wide health problem. Approaches to pharmacologically reduce drug craving or to support abstinence have been tested, but more effective therapies are needed. The mode of action of cocaine is a relatively equal inhibition of the serotonin- (5-HT), dopamine- (DA), and norepinephrine (NE) re-uptake transporters, and consequent neuroplasticity of all three monoamine transporters should be considered in relation to treatment approaches. However, research to date on cocaine addiction has primarily focused on normalizing DA signaling. We hypothesize that an improved understanding of the consequences of 5-HT transporter (SERT) inhibition may shed light on the complexities of psychostimulant drug actions and possibly open new treatment options for cocaine abusers. To pursue this objective, we implemented a novel knock-in (KI) mouse model wherein the Slc6a4 gene has been mutated to eliminate high-affinity binding of cocaine. Prior in vitro studies revealed that the SERT substitution I172M results in a loss of high-affinity recognition for cocaine without an impact on 5-HT transport function (Henry et al., J Biol Chem, 2006). Subsequent generation and analysis of the SERT M172 knock-in mouse validated these findings in vivo (Thompson et al., PNAS, 2011).

Methods: Initial evaluations of the SERT M172 model were performed with mice on a 129S6/S4 background. To establish a genetic background more common to studies of cocaine action, we moved the allele to a C57BL/6J background through speed-congenic methods. We then validated the loss of high-affinity SERT antagonism using synaptosomal studies in vitro and through microdialysis studies in vivo. For in vitro studies, midbrain synaptosomes were incubated with [3H]5-HT (50 nM) for 10 min in the presence or absence of increasing concentrations of cocaine, followed by filtration on a Brandel tissue harvester and scintillation spectrometry. For in vivo changes, we inserted a microdialysis probe (Synaptech, MI, 3 mm membrane length, 20,000 Da cut-off) into the dorsolateral hippocampus (stereotaxic coordinates -3.28 AP, 2.8 ML, -1 DV from bregma) and tissue was perfused with aCSF at 1 μL/min. 5-HT levels were determined in collected dialysates before and after 20 mg/kg cocaine i.p. using HPLC-EC methods. To assess the impact of genotype on cocaine action, we performed c-Fos immunohistochemistry on brain sections prepared from animals sacrificed 120 min post 20 mg/kg i.p. injections. The number of c-Fos positive nuclei in different brain regions was quantified with ImageJ. Finally, locomotor sensitization was assessed following saline or repeated injections of cocaine. Saline (i.p.) was administered on the first testing day, and 15 mg/kg cocaine i.p. was administered on the second, third and last testing day. Three days of home-cage treatments occurred between the second and third testing day and 14 days of forced abstinence before the last testing day.

Results: Ex vivo synaptosomal uptake experiments demonstrated a 60-fold reduction in potency for 5-HT uptake inhibition by cocaine at SERT M172 as compared to WT, similar to the shift in inhibitory potency for cocaine observed in SERT M172 transfected cells (Henry et al., J Biol Chem, 2006). Consistent with these findings, WT mice demonstrated a>450% increase in microdialysate 5-HT over baseline levels, whereas no increase in 5-HT was observed in the KI mice. Acute cocaine administration resulted in genotype-independent elevations in c-Fos expression in multiple brain regions, including the nucleus accumbens. Interestingly, the piriform cortex and the prelimbic cortex, regions related to conditional components of cocaine addiction, demonstrated higher levels of c-Fos activation in KI vs. WT mice. No genotype differences were observed with respect to locomotor activation or sensitization measures.

Conclusions: Synaptosomal experiments and microdialysis studies confirmed the utility of the SERT M172 mouse model for the study of SERT-dependent cocaine actions in vitro and in vivo. Basal locomotor activation and sensitization to repeated drug action in cocaine-treated WT and KI mice revealed no genotype differences, consistent with a predominant DA-dependence of these measures. Our c-Fos studies identified two brain regions, the prelimbic and the piriform cortex, where enhanced 5-HT signaling may be of particular importance in coordinating the CNS actions of cocaine. Both regions receive significant serotonergic innervation and communicate with subcortical nuclei involved in stress, fear and anxiety responses. Thus, the SERT M172 model may be particularly advantageous in determining the contribution of SERT-blockade induced 5-HT signaling to stress-modulated alterations in cocaine action. Finally, transcriptome-wide gene-expression studies are underway to broadly elucidate molecular pathways involved in the acute and chronic actions of cocaine. A better understanding of the impact of 5-HT in drug abuse may ultimately lead to improved therapies to support abstinence.

Keywords: cocaine, serotonin, SERT, c-Fos.

Disclosure: Nothing to Disclose.

T32. Ceftriaxone Requires Both xCT and GLT-1 Up-regulation in the Nucleus Accumbens to Attenuate the Reinstatement of Cocaine-seeking and Alter Ampa Receptor Subunit Composition

Lori Knackstedt*, Kathryn Reissner

University of Florida, Gainesville, Florida

Background: Ceftriaxone is a beta-lactam antibiotic which increases the expression and function of the glutamate transporter GLT-1 and of system xC-, which exchanges extracellular cysteine for intracellular glutamate. Basal glutamate levels in the nucleus accumbens are largely controlled by system xC-, and a decrease in its activity is a contributing cause of the altered glutamate homeostasis observed in this brain region following cocaine self-administration in rats. The catalytic subunit of xC- is xCT, and we have demonstrated that expression of xCT and GLT-1 is decreased in the nucleus accumbens core following cocaine self-administration. We have also shown that ceftriaxone attenuates cue- and cocaine-primed reinstatement while restoring levels of both xCT and GLT-1 in the nucleus accumbens core. At this time it is not known if an increase in both transport systems is essential for the attenuation of cocaine reinstatement by ceftriaxone. Here we used a morpholino antisense strategy to decrease the expression of xCT and GLT-1 protein and examined the relative importance of these two proteins in mediating the reinstatement of cocaine-seeking. Altered glutamate homeostasis in the accumbens has also been shown to change post-synaptic glutamate receptor function and expression. Thus we assessed the nucleus accumbens surface expression of AMPA receptor subunits and total protein expression of mGluR5 and the subunits of AMPA and NMDA receptors.

Methods: Rats were trained to self-administer cocaine for two weeks in an operant chamber and then experienced extinction training for three weeks. Subsequently, rats were treated with systemic ceftriaxone (200 mg/kg IP) or vehicle and intra-accumbens active or control morpholino.

Results: We found that rats treated with ceftriaxone in the presence of xCT and GLT-1 antisense did not show attenuated cue-primed reinstatement, indicating that both proteins are critical to the actions of ceftriaxone. We also found a trend for a reduction in surface GluR2 expression in cocaine animals that was reversed by ceftriaxone and not affected by xCT or GLT-1 knockdown. Cocaine self-adminstration produced an increase in surface GluR1 expression that was normalized by ceftriaxone; the effect of ceftriaxone was prevented by knocking down xCT and GLT-1 individually.

Conclusions: These data support the importance of increasing both GLT-1 and xCT expression in the attenuation of cocaine reinstatement.

Keywords: cocaine, glutamate, GLT-1, AMPA.

Disclosure: Nothing to Disclose.

T33. Juvenile Onset of Stereotypy with Loss of BDNF Signaling to D1R Expressing Striatal Neurons

Mary Kay Lobo*, Michel Engeln, Ramesh Chandra, Ashley La

University of Maryland Medical School, Baltimore, Maryland

Background: Dysfunction of basal ganglia output-pathways arising from the two striatal medium spiny neurons (MSN) subtypes is implicated in the repetitive, non-functional movements that characterize stereotyped disorders including Tourette Syndrome (TS). Despite this proposed dysfunctional role for MSN subtypes in TS and other developmental neurological disorders, characterized by stereotypic movements, there is minimal information on the role of MSN subtypes in these disorders. We have a mouse model in which, a subset of mice display involuntary stereotypic behaviors during juvenile and early adult ages. These mice, D1-Cre-flTrkB mice, have a deletion of TrkB, the receptor for BDNF, in dopamine receptor 1 (D1)-MSNs. Additionally, we previously demonstrated that these mice display a decrease in many GABA-A subunits in the striatum which is accompanied by decreased inhibition in striatal D1-MSNs. This is consistent with recent studies implicating a role for a dysfunctional GABAergic system in TS patients.

Methods: To investigate a role for D1-MSNs in mediating stereotypic behaviors we first examine repetitive behaviors in D1-Cre-flTrkB mice that display stereotypy (S), D1-Cre-flTrkB mice that display no stereotypy (NS), and D1-Cre control mice. Stereotypy including the number of complete turns, head tics, rearing, and grooming was examined for 15minutes once a week from PND21 (3 weeks) to PND56 (8 weeks). Since the transcription factor, early growth response 3 (Egr3) is regulated by BDNF and in turn transcriptionally regulates a subset of GABA-A subunits that are decreased in D1-Cre-flTrkB mice we next examine Egr3 mRNA in striatum of D1-Cre-flTrkB (S), D1-Cre-flTrkB (NS), and control mice. Finally, we use chromatin immunoprecipitation to examine Egr3 binding to GABA-A subunit target genes in striatum of these mice.

Results: We found that D1-Cre-flTrkB (S) mice display significantly more complete turns in a 15minute period at all ages (PND21- PND56) compared to D1-Cre-flTrkB (NS) and D1-Cre control mice. D1-Cre-flTrkB (S) mice display head tics, with a trend for declined head tics from juvenile to adult ages. Head tics were not observed in the other groups. Additionally, preliminary data demonstrate that Egr3 is decreased in striatum of D1-Cre-flTrkB mice. We continue to investigate if this change is specific to D1-Cre-flTrkB (S) mice and if transcriptional regulation of GABA-A subunits by Egr3 is decreased in these mice.

Conclusions: Our data demonstrate a role for D1-MSNs in stereotyped behaviors characteristic of TS. Our findings of reduced GABA-A subunits and reduced inhibition in striatal D1-MSNs in D1-Cre-flTrkB mice implicate a role for altered GABAergic signaling in D1-MSNs in stereotypy behaviors. Additionally, our preliminary data demonstrating a decrease in Egr3 in D1-Cre-flTrkB mice implicate that GABAergic dysfunction in these mice is potentially regulated through dysfunctional BDNF-TrkB-Egr3 signaling. Our ongoing studies can provide insight into the transcriptional mechanisms underlying stereotypy in the D1-Cre-flTrkB (S) mice.

Keywords: striatum, stereotypy, medium sping neuron, BDNF.

Disclosure: Nothing to Disclose.

T34. Abnormal Shift in ErbB4 Splicing is Associated with Reduced Parvalbumin mRNA Levels in Layer 4 of the Dorsolateral Prefrontal Cortex in Subjects with Schizophrenia

Daniel Chung*, Dominique Arion, David Lewis

University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania

Background: Dysfunction of the dorsolateral prefrontal cortex (DLPFC) in schizophrenia (SZ) is associated with markers of lower parvalbumin (PV)-positive interneuron activity, which might reflect impaired excitatory synapse formation in PV interneurons. ErbB4, a receptor tyrosine kinase, plays an essential role in the formation of excitatory synapses onto PV interneurons. Postmortem studies demonstrated that total ErbB4 mRNA levels are unaltered, but levels of two minor splice variants of ErbB4, JM-a and CYT-1, are increased in total DLPFC gray matter in SZ. These findings suggest that dysregulated splicing, and not altered transcription, of ErbB4 mRNA is associated with SZ. In human DLPFC, ErbB4 is expressed predominantly in PV interneurons in layer 4, but in layer 2 the largest proportion of ErbB4-positive cells are calretinin (CR)-positive interneurons, a subclass of interneuron not affected in SZ. Thus, in order to understand the role of ErbB4 in SZ, it is necessary to determine if the abnormal shift in ErbB4 splicing is associated selectively with PV interneurons. In this study, we tested the hypothesis that in the DLPFC of SZ, ErbB4 splicing is abnormal in layer 4 where PV interneurons predominate but not in layer 2 where CR interneurons predominate.

Methods: From DLPFC of 39 matched pairs of SZ and control subjects, layers 2 or 4 were selectively captured by laser microdissection and RNA was isolated. qPCR assay was performed using primers for three house-keeping genes (beta-actin, GAPDH, cyclophilin A), two layer enriched markers (CR and PV), four ErbB4 splicing variants (JM-a, JM-b, CYT-1,CYT-2) and Pan-ErbB4. The expression levels of ErbB4 splicing variant in each layer were normalized to the expression level of Pan-ErbB4 in that layer to control for laminar variability in the transcription of total ErbB4 mRNA levels across subjects. ANCOVA models were used to test the effect of diagnosis on the levels of ErbB4 splicing variants in each layer. Correlation analyses were performed to assess the relationship between shifts in ErbB4 splicing and the levels of PV or CR mRNA levels in a layer-specific manner. Shifts in ErbB4 splicing at each locus were represented by the ratio of minor variant to major variant expression levels.

Results: Levels of CR mRNA were 10X fold higher than PV mRNA levels in layer 2, whereas in layer 4 PV mRNA levels were 10X fold greater than CR mRNA levels, indicating that microdissection of layers 2 or 4 captured samples enriched for CR or PV interneurons, respectively. In SZ, CR levels were not altered in either layer, whereas in layer 4 PV levels were significantly 19.5% lower, consistent with prior reports. Levels of ErbB4 CYT-1 mRNA were increased selectively in layer 4 (19.3%) but not in layer 2 of subjects with SZ. CYT-2 mRNA levels did not differ between subject groups in either layer. There was a 22% increase in JM-a levels selectively in layer 4 but not in layer 2 of SZ subjects. JM-b mRNA levels were lower by 15% in layer 2 and by 17% in layer 4. CR mRNA levels did not correlate with the ratio of either JM variants or CYT variants in layer 2. A strong negative correlation was observed between PV mRNA levels and the ratio of JM variants, but not CYT variants, in layer 4 of SZ (R=-0.502, p=0.001).

Conclusions: The present study demonstrates (1) selective enrichment of CR or PV interneurons in layers 2 or 4, respectively, (2) higher levels of ErbB4 CYT-1 and JM-a mRNAs selectively in layer 4 of subjects with SZ and (3) a negative correlation between PV mRNA and the ratio of JM variants selectively in layer 4 of SZ subjects. These findings suggest that SZ is associated with an abnormal shift in ErbB4 splicing preferentially in PV and not CR interneurons. As PV mRNA expression depends on excitatory inputs onto PV interneurons, these findings support the interpretation that a higher ratio of JM-a to JM-b is associated with lower excitatory drive onto, and thereby lower activity of, PV interneurons in SZ.

Keywords: schizophrenia, parvalbumin excitability, ErbB4 pathway.

Disclosure: Nothing to Disclose.

T35. The L-Type Calcium Channel Genes CACNA1C and CACNA1D Are Associated with Cellular Circadian Rhythm Abnormalities in Bipolar Disorder

Michael McCarthy*, Heather Wei, Melissa Le Roux, John Kelsoe, David Welsh

University of California at San Diego, San Diego, California

Background: Bipolar Disorder (BD) is strongly influenced by genetic factors, and because it is characterized by altered circadian rhythms in sleep and activity, the “clock genes” that govern circadian rhythms have been studied as candidate susceptibility genes. While genome wide association studies have identified genetic variants that are strongly associated with BD, clock genes are not among the risk loci most reliably identified. Instead, single nucleotide variants affecting L-type calcium channel genes (e.g. CACNA1C, CACNA1D) have been linked with BD, providing indications that calcium signaling is critically affected in the illness. Calcium is involved with many functions, including bidirectional interactions with the clock. Therefore, circadian rhythm abnormalities in BD may be related to alterations in calcium signaling, and not to genetic abnormalities in clock genes. In previous work, we found that lithium treatment of human fibroblasts cultured from psychiatrically healthy donors increased the amplitude of circadian rhythms in gene expression, but that this effect was attenuated in cultures from donors with BD. Based on this finding, we have conducted follow-up studies to examine the potential of lithium-induced circadian amplitude modulation as a diagnostic marker of BD and therapeutic response to lithium. We hypothesized that cellular circadian rhythms in BD are regulated by calcium, and that BD-associated variation in calcium channel genes would affect circadian rhythms. We verified key aspects of these hypotheses, establishing an important link between BD-associated calcium channel variants, and circadian rhythms in BD.

Methods: Circadian rhythms in gene expression were measured using a bioluminescent reporter gene expressed under the control of the Per2 promoter (Per2::luc) to count photoemissions at 10 min intervals over 5 consecutive days. Rhythm parameters (amplitude, period and phase) were estimated using least squares fitting to a sine wave. For studies of mouse fibroblasts, NIH3T3 cells were stably transfected with Per2::luc and grown under hygromycin selection. For human studies, fibroblast cell lines from control and BD subjects were transduced with Per2::luc using a recombinant lentiviral vector. For drug studies, cells were grown in the presence or absence of lithium, calcium channel antagonists or both for 48 hr prior to luminometry. For gene knockdown studies, cells were transfected with siRNAs targeting specific transcripts 48hr prior to luminometry. To examine the effects of genetic variants on rhythm parameters, human cell lines from BD patients and controls were genotyped by PCR for two common calcium channel single nucleotide variants in CACNA1C, and CACNA1D previously associated with BD. Rhythm parameters were then re-analyzed as quantitative genetic traits using a dominant model. CACNA1C and CACNA1D expression was measured in human cells collected at regular time intervals using RT-PCR.

Results: In NIH3T3 cells, depolarization with potassium increased the rhythm amplitude, an effect that was attenuated by the L-type calcium channel antagonist diltiazem, indicating the presence of voltage gated calcium channels. Pre-treatment of NIH3T3 cells with diltiazem also attenuated the amplitude response to lithium. Similar effects were observed with a chemically distinct calcium channel antagonist, verapamil. Knock down of CACNA1C expression strongly increased rhythm amplitude. Knockdown of CACNA1D increased rhythm amplitude, but only transiently. In human fibroblasts, a genetic variant in CACNA1C previously linked to BD (rs4765913) was associated with rhythm effects of lithium. Those with the major allelic variant showed an increase in amplitude after lithium, while carriers of the minor, BD-associated allele did not. There were no significant associations of CACNA1D variants with any rhythm parameter. In a time course analysis of gene expression, CACNA1C expression was rhythmic in cells from controls, but non-rhythmic in cells from BD patients carrying the minor allele of rs4765913. The expression of CACNA1D was also rhythmic, but in antiphase to CACNA1C.

Conclusions: Our data reveal that L-type calcium channels in general, and CACNA1C in particular, affect signaling pathways that regulate circadian rhythms, thereby linking one of the best supported genetic associations in BD to a behavioral feature that is central to the illness, but poorly explained by previous candidate gene analyses of the clock. Our data suggest that calcium channels modulate some aspects of lithium’s effects on rhythms, a finding that could elucidate some features of lithium’s therapeutic mechanism of action. Our results also indicate that the functional relationship among calcium channel subtypes is complex, with individual differences in the temporal regulation of CACNA1C and CACNA1D. Based on our results, we conclude that L-type calcium channel modulators may be prime candidates for future drug development to regulate abnormal circadian rhythms in BD.

Keywords: calcium, circadian rhythm, bipolar disorder, lithium.

Disclosure: Nothing to Disclose.

T36. Mechanisms Mediating Circadian Gene Effects on Anxiety-like behavior: Focus on NPAS2 & GABAA

Angela Ozburn*, Joseph Kern, Puja Parekh, Ryan Logan, Zheng Liu, Kush Purohit, Yanhua Huang, Colleen McClung

Portland VA Medical Center, Portland, Oregon

Background: Clinical and pre-clinical studies have provided strong evidence that circadian rhythms and the genes that make up the molecular clock play a key role in the expression of mood-related symptoms in psychiatric disorders. These studies have revealed that circadian rhythms contribute to differences in mood state, such as anxiety and depression. Abnormal rhythms are strongly associated with psychiatric diseases like seasonal affective disorder, bipolar disorder, and major depression. More research is needed to identify mechanisms by which circadian genes in specific brain regions regulate mood, specifically anxiety. Npas2 is a core component of the molecular clock that acts as a transcription factor and is highly expressed in reward- and stress-related brain regions such as the nucleus accumbens (NAc). The NAc, primarily composed of GABAergic medium spiny neurons (MSNs), is a significant point of convergence for this circuitry. Npas2 expression is enriched specifically in Drd1 containing MSNs that make up the so-called “direct” pathway. Variations in Npas2 are associated with seasonal affective disorder and major depressive disorder. Here, we determined the effect of a chronic stress paradigm on NAc Npas2 expression in mice, characterized the behavioral phenotype of mice with a null mutation in Npas2 and mice with reduced NAc Npas2 expression (via RNAi), validated GABAA subunits as novel transcriptional targets of NPAS2, and assessed the effects of NAc Npas2 knock down on inhibitory neurotransmission.

Methods: C57BL/6 mice were subjected to a chronic unpredictable mild stress or control handling protocol and sacrificed at 6 time points over a 24 hr period (n=6/group/timepoint). Quantitative PCR was performed to determine NAc Npas2 expression (relative to Gapdh expression). We measured anxiety-like behaviors and motor coordination in Npas2 knockout and wild type mice (n=15-25/group). We determined if Npas2 knock down in the NAc (achieved via stereotaxic delivery of AAV2-Npas2 shRNA, or a scramble control sequence with no known target) is sufficient to produce the reduced anxiety-like behaviors seen in the knockout mice (n=13-19/group). Altered GABAA subunit composition and possibly altered inhibitory synaptic transmission may occur with reductions in Npas2 (since Npas2 knockout mice have reduced diazepam sensitivity and several GABAA subunit genes were recently identified as transcriptional targets of NPAS2). Using chromatin immunoprecipitation, we confirmed that NPAS2 binds genes encoding GABAA subunits (alpha 1, 2, 4, 5; beta 2, 3). In order to assess the effect of Npas2 knockdown on inhibitory neurotransmission, we performed whole-cell patch clamp recordings on NAc MSNs from mice treated with AAV-Npas2 shRNA or scramble control (mIPSCs; n=12-13 cells/group from 4-5 mice/group). Diazepam potentiates GABA-receptor mediated currents by binding specific subunits of GABAA receptors and promoting the binding of GABA, which in turn increases total conduction of chloride. Therefore, we measured the peak amplitude of the evoked IPSC at baseline and following diazepam bath application (10μM) in both scramble and Npas2-shRNA treated MSNs (n=10-12 cells/group from 4 mice/group).

Results: Chronic unpredictable mild stress resulted in significantly elevated Npas2 expression in the NAc for all time points (as compared with handled controls; p<0.0001). Npas2 knockout mice exhibited significantly decreased anxiety-like behaviors as compared with wild-type mice. Npas2 knockouts exhibited an increased percent time in open arms of the elevated plus maze (p<0.05), reduced latency to explore the light side of the light/dark box (p<0.05), and an increase in the distance traveled in the center of the open field arena (p<0.05). Further, Npas2 knockout mice were resistant to the motor incoordinating effects of diazepam (rotorod assay; p<0.05). Npas2 knock down in the NAc was sufficient to produce reduced anxiety-like behaviors. Npas2 knock down resulted in an increased percent time in open arms of the elevated plus maze (p=0.05), reduced latency to explore the light side of the light/dark box (p<0.05), and an increase in the velocity of movement in the center of the open field arena (p<0.05) as compared with scramble). We found a significant decrease in mIPSC amplitude, but not frequency, in MSNs where Npas2 has been knocked down (p<0.05), suggesting Npas2 knock down produces a postsynaptic modification of MSNs. We found that diazepam application reliably increased the average IPSC peak amplitude in scramble control cells by approximately 20%, while this increase in current amplitude was noticeably absent in Npas2 shRNA infected cells (p<0.05).

Conclusions: There is strong evidence that circadian genes play an important role in the expression of mood-related symptoms in psychiatric disorders. We found that NPAS2 modulates anxiety-like behaviors via GABAA receptors using a combination of molecular and behavioral approaches. This may be part of a homeostatic mechanism by which stress increases NPAS2 dependent transcription of specific GABAA receptor subunits, thus altering the localization and function of the receptors and ultimately altering inhibitory neurotransmission (perhaps specifically in Drd1 MSNs).

Keywords: anxiety, circadian, GABAA, NPAS2.

Disclosure: Nothing to Disclose.

T37. Using hIPSCs to Model Disease Mechanisms in Schizophrenia

Talia Atkin*, Yuchen Qi, Ziyi Sun, Sander Markx, Joseph Gogos, Mark Tomishima, Lorenz Studer, Maria Karayiorgou

Columbia University/New York State Psychiatric Institute, New York, New York

Background: Schizophrenia is a severely debilitating disorder affecting 1% of the population worldwide with a heavy social burden and high economic cost. The causes of schizophrenia are poorly understood, however genetic studies have begun to shed light on the underlying pathology. Microdeletions on human chromosome 22 (22q11.2) constitute one of the strongest known genetic risk factors for schizophrenia. Microdeletions in this region occur de novo, are found in 1% of schizophrenia cases worldwide, and are associated with high penetrance (25-30%) for schizophrenia. Animal models have suggested mechanisms by which this deletion can contribute to the disease state but whether these findings translate to humans remains to be explored. Production of human induced pluripotent stem cells (hIPSCs) via direct reprogramming is a tractable approach, which promises to be a facile source of patient-derived cell lines.

Methods: We have obtained and derived hIPSC lines from schizophrenia patients carrying 22q11.2 microdeletions and unaffected siblings, not carrying the microdeletion. The patients are currently enrolled in our human genetic studies and all meet full diagnostic criteria for schizophrenia. The 22q11.2 deletion has been confirmed by fluorescent in situ hybridization (FISH) and Multiplex Ligation-dependent Probe Amplification (MLPA) for all patient participants and excluded for all sibling control participants. We have utilized a small molecule inhibition method to direct cortical neuron specification of hIPSC lines to allow the investigation of the impact of genetic risk factors on human neuronal populations affected in schizophrenia. We use these human derived neurons to investigate disease related phenotypes.

Results: We have successfully generated forebrain neurons confirmed by Tuj1+staining in up to 60% of cells. Furthermore we have confirmed cortical neuronal identity using markers including Tbr1 and Reelin. Using electrophysiology we have demonstrated the generation of functional neurons showing depolarization activated transient-inward Na+currents and sustained-outward K+currents, that fire repetitive trains of action potentials upon somatic current injection. In these neurons we have found that the 22q11.2 deletion confers reduced neuronal complexity, corresponding with neuronal phenotypes we have observed in neurons from the brains of 22q11.2 deletion mouse models. Comparison of the genetic profile via real-time PCR revealed alterations in the expression of a number of mRNAs and decreases in miRNA expression levels in the disease derived lines compared with neurons derived from deletion negative healthy sibling controls. Finally, in addressing developmental properties of these patient derived lines we have identified a delayed maturation profile in the disease lines compared with the control lines.

Conclusions: These findings utilize human derived neurons to investigate molecular mechanisms underlying schizophrenia. Our results validate and extend findings from mouse models, therefore, supporting the utility of both mouse and human cellular models for better understanding disease pathology. Our findings demonstrate altered neuronal development in neurons derived from 22q11.2 deletion patients and identify potential molecular mechanisms by which these changes might occur. Understanding the mechanisms by which the 22q11 deletion can contribute to disease pathology in neurons derived from humans suffering from schizophrenia provides us with highly translatable avenues for treatments in 22q11.2 linked schizophrenia and for treatment of schizophrenia population wide.

Keywords: schizophrenia, hIPSCs, 22q11.

Disclosure: Nothing to Disclose.

T38. Pre-existing and Diet-induced Alterations in Striatal Function in Preclinical Models of Obesity

Carrie Ferrario*, Cameron Nobile, Peter Vollbrecht, John Corthell, Luis Lopez-Santiago, Paula Goforth

University of Michigan Medical School, Ann Arbor, Michigan

Background: Alterations in the function of mesocorticolimbic circuits, particularly in the nucleus accumbens (NAc) contribute to aberrant motivation for drug and food rewards (Dagher 2009; Carr et al. 2011; Avena et al. 2012; Volkow et al. 2013). Interestingly, human neuro-imaging studies suggest that the NAc response to food-cues is enhanced in obese people, even prior to the development of obesity (Rothemund et al., 2007; Stoeckel et al. 2008; Demos et al., 2012). This suggests that pre-existing differences in NAc function may contribute to over-eating that drives much of obesity. The NAc receives convergent glutamate and dopamine input. AMPA-type glutamate receptors provide the main source of excitation to the NAc, and dopaminergic activity can modulate glutamatergic transmission (Wolf 2010). However, few studies have examined NAc function in preclinical models of obesity. Thus, we have begun a series of studies to determine whether there are pre-existing and/or diet-induced alterations in NAc function and glutamatergic transmission in preclinical models that capture individual susceptibility to obesity.

Methods: Two models of individual susceptibility to obesity were used: outbred male Sprague-Dawley rats given free access to a sweet and fatty “junk-food” diet and male rats selective bred for their propensity or resistance to diet induced obesity (Levin 1997; bred in house). The “junk-food” diet consisted of a mash of standard lab chow, potato chips, chocolate chip cookies, peanut butter and powdered chocolate milk. Selectively bred rats were used to examine pre-existing differences. Alterations in mesolimbic function were determined by examining behavioral responses to systemic cocaine administration. NAc glutamatergic function was assessed using whole-cell patch clamp recordings of medium spiny neurons in the NAc of adult rats and biochemical approaches.

Results: We found junk-food diet-induced and pre-existing differences in striatal function. Specifically, outbred obese rats showed more robust cocaine-induced sensitization after junk-food diet deprivation than non-obese rats given the same diet. This was associated with a selective increase in GluA1, but not GluA2 surface protein expression in the NAc. Further, in a separate set of rats, junk-food diet followed by deprivation increased AMPA-mediated synaptic transmission in obese vs. non-obese rats. In addition, AMPAR-mediated transmission was decreased by the CP-AMPAR blocker NASPM in obese-rats. In selectively bred rats, pre-existing differences in sensitivity to the locomotor activating effects of cocaine and NAc transmission were found. Specifically, obesity-prone rats showed a sensitized response to the acute locomotor-activating effects of cocaine compared to obesity-resistant rats without any diet manipulation. Interestingly, preliminary data suggest that intrinsic excitability of medium spiny neurons in the NAc may be enhanced, whereas synaptic glutamatergic transmission may be decreased in obesity-prone vs. obesity resistant rats. This is opposite to decreases in excitability that are normally associated with sensitization after repeated cocaine exposure. Studies are currently underway to examine effects of junk-food diet in selectively bred rats.

Conclusions: Our data suggest that both pre-existing and diet-induced alterations in NAc function occur in rodents that are susceptible to diet induced obesity. Our data in outbred rats suggest that there are individual differences in the extent to which sugary, fatty foods increase transmission via CP-AMPARs. This is particularly interesting in light of the role for CP-AMPARs in cue-induced drug-seeking behavior (Loweth et al., 2013) and our recent work suggesting that motivation for sucrose-cues is enhanced in obese rats. Further, our data from selectively bred rats suggests that there are pre-existing differences in striatal function that may contribute to effects of junk-food diet exposure in obesity-prone individuals.

Keywords: Nucleus Accumbens, AMPAR, obesity, motivation.

Disclosure: Nothing to Disclose.

T39. Targeting Corticotropin Releasing Factor (CRF) Projections from the Bed Nucleus of the Stria Terminalis (BNST) Using Cell-Type Specific Neuronal Tracing Studies in Mice and Rats

Joanna Dabrowska*, Donald G. Rainnie

The Chicago Medical School at Rosalind Franklin University, North Chicago, Illinois

Background: The bed nucleus of the stria terminalis (BNST) is known to play a critical role in mediating the behavioral and autonomic responses to stressors. In addition, growing evidence suggests that BNST is specifically implicated in mediating of negative affect following chronic stress exposure or addiction withdrawal. The oval and fusiform nuclei of the anterolateral cell group of the BNST (BNSTALG) contain cell bodies that synthesize the stress hormone, corticotropin releasing factor (CRF). Although afferent fibers originating from the BNSTALG have been shown to innervate several key structures of the neuroendocrine and central autonomic system, the question remains as to whether, some of these fibers are CRF-positive, and what neuronal populations they project to.

Methods: To directly address this question, we injected a “floxed” anterograde tracer (rAAV5/EF1a-DIO-mCherry) into the oval nucleus of the BNSTALG of CRFp3.0CreGFP transgenic mice, which express a green fluorescent protein (GFP) under the control of the CRF promoter. Two weeks after the AAV injection, serial sections were cut through the entire brain and analyzed for the presence of dual-labeled fibers (GFP-mCherry) in potential projection sites using confocal microscopy. To further explore whether CRF neurons in the rat BNSTALG send analogous projections, we injected the oval nucleus with an AAV in which the human synapsin promoter-5 (hSyn) drives GFP expression. Six weeks later serial sections were cut from the entire brain, stained with an anti-CRF antibody, and visualized with an Alexa 586 secondary antibody. Sections were analyzed for the presence of dual-labeled fibers (green-red). Finally, to determine what populations of neurons are innervated by the CRF projections originating in the BNST (double-labeled green-red), we have performed triple-immunofluorescence protocol, in which we have marked midbrain sections with tyrosine hydroxylase (TH), hypothalamus sections with oxytocin (OT), and raphe nucleus sections with serotonin transporter (5-HTT) antibodies and visualized local neurons with a blue marker, Alexa-Fluor 350.

Results: Our results show high GFP-mCherry co-expression in cell bodies of the oval nucleus of the BNSTALG and in several terminal fields in the mouse brain. Notably, dual-labeled fibers were observed in the dorsal raphe nucleus (DRN), the vetral tegmental area (VTA), as well as in the paraventricular nucleus of the hypothalamus (PVN). Injection of rAAV5-hSyn-GFP into the rat BNSTALG showed robust somatodendritic GFP expression in the oval nucleus of the BNSTALG. Dual-immunofluorescence studies revealed that the subpopulation of GFP-positive neurons in the BNST co-expressed CRF (dual-labeled neurons). AAV5-hSyn-GFP also resulted in robust GFP expression in fibers and terminal fields in the rat brain. Analysis of dual-labeled fibers confirmed the presence of CRF-GFP fibers in several divisions of the PVN: namely the anterior (PaAP), and the medial parvicellular (PaMP) divisions, the ventral (PaV), the posterior (PaPO), and PaLM-lateral magnocellular division. CRF-GFP terminals were also seen in the VTA and the central nucleus of the amygdala. The double-labeled CRF fibers originating from the BNST were also shown to make perisomatic contacts with the local oxytocin neurons in the PaV.

Conclusions: The results of our study indicate that CRF neurons in oval nucleus of the BNSTALG send projections not only to the regions critically involved in the neuroendocrine and autonomic regulation, but also to the centers modulating reward and affective behavior. Our results also show that the neuroanatomical circuit between the CRF neurons in the BNST and oxytocin neurons in the hypothalamus is fully reciprocal, as CRF projections from the BNST innervate oxytocin neurons in the hypothalamus and vice versa. This work was supported by Grant Number R00MH-096746 from the National Institute of Mental Health and RFUMS start-up funds to JD and R01MH-072908 to DGR from the National Institute of Mental Health.

Keywords: bed nucleus of the stria terminalis, hypothalamus, CRF, neuronal tracing.

Disclosure: Nothing to Disclose.

T40. Human-like Relapse Vulnerability in the Rat: Generating and Phenotyping a Slc7a11 Knockout Rat

David Baker*, SuJean Choi, Brian Maunze, Nicholas Raddatz, Linghai Kong, John Mantsch, Aron Geurts

Marquette University, Milwaukee, Wisconsin

Background: Drug addiction is a complex, multi-faceted disorder exemplified by prolonged periods of heightened relapse vulnerability resulting from stress or exposure to drug-related stimuli, including small amounts of the abused drug. Compulsive drug use appears to involve, at least in part, persistent changes in excitatory signaling in corticostriatal circuits. A critical obstacle in attempts to understand and treat addiction is the degree to which glutamate is regulated by a poorly understood network of receptors, membrane-expressed enzymes, transporters, and release mechanism expressed on neurons and astrocytes. For example, cystine-glutamate exchange by system xc- is a form of nonvesicular glutamate release that appears to be disrupted in drug addiction, yet the potential for altered system xc- activity to contribute to the pathophysiology of drug addiction is unknown. To explore the possible link between reduced system xc- activity and heightened relapse vulnerability, we mutated the Slc7A11 gene to produce an Sxc loss of function rat.

Methods: Transgenic rats were developed by mutating the Slc7A11 gene using Zinc-Finger Nucleases. Cystine-glutamate exchange by system xc- was assessed by measuring the uptake of 14C-cystine in tissue punches obtained from the prefrontal cortex and/or the striatum. Behavioral measures included the rate acquisition of cocaine self-administration, cocaine intake during 12 daily sessions, rate of extinction, and levels of drug-seeking following administration of a low dose of cocaine (3 mg/kg, IP).

Results: Two distinct mutant lines were generated. The first strain (M1xc) involves 9 base pair changes in exon 2 leading to a frameshift and anticipated truncation of the xCT protein; the second line (M2xc) has a deletion of 39 base pairs corresponding to the majority of the 3rd transmembrane domain. Although both M1 and M2xc displayed a complete lack of Sxc- mediated uptake of 14C-cystine, behavioral testing to date has involved only M2xc rats. We did not obtain an effect of genotype on the rate of acquisition of cocaine self-administration (mean+SEM: 12+2.7 sessions in WT versus 10+3.8 in M2xc; p>.05) or the levels of cocaine consumed during 12 self-administration training (mean+SEM: 248+21 mg/kg in WT versus 218+13 in M2xc; p>.05). However, M2xc rats displayed a significant escalation of drug intake across the 12 sessions (T(12)=4.4, p<.001) that was not observed in the WT controls (T(9)=0.189, p>.05). In addition, M2xc rats exhibited higher levels of responding during extinction training (F(1,21)=5.19, p<.05). Finally, M2xc rats also displayed augmented cocaine-induced drug-seeking on the reinstatement test day (T(21)=4.29, p>.001). Remarkably, the heightened drug seeking in the M2xc rats resulting from the single injection of the low dose of cocaine (3 mg/kg, IP) remained significantly elevated for at least four days (F(1,20)=7.25, p<.05).

Conclusions: Attempts to understand the neural basis of drug-addiction have often focused on characterizing drug-induced changes in glutamate signaling. While models of pathological excitatory signaling have primarily incorporated synaptic mechanisms, it is becoming apparent that glutamate is regulated by a number of extra-synaptic processes, including nonvesicular glutamate release from astrocytes. In these studies, we find that that deletion of system xc- activity, which is often expressed on astrocytes, resulted in enhanced drug seeking evident during multiple phases of the self-administration/reinstatement paradigm. Of particular importance, M2xc rats displayed heightened drug seeking following a single injection of a low dose of cocaine that persisted for days, which is reminiscent of relapse vulnerability displayed by human substance abusers. By implicating system xc- in addiction, these and related studies are expanding the search for the biological basis of pathological excitatory signaling to include molecules expressed by the most abundant cell in the brain, astrocytes. Work supported by DA035088.

Keywords: cocaine, reinstatement, glutamate, astrocytes.

Disclosure: Promentis Pharmaceuticals.

T41. Growth Arrest and DNA Damage-Inducible 45-Beta (Gadd45b) and Neuronal Activity-dependent DNA Demethylation

David Gavin*, Handojo Kusumo, Rajiv P. Sharma, Marina Guizzetti, Alessandro Guidotti, Subhash C. Pandey

University of Illinois at Chicago/Jesse Brown VA Medical Center, Chicago, Illinois

Background: Abnormalities in DNA methylation have been reported in a variety of mental illnesses both site specifically as well as globally including in schizophrenia. If not for the ability to actively remove DNA methyl groups neuronal identity would become fixed, thereby making underlying pathological processes entrenched and immutable. Over the last decade there has been increasing support for a DNA demethylation pathway involving base excision repair mechanisms, whereby Tet enzymes convert 5-methylcytosine (5MC) to 5-hydroxymethylcytosine (5HMC) which is later removed by a process involving cytosine deaminases and thymine or uracil glycosylases. Prior reports suggest Gadd45b is a key coordinator in this active DNA demethylation pathway. However, many of these studies were performed in vivo where it is difficult to distinguish active DNA demethylation that non-dividing neurons use exclusively, from passive demethylation that can occur in dividing cells such as glia. We have previously reported abnormalities in GADD45b expression, GADD45b binding to the BDNFIX promoter, and increased 5MC and 5HMC at the BDNF IX promoter in psychotic subjects’ postmortem brain samples. We examined DNA demethylation in a homogeneous population of non-dividing mouse primary neurons using the fundamental characteristic of the neuron known to induce DNA demethylation, neuronal depolarization.

Methods: Mouse E18 primary cortical neuronal cultures were prepared based on a previously published protocol. We pharmacologically depolarized neurons using NMDA. Small interfering RNAs (siRNAs, 25 nt) were designed against Gadd45b mRNA. Target gene expression in neurons was normalized to Hypoxanthine Phosphoribosyltransferase 1 (Hprt1). Methylated DNA Immunoprecipitation (MeDIP) or hydroxymethylated IP (hMeDIP) samples were measured using qPCR. IP samples were normalized to input.

Results: In mouse E18 primary cortical neuronal cultures NMDA depolarization induces Gadd45b, Bdnf IV and IXa mRNA expression, while reducing Tet1, Dnmt3a, and Dnmt3b expression. This increase in Bdnf mRNA is accompanied by a significant reduction in 5HMC at both Bdnf IV and IXa promoters, and a reduction in 5MC at the Bdnf IXa but not the IV promoter in primary neuronal cultures. In Gadd45b siRNA transfected neurons we find that Bdnf IXa mRNA expression is not increased following NMDA depolarization, and there is a lack of decrease in Tet1 expression. Moreover, in Gadd45b siRNA transfected neurons NMDA fails to reduce 5MC at the Bdnf IXa promoter, and does not merely fail to reduce 5HMC at the Bdnf IXa promoter but actually leads to a significant increase.

Conclusions: Our results indicate that Gadd45b is involved in the removal of 5MC and 5HMC at Bdnf promoters. We speculate that knockdown of Gadd45b leads to an increase in 5HMC following neuronal depolarization because activity-dependent Tet1 reduction is impaired and because the removal of 5HMC is compromised by the lack of Gadd45b coordinating base excision repair mechanisms. These results contribute to our understanding of DNA demethylation in neurons, and indicate possible targets for mental illnesses in which abnormalities in these pathways have been reported such as schizophrenia.

Keywords: CpG Methylation, Epigenetic, Histone, Brain neurotrophic factor.

Disclosure: Nothing to Disclose.

T42. β2-Subunit Containing and α7 Nicotinic Receptors in the Amygdala Regulate Mood and Social Stress Resilience

Yann Mineur*, Sam Blakeman, Gianna Fote, Sonya Zhou, Jessica Xia, Syliva Newbold, Marina Picciotto

Yale University School of Medicine, New Haven, Connecticut

Background: Blockade of acetylcholine esterase (AChE), the primary degradative enzyme for acetylcholine (ACh), can induce symptoms of depression in human subjects. In mice, increasing cholinergic tone in the hippocampus can also induce depression-like phenotypes and decreased resilience to social stress. Both animal and human studies have suggested that compounds that alter signaling of nicotinic acetylcholine receptors (nAChRs) can have antidepressant-like properties. Pre-clinical studies have further shown that nicotine can act at both β2 subunit-containing (β2*) high affinity nicotinic acetylcholine receptors (nAChRs) and α7 type nAChRs to affect depression-like symptoms. Further, different nicotinic compounds with antidepressant-like effects converge to decrease c-fos expression (a measure of neuronal activity) in the basolateral amygdala. We therefore determined whether local manipulation of nAChR activity in amygdala was sufficient to mediate the antidepressant-like effects of nicotinic agents.

Methods: We infused the nicotinic antagonist mecamylamine into the amygdala to determine whether blocking nAChRs in this brain region would be sufficient to induce antidepressant-like effects. We then delivered small hairpin RNAs (shRNAs) into the amygdala using an adeno-associated viral vector (AAV) to down-regulate expression of the β2* or α7 nAChR subunit locally. After shRNA expression and nAChR knockdown, mice were tested in the elevated plus maze and light/dark box tests (for anxiety-like behavior), the tail suspension and forced swim tests (tests of antidepressant efficacy) and the social defeat paradigm (a test of social stress response). A separate group of β2* or α7 nAChR knock down animals was also tested for response to the AChE antagonist physostigmine, a treatment used to induce a depression-like phenotype. Anatomical controls were also performed by knocking down β2* or α7 nAChR subunits in the prefrontal cortex or the hippocampus. Following behavioral testing, localization of viral expression was verified, and c-fos expression was quantified to identify changes in neuronal activity following shRNA-mediated knockdown.

Results: Local infusion of mecamylamine or shRNAs targeting either the β2* or the α7 nAChR subunit into the amygdala induced robust antidepressant–like effects across mouse models of antidepressant efficacy. Further, knockdown of either the β2* or the α7 nAChR subunit also decreased anxiety-like behavior, increased stress-resilience in the social defeat paradigm and reversed the effects of physostigmine administration. Further, the antidepressant-like effects observed following systemic injection of mecamylamine were significantly occluded by knockdown of the β2 nAChR subunit in the amygdala. When a similar strategy was applied in the prefrontal cortex and the hippocampus, no significant effects were seen in these behavioral tests.

Conclusions: The current set of results suggest that decreasing signaling through α7 or β2* nAChRs in the amygdala results in an antidepressant-like response, likely due to a decrease in tonic ACh signaling in this structure. Blocking nicotinic signaling in the amygdala also reversed the effects of physostigmine in tests of anxiety- and depression-like behaviors. This suggests that limiting cholinergic signaling in the amygdala alone can reverse the depression-like effects induced by heightened cholinergic signaling across brains regions. Conversely, blockade of nicotinic neurotransmission in the PFC and hippocampus was not effective in altering anxiety- and depression-like behaviors consistently. The current results suggest that changes in α7 and β2* nAChR activity in the amygdala are important for mood regulation and that decreases in amygdala ACh signaling may be an important biomarker for antidepressant effects of cholinergic medications.

Keywords: Amygdala, nicotinic acetylcholine receptors, depression, stress.

Disclosure: Nothing to Disclose.

T43. GABAergic Remodeling in the Alzheimer's Disease Brain

Agenor Limon*, Jorge Reyes-Ruiz, Ricardo Miledi

University of California at Irvine, Irvine, California

Background: It has been accepted for long time that cholinergic and glutamatergic systems are severely disrupted in Alzheimer’s disease (AD); GABAergic neurotransmission, in contrast, had been generally though as well preserved. Although new evidence from animal models and human postmortem tissue suggest that there are compensatory changes in the AD brain, there is little information about how those changes affect the physiology and pharmacology human native GABA receptors. Since early stages of AD are associated with hyperexcitability in affected brain areas, understanding the functional changes of inhibitory receptors in the AD brain may be a potent tool in the prevention and design of new pharmacological therapies for early and late AD.

Methods: To gain information about the electrophysiology and pharmacology of human native GABA receptors, we first measured gene and protein expression of principal GABA receptors subunits expressed in the temporal cortex of AD postmortem brains and appropriate controls. We then microtransplanted cell membranes, isolated from the same brain area and from the same cohort, into Xenopus oocytes and recorded the electrophysiological activity of GABA receptors by two electrode voltage clamp methods. The experiments were conducted in accordance with the National Institute of Health Guide for the care and Use of laboratory Animals and after the approval of the appropriate UCIrvine committee.

Results: We found an age-dependent reduction of GABA ion currents in the AD brain. This reduction was larger when membranes were obtained from younger cases of AD, and as the age of the AD-brain increased, reductions of GABA currents with aging in control brains made the reduction in AD less conspicuous. We did not observed changes in the current-voltage relationships of the GABA currents in AD. However, we found that GABA currents from AD brains have a faster rate of desensitization than those from non-AD brains. Furthermore, GABA receptors from AD brains were slightly, but significantly, less sensitive to GABA than receptors from non-AD brains (P=0.01 Nested ANOVA, brains nested within diagnosis). The reduction of GABA currents in AD was to the same extent, if no more, than the reduction of currents elicited by non-NMDA glutamate receptors. The reduction of GABA currents was also associated with reductions of the principal GABA receptor subunits normally present in temporal cortex measured by qRT-PCR. Pair wise analysis of the transcripts within control and AD groups and analyses of the proportion of GABA receptor subunits altogether found down regulation of alpha1 and gamma1 subunits in AD. In contrast, the proportions of the transcripts for alpha2, beta1 and gamma1 were up-regulated in the AD brains. Additionally, we found that the linear correlation between gene expression of gephyrin and gamma2 subunits found in controls was missing in AD brains.

Conclusions: By utilizing an integrative approach between gene expression and an innovative approach recording the activity of native and disease-affected receptors we have found that GABA receptors are severely affected in AD. The observed changes were not limited to less magnitude of GABA responses but also to physiological and pharmacological changes of native receptors. The faster desensitization and lower sensitivity of microtransplanted GABA receptors from AD brains suggest that the functional efficacy of the inhibitory signaling in AD is impaired; these changes may thus promote hyperexcitablity of affected brain areas and produce paradoxical responses to clinically used gabaergic medications. Since in normal conditions gephyrin participates in the clustering of gamma2-containing GABA receptors at inhibitory synapses the lack of correlation between gamma2 and gephyrin points towards dysfunctional spatial rearrangements of GABAergic synapses in AD brains. Our results indicate that GABAergic signaling is a potential target for pharmacological intervention in AD.

Keywords: Alzheimer's disease, inhibition, synapses.

Disclosure: Nothing to Disclose.

T44. Investigating the Role of Nacore Astrocytes in Reinstated Methamphetamine Seeking

Michael Scofield*, Heather Boger, Peter Kalivas, Carmela Reichel

Medical University of South Carolina, Charleston, South Carolina

Background: Glial cells of the central nervous system directly influence neuronal activity by releasing neuroactive species, including glutamate, cytokines chemokines and neurotrophic factors. Methamphetamine (meth) -induced reductions in extracellular glutamate in the nucleus accumbens core (NAcore) produce alterations in synaptic plasticity linked to relapse vulnerability. Moreover, meth exposure is potently neurotoxic, inducing activated microglia, blood brain barrier disruption and enhanced neuroinflammation, which have been linked to drug abuse related behaviors. As such, ample evidence supports a glial involvement in the neural substrates of meth addiction. Here we target glial cells as an approach to pharmacotherapy for the treatment of meth addiction by promoting glial glutamate release and modulating glial cytokine release.

Methods: We transduced NAcore astrocytes with an AAV viral vector encoding the hM3Dq (Gq) designer receptor exclusively activated by a designer drug (DREADD), under control of the glial fibrillary acidic protein (GFAP) promoter. We then quantified DREADD-mediated NAcore glutamate release following intracranial or systemic administration of CNO and validated the glial origin of release using biosensors. For reinstatement testing, rats underwent 10 days of daily meth or sucrose self-administration and were tested for reinstatement of drug seeking precipitated by conditioned cues following 14 days of extinction training. Tests were counterbalanced between clozapine-N-oxide (CNO) and vehicle groups. In a separate experiment, rats received intracranial microinfusion of ibudilast or vehicle during the last 5 days of extinction of meth seeking. Ibudilast, is a glial modulator that inhibits phosphodiesterase activity and the release of proinflammatory cytokines, while promoting release of anti-inflammatory cytokines.

Results: Administration of CNO in GFAP-Gq-DREADD transduced animals increased extracellular NAcore glutamate levels in vivo. The origin of released glutamate was validated by the absence of CNO-mediated release in mice expressing a dominant-negative SNARE variant also driven by the GFAP promoter. Systemic administration of CNO inhibited cue-induced reinstated meth, but not sucrose seeking. Furthermore, bilateral microinfusion of ibudilast into the NAcore reduced cued meth seeking when compared to vehicle controls. The importance of cytokine regulation or phosphodiesterase inhibition with respect to reinstated meth seeking and distinguishing which cytokines were regulated by ibudilast in the NAcore remains to be determined.

Conclusions: Here, we present data supporting Gq-DREADD-mediated exocytotic release of glutamate from astrocytes. Transient enhancement of glial glutamate release inhibited meth, but not sucrose seeking. Further, glial modulation of cytokine release and phosphodiesterase inhibition blocked reinstated meth seeking. Combined, these studies indicate that glial cells could serve as potential pharmaceutical targets for the treatment of relapse to meth addiction. Taken together, our data indicate that aberrant levels of homeostatic glutamate and the inflammatory response to the neurotoxic effects of meth contribute to the enduring propensity for relapse. Further study of how meth exposure alters glutamate homeostasis and the neuroinflammatory environment will provide insight into potential therapeutic strategies designed to prevent relapse.

Keywords: Astrocyte, DREADD, Meth, NAcore.

Disclosure: Nothing to Disclose.

T45. Adolescent Corticosteroid Exposure and TrkB Activity Regulate Action Selection and Depression-like Behavior in Adulthood

Shannon Gourley*

Emory University School of Medicine, Atlanta, Georgia

Background: Prolonged exposure to stressors or the primary stress hormone corticosterone (CORT; cortisol in humans) impairs the ability of rodents to select actions based on their consequences, resulting in maladaptive habits. It is widely appreciated that adolescence may be a period of vulnerability to stressors, thus we hypothesized that subchronic exposure to CORT during adolescence may bias decision-making towards maladaptive habits, and induce depression-like behavior, even in adulthood.

Methods: We exposed mice to subchronic CORT during early adolescence and adulthood and assessed decision-making strategies in adulthood using instrumental conditioning degradation and outcome devaluation. Sucrose consumption and progressive ratio assays were used to assess depression-like behavior. We then attempted to recapitulate the effects of prolonged CORT exposure using targeted amygdalo-hippocampal silencing of the trkB receptor. Finally, we aimed to reverse behavioral deficits by administering the trkB agonist, 7,8-dihydroxyflavone, during adolescence and testing behavioral outcomes in adulthood.

Results: Mice with a history of adolescent CORT exposure rapidly developed stimulus-response habits, while subchronic exposure in adults had no effects. Viral-mediated expression of an inactive form of the high-affinity BDNF receptor trkB in the ventral hippocampus and central nucleus of the amygdala (but not basolateral nucleus) recapitulated the effects of adolescent CORT exposure, while 7,8-dihydroxyflavone corrected behavioral deficiencies. Importantly, adolescent 7,8-dihydroxyflavone treatment also had antidepressant-like effects that persisted into adulthood, even despite the cessation of treatment in late adolescence.

Conclusions: We report that subchronic CORT during early adolescence results in a bias towards habit-based decision-making in adulthood. Over-expression of inactive trkB within amygdalo-hippocampal networks recapitulates these effects. Behavioral deficiencies, including depression-like amotivation, can be rescued by 7,8-dihydroxyflavone, with protective effects persisting well beyond the period of active drug treatment. We discuss a model wherein trkB binding in the ventral hippocampus could regulate BDNF secretion and trkB binding in the amygdala and thereby coordinate the regulation of motor output. Clarification of these possibilities could critically elucidate mechanisms of goal-directed decision-making and adolescent-emergent stressor-related depression.

Keywords: adolescence, BDNF, action, depression.

Disclosure: Nothing to Disclose.

T46. Telomere Dysregulation in the Hippocampus of a Rat Genetic Model of Depression. Normalization by Lithium Treatment

Aleksander Mathe*, Yabin Wei, Lena Backlund, Lina Martinsson, Gregers Wegener, Catharina Lavebratt

Karolinska Institutet, Stockholm, Sweden

Background: Telomeres are repeated DNA sequences at the end of each chromosome maintained by the enzyme telomerase protecting it from shortening and fusing with other chromosomes. Shortening of leukocyte telomeres (LT) is associated with aging and the process is accelerated by a number of chronic diseases and in major depressive disorder (MDD). Several drugs used in treatment of brain disorders modify telomere length (TL) and recently lithium was found to protect against LT shortening in bipolar disorder patients (Martinsson L et al 2013). For obvious reasons the relationship between telomeres in blood and brain is not known; the question if telomeres are changed in brain of depressed patients or if lithium affects telomeres in brain remains unanswered. In view of these facts, we employed a reverse translational approach by using a rodent genetic model of depression, the Flinders Sensitive Line (FSL) and treated the rats with lithium.

Methods: (1). Animals and lithium treatment. Male FSL and FRL rats were kept under controlled conditions of temperature, humidity and daylight cycle (12:12h). Chow and tap water were available ad libitum. Rats were assigned to a 6 week treatment with 2.19g Li2SO4/kg or vehicle admixed to chow. Following euthanasia, brains were harvested, hippocampi dissected and stored at -80°C. (2).Telomerase activity was detected by real-time telomeric repeat amplification protocol (Hou M et al 2001). Hippocampi were lysed in CHAPS buffer and the total protein measured by the Pierce BCA Protein Assay Kit. A HeLa cell line was a telomerase-positive and CHAPS buffer and heat-inactivated samples were negative controls. The reaction was performed on ABI PRISM 7900 HT Sequence Detection System. (3).DNA/RNA extraction and reverse transcription. Genomic DNA and total RNA were extracted with an AllPrep DNA/RNA/miRNA Universal Kit and concentrations determined using the NanoDrop ND-1000 Complementary DNA was synthesized using the SuperScript III First-Strand Synthesis System for RT-PCR. (4).Gene expression. Amplification of target and reference genes was assessed using qRT-PCR; amplifications were performed in triplicates with Power SYBR Green on an ABI PRISM 7900 HT Sequence Detection System. Two reference genes were Gapdh and cyclophilin A. (5). Relative TL of the rat hippocampal DNA was determined according to Cawthon RM protocol (2002). Triplicate DNA samples were used both for the telomeres and the single-copy gene (ribosomal protein L30, Rpl30) qRT-PCR amplified by Power SYBR Green on an ABI PRISM 7900 HT Sequence Detection System.

Results: (1). FSL rats had shorter TL compared to the FRL (P=0.038). Since the expression levels of full-length Tert, in most cases, correlate with telomerase activity, we determined the expression of the full-length Tert. Tert levels were reduced in the FSL compared to the FRL rats (P=0.023). Consistent with the downregulation of Tert expression, telomerase activity was lower in the depressed FSL (P=0.041). (2). Lithium treatment compared to the vehicle treated group increased Tert expression in the hippocampi of the FSL (P=0.012). In line with the Tert upregulation, telomerase activity was also increased in the FSL-Li group (P=0.015). Additionally, we explored if lithium influenced the TL in the FSL rats; the point estimate of the TLwas increased but it did not reach the statistical level of significance.

Conclusions: Two novel findings on telomere regulation are: i. hippocampal telomeres were shorter in FSL vs the controls. In accordance with this finding, the Tert expression and the telomerase activity were reduced; ii. the aberrant Tert expression and telomerase activity were reversed by lithium. To clarify if telomeres are also shorter in the brain of depressed subjects we used a reversed translational strategy and examined brains of a rat model of depression. Since hippocampus plays pivotal roles in CNS we assessed telomere length in that region and found that TL was shorter. As telomerase is the main determinant of TL we studied that enzyme and found that its activity was also decreased. Limitations: we did not perform behavioral tests and do not have evidence that enhanced telomerase activity is associated with an antidepressant effect. Project strenght: we show for the first time i. telomeres are shorter, ii.Tert expression is reduced, iii.telomerase activity is decreased in a rat depression model. In summary, this is the first report showing telomere dysregulation in brains of a well defined rat depression model as well as restorative effects of lithium.

Keywords: lithium treatment, depression, telomeres, animal model.

Disclosure: Nothing to Disclose.

T47. Metabotropic NMDA Receptor-dependent LTD is Independent of GluN2 Subunit Composition

John Gray*

University of California, Davis, Davis, California

Background: NMDA receptors (NMDARs) are ionotropic glutamate receptors that play crucial roles in synaptic plasticity and neuronal development. Disturbances in NMDAR function have been implicated in a broad range of neuropsychiatric disorders, including schizophrenia. Provocative emerging evidence supports a role for NNMDAR-mediated signaling in the absence of ion flux through the receptor. This non-ionotropic (or metabotropic) signaling is thought to be due to agonist-induced conformational changes in the receptor, independent of channel opening. A recent study demonstrated an essential metabotropic NMDAR function in long-term depression (LTD), which relies on glutamate binding but is independent of the co-agonist D-serine and ion influx through the receptor. However, this finding has been directly refuted to two other laboratories. Here, I sought to independently examine the validity of this metabotropic NMDAR-dependent LTD. Further, given that the functional and regulatory properties of NMDARs are largely determined by their GluN2 subunit composition, the subunit-dependence of potential metabotropic NMDAR signaling was examined.

Methods: The hippocampal CA3-to-CA1 synapse is a model excitatory synapse for delineating the mechanisms of synaptic plasticity. Using transgenic mice with conditional KO alleles for GluN2A (GRIN2A) and GluN2B (GRIN2B), I eliminated the target NMDAR subunit in a small subset of hippocampal neurons by transcranial stereotactic injection of neonatal (P0) mice with a recombinant adeno-associated virus expressing a Cre-GFP fusion protein (rAAV1-Cre-GFP). Acute hippocampal slices were then made from these mice from P17-P24 and simultaneous whole-cell patch clamp recordings were obtained from transfected and neighboring control cells. LTD was induced by 1Hz stimulation of Schaffer collaterals in the presence of the competitive NMDAR glycine-site antagonists 7-chlorokynuerenic acid (7CK) or L-689,560 (L689) and/or the glutamate-site antagonist AP5.

Results: Cre expression in the conditional KO mice provided a pure population of synaptic GluN2A- or GluN2B-containing NMDARs. Using standard LTD induction protocols, I confirmed the finding that LTD was still present in the presence of NMDAR glycine site antagonists but was blocked by the glutamate site antagonist AP5. These results provide independent confirmation of the metabotropic NMDAR-dependent LTP. Because glutamate binding to the GluN2 subunit is required for this metabotropic effect, presumably through conformational changes in the GluN2 subunit, it was hypothesized that there might GluN2 subunit-specific requirements. Surprisingly, I found that the metabotropic NMDAR-dependent LTD was still present when the NMDAR composition was purely GluN2A or GluN2B, suggesting a lack of subunit specificity for this effect.

Conclusions: Synaptic NMDARs are part of a large macromolecular complex of signaling and scaffolding molecules. Emerging evidence points to a role for channel-independent, agonist-based NMDAR signaling, likely through conformational and compositional changes in this macromolecular complex. This finding opens the door for a reexamination of NMDAR-mediated processes and may yield new insights into the pathophysiology of schizophrenia and other neuropsychiatric disorders. Indeed, numerous lines of evidence have implicated a dysregulation or hypofunction of NMDARs in schizophrenia. One possible cause of this hypofunction may be a reduction in the levels of D-serine, a requisite co-agonist of NMDARs. A reduction in synaptic D-serine levels might lead to the under-activation of the NMDAR glycine site and bias synaptic plasticity towards LTD and subsequent dendritic spine loss. Future studies will examine the molecular mechanisms involved in metabotropic NMDAR signaling. This emerging new NMDAR biology has potential to facilitate identification of novel and more specific targets for the treatment of schizophrenia.

Keywords: synaptic plasticity, D-serine, glycine, NMDA.

Disclosure: Nothing to Disclose.

T48. Abnormal Subcellular Localization of GABA(A) Receptor Subunits in Schizophrenia

Toni Mueller*, Colton Remedies, Vahram Haroutunian, James Meador-Woodruff

University of Alabama at Birmingham, Birmingham, Alabama

Background: Schizophrenia is a chronic neuropsychiatric illness which presents with a diverse constellation of symptoms and has variable deficits across multiple brain regions, neurotransmitter systems, and cell types. One of the most consistent findings in schizophrenia research is decreased expression of GAD67, an enzyme necessary for the production of the neurotransmitter, GABA. Recently, we have found alterations in the post-translational processing of some GABA(A) receptor (GABAAR) subunits; specifically, we have reported abnormalities of N-glycosylation of the α1, β1, and β2 GABAAR subunits in schizophrenia. In conjunction with other studies demonstrating changes in receptor stoichiometry at the synapse, these data suggest post-synaptic GABA deficits may also occur in schizophrenia. N-glycosylation is known to play an important role in protein folding and assembly, endoplasmic reticulum (ER) quality control mechanisms, and forward trafficking of proteins from the ER to the plasma membrane. To ascertain if these processes are disrupted in schizophrenia, our current study was designed to examine the subcellular localization of GABAAR subunits which we have previously found to be abnormally N-glycosylated. We hypothesized that abnormal N-glycosylation may be associated with altered subcellular localization of GABAAR subunits, reflecting disturbances of intracellular receptor trafficking and/or targeting in schizophrenia.

Methods: Samples of gray matter from the full cortical thickness of the left superior temporal gyrus (Brodmann area 22) of 16 schizophrenia and 14 comparison subjects were obtained from the Mount Sinai Medical Center brain collection. Subcellular fractionation was performed using nitrogen cavitation, differential sucrose gradient centrifugation, and Triton solubilization. This protocol yields light membrane/cytosol (L/C), ER, and synapse (SYN) enriched fractions. It also yields a fraction containing markers for mitochondria, extrasynaptic membranes, and other membrane and vesicle-associated proteins; this triton-soluble fraction, hereafter referred to as the “other membrane” (OTH) fraction, does not contain nuclear or excitatory synaptic markers. Prepared samples of total homogenate as well as the L/C, ER, SYN, and OTH fractions were loaded in 4-12% Bis-Tris gels and run using standard SDS-PAGE and semi-dry transfer methods; each blot had samples from 1 schizophrenia and 1 comparison subject and was probed with primary antibodies against VCP, as a loading control; gephyrin, as a synaptic and extrasynaptic marker; DNAJC4, as a cytosolic marker; JM4, as an ER marker; and each of the α1, α2, β1, β2, and γ2 GABAAR subunits, as well as the appropriate secondary antibodies. Western blots were analyzed using Image Studio software. The relative abundance of proteins was calculated for each fraction by measuring the signal intensity of each target subunit normalized to the signal intensity of VCP, signal intensity of VCP-normalized fraction marker (DNAJC4, JM4, and gephyrin, respectively), and VCP-normalized signal intensity of the target in total homogenate.

Results: β2 GABAAR subunits in the ER are increased in schizophrenia. The β2 GABAAR subunit is visualized as multiple isoforms in a fraction-specific manner, with bands at approximately 52 and 50kDa in total homogenate; 50kDa in the L/C fraction; 50 and 48kDa in the ER fraction; 52, 50, and 48kDa in the SYN fraction; and 52, 50, and 47kDa in the OTH fraction. The mean ER signal intensity of all β2 isoforms (β2-ALL) was 92% higher in schizophrenia (t(28)=2.3, p=0.03), and the mean ER signal intensity of the primary 50kDa β2 isoform (β2-50kDa), which is expressed in all subcellular fractions, was 117% higher in the ER in schizophrenia (t(28)=2.5, p=0.02). There was no difference in the relative expression of the 48kDa β2 isoform (β2-48kDa) in the ER and no difference in the relative expression of β2-ALL, β2-50kDa, β2-48kDa, or 52kDa β2 (β2-52kDa) GABAAR subunit isoforms in the SYN fraction between schizophrenia and comparison subjects. The ratio of β1 to β2 GABAAR subunit expression is decreased in the ER in schizophrenia. There was no difference in the relative amount of the β1 GABAAR subunit expressed in the ER or SYN fractions between schizophrenia and comparison subjects. However, the ratio of β1:β2-50kDa signal intensities was less in the ER in schizophrenia (t(27)=2.8, p=0.01) and trended toward less in the SYN fraction (U(14, 15)=62, p=0.06). The ratio of β1:β2-52kDa signal intensities in the SYN fraction also trended toward less in schizophrenia (t(27)=2.0, p=0.06), while the ratio of β1:β2-48kDa signal intensities did not differ between groups in either the ER or SYN fractions. The ratio of β1:β2-ALL GABAAR subunit signal intensities was significantly reduced in the ER in schizophrenia (t(27)=2.6, p=0.02) but was not different in the SYN fraction.

Conclusions: This study demonstrates that the subcellular localization of the β2 GABAAR subunit, which is known to have altered N-glycosylation in schizophrenia, is abnormal in schizophrenia. Previously, it has been shown that less N-glycosylated β subunits are preferentially incorporated into intact receptors exiting the ER. Thus, the increase in ER-localized β2 GABAAR subunit protein expression and decrease in the ratio of β1:β2-ALL and β1:β2-52kDa in the ER is consistent with our previous finding in schizophrenia that there is less immature N-glycosylation of the β1 GABAAR subunit and altered total N-glycosylation of the β2 GABAAR subunit. Together, this suggests that the less glycosylated β1 subunit may be preferentially incorporated into intact GABAARs trafficked to the synapse, while the β2 subunit is retained in the ER. While N-glycosylation abnormalities of several post-synaptic neurotransmitter receptor and transporter subunits have been established, these current findings suggest a mechanistic consequence of previously identified alterations which may more directly contribute to the pathophysiology of schizophrenia.

Keywords: N-glycosylation, trafficking, endoplasmic reticulum.

Disclosure: Nothing to Disclose.

T49. Enhancement of Stress Resilience Through Hdac6-mediated Regulation of GR Chaperone Dynamics

Olivier Berton*, Jeanine Jochems

University of Pennsylvania Perelman School Medicine, Philadelphia, Pennsylvania

Background: Acetylation of Hsp90 regulates downstream hormone signaling via the glucocorticoid receptor (GR), but the role of this molecular mechanism in stress homeostasis remains poorly understood. Here we tested whether acetylation of Hsp90 in the brain predicts and modulates the behavioral sequelae of a mouse model of social stress.

Methods: Mice subjected to chronic social defeat stress (CSDS) were stratified into resilient and vulnerable subpopulations based on social avoidance. HPA axis function was probed using a DEX/CRF test. Hsp90 acetylation, Hsp90-GR interactions and GR translocation were measured using CoIP and western blot. To manipulate Hsp90 acetylation, we pharmacologically inhibited Hdac6, a known deacetylase of Hsp90 using a novel selective inhibitor (ACY738) or overexpressed a point-mutant that mimics the hyperacetylated state of Hsp90 at lysine residue K294.

Results: Lower levels of acetylated Hsp90, higher GR-Hsp90 association and enhanced GR translocation were observed in DRN of vulnerable, but not resilient mice after CSDS. Administration of ACY-738, a brain penetrant Hdac6-selective inhibitor, led to Hsp90 hyperacetylation in brain and in neuronal culture. ACY-738 increased the association of Hsp90 with FKBP51, reduced association with FKBP52 and inhibited hormone-induced GR translocation in cell-based assays. These cellular effects were replicated by acetylation-mimic point-mutant Hsp90. in vivo, ACY-738 promoted resilience phenotype after CSDS. Selective viral overexpression of the acetylation-mimic mutant of Hsp90 selectively in raphe serotonin neurons was sufficient to reproduce the effect of ACY-738 in CSDS.

Conclusions: Hyperacetylation of Hsp90 is a predictor and molecular determinant of stress resilience in mice. Brain-penetrant Hdac6 inhibitors that increase Hsp90 acetylation and modulate GR chaperone dynamics offer a promising strategy to curtail deleterious socioaffective effects of stress and glucocorticoids.

Keywords: social defeat, HDAC, GR chaperone, FKBP5.

Disclosure: O. Berton is inventor on pending patent application #61713014 “Pyrimidine hydroxyamide compounds as protein deacetylase inhibitors and methods of use thereof.

T50. A Fluorescence-based Preclinical Marker for Antidepressant Efficacy

Mark Rasenick*, Jeff Schappi, Andrew Czsyz

University of Illinois College of Medicine, Chicago, Illinois

Background: Numerous studies suggest that chronic, but not acute, antidepressant treatment increases physical coupling between Galphas and adenylyl cyclase (AC) resulting in increased cAMP signaling. Biochemical studies have demonstrated that this is due, at least in part, to the translocation of Galpha s from lipid rafts into non-raft membrane fractions where it is more effectively coupled to AC. Although this effect is highly reproducible, 4°C lipid raft extractions require millions of cells and are criticized because they may not reflect the same physiology as living cells (whether or not detergents are used to extract lipid raft fractions). We have developed a screen for antidepressant action based on Fluorescence Recovery After Photobleaching (FRAP) that is reliable, accurate and high throughput.

Methods: A monomeric variant of GFP-Galpha s was constructed and transfected into C6 glioma cells. Stable GFP-Gslpha s C6 closed were selected by Fluorescence Activated Cell Sorting and treated with 11 different antidepressant compounds (50 nM to 10 μM) for 1 hour to three days. Cells were photobleached on a patch of membrane with constant size, and FRAP was measured in a Zeiss 710 Meta confocal microscope. GFP-Gsalpha variants with modified acylation sites and well as GFP-Galpha i1 were also constructed. Stable C6 cell lines were made with these and clones selected as above.

Results: Every antidepressant tested retarded GFP-Galpha s FRAP after chronic, but not acute treatment. Several compounds lacking direct antidepressant activity (r-citalopram, amphetamine, ziprasidone, diazepam) did not show this effect. Removal of membrane anchoring domains from GFP-Galpha s increased the speed of fluorescence recovery. Adenylyl cyclase and other proteins with multiple transmembrane domains showed extremely slow fluorescence recovery. Disruption of lipid rafts or the membrane cytoskeleton increased FRAP for GFP-Gslpha s or GFP-Galpha i1, but antidepressant treatment had no effect on FRAP of GFP-Galpha i1 or GFP-Gslp[ha S acylation mutants.

Conclusions: Initially, results obtained were contrary to the hypothesis that removal of proteins (including G proteins) from lipid rafts would speed their recovery from photobleaching. It now appears that increased association of GFP-Galpha s with adenylyl cyclase, following from chronic antidepressant treatment, results in slower fluorescence recovery. Fluorescence recovery of GFP-Gsalpha s proteins with altered acylation and GFP-Galpha i1 are not changed by antidepressant treatment, since they do not show increased association with adenylyl cyclase. The sensitivity and specificity of GFP-Gslpha s photobleaching and recovery make it an ideal platform for identification of compounds with potential antidepressant properties using a high content screening system. Furthermore, this mechanism might be adopted to predict response of individual patients employing a screen using the patient's own blood cells and an array of antidepressant drugs.

Keywords: depression, G protein, biomarkers, GPCR.

Disclosure: Research Support, Eli Lilly; Equity and Board Member, Pax Neuroscience.

T51. A Pilot Study of Soluble Epoxide Hydrolase Activity in Eating Disorders

Pei-an Shih*, Christophe Morisseau, Jun Yang, Bora Inceoglu, Ursula Bailer, Ashley Van Zeeland, Andrew W. Bergen, Pierre Magistretti, Wade Berrettini, Katherine Ann Halmi, Nicholas Schork, Bruce D. Hammock, Walter Kaye

University of California at San Diego, La Jolla, California

Background: Soluble epoxide hydrolase (sEH) converts epoxides from polyunsaturated fatty acids (PUFA) to the corresponding diols. We previously showed that cytochrome P450 epoxide metabolites of PUFAs are associated with anorexia nervosa (AN), and observed higher diolepoxide ratios in AN patients suggesting that sEH activity should be elevated in AN patients compared to controls. Separately, our earlier exon sequencing study which covered exon 14 to 3’-UTR region of the gene coding for sEH, EPHX2, showed both rare and common variants in this gene region contribute to AN susceptibility. Of the common variants associated with AN, the high-expressor allele, by in silico sequence read analysis, of a 3’-UTR variant (rs1042064) was found to be less frequent in AN compared to controls (p=2.38E-02), suggesting that sEH activity could be lower in AN patients. To resolve this discrepancy, we directly measured ex vivo sEH activity in buffy coats using t-DPPO, a well characterized surrogate substrate for sEH, in 15 recovered AN, 3 recovered bulimic nervosa patients (BN), and 5 healthy control women to assess the association between sEH activity and eating disorders (ED).

Methods: Buffy coat used for sEH assay was extracted from the whole blood collected from 15 recovered AN, 3 recovered BN, and 5 age- and sex-matched healthy control women. Recovery was defined as the offset of ED symptoms if the subjects experienced at least one year without symptoms of eating disorder and inappropriate compensatory behaviors. Student’s T, ANOVA, and correlation coefficient tests were applied to assess association between predictors (diagnosis and AN characteristics) and sEH activity.

Results: Interestingly, the activity of sEH was not associated with BMI, lifetime lowest BMI, ED types, or temperament traits of perfectionism, novelty seeking, harm avoidance, depression, trait anxiety in either ED subjects or controls. However, sEH level was higher in ED subjects compared to controls (0.012 versus 0.007 nmol.min-1.mg-1, p=0.05). sEH activity was significantly and inversely correlated with age (r=-0.527, p=0.016). Surprisingly, sEH level was significantly associated with state anxiety only in controls (r=-0.98, p=0.0013) but not in ED subjects (r=0.04, p=0.87). This association is consistent with our previous finding using the oxylipin ratio as plasma biomarkers of sEH activity.

Conclusions: This pilot study provides evidence that sEH activity is a factor in ED risk and is higher in ED patients. We obtain biochemical evidence that suggests that the EPHX2 3’-UTR variant rs1042064 may not play a key role in modulating sEH activity in ED subjects. We observed that sEH has an influence on state anxiety in healthy individuals. In contrast, patients with ED express higher activity of sEH and have lower levels of plasma epoxy fatty acids which may eliminate sEH’s influence on anxiety perception in ED. Strengths of this study include a specific hypothesis, and detailed patient and metabolic assessments; limitations include a modest sample of ED patients and a small sample of controls. Although it is yet unclear what the consequences of elevated sEH on brain and ED prognosis are, the serious nature of ED and these findings justify the need to further investigate the role of sEH in human psychopathophysiology.

Keywords: Soluble Epoxide Hydrolase, Eating Disorders, Metabolomics, Fatty Acids.

Disclosure: Nothing to Disclose.

T52. Nuclear Factor κB Activity is Increased by Alcohol Place Conditioning

Britessia Smith, Camilla Karlsson, Faaz Rehman, Abbey Borich, Jenica Tapocik, Markus Heilig, Jesse Schank*

University of Georgia, Athens, Georgia

Background: Nuclear factor κB (NFkB) is a transcription factor that is commonly associated with the innate immune system, and it is activated in response to infection and inflammation. This transcription factor has many downstream targets, the most studied of which are inflammatory mediators such as cytokines. More recently, NFkB has gained attention as an important mediator of more complex psychiatric phenomena like stress and addiction. For example, NFkB has been shown to mediate behavioral and neuroanatomical phenotypes resulting from social defeat stress, chronic unpredictable stress, and repeated cocaine administration. It has also been demonstrated that NFkB activity is stimulated by alcohol exposure. Much of the research in regards to the role of NFkB in alcohol responses has focused on intoxication-induced neurotoxicity, and has generally used large doses of alcohol. Few studies thus far have explored the role of NFkB in alcohol reward, reinforcement, or voluntary consumption. Notably, however, it has been reported that the NFkB stimulator lipopolysaccharide can induce a long-lasting increase in voluntary alcohol consumption.

Methods: In these studies, we use the standard rodent model of conditioned place preference to assess the activity of NFkB in response to doses of alcohol that have rewarding properties. Given that NFkB is stimulated in response to alcohol in general, and that this transcription factor has been shown to mediate the rewarding properties of other drugs such as cocaine and morphine, we hypothesized that NFkB activity would be stimulated in response to alcohol place conditioning. To measure this we used a line of transgenic mice that express the LacZ gene under the direction of a promoter that has NFkB response elements. In these animals, post-hoc staining for β-gal identifies cells in which NFkB has been activated.

Results: After place conditioning for alcohol with a dose of 2 g/kg, which induces a strong place preference, we measured β-gal staining in a panel of brain regions including the nucleus accumbens (NAC), amygdala, ventral tegmental area, bed nucleus of the stria terminalis, lateral septum, and dorsal raphe nucleus (DR). We observed significantly increased β-gal staining specifically in the NAC shell and DR in the alcohol-treated group.

Conclusions: These results suggest that, in addition to its role in alcohol-induced neurotoxicity, NFkB may be involved in the development of alcohol reward. Future studies will use double labeling strategies to determine the type of cells that show this increased NFkB activity and will assess the effect NFkB inhibition on alcohol place preference.

Keywords: Alcohol, Transcription Factor, Reward, NF kappa B.

Disclosure: Nothing to Disclose.

T53. Increased CSF Matrix Metalloproteinase-9 (MMP-9) and Reduced White Matter Integrity with Increasing Age in Late-life Major Depression

Nunzio Pomara*, Chelsea Reichert, Sang Han Lee, Jay Nierenberg, Matthew R. Halliday, Abhay P. Sagare, Blas Frangione, Berislav V. Zlokovic

Nathan S. Kline Institute/New York University School of Medicine, Orangeburg, New York

Background: Inflammatory factors, oxidative stress and neurovascular unit dysfunction with blood-brain barrier (BBB) hyperpermeabilty have been implicated in the pathogenesis of major depressive disorder. Several lines of evidence also indicate that increased expression of various proteolytic enzymes but especially matrix metalloproteinase (MMP-9), mediate the basal lamina degradation and BBB damage. However, in spite of these observations, we found no studies that have examined CSF MMPs in depression. These considerations prompted us to determine if elderly cognitively-intact individuals with a DSM-IV diagnosis of late-life major depression (LLMD) differ from healthy controls with respect to CSF MMP-9 levels. Additionally, since Diffusion Tensor Imaging (DTI)-derived metrics have been found to be sensitive to subtle brain changes associated with inflammatory markers and CSF MMP-9 elevation in non-psychiatric disorders, we employed DTI to examine the relationship between CSF MMP-9 and white matter integrity.

Methods: CSF was obtained from 47 older subjects with intact cognition having an MMSE score of at least 28 and without significant white matter hyperintensities (WMHI, 28 LLMDs, 19 Healthy controls). MMP-9 was determined using enzyme-linked immunoassays. Magnetic Resonance Imaging (MRI) scans were performed to assess fractional anisotropy (FA), a measure of white matter integrity. Each of the DTI measures was averaged over the entire white matter compartment in images transformed into MNI space. DTI interpretations of FA included mean trace diffusivity (TR), mean radial diffusivity (LR), and mean axial diffusivity (L1). Pearson’s correlations were calculated to examine associations between MMP-9 and these global DTI measures.

Results: There were no significant group differences in CSF MMP-9 levels. MMP-9 was associated with decreased FA (r=−.35, p=.04) and with increased TR (r=+.44, p=.01), L1 (r=+.42, p=.01), and LR (r=+.42, p=.01) which are indicative of greater white matter pathology in the entire group irrespective of diagnosis. However, separate analyses in individuals over age 65 revealed that only the subjects with LLMD (n=14) showed a significant negative correlation between CSF MMP-9 and FA (r=-.85, p=.001).

Conclusions: Higher CSF MMP-9 levels in healthy elderly, especially in those with late-life depression and older than 65 years of age are associated with subtle loss of white matter integrity in the absence of increased of macrovascular lesions such as WMHI, suggesting they are associated with a nonvascular mechanism in LLMD. Future studies should determine the basis for these MMP-9-associated effects and perhaps whether they are related to BBB breakdown and increased age-related oxidation and inflammation in depression.

Keywords: Matrix Metalloproteinase-9 (MMP-9), Cerebrospinal Fluid, Diffusion Tensor Imaging, Late-life Depression.

Disclosure: Nothing to Disclose.

T54. Age-related Changes in Cell Adhesion Molecule, Progenitor Cells, And Vascularity in Human Hippocampus in Major Depression

Maura Boldrini*, Adrienne Santiago, Tanya Butt, Andrew Dwork, Gorazd Rosoklija, Victoria Arango, Rene' Hen, J. John Mann

Columbia University/New York State Psychiatric Institute, New York, New York

Background: Neurogenesis decreases with age in mice faster than in human brain even after adjustment for lifespan. Untreated major depression (MDD) is associated with fewer mature granule cells and that effect may be due to impairment in neurogenesis or cell survival. We examined postmortem the relationship of age to nestin-immunoreactive (IR) neural progenitor cells (NPCs) and capillaries, polysialylated neural cell adhesion molecule (PSA-NCAM)-IR and Ki-67-IR cells, and nuclear antigen (NeuN)-IR mature granule neurons (GNs) in the dentate gyrus (DG) of subjects without psychopathology or treatment (controls) compared with subjects with untreated MDD.

Methods: Frozen hippocampal sections at a 2-mm interval throughout the rostro-caudal extent of the DG from 16 controls, age 17-79 yrs., and 16 subjects with DSM-IV MDD, age 17-83 yrs., were fixed and processed for immunohistochemistry, immunohistofluorescence and stereology. Clinical Measures included a psychological autopsy and the SCID I or NP (Non-Patient edition) and II, the Global Assessment Scale (GAS) and the St. Paul-Ramsey Life Experience Scale (SPRS). Developmental history and recent medication history were obtained. Cause of death was recorded. Neuropathological data were obtained by brain examination. Fluorescent double/triple-labeled cells and capillaries were analyzed with 60x oil-immersion objective on a confocal microscope. Co-labeling was confirmed by acquiring z-stacks at 1-m intervals and creating orthogonal views. Linear regression analysis and T-tests for independent samples were performed using SPSS, p<0.05 was defined as the significance level. Data were expressed as mean+/- SEM.

Results: Triple-labeling immunohistofluorescence verified the vascular localization of nestin showing nestin immunostaining on capillaries that were immunoreactive for the marker of endothelial cells CD-31 (or platelet/endothelial cell adhesion molecule, PECAM) and for collagen-IV, a marker of vascular basement membranes. Confocal images of cells stained with PSA-NCAM, doublecortin, and DAPI showed the co-localization of PSA-NCAM and doublecortin in immature neuroblasts in human brain. Mitotic cell number did not decrease with age in DG in controls or MDD. Nestin-IR NPC density (cells/mm3) decreased with older age (r=-888, p=.001) in control subjects, but not in MDD, selectively in anterior DG. NPC density declined (r=-.740, p=.014) with increasing age in the posterior DG in MDD, but not in anterior or mid DG. Nestin-IR capillary length (r=-.602, p=.038), number of capillaries per DG cubic mm (r=-.644, p=.007), and number of bifurcations per capillary (r=-.613, p=.034) decreased with aging in the whole DG and capillary area decreased with aging in posterior DG (r=-.992, p=.008) in controls. In MDD we did not find correlations between capillary measures and age. PSA-NCAM-immunoreactive cells with multipolar (r=-.797, p=.010), unipolar (r=-.817, p=.007), or round morphology (r=-.717, p=.030) decreased with aging in anterior DG in controls, but not in MDD. NeuN-immunoreactive GN number or density (cells/mm3) in the DG did not change with aging in controls or MDD.

Conclusions: Fewer PSA-NCAM-IR cells and NPCs in anterior DG and fewer capillaries in the whole DG with aging in controls may contribute to age-related memory and pattern recognition decline. Cellular and vascular changes associated with older age in anterior DG were not present in MDD, perhaps due to the cell loss already associated with MDD neuropathology.

Keywords: dentate gyrus, neurodegenerative, aging, neuroplasticity.

Disclosure: Nothing to Disclose.

T55. Transcriptomics of Nerve Injury: Axotomy-induced Changes in Sensory-motor Circuits at the Spinal Level Analyzed Using RNA-Seq

Michael Iadarola*, Samridhi Goswami, James Klimavicz, Jacklyn Gross, Andrew Mannes

National Institutes of Health, Bethesda, Maryland

Background: Many neurological and, potentially, psychiatric disorders involve damage to a neuronal population or neuroanatomical component. We address some of the underlying processes through an examination of molecular responses to axonal damage within multiple neuronal compartments using RNA-Seq to obtain comprehensive, quantitative measures of gene transcription. Axotomy of peripheral nerve captures many important PNS and CNS response features when analyzing ventral horn (VH), containing axotomized motor neurons, dorsal root ganglion (DRG) containing axotomized sensory neurons, and dorsal horn (DH) containing no axotomized neurons. Some of the important differential elements are the presence or absence of synapses, types of synaptic inputs, and glial populations. The VH contains alpha and gamma motor neurons and the axons of these neurons are cut completely by the peripheral axotomy. The VH is also notable in that the axotomized neurons receive synapses. DRG neurons are pseudounipolar, do not receive synapses, and only the peripheral portion of the axon is cut. Neurons of the DH receive aberrant input from the primary afferent neurons but none of the DH neurons are axotomized. This variety of distinguishing features allows specific and broad comparisons between anatomical compartments and molecular targets. We also sampled over multiple time periods to determine acute and sustained transcriptional modulation. A range of alterations was observed designed to meet multiple cellular compensatory, regenerative, and tissue remodeling processes.

Methods: RNA-Seq provides precise quantification of transcriptomes at maximal resolution and dynamic range. Rat L4-L5 DRG, and corresponding lumbar DH and VH were obtained at 0, 1, 3, 10, 30 and 90 days after axotomy. Libraries were constructed from polyA+mRNA and sequenced on an Illumina HiSeq2000 to a depth of 40 million reads/sample at 100 bases/read. The RNA-Seq data were mapped with RUM and statistical analyses performed using DESeq. We examined gene changes from several viewpoints: alterations within individual or overlapping anatomical compartments, clustering based on temporal profile (e.g., peak day 1, peak day 3, etc.), initial pattern of change between 0, 1 and 3 days to focus on modes of transcriptional control, and whether the change is increased, decreased and sustained at day 90. Because the data are quantitative we filtered out genes that were not substantially altered in a quantitative sense. Our general tendency was to be more inclusive at the initial stages but to increase statistical thresholds as analyses progressed.

Results: Each tissue contained approximately 13,000 transcripts. Expression varied from 1 to>3,000 reads/kilobase of transcript/million bases sequenced (RPKM); this unit normalizes for transcript length and depth of sequencing. Axotomy of rat sciatic nerve significantly modulated 1700 genes with an adjusted fold difference>2 and above. Significantly regulated genes could be subdivided into two categories: induced (basal expression below 1 RPKM) and upregulated (basal expression>1 RPKM). 60% of the significantly regulated genes occurred only in the DRG. The remainder were split into VH or DH or combinations of the three. Clustering within the anatomical compartments disclosed 6 temporal patterns: Peak at Day1, Day 3, Day 10, Bimodal, Transient Decrease, and Delayed Increase. At day 1 multiple regulated genes coded for secreted peptidase inhibitors and modulators of transcription. VH RNA-seq disclosed 200 genes regulated at 1.5-fold or more. VH contained a strong enrichment for mitotic and cell cycle transcripts at Day 3. The RNA-Seq DH dataset contained 400 genes regulated>1.5-fold. Many DH Day 3 genes also were involved in mitosis. It is unclear however, whether the mitotic changes are due to the same processes in the two tissues.

Conclusions: Diverse sets of regulated genes were identified, some tissue specific, some shared, however, the same gene may be expressed at a different level and regulated at different times and degrees in the three tissues. We associate the early activation of secreted peptidase and transcription inhibitor genes in DRG as consistent with cellular stabilization and inhibition of extracellular matrix modification. Sustained changes in structural genes at 90 days may indicate permanent modifications. Ventral horn mitotic signatures are consistent with biological indicators of microglial proliferation. Microglia have been shown to displace synaptic terminals from motoneurons, suggesting that a process of "neuronal isolation" is instituted during the repair period. These examples show that understanding the molecular changes after axotomy at a fine level of molecular resolution is key to understanding the pathophysiological consequences of nerve injury and the ensuing regeneration and repair mechanisms. These data may also illuminate mechanisms operating in other CNS injury and degenerative disorders.

Keywords: Transcriptome, RNA-Seq, Axotomy, Regeneration.

Disclosure: Nothing to Disclose.

T56. Contributions of Inflammatory Cytokine Signaling to the Enduring Effects of Early-Life stress: A Serotonin Connection?

Nicole Baganz*, Jarrod Smith, Lise Harbom, Matthew Robson, William Hewlett, Randy Blakely

Vanderbilt University Medical Center, Nashville, Tennessee

Background: Early-life stress (ELS) during childhood can have an enduring, negative impact on the quality of life and social interactions that can contribute to lifelong risk for cognitive disability and mood disorders, such as major depression. Foremost among the pharmacological treatments for adult mood disorders are agents that block the serotonin (5-HT) transporter and thus potentiate 5-HT signaling. We hypothesize therefore that mechanisms that impact 5-HT signaling in general, and SERT specifically, may be altered by ELS, leading to the therapeutic benefit of such medications. There is a growing body of evidence linking ELS to alterations in adult immune function. Our lab and others have shown that signaling pathways relevant to immune system signaling regulate SERT, expression, trafficking or function. In particular, our lab has shown that presynaptic interleukin-1β (IL-1β) signaling via p38 MAPK-linked pathways (Zhu et al, 2006), as well as peripheral immune system activation induced by the bacterial endotoxin, lipopolysaccharide (LPS) (Zhu et al, 2010), leads to a rapid stimulation of SERT activity. Moreover, the depressive-like behavior of mice induced by LPS is lost in interleukin-1 receptor (IL-1R) knockout (KO) mice or when animals are systemically treated with a p38 MAPK inhibitor. Here, we report the results of ongoing studies designed to test the hypothesis that ELS is associated with long-lasting immune system activation and that adult depression-like behavior may depend upon IL-1R/p38 MAPK signaling in 5-HT neurons.

Methods: C57Bl/6J and IL-1R knockout (IL-1R KO) mice were purchased from Jackson Laboratories, with the latter animals bred as heterozygotes to generate IL-1R+/+and -/- mice. In our MS model, we imposed a daily 3 hr separation of C57Bl/6 wildtype (WT) and IL-1R KO pups littermate controls from their mother from P1-P14. Male mice were weaned normally at P21 and, at 8-10 weeks old, were tested in behavioral assays, namely the tail suspension and forced swim tests, as well as open field and elevated plus maze. Brains and spleens were subsequently harvested for evaluation of cytokine levels via qPCR, and plasma corticosterone (CORT) levels were measured using ELISA assays. To ascertain whether IL-1R mediated signaling specifically within 5-HT neurons is responsible for MS phenotypes, we have generated IL-1Rflox/flox mice, congenic on a C57BL/6J background. IL-1Rflox/flox animals were crossed to B6.C-Tg (CMV-Cre) animals to evaluate efficiency of excision of the floxed IL-1R allele.

Results: As others have demonstrated, we detected long-lasting consequences on behavior, including heightened depressive- and anxiety-like behavior of adult animals subjected to MS. Additionally, we found that MS led to adult elevations in mRNA levels of multiple inflammatory signaling molecules (e.g. IL-1β and IL-6) in both the brain and periphery. Remarkably, constitutive IL-1R KO mice were found to lack the anxiety and depressive symptoms of their WT, MS treated littermates. Moreover, the MS-induced elevations in inflammatory cytokine levels observed in WT animals were also absent in IL-1R KOs. Interestingly, plasma CORT levels from MS mice were unaffected by genotype. Introduction of loxP sites in the Il1r1 gene resulted in no impact on endogenous IL-1R mRNA expression and supported efficient excision by CMV-Cre.

Conclusions: As IL-1β signaling can modulate SERT activity, effects that are lost in the IL-1R KO, and SSRI treatments can reverse MS-induced changes in behavior (Yoo et al, 2013), we hypothesize that one determinant of MS-induced behavioral changes may involve IL-1β alterations in SERT expression and/or function. In prior studies (N. Baganz et al, manuscript in prep), we have shown that raphe-specific elimination of the IL-1β signaling enzyme p38 MAPKα prevents the induction of anxiety and depressive behavior following peripheral LPS treatments. Here move our efforts to an ethologically valid model of ELS, and obtain evidence that MS induces lifelong elevations, centrally and peripherally, in pro-inflammatory cytokines and that enduring behavioral changes require expression of the IL-1R. To our knowledge, these studies are the first to demonstrate MS induced elevations in pro-inflammatory cytokines as well as a requirement for their signaling potential in ELS induction of depressive behavior. Moreover, we developed a new transgenic model that allows temporal/regional control of IL-1R deletion to determine the role of IL-1R signaling in 5-HT neurons in response to MS.

Keywords: cytokine, serotonin, depression, stress.

Disclosure: Nothing to Disclose.

T57. The Induction and Expression of Conditioning by Amphetamine Are Differentially Regulated by Nucleus Accumbens Cyclin-dependent Kinase 5

Paul Vezina*, Bryan F Singer

University of Chicago, Chicago, Illinois

Background: The induction of contextual drug conditioning involves the repeated intermittent administration of a drug in association with a complex of environmental stimuli. To test for the expression of conditioning, animals are returned to the context some time later and tested in the absence of the drug. The proline-directed serine/threonine kinase cyclin-dependent kinase 5 (Cdk5) has been implicated in drug abuse, yet the effects of inhibiting Cdk5 in the nucleus accumbens (NAcc) on the induction and expression of drug conditioning have yet to be examined. In a number of recent reports assessing the contribution of Cdk5 signaling to Pavlovian and instrumental conditioning, knock-out, knock-down, or transgenic mice as well as long-lasting lentiviral-mediated gene transfer were used to manipulate Cdk5 and its phosphorylation target kalirin-7. However, because these manipulations spanned the induction and expression phases of conditioning, the results obtained were difficult to interpret as it was impossible to ascertain whether they were due to effects on induction, expression, or a combination of the two. We thus tested the effects of the Cdk5 inhibitor Roscovitine in two separate experiments.

Methods: In one, vehicle or Roscovitine (40 nmol/0.5μl/side) were administered to the NAcc 30 minutes before the drug exposure injections during contextual locomotor conditioning by amphetamine. In the other, vehicle or Roscovitine were administered 30 minutes before testing for contextual conditioning following a saline injection.

Results: Inhibiting Cdk5 in the NAcc during amphetamine exposure prevented the induction of contextual locomotor conditioning. Conversely, inhibiting NAcc Cdk5 on the test for conditioning enhanced the expression of the conditioned locomotor response.

Conclusions: Together, these findings indicate that Cdk5 in the NAcc exerts different effects on the induction (enables it) and expression (suppresses it) of contextual locomotor conditioning by amphetamine. They are consistent with the findings of others suggesting that different neuronal mechanisms underlie the acquisition and expression of excitatory conditioning. Importantly, these different actions of Cdk5 require that caution be exercised when interpreting the results of reports using preparations that confound effects on induction and expression of conditioning.

Keywords: drug conditioning, Cdk5, locomotion, roscovitine.

Disclosure: Nothing to Disclose.

T58. Is SERT All There Is? Genetic Dissection of the Actions of Acute and Chronic Serotonin Selective Reuptake Inhibitors in the SERT M172 Mouse

Alex G. Nackenoff*, Randy D. Blakely

Vanderbilt University Medical Center, Nashville, Tennessee

Background: Depression is one of the most common and burdensome disorders worldwide. The most prescribed antidepressants are serotonin (5-HT) selective reuptake inhibitors (SSRIs), believed to provide therapeutic benefit by antagonizing the 5-HT transporter (SERT) and elevating extracellular 5-HT levels. Significant—yet indirect—evidence supports this conclusion. However, multiple SSRIs have been found to interact with a sizable number of non-trivial targets at physiologically relevant concentrations and several of these, when manipulated specifically, can induce some of the same biochemical and behavioral effects associated with SSRI administration. Just as depression is a complex phenotype, the therapeutic (and untoward) actions of SSRIs are multi-dimensional and the degree to which these are dependent on altered 5-HT signaling is difficult to assess through pharmacological approaches alone. In order to separate the SERT dependent effects of SSRIs from their actions at other targets, we developed a knock-in mouse line that expresses a point mutation in the SERT SSRI binding site, converting an Ile at amino acid 172 to Met (SERT M172) (Thompson et al, PNAS 2011). The M172 substitution is benign with respect to SERT protein expression and 5-HT transport activity in vitro and in vivo, but confers a marked reduction in sensitivity (10-1000 fold) to many SERT antagonists. In our prior studies, where SERT M172 was expressed on a 129S6/S4 background, WT mice on the same background demonstrated meager or anomalous responses to SSRIs in acute tests of antidepressant action (tail suspension test (TST) and forced swim test (FST)). Although insensitivity was noted for citalopram and fluoxetine in biochemical, physiological, and behavioral assays, deeper analyses of SSRI action, particularly with chronic administration, required moving to a more favorable genetic background. Here we evaluate the acute and chronic actions of SSRIs using biochemical, cellular, and behavioral endpoints.

Methods: Through standard breeding efforts, we established the SERT M172 allele on a congenic, C57BL/6J background. In these animals, we first repeated our biochemical studies, assessing genotype effects on SERT protein levels, 5-HT uptake kinetics, and biogenic amine levels. Next, we dosed animals acutely (20 mg/kg i.p.), monitoring genotype effects on performance in the TST and the FST. Finally, using the same dosage regimen, we assessed the impact of chronic SSRI administration (20 mg/kg p.o. for 4 weeks) on behavior in the Novelty Induced Hypophagia (NIH) test, monitoring both latency to approach and consumption of liquid vanilla Ensure™. Finally, we assessed the effects of chronic SSRI dosing (20 mg/g p.o. for 4 weeks) on stem cell proliferation rates in the subgranular zone of the hippocampal dentate gyrus, using immunohistochemistry approaches.

Results: SERT M172 mice display normal SERT protein levels, 5-HT transport kinetics, and basal 5-HT levels, confirming that the SERT I172M substitution appears benign with respect to 5-HT homeostasis. As expected, WT mice demonstrated reduced immobility in the TST and FST tests in response to fluoxetine, citalopram and paroxetine as compared to vehicle-injected animals. In contrast, immobility responses in both tests were unaffected by fluoxetine and citalopram in the SERT M172 animals. In keeping with a lack of effect of the SERT M172 allele on paroxetine sensitivity, paroxetine significantly reduced immobility in the SERT M172 mice, equivalent to that seen in WT animals. In the NIH test, which is sensitive to chronic but not acute SSRI administration, both delay in approach and consumption of Ensure™ were unaffected by fluoxetine and citalopram in the SERT M172 mice, whereas paroxetine reduced approach delay and elevated consumption. Similarly, the positive effect of fluoxetine and citalopram on hippocampal stem cell proliferation rate was lost in the SERT I172M mice whereas genotype had no impact on stem cell production rate induced by paroxetine.

Conclusions: The SERT M172 mice permit a clear demonstration that SERT, and enhanced extracellular 5-HT, is essential for the immobility reducing behavioral actions of acute administration of citalopram and fluoxetine in the TST and FST, and not due to off-target activity. These data also provide the first direct evidence that the chronic actions of these SSRIs in the NIH test and in stem cell proliferation depend solely on their SERT antagonism. Studies are now underway to evaluate the SERT dependence of mixed-action antidepressants, where the SERT M172 model should allow for the contribution of SERT to be assessed, as well as to elucidate the spectrum of antidepressant-induced changes in gene expression that can be attributed directly to the transporter and while not required for their actions in the aforementioned tests, may provide auxiliary support for their actions, or contribute to side effects.

Keywords: Antidepressants, SSRI, Depression, Serotonin.

Disclosure: Nothing to Disclose.

T59. Combined Treatment with Ketamine and Melatonin Promotes Neurosphere Formation of Human Neuronal Precursors

Gloria Benitez-King*, Carlos Berlanga, Salvador Alarcón-Elizalde, Jiabei Liu, Margarita Dubocovich

Instituto Nacional de Psiquiatría Ramón de la Fuente Muñíz, Ciudad de México, Mexico

Background: Major depressive disorder (MDD) is a serious, recurrent and disabling psychiatric illness, projected to be the leading cause of disability worldwide in 2020. Current available antidepressants are associated with frequent relapses, low remission rates and delayed responses (Willner et al., Neurosci Biobehav Rev 37: 2331, 2013). Ketamine (KET), a noncompetitive N-methyl-D-aspartate (NMDA) receptor antagonist causes a rapid antidepressant response in a time lapse of hours. A sub-anesthetic dose of KET enhances dendrite spine formation (Duman, Depress Anxiety 31:291, 2014). Similarly, melatonin (MEL) synergizes anti-depressant actions of serotonin reuptake inhibitors and increases new neuron formation and dendritogenesis in the hippocampal formation of adult rodents (Domínguez-Alonso et al., J Pineal Res 52:427, 2012; Liu et al., J Pineal Res 54:222, 2013; Ramírez-Rodríguez et al., J Pineal Res 56:450, 2014). The olfactory neuroepithelium is generated throughout life and offers a model to investigate the effects of drugs on neuronal differentiation and neurogenesis. It is formed by globose and sustentacular cells and olfactory neurons. Globose cells are stem cells that in culture actively proliferate forming nestin positive sphere-like clusters (neurospheres) (Calof et al., J Neurobiol 36:190, 1998). Neuroepithelial cells are embryologically related to the limbic region and show a gene expression profile similar to that of central nervous system neurons (Horiuchi et al., Neurosci Res 77:247, 2013). Recently we demonstrated that MEL increases axonal formation through increases in MT1 and MT2 melatonin receptor expression in a human neuronal precursor clone obtained by a non-invasive exfoliation procedure (G.B-K unpublished). The aim of this study was to investigate the effect of combined treatment with KET and MEL on the neurodevelopment of cloned human neuronal olfactory neuroepithelium precursors.

Methods: All experiments were conducted after obtaining written consent of the participant and approved by the Instituto Nacional de Psiquiatría Ethics Committee. Cloned neuronal precursors obtained as described (Othman et al., Biothechnic & Histochem 80:189, 2005) with identified neuronal lineage by tubulin Beta III and nestin staining were cultured up to passage 5, and plated on glass cover slips at 10 000 cells/cm2. After 4 days in culture, cells were incubated with vehicle or various concentrations of KET (10-11-10-5M) for 24 h. Parallel cultures were cultured with 10-7M KET from 12 to 72h or with KET10-7M for 48h followed by MEL treatment for 6h (10-11-10-5M). Cells were fixed and stained with either an anti-MAP2 antibody to label dendritic spines, or with an anti-nestin antibody to label the globose cells also referred as neuronal progenitor cells. Microfilaments and nucleus were labeled with rhodamine-phalloidin and DAPI respectively. Images were observed with an epifluorescence Nikon microscope, equipped with a Nikon OS-2Mu Digital Sight camera, and acquired with software from Nikon.

Results: KET (10-11-10-5 M) increased the number of olfactory neuronal cells with dendritic spines with an EC50 of 10-9M after an optimal time of treatment of 48h. Cloned cells incubated with KET 10-7M for 48h and MEL 10-11M or 10-9 M for 6h showed an increased number of nestin-positive globose cells, which are the mitotic phenotype of neuronal progenitors present in the neuroepithelial tissue. However, an increased number of neurospheres were stained with the anti-nestin antibody in cultures incubated with KET-MEL at all concentrations tested. As expected, dendritic spines were clearly present in cells incubated with KET and 10-9 M MEL.

Conclusions: KET increased the number of cells with dendritic spines and neurosphere formation. MEL increased axonal formation and synaptogenesis (G.B-K unpublished). An increased dendritic spine and neurosphere formation was also evident in cells treated with the KET-MEL combination. Our results suggest that human neuroepithelium olfactory neuronal precursors could be useful models for pharmacological screening of antidepressants and assessment of neurogenesis, axonal and dendrite formation and synaptogenesis. Combination treatment with KET-MEL could be useful to reestablish altered neuronal circuitry in patients with MDD. Supported by SEP/Conacyt México Grant No 178075.

Keywords: Ketamine, Melatonin, Neuroepithelium, Neurospheres.

Disclosure: Nothing to Disclose.

T60. Altered Subventricular Zone Niche in Schizophrenia Patients with Immune Activation

Samantha Fung, Guy Barry, Borris Guennewig, Vibeke Catts, Dominik Kaczorowski, Cyndi Shannon Weickert*

Neuroscience Research Australia, Sydney, Australia

Background: Changes characteristic of immune system activation are found in the prefrontal cortex of 40% of people with schizophrenia. Further, these “inflammatory” hallmarks are associated with exaggerated interneuron pathology reflected in lower interneuron mRNAs, elevated white matter neuron density and possibly with elevated cell death markers. Taken together, these observations would be consistent with cortical tissue damage and an ensuing repair attempt by endogenous cortical interneuron precursors. Thus, we hypothesized that molecular changes consistent with regenerative neurogenesis could be detected in the birthplace of new interneurons [subventricular zone (SVZ)] in adults with schizophrenia with elevated cortical cytokines.

Methods: Next generation sequencing (Illumina HiSeq platform) was performed on SVZ RNA from postmortem cases (7 controls and 7 people with schizophrenia). Gene expression changes were confirmed with qPCR analysis on SVZ cDNA from 16 controls and 19 individuals with schizophrenia, and correlations with cortical mRNAs were performed using previously published data indicative of neuroinflammatory state derived from the dorsolateral prefrontal cortex from the same cohort.

Results: In the SVZ, 87 genes were differentially expressed between individuals with schizophrenia with high cortical inflammation compared to low (FDR<0.05), many of these were represented in inflammatory-related gene ontology groups. For example, the immune system regulator suppressor of cytokine signaling 3 (SOCS3), which also promotes neurogenesis, was increased in the SVZ of the high inflammation group relative to low (logFC=-3.99, FDR<0.001). We also found that sialyltransferase 8 (ST8SIA1), expressed during neuronal differentiation to promote glycosylation of NCAM (PSA-NCAM), was increased in the high inflammation SVZ (logFC=-0.77, FDR=0.028). These changes were confirmed by qPCR (SOCS3 mRNA increased 56.1%, t(17)=2.64, p=0.017; and ST8SIA1 mRNA increased 34.9% in high vs low inflammation schizophrenia, t(17)=2.18, p=0.044). Furthermore, in schizophrenia, SVZ ST8SIA1 mRNA also positively correlated with cortical cell death pathway genes TNFSF13 (r=0.643, p=0.003) and FASR (r=0.624, p=0.003), and SVZ SOCS3 mRNA negatively correlated with DLPFC somatostatin (r=−0.62, p=0.005) and NPY (r=-0.63, p=0.004) mRNAs, which we have reported to be reduced in schizophrenia. SOCS3 also positively correlated with the mitotic marker EGFR mRNA in the SVZ in schizophrenia (r=0.56, p=0.013).

Conclusions: We find molecular changes consistent with an induction of new cell birth and differentiation in the SVZ of people with schizophrenia who also exhibit high expression of cortical cytokines. Further, SVZ neurogenic marker increases were related to increased cortical cell death and reduced cortical interneuron markers in schizophrenia. These findings support the hypothesis that there is an active neurogenic response to cortical neuronal damage and neuroinflammation in the brains of some people with schizophrenia.

Keywords: schizophrenia, subventricular zone, interneuron, neurogenic niche.

Disclosure: Nothing to Disclose.

T61. Epigenetic Regulation of Serotonin (5-HT) 5-HT2A:5-HT2C Receptor Balance in Maladaptive Impulsivity

Noelle C. Anastasio*, Aaron L. Miller, Richard B. Pyles, F. Gerard Moeller, Divya Ramesh, Lawrence C. Sowers, Kathryn A. Cunningham

University of Texas Medical Branch of Galveston, Galveston, Texas

Background: Impulsivity factors into the multifaceted determinants that underlie the etiology of substance use and obesity/eating disorders. Prevention and treatment of these chronic pathological maladies could be greatly advanced by a more comprehensive understanding of the biology of impulsivity. A myriad of epigenetic, transcriptional, translational and topological mechanisms govern neural function. Considerable evidence indicates 5-HT systems play a role in impulsivity through the 5-HT2A receptor (5-HT2AR) and 5-HT2CR within the medial prefrontal cortex (mPFC). We tested the hypothesis that impulsivity is dynamically controlled by a 5-HT2AR:5-HT2CR homeostasis and that epigenetic factors may contribute to individual differences. There are no direct analyses linking DNA methylation of 5-HT2AR or 5-HT2CR genes to impulsivity in humans or rodent models of impulsivity. If methylation patterns mediate a gene x environment interaction to control impulsivity, we expect that that high impulsivity will track with lower and higher methylation of 5-HT2AR and 5-HT2CR, respectively.

Methods: We investigated the relationship among blood lymphocyte 5-HT2AR and/or 5-HT2CR DNA methylation patterns and a self-report measure of impulsivity (Barratt Impulsiveness Scale; BIS-11). Lymphocytes were collected from blood samples from healthy control (n=10) and cocaine-dependent (n=10) individuals and genomic DNA was extracted. Impulsivity was evaluated in an outbred rat population using the 1-choice serial reaction time task; medial prefrontal cortex (mPFC) was extracted from rats characterized as high or low impulsive and processed for DNA, RNA or protein analyses. Genotyping for single nucleotide polymorphisms in the 5-HT2AR [e.g., A(-1438)G, T102C] and 5-HT2CR [C23S, C(-759)T, G(-697)C] was performed on human lymphocyte samples only. The bisulfite-treated DNA from both human and rat samples was subjected to epigenetic PyroMark sequencing to determine the amount of methylated cytosine residues of the 5-HT2AR and 5-HT2CR genes. Specifically designed primers that amplified eight overlapping fragments of the 5-HT2AR promoter and approximately 62 CpG sites were employed. For the 5-HT2CR gene, a Qiagen GeneGlobe assay was employed. We employed immunoblots to detect mPFC synaptosomal protein expression of the 5-HT2AR and 5-HT2CR.

Results: There was no association of methylation patterns of the 5-HT2AR between control and cocaine-dependent individuals; % methylation at sites -1439, -1420, and 102 associated with genotype independent of cocaine dependence status. There was a positive correlation between % methylation at site -1439 of the 5-HT2AR and total BIS-11 scores (r=0.629, p=0.016) independent of cocaine dependence status. The 5-HT2CR gene was not detectable in blood lymphocyte samples. Levels of impulsivity in rats positively correlated with mPFC 5-HT2AR protein levels (R2=0.21;p<0.01), inversely correlated with mPFC 5-HT2CR protein levels (R2=0.22;p<0.01) and positively correlated with the mPFC 5-HT2AR:5-HT2CR ratio (R2=0.31;p<0.01). Methylation of 5-HT2AR did not associate with impulsivity in rats; 5-HT2CR was methylated to a greater degree in high vs. low impulsive rats (p<0.05).

Conclusions: These translationally-focused analyses suggest that variations in the methylation patterns of the 5-HT2R genes may represent putative biomarkers of individual differences of impulsivity. Further, these data indicate that there is an interactive relationship between mPFC 5-HT2AR and 5-HT2CR that may drive impulsive phenotypes. Methylation patterns of these genes may mediate a gene x environment interaction to control inherent impulsivity. Thus, we have uncovered potential neurobiological mechanisms of importance in driving inherent impulsivity.

Keywords: impulsivity, serotonin 2 receptor, epigenetics, translational.

Disclosure: Dr. Moeller is a consultant for Boehringer-Ingelheim. Dr. Cunningham is a consultant for Arena Pharmaceuticals and an editor of Neuropsychopharmacology Reviews for which she receives compensation from the American College of Neuropsychopharmacology. All other authors report no biomedical financial interests or potential conflicts of interest.

T62. Environmental Enrichment Paradigm Identifies GSK3 as a Target Gene for Protection Against Psychiatric Disorders

Miroslav Nenov, Yafang Zhang, Elisabeth Crofton, Federico Scala, Marcello D'Ascenzo, Thomas Green, Fernanda Laezza*

University of Texas Medical Branch, Galveston, Texas

Background: Vulnerability to psychiatric disorders is intimately interlinked with environmental factors that can predispose or protect individuals against these afflictions. By turning on and off specific genetic programs in the mesocortico-limbic circuit, enriched environmental conditions (EC) can exert protective effects against depression- and addiction-related behaviors, opposing maladaptive neuronal plasticity that accounts for aberrant behavior. Yet, how these genetic programs translate into molecular and cellular changes at the circuitry level is still poorly understood. In search for EC-induced cellular pathways relevant for affective disorders and addiction, we conducted an unbiased transcriptomic study in the nucleus accumbens (NAc) of rats raised in EC and found that glycogen-synthase 3 (GSK3) and the voltage-gated Na+(Nav) channel Nav1.6 mRNA levels were significantly decreased when compared to isolated rats (IC). GSK3 is a complex enzyme that acts as a converging node of neuronal activity through a downstream cascade of phosphorylation, and dysfunctional activity of GSK3 associates with psychiatric disorders and additive behaviors. Nav1.6 is a fundamental molecular determinant of neuronal excitability, responsible for rhythmic oscillations, adaptations and plasticity in the NAc. Based on this evidence, we postulated that changes in GSK3 expression in the NAc might coordinate changes in neuronal excitability through Nav1.6 with effects on depression-, anxiety- and addiction-related behavior. With a combination of viral vector genetic silencing, patch-clamp electrophysiology and behavioral studies, we provide exciting data demonstrating that genetic silencing and/or pharmacological suppression of GSK3 in the NAc leads to reduced firing and suppression of Na+-persistent currents in medium spiny neurons, increasing depression-like and addiction-related behavior while reducing anxiety-like behavior in multiple behavioral models.

Methods: Next generation RNA sequencing was used to identify EC-related genes in the NAc. Rats were assigned either to a singly-housed isolated control condition (IC) with no social contact or novelty, or a group-housed enriched condition (EC) with social contact, exercise and novelty. EC and IC rats self-administered saline or cocaine for 14 days. Hairpins targeting GSK3beta have been designed, sub-cloned into an appropriate AAV2 vector, and validated for functional knockdown in vitro and in vivo. Whole-cell patch-clamp and loose-patch recordings in visually identified medium spiny neurons were performed in acute NAc shell slices. Slices were either injected with AAV-shRNA against GSK3 and/or control vector, or exposed to the GSK3 inhibitor CHIR99021. Depression-like and anxiety-like behaviors along with addiction-related behaviors were evaluated in multiple behavioral models of AAV-transduced rats.

Results: Bioinformatics analysis of regulated transcripts indicates that EC leads to decreased transcript levels of several proteins involved in the PI3K/Akt signaling pathway compared to the IC group. Intriguingly, we find that GSK3beta and SCN8A (coding for Nav1.6) were among the regulated transcripts. In NAc shell acute brain slices, medium spiny neurons expressing AAV-GSK3beta shRNA exhibited a decrease in firing rates estimated with loose-patch recordings compared to AAV control (2.4±0.6, n=14 versus 3.6±0.4; n=16 in AAV control; p<0.009), while bath application of the GSK3 inhibitor, CHIR99021 (1uM) significantly suppressed evoked action potential firing of medium spiny neurons (12±0.5 maximum number of spikes, n=6, p<0.05) compared to vehicle treated cells (17±1 maximum number of spikes, n=7). Furthermore, CHIR99021 suppressed Na+persistent currents in the NAc shell (3±0.25 pA/pF, n=12 in CHIR99021 compared to control 2.1±0.15 pA/pF, n=10, p<0.05) supporting a role of Nav1.6 as a GSK3 target. Silencing GSK3beta in the NAc shell increases depression-like and addiction-related behavior while reducing anxiety-like behavior in multiple behavioral models.

Conclusions: Taken together, these findings provide evidence for a new complex role of GSK3 in regulating mesocortico-limbic circuits through an effect on Nav channels that plays a significant role in controlling neuronal firing and behaviors. As such, we provide new knowledge of the cellular and molecular mechanisms underlying protection against and predisposition toward affective disorders and addiction in the reward circuit. Support: R01MH095995 (FL), R01DA029091 (TAG).

Keywords: ion channels, kinases, excitability, nucleus accumbens.

Disclosure: Nothing to Disclose.

T63. Antidepressant Effects of Pachyman, a Natural Ingredient, via Alteration of Microglial Cytokine Expression in Social Defeat Stress Models

Koki Ito, Atsushi Saito, Michael Ballinger, Jed Fahey, Paul Talalay, Atsushi Kamiya*

Johns Hopkins University School of Medicine, Baltimore, Maryland

Background: There are a number of clinically effective treatments for stress-associated mental disorders, especially depression. Nonetheless, a large portion of those afflicted exhibit treatment-resistance to first-line treatments (e.g., SSRIs), which calls for novel interventions based on pathological mechanisms. Alterations in immune and inflammation processes, including changes in expression of pro- and anti-inflammatory cytokines are observed in patients with depression. Consistently, recent preclinical studies indicate that chronic stress induces intracerebral activation of immune and inflammation systems that are largely regulated by immune cells, such as microglia. Notably, several natural products, such as constituents of traditional Eastern medicines have been empirically used for treatment and prevention of depressive symptoms, although there is almost no mechanism-based evidence for the effectiveness of most traditional medicines. Thus, understanding the mechanistic link between their antidepressant and anti-inflammatory effects may open a new window for identification of novel preventive and/or treatment targets for depressive symptoms. Pachyman, 1, 3-β-Glucans is a polysaccharide extracted from Poria cocos used as a main ingredient in many traditional Eastern medicines, and is reported to have immune/inflammatory effects. In this study, we examine the effects of pachyman on stress-associated behavior and explore its underlying mechanism, with focus on microglial immune/inflammation changes.

Methods: We examined whether chronic oral-administration of pachyman prevents observed microglial immune changes and depressive behaviors by using the chronic social defeat mouse model. The effect of pachyman on sociability was assessed by the social interaction test using mice subjected to chronic social defeat, with administration of pachyman. The effect of pachyman on cytokine expression in microglia, serum, and brain tissue, including the prefrontal cortex and hippocampus in the social defeat stress model as well as in lipopolysaccharide-stimulated murine microglial cells was assessed.

Results: We found that chronic treatment of pachyman prevented social avoidance phenotypes induced by social defeat stress. Notably, pachyman strongly suppressed microglial expression of a specific cytokine in the prefrontal cortex, but not in the hippocampus. Interestingly, we observed that pachyman was taken up by microglial cells under in vitro conditions, suggesting that pachyman may have antidepressant effect via direct targeting microglia to modulate cytokine expression.

Conclusions: We currently examine whether chronic oral-administration of pachyman goes through the blood brain barrier to target microglia under social defeat stress conditions. We also examine the molecular mechanisms of antidepressant action of pachyman, which may provide a basis for identifying novel drug targets for the prevention and treatment of stress-associated human behaviors. From its use over the past centuries, several Eastern medicines that contain pachyman as a main ingredient, are empirically known to have a therapeutic effect on neuropsychiatric conditions, including depressive symptoms. Nonetheless, underlying mechanisms of its action are largely unknown. Thus, exploring the effect of pachyman may provide a unique opportunity to uncover new targets for pharmacotherapy of stress-associated psychiatric conditions, including depression and can also accelerate the understanding of pathophysiology of these conditions.

Keywords: pachyman, depression, social defeat stress, microglia.

Disclosure: Nothing to Disclose.

T64. Persistent Memory Deficits and Histone Modification after Systemic Inflammatory Event

Natalie Tronson*, Elissa Donzis, Natalie Nevárez

University of Michigan, Ann Arbor, Michigan

Background: Inflammatory events including myocardial infarction, illness or major surgery, commonly lead to cognitive deficits and depression-like behavior lasting months or years after the event. These clinical observations paralleled in observed in animal models of myocardial infarction and inflammation. The mechanisms mediating such persistent effects remain unknown. Rodent models of MI have previously been reported to result in depression like behavior and enhance fear conditioning soon after the infarction, and mechanisms mediating these alterations have been hypothesized to be due to acute cytokine-dependent signaling. In addition, we have previously demonstrated impaired context fear conditioning and increased depression-like behavior 8 weeks following surgically induced myocardial infarction in mice. These alterations in mood and memory persist beyond the duration of cytokine activity in the brain after MI. One candidate mechanism for mediating such persistent effects is epigenetic changes as a consequence of cytokine-dependent signaling. Dysregulation of histone acetylation has been shown to mediate memory deficits during aging, and similar mechanisms may be triggered by inflammatory signaling. Here, we hypothesize that although the initial cytokine signaling would be resolved within 8 weeks after myocardial infarction, changes in histone modifications and would persist, mediating the lasting impairments after a transient inflammatory event.

Methods: We used a surgical model of heart attack (myocardial infarction, MI) in male and female mice to determine the sustained effects of a systemic inflammatory event on fear-associated memory, histone modifications, and intracellular signaling mechanisms of memory. Fear conditioning consisted of 3 minute exposure to context followed by a single 0.8mA, 2sec food shock. Mice were randomly assigned to one of three surgical conditions: cryo-injury MI, Sham surgery, or Non-operated. 8 weeks after surgery, half of each group was randomly assigned to be fear conditioned or serve as untrained controls. One hour after fear conditioning hearts and hippocampi were dissected, and blood serum collected, from all mice. We used multiplex cytokine analysis to determine cytokines in the hippocampus and serum 8 weeks after MI. In addition, we conducted western blotting to determine persistent alterations in histone acetylation, phospho-acetylation and methylation, as well as dysregulation of signal transduction as a consequence of MI.

Results: Eight weeks after MI surgery, both male and female mice exhibited impairments in fear conditioning compared with non-operated controls. In addition, female but not male Sham-operated mice exhibited deficits in context fear conditioning. There were no differences in peripheral or hippocampal cytokine level at this time. In contrast, we observed dysregulation of ERK signaling, Arc activation, and increased histone acetylation and phospho-acetylation in the hippocampus eight weeks after MI.

Conclusions: These data show a persistent dysregulation of intracellular signaling and epigenetic regulation in the hippocampus after a systemic inflammatory event. These results are consistent with findings demonstrating that memory-impairing treatments, such as chemotherapy, or normal aging are associated with increases in histone acetylation and decreases in histone deacetylase activity. The causal role of histone modifications and dysregulated signal transduction is under further investigation. These findings identify potential mechanisms in the brain that may mediate lasting changes in mood and cognition after a systemic inflammatory event, and suggest novel targets for prevention and treatment of persistent cognitive deficits after MI, illness or major surgery.

Keywords: cytokine, epigenetic, memory, fear conditioning.

Disclosure: Nothing to Disclose.

T65. Neuroserpin Protects Parvalbumin Interneurons Against Perineuronal Net Degradation for Cortical Stability

Noreen Bukhari, Poromendro Burman, Michael Demars, Ayan Hussein, Hirofumi Morishita*

Icahn School of Medicine at Mount Sinai, New York, New York

Background: Schizophrenia is a devastating disorder with typical onset during adolescence. One of the major underlying pathophysiological changes is the developmental dysfunction of cortical inhibitory GABA neurons expressing parvalbumin (PV). PV interneurons are known to increasingly being enwrapped by perineuronal nets (PNNs) of extracellular matrix to stabilize the cortical circuits across the developmental critical period. Recent studies reported that PNNs are reduced in the human schizophrenic brains. Identification of endogenous protective mechanisms of PV interneurons against the degradation of PNNs will provide novel therapeutic targets for the prevention and intervention of psychiatric conditions. PNNs are primarily composed of chondroitin sulfate proteoglycans (CSPGs), which are key targets of serine protease tissue plasminogen activator (tPA). Here we tested the hypothesis that an endogenous inhibitor of tPA called Neuroserpin (or Serpini1) normally acts as a protective shield for PV interneurons against the degradation of PNNs to maintain cortical stability. Importantly, de novo damaging mutation of Neuroserpin gene was recently identified in persons with schizophrenia (Gulsuner et al., 2013).

Methods: We employed mouse visual cortex as a well-established model of developmental critical period for cortical maturation. The impact of tPA-dependent proteolysis and its inhibition by Neuroseprin to the degradation of PNNs was examined by immunohistochemistry for PNNs of PV interneurons in Neuroserpin knock out (KO) mice. The experience-dependent regulation of tPA activity and Neuroserpin expression were examined by an amidolytic assay of brain homogenates from visual cortex, in situ hybridization and immunohistochemistry. To examine the functional role of Neuroserpin on cortical stability, adult Neuroserpin KO mice were subjected to 4-day monocular deprivation and then to in vivo extracellular recordings to assess functional plasticity in vivo. The role of Lynx1 as an upstream regulator of Neuroserpin was also examined by analyzing the expression of Neuroserpin and PNNs as well as visual cortex plasticity in Lynx1KO mice with an without neuronal and PV-cell specific viral overexpression of Neuroserpin in visual cortex.

Results: In concert with the increased expression of PNNs around PV interneurons from juvenile to the adult, we found that the expression of Neuroserpin remains elevated in the adult visual cortex in mark contrast to the juvenile period when neuroserpin reduces specifically in PV interneurons in an experience-dependent manner. Strikingly, the removal of Neuroserpin in the adult induced the degradation of PNNs and unmasked robust experience-dependent cortical plasticity, suggesting that Neuroserpin protects PV interneurons against tPA-dependent proteolysis to limit adult plasticity. In search of upstream mechanisms that limit experience-dependent reduction of Neuroserpin in the mature cortex, we turned to Lynx1, which is enriched in PV-cells and increases its expression after the critical period to limit cortical plasticity (Morishita et al 2010). We found that adult LynxKO mice showed reduced Neuroserpin expression, elevated tPA activity, and reduced PNNs after monocular deprivation as in the juvenile wild-type mice. Consistently, the elevated adult plasticity in Lynx1KO mice was blocked by pan-neuronal or PV interneuron-specific viral overexpression of Neuroserpin in visual cortex, suggesting Lynx1 as a gate to limit the experience-dependent reduction of Neuroserpin and PNNs in PV-cells and subsequent functional plasticity in the adult.

Conclusions: Collectively, our study suggests Neuroserpin as a novel protective shield for PV interneurons from the degradation of PNNs as well as a novel negative regulator of cortical plasticity to maintain cortical stability in the adult brain. As Neuroserpin is a part of the core transcriptome network consisting of de novo mutations in Schizophrenia (Gulsuner et al 2013), Neuroserpin may serve as an attractive therapeutic target to enhance the endogenous protective ability of PV interneurons from risk factors such as oxidative stress to maintain the cortical stability.

Keywords: Neuroserpin, Parvalbumin, Perineuronal nets, cortical plasticity.

Disclosure: Nothing to Disclose.

T66. Transcriptomic Effects of Antidepressant Treatment and Glucocorticoid Receptor-overexpression on the Maturational Status of Brain Cells in Mice

Tsuyoshi Miyakawa*, Hisatsugu Koshimizu, Koji Ohira, Hideo Hagihara, Rika Takeuchi, Keizo Takao

Institute for Comprehensive Medical Science, Fujita Health University, Toyoake, Japan

Background: The development and maturation of the brain has long been believed to be a one-way process. In previous studies on mouse models of psychiatric disorders, we proposed pseudo-immature brain cell states as intermediate phenotypes in these disorders (Yamasaki et al., Mol. Brain, 2008; Takao et al., Neuropsychopharmacol., 2013). Recent studies, including our own, have shown that some brain cells repeatedly undergo rejuvenation and maturation in response to changes in the external environment, such as treatment with antidepressant [fluoxetine (FLX)] (Kobayashi et al., PNAS, 2010; Karpova et al., Science, 2011), pilocarpine-induced seizure (Shin et al., Bipolar Disord., 2013), and physiological stimulation (Donato et al., Nature, 2013). Maturation failures in the brain have also been identified in some regions and cell types in patients with psychiatric disorders, including schizophrenia and bipolar disorder (Barreta et al., Neuropharmacol., 2012; Gandal et al., PLoS ONE, 2012; Walton et al, Transl. Psychiatry, 2012; Hagihara et al., Mol. Brain, 2014).

 Overexpression of glucocorticoid receptor (GR) in the forebrain during early life and throughout the lifetime causes increased depression-like and/or anxiety-like behaviors in mice (Wei et al., PNAS, 2004; Biol. Psychiatry, 2012), suggesting that mice overexpressing GR (GRov mice) represent a potential animal model for mood disorders, such as depression and anxiety disorder. Overexpression of GR has been shown to cause an aging-like neuroendocrine phenotype and mild cognitive dysfunction in young mice (Wei et al., J. Neurosci., 2007), raising the possibility that GRov mice may have an "over-matured" brain status.

In this study, to further characterize changes in cell maturity in brains treated with antidepressants and brains from a potential animal model for mood disorders, we compared genome-wide gene expression in the medial frontal cortex (mFC) and dentate gyrus (DG) of FLX-treated mice and the DG of GRov mice with those of the corresponding regions in normally developing brains of wild-type (WT) mice.

Methods: Mice (9 weeks old) were subcutaneously administered either antidepressant or vehicle pellets in the dorsal interscapular region for 3 weeks (Ohira and Miyakawa, Mol. Brain, 2011). All animal experiments were approved by the Institutional Animal Care and Use Committee of Fujita Health University. For the mFC and DG of FLX-treated mice (12 weeks old) and mFC of developing mice (2 and 12 weeks old), GeneChip analyses were conducted using Mouse Genome 430 2.0 Array and GeneChip Scanner 3000 (Affymetrix, Santa Clara, CA). Microarray data of the DG of GRov mice are available in the NextBio database (ID: GSE30187). These microarray data sets were then analyzed with NextBio (NextBio, Cupertino, CA), a repository of analyzed microarray data sets that allows investigators to compare gene expression patterns (signatures) among mouse lines (Kupershmidt et al., PLoS ONE, 2010; Takao and Miyakawa, PNAS, 2014). NextBio compares the signatures in publicly available microarray databases with a signature provided by the user, using a “Running Fisher” algorithm. The overlap P-value—direction of the correlation between two given gene signature sets—and the P-values between subsets of gene signatures, is calculated with the algorithm.

Results: We compared genome-wide gene expressions in the FLX-treated mFC and DG with those of the corresponding regions of normally developing brains in WT mice (mFC, P14 compared to P84; DG, P14 compared to P30). We found that the mFC in adult mice treated with FLX resembled juvenile mFC in terms of its genome-wide expression profile (overlap P-value: 2.3 × 10-9). The common gene alterations between the FLX-treated mFC of adult mice and the mFC of normally developing mice might be partly due to maturational abnormalities in fast-spiking interneurons, astrocytes, and oligodendrocytes. Similar results were observed in the DG (overlap P-value: 2.0 × 10-65). On the other hand, the DG of GRov mice exhibit an opposite expression pattern of immaturity and maturity marker genes when compared to the developing DG (P14 compared to P30) in WT mice (mice with GR overexpression during early life, overlap P-value: 2.5 × 10-18; mice with GR overexpression during the entire life, overlap P-value: 2.6 × 10-9).

Conclusions: In this study, we demonstrated (1) that chronic FLX treatment induces pseudo-immaturity of the mFC and DG of adult mice with respect to gene expression patterns, and (2) that maturation status of the DG in GRov mice may be changed toward "over-maturity". Based on these results, it is tempting to speculate that FLX treatment may change the status of brain cells in mood disorders from over-maturity to normal state or immaturity, which might be a mechanism underlying the antidepressant effect of FLX.

Keywords: antidepressant, mood disorder, glucocorticoid receptor, gene expression.

Disclosure: Tsuyoshi Miyakawa is an advisor/consultant for Astellas Pharma Inc., which provided partial support for these studies.

T67. vmPFC Glutamate Correlates of Cocaine Craving During Protracted Withdrawal

Christina Shin, Michela Serchia, John Shahin, Anna Agaronova, Karen Szumlinski*

University of California at Santa Barbara, Santa Barbara, California

Background: Re-exposure to drug-associated cues, even following protracted abstinence, can elicit intense drug craving and trigger relapse to drug-taking. In both humans and animal models of addiction, cue-elicited drug craving increases in a time-dependent manner during drug abstinence - a phenomenon termed “incubation of craving”. The neural substrates of this phenomenon are not fully understood but may involve a sensitization of cue-elicited neurotransmitter release within the ventromedial prefrontal cortex (vmPFC).

Methods: To test this hypothesis, male Sprague-Dawley rats were trained to lever-press for cocaine (0.25 mg/infusion; 6 h/day) or sucrose pellets (45 mg) and in both cases, reinforcer delivery was signaled by a tone-light compound stimulus. A control group was allowed the opportunity to lever-press for the compound stimulus only. After 10 consecutive days of self-administration training, animals were left undisturbed for either 3 or 30 days at which time, rats underwent an in vivo microdialysis session in the operant chamber, during which animals were allowed to respond for presentation of the compound stimulus, in the absence of reinforcer delivery.

Results: Cocaine-trained animals exhibited a time-dependent intensification of cue-reinforced responding that was selective for the lever that previously delivered cocaine. In contrast, the behavior of sucrose and control animals during this test was not lever-selective. Engaging in cocaine-seeking elicited a withdrawal-dependent sensitization of vmPFC glutamate release, but a dissipation of cue-elicited dopamine release. In contrast, the changes in vmPFC glutamate and dopamine elicited by sucrose-paired or neutral cues did not vary with the passage of time during withdrawal.

Conclusions: These data provide novel evidence that an incubation of cue responsiveness within vmPFC glutamate is a biochemical correlate of incubated cocaine craving. These argue that pharmacotherapeutic strategies that blunt corticofugal glutamate responsiveness to cocaine-paired cues as a strategy for facilitating addiction recovery.

Keywords: craving, glutamate, cocaine, prefrontlal cortex.

Disclosure: Nothing to Disclose.

T68. Adjunctive GLYX-13 Induces Prolonged Efficacy in Subjects with Major Depressive Disorder (MDD)

Ronald Burch, Sheldon Preskorn, Lee Bastin, Wen Yu, Jeffrey Burgdorf*, Joseph Moskal

Northwestern University, Evanston, Illinois

Background: GLYX-13, a NMDA receptor functional partial agonist, has demonstrated antidepressant activity following administration of single doses to subjects with MDD.

Methods: GLYX-13 was administered over 12 weeks to subjects who had responded inadequately to another antidepressant agent. Subjects continued taking the other antidepressant during the entire course of the study. The study was divided into 3 parts. During the first 6 weeks, subjects were randomized to receive GLYX-13 at 5 mg/kg IV or 10 mg/kg IV. Subjects returned to the clinic weekly. If subjects had reached clinical response (HDRS-17 reduced>50% from baseline) to GLYX-13 administered the previous week, placebo was administered weekly until relapse (HDRS-17 reduced<50% from baseline). At the 6th week of evaluation, subjects who had achieved response to GLYX-13 at some visit were assigned to weekly or biweekly dosing based on time to relapse during placebo administration and randomized to continue receiving GLYX-13 or placebo (randomized withdrawal) for a subsequent 6 weeks of dosing. At the end of the randomized withdrawal period, subjects received placebo injections for 4 weeks. Subjects were blind to the treatments. Third party evaluators blind to treatment and protocol were utilized. Treatment dose and interval were calculated by an interactive web-based response system based on a mathematical algorithm; site personnel were blinded to treatment.

Results: Of the subjects, 67% were female, 33% male, median age was 50 years, with median diagnosis of MDD 18 years on average. Baseline HDRS-17 score was 23.2-24.1 across groups. During the 6 week adaptive dose interval period 53% of 368 subjects reached response. Approximately 67% of responders relapsed within 2 weeks following cessation of GLYX-13 and were assigned to weekly dosing during the randomized withdrawal period whereas 33% of subjects relapsed more slowly and were assigned to biweekly dosing. At the end of the randomized withdrawal period, 65% of subjects who received biweekly doses of GLYX-13 achieved response and 45% achieved remission (defined as HDRS score<7). Percent of subjects who achieved remission in the 5 mg/kg weekly group was not statistically different from the biweekly dose groups whereas only 42% of subjects who received 10 mg/kg weekly reached remission. Site investigators blinded to treatment groups and dose level also evaluated subjects using the CGI-S. CGI-I at baseline was 4.3 – 4.6 across groups. At the end of the randomized withdrawal period, CGI-I scores were reduced by 2.5+0.12 in the biweekly groups 2.5+0.14 in the 5 mg/kg IV weekly dose group, but was reduced only 1.6+0.15 in the 10 mg/kg IV group . HDRS-17 scores did not return toward baseline in the subjects randomized to placebo during the second 6 week period, and during the 4 week washout phase following the randomized withdrawal phase, HDRS-17 scores remained at their low level in subjects who had been receiving GLYX-13 as well as the subjects who had been receiving placebo. Thus, in subjects who received placebo for 10 weeks after attaining response to GLYX-13 maintained low HDRS-17 scores.

Conclusions: Adjunctive GLYX-13 caused reduction in HDRS-17 scores in subjects with MDD that had responded inadequately to another antidepressant agent. Following the first few doses, response relapsed over a week or more but as treatment continued, decrease in HDRS-17 in response to each dose of GLYX-13 progressively decreased such that following 6 weeks of dosing scores were reduced from 23.5+0.34 at baseline to 10.3+0.65 in responders. Maximum reduction of HDRS-17 scores in all dosing groups was apparent by week 10 (week 3 of randomized withdrawal). Following 6 weeks of dosing, withdrawal of GLYX-13 was not associated with return of HDRS-17 score toward baseline for up to 10 weeks. Whether this long-term efficacy was due solely to GLYX-13 (which induces synaptic plasticity for several weeks following a single dose) or was instead due to GLYX-13 causing some other effect that allowed the background antidepressant to maintain efficacy after withdrawal of GLYX-13 remains to be evaluated in subsequent studies.

Keywords: NMDA Receptor, GLYX-13, Antidepressant, Major Depressive Disorder.

Disclosure: Ronald Burch, Lee Bastin, Wen Yu, and Joseph Moskal are employees of Naurex, Inc. Sheldon Preskorn is a clinincal trial design consultant for Naurex, Inc. Jeffrey Burgdorf is a consultant for Naurex, Inc.

T69. Cariprazine Monotherapy for the Treatment of Bipolar I Depression: Results: of an 8-Week, Double-blind, Placebo-controlled Study

Joseph R. Calabrese*, Suresh Durgam, Alan Lipschitz, Hua Guo, Willie Earley, István Laszlovszky, György Németh

Case Western Reserve University School of Medicine, Cleveland, Ohio

Background: Although bipolar I depression (BD) is associated with substantial disease burden and economic costs, relatively few effective pharmacological treatments are available. BD is estimated to affect 1% of the general US population (Merikangas et al 2007). Compared to other mental illnesses, BD is more likely to be accompanied by lifetime co occurring anxiety, substance abuse, and medical conditions (Merikangas et al 2007, Kessler et al 2005), which increase standardized mortality ratios to 1.6-2.1 (Osby et al 2001). Goldstein and colleagues (2009) reported high rates of metabolic syndrome in BD and increased risk of cardiovascular disease with onset occurring more than a decade earlier than in control subjects. Long-term functional outcomes are poor, with frequent recurrences of severe symptoms, persistent subthreshold symptoms between episodes, and persistent functional impairment and lost productivity. In an attempt to address these unmet needs, cariprazine, a dopamine D3 and D2 receptor partial agonist with preferential binding to D3 receptors, has been studied in acute manic and mixed episodes, and more recently in this Phase II study of BD (NCT01396447).

Methods: Patients meeting DSM-IV criteria for BD and a Montgomery-Åsberg Depression Rating Scale (MADRS) total score of ≥20 were randomized to 8 weeks of double-blind treatment with placebo (n=148) or 1 of 3 fixed-doses of cariprazine (0.75 mg/d, n=143; 1.5 mg/d, n=147; 3 mg/d, n=146). Primary and secondary assessments were the MADRS and the Clinical Global Impressions-Severity (CGI-S) scale, respectively. Although the double-blind treatment period was 8 weeks in total duration, primary and secondary efficacy outcomes were prospectively defined as mean score change from baseline to Week 6. Additional efficacy outcomes included mean change from baseline in the 17-item Hamilton Depression Rating Scale (HAMD17) total score, treatment response (≥50% improvement in MADRS total score), and symptom remission (MADRS total score ≤10; HAMD17 total score ≤7). Least squares mean differences (LSMDs) between cariprazine and placebo for mean score changes were analyzed using an MMRM approach, with P values for primary and secondary efficacy parameters adjusted for multiple comparisons. Cohen’s d effect size estimates were calculated post hoc for the primary analysis. Response and remission were analyzed using odds ratios (ORs) based on a logistic regression model. Safety assessments included treatment-emergent adverse events (TEAEs), vital signs, clinical laboratory tests, electrocardiograms, extrapyramidal symptom scales, and the Columbia-Suicide Severity Rating Scale.

Results: After 6 weeks of treatment, patients receiving cariprazine 1.5 mg/d compared with placebo had significantly greater mean improvement in MADRS total scores (LSMD [95% CI]: 4.0 [-6.3, -1.6], adjusted P=.003). MADRS reductions in the cariprazine 0.75 and 3 mg/d groups were not significantly different than placebo (LSMD [95% CI]: 0.75 mg/d, -1.9 [ 4.3, 0.5], adjusted P=.13; 3 mg/d, -2.5 [-4.9, -0.1], adjusted P=.11). Effect sizes for the 0.75, 1.5, and 3.0 mg/d groups were 0.20, 0.42, and 0.26, respectively. Patients receiving cariprazine 1.5 mg/d also had significantly greater improvement relative to placebo in CGI-S scores (LSMD [95% CI]: -0.4 [-0.6, -0.1], adjusted P=.01). CGI-S changes in the 0.75 mg/d (LSMD [95% CI]: 0.1 [-0.4, 0.1], adjusted P=.30) and the 3.0 mg/d (-0.3 [-0.5, -0.0], adjusted P=.11) groups did not differ significantly from placebo. Mean improvement in HAMD17 total scores at Week 6 in the 1.5 mg/d (LSMD [95% CI]: -2.7 [-4.4, -1.0], P=.002) and 3.0 mg/d (-2.2 [ 3.9, 0.5], P=.01) groups were significantly greater than placebo. A significantly greater percentage of patients in the cariprazine 1.5 and 3.0 mg/d groups relative to placebo achieved MADRS response (≥50% improvement) at Week 6 (placebo, 31.9%; 0.75 mg/d, 38.6%, OR=1.4, P=.23; 1.5 mg/d, 49.7%, OR=2.1, P=.002; 3 mg/d, 44.8%, OR=1.7, P=.02). Significantly greater rates of remission at Week 6 were also seen in the cariprazine 1.5 mg/d group relative to placebo using both MADRS (placebo, 19.9%; 1.5 mg/d, 36.6%, OR=2.4, P=.002) and HAMD 17 remission criteria (placebo, 15.6%; 1.5 mg/d, 30.3%, OR=2.3, P=.004). The percentage of patients meeting MADRS or HAMD17 remission criteria at Week 6 in the cariprazine 0.75 (23.6% and 20.0%, respectively) and 3.0 mg/d (27.6% and 21.4%, respectively) groups were not significantly different than placebo (19.9% and 15.6%, respectively). The only TEAE that occurred in any cariprazine group at an incidence of ≥5% and twice the rate of placebo was akathisia (placebo, 1.4%; 0.75 mg/d, 2.8%; 1.5 mg/d, 4.8%; 3.0 mg/d, 14.4%). Most incidences (94.1%) of akathisia were mild (38.2%) or moderate (55.9%) in intensity and did not result in study discontinuations. Mean changes in metabolic and other laboratory parameters were generally small and similar among treatment groups.

Conclusions: In adults with depressive episodes associated with BD, cariprazine 1.5 mg/d demonstrated significantly greater efficacy than placebo on both the primary and secondary endpoints, MADRS and CGI-S, respectively. Cariprazine was safe and generally well tolerated.

Keywords: cariprazine, bipolar depression, dopamine.

Disclosure: Supported by funding from Forest Laboratories, Inc., a subsidiary of Actavis plc, and Gedeon Richter Plc. J. Calabrese has served on advisory boards of Forest Laboratories, Inc. and Gedeon Richter Plc. S. Durgam, A. Lipschitz, H. Guo, and W. Earley are employees of Forest Research Institute, a subsidiary of Actavis plc. I. Laszlovszky and G. Németh are employees of Gedeon Richter Plc.

T70. Dual Orexin Receptor Antagonist E2006 Shows Efficacy on Sleep Initiation and Maintenance in Phase 2 Study

Andrew Satlin*, Patricia Murphy, Margaret Moline, Colin Orford, Luigi Giorgi, Kate Bradshaw

Eisai, Woodcliff Lake, New Jersey

Background: While many medications are used, approved and off-label, to treat insomnia disorder, an unmet medical need remains for a medication that leads to both rapid sleep induction and sustained sleep, without daytime residual sleepiness or aberrant nocturnal behaviors. Most approved treatments for insomnia act diffusely to cause sedation; a better approach may be to counteract hyperarousal by decreasing activity of wake-promoting systems. This is the presumed therapeutic mechanism of the dual orexin receptor antagonists (DORA) under development. This report presents the results from a Phase 2 proof-of-concept/dose range-finding study of E2006, a novel DORA. Study objectives were to identify a dose(s) of E2006 that balanced efficacy (change from baseline (BL) on objective polysomnographic (PSG) sleep parameters) and safety (subjective and objective next-day residual sleepiness) at the beginning and end of 15 days of treatment with E2006 vs placebo.

Methods: E2006-G000-201 was a US-based, multicenter, randomized, double-blind, placebo-controlled, parallel group, dose response study of E2006 efficacy and safety in adults and elderly subjects with insomnia disorder per DSM-5. A Bayesian adaptive design tested 6 doses of E2006 (1, 2.5, 5, 10, 15, 25 mg) or placebo. Subjects underwent 8-hr PSG for 2 consecutive nights starting at median habitual bedtime (MHB) calculated from a sleep diary completed for 7 days prior to the first screening/BL PSG night. After BL, eligible subjects were randomized (double-blind) to receive E2006 or placebo for 15 nights. All subjects then received single blind placebo for 2 nights to assess for rebound insomnia. Study drug was administered 30 m prior to MHB each night in the clinic, and PSGs were recorded for 8 h starting 30 m postdose on Days 1&2 and Days 14&15. Next-day residual sleepiness was assessed on clinic mornings with: Karolinska Sleepiness Scale (KSS), digit symbol substitution (DSST) and Reaction Time (RT). Subjects took E2006 or placebo 30 m before anticipated bedtime while at home during the treatment period. Sleep diaries were completed each morning. Safety was monitored throughout the study, including treatment emergent adverse events (TEAEs), ECGs, vital signs, clinical hematology, and blood chemistry tests. Sleep parameters were obtained for each PSG and averaged in pairs for the first 2 and last 2 treatment nights. Efficacy measures included: Sleep efficiency (SE) - total sleep time/time in bed X 100 (1□ efficacy endpoint); Latency to Persistent Sleep (LPS) - minutes from lights off to the first 30 s epoch of 20 consecutive epochs of non-wake; Wake After Sleep Onset (WASO) - minutes of wake from LPS to lights on.

Results: 616 subjects were screened, with 291 randomized (63.5% F, mean age 48 y). Mean (SD) BL Insomnia Severity Index score was 20±3, indicating moderate to severe insomnia. There were no important differences between treatment groups in any demographic parameter. At BL, overall mean SE was 65±12%; median LPS was 67±42 m; mean WASO was 109±43 m. 94.5% of E2006 subjects and 91.1% of placebo subjects completed the study. Only 1 subject (E2006 25 mg) discontinued due to an adverse event. Following treatment on Days 1&2, the LS mean treatment difference between E2006 and placebo for change from BL in SE was statistically significant for all groups (1 mg: 4.6; 2.5 mg: 4.4; 5 mg: 5.7; 10 mg: 8.1; 15 mg: 10.1; 25 mg: 10.1%, respectively) suggesting a dose-related increase in efficacy but with overlapping confidence intervals (CI). LPS decreased across all treatment groups, with change from BL ranging from –43 m for 1 mg to – 50 m for 25 mg compared to a placebo change of 23 m, with statistically significant decreases at 2.5 mg and higher. WASO also decreased in all treatment groups, with statistically significant decreases at 10 mg and higher compared to placebo. As with SE, CI overlapped among E2006 doses for LPS and WASO. For those doses that showed efficacy for sleep onset or sleep maintenance at Days 1&2, the initial treatment benefits were maintained over the 15-day treatment period. The changes from BL on subjective (KSS) and objective (DSST, RTI) assessments of next-day residual sleepiness were of small magnitude, and with few exceptions were not statistically different from placebo for E2006 treatment groups. The one notable exception was at 25 mg, where KSS showed a small but significant increase 1 h after waking both on Days 2&3 (LS mean difference 0.47; p<0.04) and Days 15&16 (0.68; p<0.01). TEAEs were more common in the E2006 groups, with higher rates (≥5% & 2x placebo) of somnolence, sleep paralysis, nightmares, and abnormal dreams. Somnolence appeared to be the only dose-related TEAE; of those subjects, 83% reported these events as mild, and 17% reported them as moderate. There were 2 SAEs (one placebo, one 25 mg). All adverse events except the SAE at 25 mg were considered mild or moderate.

Conclusions: The results of this study highlight the potential of this novel DORA for the treatment of insomnia disorder. The study drug was well tolerated, with a high rate of completion and adverse events generally rated as mild to moderate. Despite the small treatment group sizes, efficacy was demonstrated for both sleep initiation and sleep maintenance. The subjective and objective measures of next-day residual sleepiness support the potential clinical utility of E2006 for insomnia disorder.

Keywords: insomnia, orexin, Bayesian.

Disclosure: All authors are full-time employees of Eisai, Inc.

T71. HIV Risk Reduction With Buprenorphine-Naloxone or Methadone: Findings from A Randomized Trial

George Woody*, Douglas Bruce, P. Todd Korthuis, Sumedha Chhatre, Maureen Hillhouse, Petra Jacobs, James Sorenson, Andrew Saxon, Sabrina Poole, David Metzger, Walter Ling

University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania

Background: Research has shown that methadone maintenance reduces opioid use and is an effective HIV risk reduction intervention. This finding has been observed when methadone patients are compared to their community counterparts who are not in treatment and when opioid use during treatment is compared to pre- and post-treatment use. Significantly lower rates of opioid use have been observed when patients with regular methadone program attendance are compared to those with poor attendance, and when patients receiving minimal ancillary services are compared to those receiving more intensive services. Consistent with these reductions in opioid use, methadone maintenance markedly reduces opioid injection and needle sharing.

Methods: Secondary analysis from a study of liver enzyme changes in patients randomized to MET or BUP who completed 24-weeks of treatment and had 4 or more blood draws. The initial 1:1 randomization was changed to 2:1 (BUP: MET) after 18 months due to higher dropout in BUP. The Risk Behavior Survey (RBS) measured past 30-day HIV risk at baseline and weeks 12 and 24.

Results: Among 529 patients randomized to MET, 391 (74%) were completers; among 740 randomized to BUP, 340 (46%) were completers; 700 completed the RBS. There were significant reductions in injecting risk (p<0.0008) with no differences between groups in mean number of times reported injecting heroin, speedball, other opiates, and number of injections; or percent who shared needles, did not clean shared needles with bleach, shared cookers, or engaged in front/back loading of syringes. The percent having multiple sex partners decreased equally in both groups (p<0.03). For males on BUP the sex risk composite increased; for males on MET, the sex risk decreased resulting in significant group differences over time (p<0.03). For females, there was a significant reduction in sex risk (p<0.02) with no group differences.

Conclusions: These findings show marked and approximately equal reductions in injection and injection related risk among participants who remained in their assigned treatment condition and completed 4 or more blood draws over 24-weeks. Reasons for the small but significant increase in amphetamine use in BUP participants are unclear and may be an incidental finding as it has not been previously reported.

Keywords: HIV, Buprenorphine-Naloxone, Methadone.

Disclosure: Nothing to Disclose.

T72. Adjunctive Raloxifene Treatment Improves Attention and Memory in Men and Women with Schizophrenia

Thomas Weickert*, Danielle Weinberg, Rhoshel Lenroot, Stanley Catts, Ruth Wells, Ans Vercammen, Maryanne O'Donnell, Cherrie Galletly, Dennis Liu, Ryan Balzan, Brooke Short, Pellen Daniel, Jackie Curtis, Vaughan Carr, Jayashri Kulkarni, Peter Schofield, Cynthia Weickert

University of New South Wales, Randwick, Australia

Background: There is increasing clinical and molecular evidence for the role of hormones and specifically estrogen and its receptor in schizophrenia. A selective estrogen receptor modulator, raloxifene, stimulates estrogen-like activity in brain and can improve cognition in older adults. The present study tested the extent to which adjunctive raloxifene treatment improved cognition and reduced symptoms in young to middle age men and women with schizophrenia.

Methods: Ninety-five patients with a diagnosis of schizophrenia or schizoaffective disorder were recruited into a dual-site, thirteen week, randomized, double-blind, placebo-controlled, crossover trial of adjunctive raloxifene treatment in addition to their usual antipsychotic medications. Symptom severity and cognition in the domains of working memory, attention/processing speed, language and verbal memory were assessed at baseline, six, and thirteen weeks.

Results: Analyses of the initial six-week phase of the study using a parallel groups design (with 39 patients receiving placebo and 40 receiving raloxifene) revealed that adjunctive raloxifene treatment showed significant improvement relative to placebo in memory and attention/processing speed. There was no reduction in symptom severity with raloxifene treatment compared to placebo. There were also significant carryover effects, suggesting some cognitive benefits are sustained even after raloxifene withdrawal. Analysis of the 13-week crossover data revealed significant improvement with raloxifene only in attention/processing speed.

Conclusions: This is the first study to show that daily, oral adjunctive raloxifene treatment at 120 mg/day has beneficial effects in attention/processing speed and memory for both men and women with schizophrenia. Thus, raloxifene may be useful as an adjunctive treatment for cognitive deficits associated with schizophrenia.

Keywords: schizophrenia, estrogen receptor, cognition, raloxifene.

Disclosure: Nothing to Disclose.

T73. Telephone Administration of the CDR – Excellent Agreement with Face-to-Face Administration

Janet Williams*, Kristin Hannesdottir, Christopher Randolph, Elizabeth Eureyecko, Jessica Langbaum, Pierre Tariot, Martin Farlow, James Galvin, Carolyn Langlois, Cynthia Hunt, Tina Olsson, Michael Poole, Christopher Weber, Peter Boehm, Elan Cohen, Lori Garzio, Robert Alexander

MedAvante, Inc., Hamilton, New Jersey

Background: Large intra-subject variability is common in measures of neurocognition such as the CDR in clinical trials of mild cognitive impairment (MCI) and mild to moderate Alzheimer’s disease (AD). This may be due to baseline score inflation and lack of standardization among site raters. Remote administration would allow independent, blinded and calibrated raters to administer this outcome measure with minimized variability and high interrater reliability. Interview-based scales including the CDR, ADCS-ADL and FAQ lend themselves to administration by telephone because they do not require observation of subjects or manipulation of objects.

Methods: A non-interventional methodologic study tested the procedural validity of telephone administration of the CDR, ADCS-ADL and FAQ compared to face-to-face administration. Thirty-one subjects with MCI and 29 with mild to moderate AD and their caregivers were identified from the clinical populations served at the three investigative sites and interviewed by trained raters either remotely by telephone or face-to-face. Central raters traveled to each of the study sites to conduct the face-to-face interviews. The order of methodologies was counterbalanced; each rater conducted an equal number of interviews in each method, blinded to severity status. Subjects with mild or moderate AD were confirmed by the site as having an MMSE score between 16 and 25.

Results: Subjects ranged in age from 61 to 92 years (M=74), with 42% female. Caregivers were primarily spouses (73%) or children (16%), reporting an average of 6.5 days per week contact with the subject and an average of 12.7 hours of contact daily. The intraclass correlation coefficients (ICC) between phone and face-to-face raters’ scores were: CDR-Sum of boxes 0.91; CDR Global Score 0.83; ADCS-ADL Total score 0.95; FAQ Total score 0.92. Using dependent samples t-tests, there were no statistically significant differences between any of the scores when subjects were assessed by telephone compared to face-to-face.

Conclusions: Remote central ratings conducted via telephone by well-trained and calibrated raters of the CDR Global score, CDR Sum of boxes, ADCS-ADL, and FAQ were found to have excellent agreement with face-to-face administration. There were no statistically significant differences between any of these four scores when subjects were assessed by telephone compared to face to face, using dependent samples t-tests. Telephone administration of these scales allows for blinding of raters to study visit and prior scores, which eliminates potential sources of rater bias. In addition, central telephone ratings can be conducted with far fewer raters, reducing interrater variability and potentially increasing signal detection. Possible limitations of the study include the relatively small sample size across just three sites in the U.S., with just six raters.

Keywords: Dementia, validity, Alzheimer's, remote administration.

Disclosure: Several authors, as indicated, are employed by MedAvante.

T74. A Study of Swedish Massage Therapy for Generalized Anxiety Disorder

Mark Rapaport*, Pamela Schettler, Becky Kinkead, Erika Larsen, Sherry Edwards

Emory University School of Medicine, Atlanta, Georgia

Background: Although complementary and alternative medicine (CAM) therapies are popular and prevalent among patients with psychiatric disorders, there have been few carefully performed clinical trials to evaluate the efficacy of these modalities. In this NCCAM-funded pilot study we compared and contrasted the acute efficacy of a carefully defined Swedish Massage Therapy (SMT) protocol vs. a carefully defined light touch (LT) control condition. We hypothesized that SMT would cause a significant decrease in clinician-rated and self-report measures of anxiety.

Methods: 40 subjects with DSM-IV defined GAD (32 women and 8 men) with a mean age of approximately 36.5 years and a baseline Ham-A of 20.05 (3.34) for SMT and 19.58 (4.90) for LT were randomized to 12 acute treatment sessions. Raters masked to treatment condition performed Ham-A ratings after each session and Ham-D-17 ratings poor to the first session and at the end of the 12th session. The subjects rated themselves after each session with the QIDS and at baseline and after the 12th session with the STAI-state scale. The statistician was masked to treatment assignment while performing the analysis.

Results: The Ham-A scores decreased -11.67 (1.09) for the SMT group and -8.41 (1.01) for the LT group; t=2.19, df=106, p=0.03, ES=-0.690. Baseline STAI was 54.23 (10.97) for SMT vs. 51.56 (11.98) for LT and decreased by 14.85 (7.05) for SMT vs. 5.81 (16.81) for LT; t=-1.95, df=21, p=0.065; ES=-0.675. The baseline Ham-D was 16.71 (5.14) for SMT and 15.29 (4.51) for LT and decreased by -9.21 (5.73) for SMT and -3.71 (7.12) for LT; t=-2.34; df=29, p=0.027; ES=-0.843. The baseline QIDS was 10.62 (3.88) for SMT and 9.63 (3.99) for LT and decreased by -5.11 (0.67) for SMT and -2.33 (0.62) for LT; t=-3.04, df=103,p=0.003; ES=-0.956.

Conclusions: This pilot study suggests that SMT may be an effective acute treatment for individuals with GAD. SMT separated significantly from LT on the primary outcome measure the Ham-A and on both of the depression measures. There was a trend toward statistical significance on the STAI. This is the first study that employed SMT as a monotherapy for GAD and these results are promising but require confirmation in a larger trial.

Keywords: GAD, alternative medicine, massage, clinical trials.

Disclosure: Nothing to Disclose.

T75. Catch Me If You Can: How a Subject Registry Combines Voluntary, Investigator-based Use at Prescreen and Sponsor-mandated Use at Screen to Reduce Duplicate Enrollment

Thomas Shiovitz*, Marlene Zarrow, Sabrina Schoneberg, Lina Seikh

California Neuroscience Research, Sherman Oaks, California

Background: Duplicate and professional subjects (“professionals”) are a significant problem, especially in CNS studies. Subject registries attempt to reduce the number of professionals enrolling in clinical trials by identifying and eliminating them prior to randomization. When subject registries are integrated into protocols, investigators are mandated to use the registry at screen and professionals become screen failures. Yet, many professionals escape detection because not all sponsors use registries in their protocols. Therefore, we have worked with investigators in high-duplicate areas to enter all potential subjects into a registry during the prescreen process, regardless of whether the site is also using the registry at screen.

Methods: Participating clinical trial sites either entered subjects into CTSdatabase (a privately available subject registry) as a mandated protocol procedure at screen, as part of an investigator collaboration during the pre-screening process, or both. We targeted sites in regions identified by sponsors as high-duplicate areas for inclusion in the collaboration (e.g. “join us to help create a duplicate-free corridor in your area”). Potential subjects presenting at participating CNS clinical trial sites between October 31, 2011 and August 1, 2014, signed an IRB-approved authorization to allow certain identifiers (gender, initials, date of birth, last 4 of SSN, height, weight) and study indication to be entered into CTSdatabase. The data set was comprised of subjects who matched enough key identifiers to be a “virtually certain” match (<1 x 10-7 likely to have matched by chance) with a subject at a second, unique site. Subjects who returned to the same site (“same-site matches”) were removed from the data set.

Results: 10,029 subjects were entered at participating sites. 1703 same-site matches were excluded from the data set. 2115 potential duplicate subjects matched a majority of key identifiers, enough to be considered virtually certain matches. Of these virtually certain matched subjects, 322 attempted to screen or prescreen at a second site within 30 days, 532 within 60 days and 1108 within 180 days of presenting at the first site. Within a 60-day period, 43% of matching subjects travelled to sites that were more than 25 miles apart. 10 matching subjects traveled to sites that were over 100 miles apart. Subjects were prevented from entering studies not only by detecting matches between two prescreening sites or between two screening sites, but also by detecting matches between sites that only used the registry at prescreen and sites that only used the registry as part of a protocol.

Conclusions: Subject registries can attempt to address the professional subject problem in several ways: tracking potential subjects pre-screen, tracking those who have signed consent but have not yet been randomized and tracking those that try to screen elsewhere while still active in a study. Ideally, sponsors will add the use of a subject registry into their protocols, mandating use by all participating sites. However, this may not show matches among non-participating sites in a given area or among non-participating sponsors. As a result, some professionals may go undetected. We have encouraged investigators in high-duplicate areas to register all prescreening subjects in an attempt to identify and eliminate professionals. Sponsors may encourage such site participation by selecting sites which use a registry at prescreen and by reimbursing for costs associated with registry use. Until a subject registry is commonplace in all protocols, professionals can be tracked at both prescreen and screen; the overlap will bring together sites and sponsors that use a registry to eliminate the largest possible number of these problematic subjects.

Keywords: Duplicate subject, Professional subject, Clinical trial, Professional Patient.

Disclosure: Thomas M. Shiovitz, M.D. has ownership interest in CTSdatabase, LLC, the subject registry used to collect the match data contained in the poster.

T76. Baclofen as a Pharmacotherapy for the Treatment of Concurrent Alcohol and Nicotine Dependence: A Double-blind, Placebo-controlled, Randomized Trial

Mehdi Farokhnia*, Steven M. Edwards, Jared Bollinger, Jonathan Amodio, William H. Zywiak, Jennifer W. Tidey, Robert M. Swift, George A. Kenna, Lorenzo Leggio

National Institute on Drug Abuse, National Institutes of Health, Bethesda, Maryland

Background: Alcohol and nicotine use frequently co-occur and up to 75% of alcoholic smokers would require treatment for both dependencies. However, there is presently no approved single treatment that addresses both addictions concurrently. Combination of alcohol and nicotine provides additive neurochemical effects, which potentiate reinforcement for both substances and results in increased dopamine release in the nucleus accumbens. By stimulating GABAB receptors on the cell bodies of dopaminergic neurons and on the terminals of glutamatergic afferent neurons, baclofen may inhibit dopamine neurotransmission and, in turn, dopamine-mediated behaviors. Different clinical and preclinical studies suggest that the GABAB receptor agonist baclofen may be an effective pharmacotherapy for alcoholism or smoking, but no studies have investigated its effect in concurrent alcohol and cigarette use, which was the aim of the present study.

Methods: In this double-blind placebo-controlled randomized clinical trial, 30 treatment-seeking alcoholic smokers were randomized to receive either baclofen (80 mg/day) or placebo for 12 weeks. A sub-group of patients (n=18) participated in an alcohol cue-reactivity experiment as well. To be eligible, participants had to have DSM-IV diagnoses of both alcohol and nicotine dependence in addition to heavy use of both alcohol (≥4 and ≥5 standard drink units per day on average for women and men, respectively) and nicotine (≥10 cigarettes per day on average) during the last 90 days before screening. Patients were visited 7 times during the treatment phase. Measurements of alcohol and nicotine use (Timeline Follow-Back) and craving (various questionnaires) were taken at each visit.

Results: Baclofen-treated patients experienced significantly more decrease in the percent days of abstinence from alcohol-tobacco co-use compared to the placebo group (p=.004). Baclofen effect was moderated by alcohol dependence severity as more severely alcohol dependent participants showed significantly greater response to that (p<.001). During the course of this study, the percent days of alcohol-tobacco co-use declined in both groups, but this decline was greater after placebo than baclofen (p<.001). Analyses of the single use of each substance did not reveal an effect of baclofen vs. placebo on duration of abstinence from alcohol or nicotine. Reduction of alcohol craving scores was not significantly different between the two groups. By contrast, baclofen significantly reduced cigarette craving measured by Questionnaire on Smoking Urges-Brief (p=0.02) and slightly reduced the Smoking Visual Analogue Scale scores (p=0.05). Additionally, baclofen slightly decreased alcohol urge (p=.05) and significantly reduced salivation (p=.001) in the cue-reactivity sub-study although no medication × cue type interaction was found.

Conclusions: Baclofen’s effects on alcohol sensitivity and nicotine reinforcement may have been responsible, respectively, for the ability of baclofen to reduce drinking and smoking in this population of alcoholic smokers. Co-administration of alcohol and nicotine results in increased dopamine release, which may be blocked by baclofen, thus resulting in its ability to promote abstinence from both substances. Here we found that baclofen was more effective in increasing the days of abstinence from alcohol-tobacco co-use in those patients with a greater degree of alcohol dependence. In summary, this study provides preliminary evidence suggesting the potential efficacy of baclofen in the treatment of alcoholic smokers. However, the mixed results and the small sample require larger confirmatory studies.

Keywords: alcohol, nicotin, baclofen, GABAB.

Disclosure: Nothing to Disclose.

T77. Alzheimer’s Prevention Registry: A Shared Resource to the Scientific Community to Facilitate Enrollment in Studies

Pierre Tariot*, Jessica Langbaum, Eric Reiman, Nellie High, Paul Aisen, Marilyn Albert, Meryl Comer, Jeffrey Cummings, Jennifer Manly, Ronald Petersen, Reisa Sperling, Gabrielle Strobel, Michael Weiner

Banner Alzheimer's Institute, Phoenix, Arizona

Background: Recruitment and enrollment into clinical trials is a major obstacle faced by researchers and study sponsors. It has been estimated that fewer than 10% of Americans participate in clinical trials, mostly due to lack of awareness about study opportunities, resulting in approximately 80% of research studies failing to meet their enrollment goals in the stated timeframes. Given the growing number of preclinical treatment trials being conducted or in the planning stages, we developed a web-based Alzheimer’s Prevention Registry (“Registry”) to help studies make enrollment more efficient and timely. The Registry, which aims to enroll at least 250,000 individuals, is intended to provide a shared resource to the AD scientific community to facilitate enrollment in preclinical studies and to complement and enhance local recruitment efforts.

Methods: Prior to creating the Registry, a national survey of 1,024 adults age 18-75 was conducted for to help guide Registry development and outreach strategy. Interested adults of all ages, with and without memory and thinking problems, are eligible to join at www.endALZnow.org. Based on lessons learned from the Arizona Alzheimer’s Research Registry and modeled after other web-based research registries, this Registry was purposely designed to have a low threshold of commitment at entry. At enrollment, individuals are asked to provide their email address; after enrollment they can complete additional contact and demographic information at their convenience and discretion. Enrollees receive regular email communication to keep them apprised of the latest news in Alzheimer’s prevention research. In addition, enrollees receive email notifications when study opportunities become available in their communities, with information on whom to contact to explore the possibility of their participation. A/B testing is used to refine messaging and collection of demographic information to increase enrollment. Beginning in Q3 2014, a Researcher/Study Opportunities Portal will be available which will allow reporting of de-identified study enrollment metrics to help demonstrate the utility and impact of the Registry.

Results: The national survey found that 60% were likely to join the Registry, with the biggest motivators to joining being to prevent themselves (73%) or a loved one (77%) from developing AD and believing that clinical trials are key to medical breakthroughs (72%). The Registry was launched in May 2012 and underwent a website re-imagination beginning in July 2013. A/B testing continues to be used to refine the website and data collection methods. As of July 2014, over 40,000 individuals have joined the Registry. 95% of enrollees have provided some additional contact and demographic information, with 85% of enrollees answering all questions. Registrants are predominantly women (78%), report a family history of dementia (72%) and have no diagnosis of cognitive impairment (95%). 36% of enrollees are between the ages of 46-60; 37% are between the ages of 61-75. To date, the most successful outreach mechanisms include news articles, TV and radio interviews, and targeted partnerships that allow interested individuals to sign up directly through an email or website. Paid online advertising and search engine marketing have a generated relatively fewer enrollments.

Conclusions: The Registry is an engaged community of individuals who want to stay abreast of the latest in Alzheimer’s news and scientific advances, and to be connected to research studies taking place in their communities. The Registry has been well-received and enrollment continues to increase; results from A/B testing and the impact that website modifications had on enrollment will be discussed. Preclinical treatment trials such as the ADCS “A4” Trial, the Takeda/Zinfandel TOMMORROW study, and the Alzheimer’s Prevention Initiative (API) APOE4 Trial will use the Registry to aid with recruitment. The Registry is open to support enrollment for additional early-stage trials. The planned Portal will enable us to report the success of Registry in facilitating enrollment into these and other studies. We continue to explore novel approaches for increasing enrollment and engagement of enrollees, as well as collaborating with researchers to help promote relevant studies taking place in their catchment areas.

Keywords: clinical trials, prevention, Alzheimer's, registry.

Disclosure: Consulting fees: Abbott Laboratories, AbbVie, AC Immune, Boehringer-Ingelheim, California Pacific Medical Center, Chase Pharmaceuticals, CME Inc., Medavante, Otsuka, Sanofi-Aventis.

T78. Treadmill Exercise Improves Fitness and Reduces Craving and Use of Cocaine in Individuals with Cocaine and Tobacco-use Disorder

Richard De La Garza*, Daisy Thompson-Lake, Colin Haile, Joel Eisenhofer, Thomas Newton, Jin Yoon, James Mahoney

Baylor College of Medicine, Houston, Texas

Background: Rodent research shows that wheel running is reinforcing, and reduces cocaine self-administration, cocaine cue-induced reinstatement, and amphetamine-induced dopamine release. Exercise has been shown to be reinforcing in humans and may be useful as a treatment for substance use disorders. For this study, we sought to evaluate the effects of treadmill exercise on basic fitness measures and objective and subjective measures of cocaine use and craving in human volunteers.

Methods: Participants (N=24) included treatment-seeking individuals with cocaine and tobacco-use disorder (cigarette smokers). Participants were randomized to either running or walking (30 min per session, 3 times per week) or sitting (control condition; same amount of time each day) for 4 consecutive weeks. The exercise intensity for Runners and Walkers was individually calculated as a percent of maximum heart rate (HR) derived from a treadmill test conducted during screening. Computerized cognitive behavioral therapy and contingency management were also given to all participants. Fitness measures included changes in body weight and resting HR. The objective measure for cocaine use was urine benzoylecgonine (BE) and subjective measures included changes in craving and motivation to quit.

Results: There were no significant differences in demographic or drug use variables among Runners (N=10), Walkers (N=7) and Sitters (N=7). On average, participants were Black (71%), male (80%), 44.7±1.1 (Mean±S.E.M) years of age, and reported using cocaine for 19.7±8.2 years and 15.3±1.6 of the last 30 days. The exercise program was very well tolerated with>90% retention rate and no treatment-related adverse events. Several metrics indicated clear distinctions among Runners vs. Walkers vs. Sitters, including mean distance covered (1.89±0.09, 1.19±0.14, 0±0 miles, respectively; p<0.0001) and calories burned (274.4±32.9, 134.6±10.5, 73.5±0; p<0.0001) during sessions. Remote physiological monitoring via Bioharness during sessions showed that the groups also differed according to mean maximum HR (p<0.0001), respiration (p<0.0001), and core body temperature (p<0.0001). Within-session changes (post-exercise minus pre-exercise) included most notably reduction in craving for cocaine (VAS scale of 100)(-10.2±5.7, −10.5±5.8,+2.4±2.9; p=0.19). Across the 4-week study, exercise improved fitness measures including reducing body weight (−11.3±8.5, −4.0±2.9,+2.7±2.3 pounds; p=0.28) and decreasing resting HR (−3.3±4.1,+5.2±2.6,+9.1±3.0 bpm; p=0.05). Though not statistically significant, exercise improved abstinence from cocaine (number of negative BE urines out of 12) (4.1±1.6, 4.9±1.9, 0.9±0.6; p=0.19) and increased self-reports of NO cocaine use in last 24h (number of days out of 12) (8.2±1.4, 9.3±0.5, 4.9±1.5; p=0.07). At the end of the study, exercise reduced daily craving for cocaine (−7.9±3.2, −6.9±3.3, −0.6±1.8; p=0.18), but unexpectedly reduced Motivation to Quit cocaine (scale of 100) (−30.2±7.9,+2.14±9.1, −9.1±9.1; p=0.04). In general, similar reductions in nicotine use and craving were observed (to be reported elsewhere).

Conclusions: To our knowledge, this is the first study to evaluate the effects of a multi-week exercise program in individuals with cocaine and tobacco-use disorder. The data clearly show significant improvements in basic fitness measures and several indices reveal non-significant reductions in cocaine and use and craving. Despite marked statistical differences between Runners and Walkers for distance covered, calories burned, and maximum HR and respiration, there were few instances where running proved more effective than walking for reducing cocaine use or craving. Taken together, the data from this study provide compelling evidence for the efficacy of exercise for improving fitness and reducing cocaine use. Funding: NIH DA030722 to RD.

Keywords: Cocaine, Exercise, Smoking.

Disclosure: Nothing to Disclose.

T79. A Pharmacogenetics Supported Clinical Trial to Delay Onset of Mild Cognitive Impairment due to Alzheimer’s Disease Using Low Dose Pioglitazone: The Tommorrow Study

Kumar Budur*, Ferenc Martenyi, Kathleen A. Welsh-Bohmer, Daniel K. Burns, Carl Chiang, Janet O'Neil, Grant Runyan, Jennifer Schuster, Donna G. Crenshaw, Michael W. Lutz, Craig A. Metz, Ann M. Saunders, Deborah Yarbrough, David Yarnall, Eric Lai, Stephen K. Brannan, Allen D. Roses

Takeda Development Center Americas, Inc., Deerfield, Illinois

Background: The increasing number of individuals at risk for developing late onset Alzheimer’s disease (LOAD) has shifted attention from treatment to delaying symptom onset. Research is focusing on early intervention in the AD cognitive decline continuum. Length variation of a polyT tract (rs10524523) located within intron 6 of the translocase of outer mitochondrial membrane 40 homolog (TOMM40) gene (chr19:44,899,792-44,899,826, human genome reference assembly GRCh38/hg38), has been associated with LOAD age of onset. TOMM40 is adjacent to, and in linkage disequilibrium with, the apolipoprotein E (APOE) gene. A biomarker risk assignment algorithm (BRAA) will identify individuals at high risk for developing mild cognitive impairment (MCI) due to AD during the duration of the trial (5 years). The BRAA incorporates an individual’s age (from 65 to 83 years) and TOMM40 and APOE genotypes.

Methods: AD and evaluate the efficacy of low-dose pioglitazone to delay onset of MCI due to AD in cognitively normal, high-risk individuals. Low-risk subjects will be assigned to placebo; high-risk subjects will be evenly randomized to either low-dose pioglitazone or placebo. The study will enroll approximately 5800 subjects (ages 65-83) and will apply operationalized criteria for MCI due to AD, a primary endpoint event. The anticipated treatment period is 4 years, ending when a target number (410 in the high-risk group) of primary endpoint diagnoses have been reached. Along with the Clinical Dementia Rating scale, key assessments that enable the diagnosis are 12 neuropsychological measures representing 5 key cognitive domains affected in early AD. A separate neuropsychological instrument validation strategy will ensure that the test measures perform consistently across cultures and languages.

Results: The study was initiated in summer 2013 and is currently recruiting in the US, UK and Australia. To date, over 5000 subjects have been screened and over 760 subjects randomized. Additional sites in Germany, Switzerland, Italy and Russia are expected to begin recruitment in early 2015.

Conclusions: The study was designed with input from international experts and finalized following discussions with US and EU regulatory authorities. It represents a unique opportunity to qualify the BRAA to stratify risk of developing MCI due to AD in the next 5 years and to explore therapeutic intervention in the earliest phase of the AD continuum.

Keywords: Biomarker risk assignment algorithm, mild cognitive impairment due to AD, cognitive domains in AD.

Disclosure: Employed by Takeda Development Center Americas, Inc., Deerfield, IL.

T80. Effects of Levomilnacipran ER on Motivation/Energy and Functioning in Adults with Major Depressive Disorder: Post Hoc Analysis of a Phase 3 Trial

Alan Lipschitz*, Carl Gommoll, Changzheng Chen, Michael E. Thase

Forest Research Institute, Jersey City, New Jersey

Background: In patients with major depressive disorder (MDD), reduced motivation and energy have been associated with greater disease severity and poorer quality of life. The Motivation and Energy Inventory–Short Form (MEI) is an 18-item instrument that has been validated in MDD patients for assessing energy-related symptoms and the effects of medications on these symptoms. Levomilnacipran extended-release (ER) is a serotonin and norepinephrine reuptake inhibitor that is approved for the treatment of MDD in adults. in vitro, this drug has shown greater potency for inhibition of norepinephrine versus serotonin reuptake. Dysregulation of the noradrenergic system may be associated with symptoms related to motivation, fatigue, and energy, which may impact functioning and the ability to perform daily activities. Results from an 8 week, double-blind, placebo-controlled, phase 3 study showed that patients with MDD who received levomilnacipran ER 40 120 mg/d experienced statistically significant mean improvements relative to placebo in MEI total score (P<.05) and subscale scores (Cognitive/Mental Energy, Social Motivation; both P<.05). A post hoc analysis of data from this study was conducted to further explore the relationship between motivation/energy and functional impairment.

Methods: Patients were stratified by median baseline MEI total score to identify those with “lower” and “higher” levels of motivation/energy (total score ≤28 and>28, respectively). In each subgroup, change from baseline to Week 8 in Sheehan Disability Scale (SDS) total and subscale scores (Work/School, Social Life, Family/Home Life) were analyzed using a mixed-effects model for repeated measures (MMRM). The percentage of patients who responded to treatment, defined as ≥50% improvement from baseline in MEI total score, was also analyzed; the odds ratio (OR) for MEI response (levomilnacipran ER vs placebo) was based on a logistic regression model. In both MEI responders and nonresponders, changes in SDS total and subscale scores were analyzed using the MMRM approach.

Results: Patients with lower motivation/energy (ie, lower MEI total scores) at baseline (n=214) showed greater functional impairment relative to patients with higher motivation/energy (n=214); baseline SDS total scores in the higher and lower motivation/energy subgroups were approximately 21 and 18, respectively. In patients with lower motivation/energy at baseline, least squares mean differences (LSMDs) between levomilnacipran ER and placebo were significant for changes in SDS total score (-3.9, P<.001) and subscale scores (Work, -1.3, P=.002; Social, 1.2, P=.002; Family, -1.5, P<.001). These differences were not significant in the subgroup with higher motivation/energy at baseline. The percentage of patients who met the criteria for MEI response was greater with levomilnacipran ER than placebo, although the between-group difference did not reach significance (58.4% vs 47.0%; OR=1.51; P=.058). In MEI responders, however, levomilnacipran ER-treated patients showed significantly greater improvement than placebo-treated patients on the SDS total score (LSMD=-2.6, P=.007) and the Work subscale score (LSMD=-1.0, P=.007); differences on Social (LSMD=-0.6, P=.095) and Family (LSMD=-0.7, P=.052) subscale scores did not reach statistical significance. Changes in SDS total and subscale scores were not significant between treatment groups in the MEI nonresponder group.

Conclusions: The results of this post hoc analysis indicate that patients with lower motivation/energy at baseline experienced significant mean improvements in functional impairment after 8 weeks of flexible-dose treatment with levomilnacipran ER 40-120 mg/d. In patients who experienced ≥50% improvement from baseline in MEI total score, there was significantly greater improvement in functional impairment with levomilnacipran ER than with placebo. Together, these results suggest that there may be a relationship between motivation/energy and functional impairment in adults with MDD, and that in patients treated with levomilnacipran ER, improvements in motivation/energy may be associated with concomitant improvements in functional impairment.

Keywords: Major depressive disorder, serotonin and norepinephrine reuptake inhibitor, MEI, levomilnacipran.

Disclosure: A. Lipschitz, C. Gommoll, and C. Chen are employees of Forest Research Institute, a subsidiary of Actavis, plc. Dr. Thase has consulted with Forest, been a member of Advisory Boards pertaining to levomilnacipran, and received research funding from Forest.

T81. Smoking Cessation Through Reduction: Does It Enhance or Diminish Successful Quitting?

Charles Wilcox*, Daniel Grosz, My-Linh Tong, Judy Morrissey, Don De Francisco, Kimberly Guevarra, Nader Oskooilar

Pharmacology Research Institute, Newport Beach, California

Background: A 2007 U.S.-based population survey reported that more than half of those motivated to quit smoking wished to use a Reduce-to-Quit approach. Similarly, U.K. statistics (2009) indicated that only 12% of smokers desiring to stop smoking were willing to do so abruptly. The effectiveness of a Reduce-to-Quit strategy using varenicline had not been previously evaluated in the context of a multi-centered placebo-controlled clinical trial.

Methods: Our research center, along with more than 70 other study sites, was involved in the enrollment of more than 1,400 subjects into this double-blind study. We are reporting (only) on data generated by, and analyzed at, Pharmacology Research Institute. There was a (3-10 day) screening phase into which 49 adult smokers were entered and five (5) were excluded. Subjects meeting all of the entry criteria were randomly assigned, on a one-to-one ratio (i.e., n=22 per group) to varenicline or placebo for a 24-week, two-stage treatment phase. During the first 12-weeks (“reduction phase”), the smokers made incremental efforts to reduce their smoking. During the subsequent 12-week treatment period (“abstinence phase”), participants were encouraged and counseled to be abstinent from smoking. Active treatment was concluded at Week-24; subjects then entered the 28-week post-treatment “follow-up” phase, eventually completing the study at Week-52. Brief, ten-minute, smoking cessation counseling sessions were integrated into each visit, beginning at baseline (Week-0). Successful cessation was pre-defined as end-exhaled carbon monoxide (CO) measurement<10 ppm, plus subject reports via the Nicotine Use Inventory (NUI).

Results: Notwithstanding the very small sample size, clinically and statistically significant results were demonstrated beginning at Week-12 (p<.01) and at numerous time points throughout the 24-week active treatment phase. The Week-28 analysis of time to “first quit” incidence demonstrated a 71% success rate for the varenicline group, compared to 32% for the placebo group (p<.01). The sustained “permanent quit” analysis also demonstrated superior efficacy for varenicline (57%) versus placebo (26%), as defined and demonstrated by a non-relapsing, successfully sustained quit outcome (p<.05). We also statistically analyzed nine baseline variables to investigate their potential impact on subjects’ cessation efforts. Three of them were strongly and positively associated with successful and sustained cessation: female gender (p<.05), fewer years of smoking (p<.01) and a higher age when first began smoking (p<.01). Additionally, the older the age of the study participant also demonstrated a statistical trend toward higher success of quitting (p=.08).

Conclusions: At our research center we saw strong clinical and robust statistical evidence that a strategy of Smoking Cessation Through Reduction can be very effective. Additionally, the aforementioned results provide compelling evidence that when varenicline is used with brief counseling sessions and a reduce-to-quit approach, it is not only effective, its efficacy may be enhanced.

Keywords: Smoking Cessation, Nicotine, Addiction, Varenicline.

Disclosure: Nothing to Disclose.

T82. Efficacy and Safety of Vilazodone in Generalized Anxiety Disorder: A Randomized, Double-blind, Placebo-controlled Trial

Michael E. Thase*, Maju Mathews, Giovanna Forero, Rene Nunez, Changzheng Chen, Carl Gommoll, Suresh Durgam

University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania

Background: It is estimated that 6.8 million adults in the US have generalized anxiety disorder (GAD). Characterized by excessive and uncontrollable worries about everyday events, GAD affects women twice as often as men and frequently occurs comorbidly with mood disorders, including depression. Vilazodone is a selective serotonin reuptake inhibitor (SSRI) and 5-HT1A receptor partial agonist that is FDA-approved for the treatment of major depressive disorder (MDD) in adults. The efficacy, safety, and tolerability of vilazodone in the treatment of GAD were evaluated in a Phase III clinical trial.

Methods: This was a multicenter, randomized, double-blind, placebo-controlled, parallel-group, fixed-dose, Phase III study to evaluate the efficacy, safety, and tolerability of vilazodone for the treatment of GAD (NCT01629966). Patients (18-70 years of age) met Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) criteria for GAD, with a Hamilton Anxiety Scale (HAMA) total score ≥20. Patients were randomized (1:1:1) to vilazodone 20 mg/day, 40 mg/day, or placebo. The study had a 7-day screening period, 8 weeks of double-blind treatment, and a 1-week, double-blind down-taper safety period. The primary and secondary efficacy parameters were change from baseline to Week 8 in HAMA and Sheehan Disability Scale (SDS) total scores, respectively. Safety assessments included adverse event (AE) recording, clinical laboratory and vital sign measures, electrocardiograms, and the Changes in Sexual Function Questionnaire (CSFQ). Safety analyses were based on the Safety Population (all patients who received ≥1 dose of study drug). Efficacy analyses were based on the Intent-to-Treat (ITT) Population (patients in the Safety Population who had a baseline and ≥1 postbaseline HAMA assessment) and were conducted using a mixed-effects model for repeated measures (MMRM); P values were adjusted for multiple comparisons using a matched parallel gatekeeping procedure.

Results: The Safety Population comprised 673 patients (placebo=221, vilazodone 20 mg=227, vilazodone 40 mg=225); mean age was approximately 40 years,>60% were women, and>70% were white. The study was completed by 81% of placebo, 77% of vilazodone 20-mg, and 71% of vilazodone 40-mg patients. A greater percentage of vilazodone 40-mg patients than placebo patients discontinued prematurely (P<.05); discontinuation due to AEs occurred in 13% of vilazodone 40-mg patients versus 5% of placebo patients (P<.05), and in 8% of vilazodone 20-mg patients. The (LS) mean change from baseline to Week 8 in HAMA total score was -11.7, 13.0, and -13.5 for placebo, vilazodone 20 mg, and vilazodone 40 mg, respectively; the LS mean difference (LSMD) versus placebo was statistically significant for the vilazodone 40-mg group ( 1.8 [P=.031]), but not the 20-mg group (-1.3 [P=.083]). Mean change from baseline to Week 8 in SDS total score was numerically greater in the vilazodone groups, but the LSMDs versus placebo did not achieve statistical significance for either vilazodone dose (20 mg: -1.4, P=.083; 40 mg: -1.5, P=.070). Treatment-emergent AEs (TEAEs) were reported in 62%, 72%, and 71% of patients in the placebo, vilazodone 20-mg, and vilazodone 40-mg groups, respectively. TEAEs reported at an incidence ≥10% for vilazodone 20 mg and 40 mg were nausea (24% and 26%), diarrhea (25% and 21%), and headache (14% and 11%). Overall mean changes in CSFQ scores from baseline to Week 8 were similar for vilazodone 20-mg and 40-mg doses (1.9 and 1.5) and placebo (1.7). For most laboratory parameters and vital signs, the incidence of shifts from normal values at baseline to high values during treatment was small and similar among groups. Potentially clinically significant changes in total cholesterol and triglyceride levels were seen in>10% of patients in all treatment groups. No QTC interval>500 msec was noted in any group.

Conclusions: Improvement in HAMA total score from baseline to Week 8 was significantly greater in the vilazodone 40 mg/day group compared with placebo, indicating greater reduction in anxiety symptoms for patients in this dose group. The safety and tolerability profile of vilazodone in this study was similar to prior vilazodone studies in patients with MDD. Changes in sexual function were similar for vilazodone and placebo.

Keywords: vilazodone, generalized anxiety disorder.

Disclosure: Supported by funding from Forest Laboratories, Inc., a subsidiary of Actavis plc. Michael E. Thase has been a consultant to, an advisory board member of, and received research funding from Forest Laboratories, Inc., a subsidiary of Actavis plc. Maju Mathews, Giovanna Forero, Rene Nunez, Changzheng Chen, Carl Gommoll, Suresh Durgam are employees of Forest Research Institute, a subsidiary of Actavis plc.

T83. An 8-week, Randomized, Double-blind, Placebo-controlled Trial of Adjunctive Ziprasidone in Patients with Major Depressive Disorder Receiving Treatment with Escitalopram

George Papakostas*, Michaela B. Swee, Lee Baer, Richard C. Shelton

Massachusetts General Hospital, Boston, Massachusetts

Background: There are several pharmacologic and clinical properties of ziprasidone, an atypical antipsychotic agent, that warrant its testing as adjunctive therapy in clinical trials of MDD. Unfortunately, however, double-blind, placebo-controlled trials of ziprasidone augmentation for MDD have not been conducted to date. The present work reports the results of an 8-week, randomized, double-blind, placebo controlled trial of ziprasidone augmentation of the selective serotonin reuptake inhibitor (SSRI) escitalopram.

Methods: This study was an 8-week, randomized, double-blind, parallel, placebo controlled trial of ziprasidone augmentation of escitalopram for MDD patients who were non-remitters following an 8 week, prospective, open-label, flexible-dose trial of escitalopram.

Results: 531 subjects were screened across three sites (Massachusetts General Hospital, University of Alabama in Birmingham, Vanderbilt University), resulting in 458 (86.2%) MDD outpatients meeting eligibility for enrollment in the first phase of the study, an 8-week, open-label trial of flexible-dose escitalopram. Of these patients, 311 (67.9%) completed the first phase of the study. 139 (98 women, mean age 44 years, mean escitalopram dose 20mg, mean HAMD17 score of 17) outpatients were randomized into phase 2 – a randomized, double-blind, adjunctive placebo-controlled trial of ziprasidone (20-80mg po BiD) augmentation of escitalopram. At the end of the study, reduction in HAMD-17 scores was significantly greater for adjunctive Ziprasidone- than adjunctive Placebo-treated patients (p<0.05). Discontinuation due to intolerance was higher among ziprasidone-treated patients (p<0.05).

Conclusions: In the present trial, Ziprasidone proved to be an efficacious adjunctive treatment strategy for MDD.

Keywords: ziprasidone, escitalopram, augmentation, adjunctive.

Disclosure: Dr. George Papakostas: Dr. Papakostas has served as a consultant, received honoraria and research support (awarded to MGH not to him) from Pfizer Inc who provided free ziprasidone and placebo for the study. Forest pharmaceuticals provided free escitalopram for the study. The study was funded by the NIMH. Dr. Richard C. Shelton: Consultant: Bristol-Myers Squibb Company; Cerecor, Inc.; Cyberonics, Inc; Forest Pharmaceuticals; Janssen Pharmaceutica; Medtronic, Inc; Naurex, Inc; Pamlab, Inc; Pfizer, Inc; Ridge Diagnostics; Shire Pic; Takeda Pharmaceuticals. Grant/research support: Assurex Health; Bristol-Myers Squibb; Elan, Corp.; Forest Pharmaceuticals; Janssen Pharmaceutica; Jazz Pharmaceuticals; Naurex, Inc.; Novartis Pharmaceuticals; Otsuka Pharmaceuticals; Pamlab, Inc.; Takeda Pharmaceuticals. Dr. Lee Baer: None.

T84. Categorical Improvement Across Mania Symptoms: Pooled Analyses of Cariprazine Phase II/III Trials

Stephen Zukin*, Kaifeng Lu, Adam Ruth, Marc Debelle, Krisztián Nagy, Suresh Durgam, Joseph R. Calabrese

Johns Hopkins University, Baltimore, Maryland

Background: Cariprazine is a potent dopamine D3 and D2 receptor partial agonist antipsychotic with preferential binding to D3 receptors. The efficacy and tolerability of cariprazine in the treatment of bipolar mania are supported by 3 Phase II/III studies. The Young Mania Rating Scale (YMRS), which was the primary efficacy measure in these studies, evaluates a broad range of mania symptoms: elevated mood, increased motor activity-energy, sexual interest, sleep, irritability, speech, language-thought disorder, thought content, disruptive-aggressive behavior, appearance, and insight. Item scores range from either 0-4 (7 items) or 0-8 (4 items) with higher scores indicating greater severity. This analysis of pooled data from the cariprazine bipolar mania trials assessed clinically relevant symptom improvement in individual YMRS items by evaluating the percent of patients that shifted from a more severe symptom category at baseline to a less severe category at end of study.

Methods: The analyses were conducted on pooled data from 3 positive, Phase II/III, 3-week, double-blind, randomized, placebo-controlled trials in patients with manic or mixed episodes associated with bipolar I disorder (NCT00488618, NCT01058096, NCT01058668). All cariprazine doses were pooled for this analysis (2 studies: flexible-dosed cariprazine 3-12 mg/day; 1 study: fixed/flexible-dosed cariprazine 3-6 mg/day or 6-12 mg/day). Shift analyses were performed for all individual YMRS items using 2 different criteria: (1) patients shifting from moderate or worse severity (defined as YMRS item score ≥2 for items scored 0-4 and ≥4 for items scored 0-8) to mild/no symptoms (defined as score<2 for 0-4 items and<4 for 0-8 items); and (2) patients shifting from marked or worse severity (defined as score ≥3 on 0-4 items and ≥6 on 0-8 items) to mild/no symptoms (defined as score<2 for 0-4 items and<4 for 0-8 items). Odds ratios (OR) with P values for cariprazine vs placebo were analyzed based on a logistic regression model with study, treatment group, and baseline values as explanatory variables.

Results: The pooled intent-to-treat population comprised 429 placebo patients and 608 cariprazine patients. The percentage of patients that shifted from moderate or worse severity at baseline to mild/no symptoms at Week 3 was significantly higher for cariprazine vs placebo for all YMRS single items: elevated mood, 48.0% vs 35.1%; increased motor activity-energy, 48.0% vs 37.0%; sexual interest, 68.3% vs 51.5%; sleep, 58.1% vs 43.4%; irritability, 75.7% vs 53.8%; speech, 72.1% vs 52.6%; language-thought disorder, 55.8% vs 39.7%; thought content, 64.3% vs 47.9%; disruptive-aggressive behavior, 78.5% vs 62.2%; appearance, 75.3% vs 58.9%; and insight, 58.7% vs 41.2%. ORs for these outcomes ranged from 1.6 (increased motor activity-energy) to 2.7 (irritability); P<.001 for all items. In patients with marked or worse symptom severity at baseline, a significantly greater proportion of cariprazine vs placebo patients shifted to mild/no symptoms at Week 3 on 9 of the 11 YMRS items: elevated mood, 43.2% vs 31.4%; increased motor activity, 44.4% vs 34.0%; sexual interest, 59.8% vs 38.5%; sleep, 55.8% vs 44.7%; language-thought disorder, 49.4% vs 34.1%; insight, 50.3% vs 35.6%; irritability, 64.9% vs 32.0%; speech, 55.8% vs 39.9%; and thought content, 53.9% vs 35.6%. ORs ranged from 1.5 (increased motor activity-energy) to 4.0 (irritability) with P<.05 for all 9 items. For appearance (OR=1.7; P=.19) and disruptive-aggressive behavior (OR=0.8; P=.68), between-group differences in the proportion of patients who shifted from marked or worse severity to mild/no symptoms were not statistically significant; however, the sample size for these items was small as very few patients had marked or worse symptomatology at baseline on these items.

Conclusions: As measured by YMRS single items, a significantly greater proportion of cariprazine-treated patients compared with placebo-treated patients shifted from moderately or markedly severe symptoms at baseline to mild/no symptoms at Week 3 These results suggest that cariprazine is associated with clinically meaningful improvements across a broad spectrum of mania symptoms.

Keywords: cariprazine, mania, dopamine.

Disclosure: Supported by funding from Forest Laboratories, Inc., a subsidiary of Actavis plc, and Gedeon Richter Plc. Stephen Zukin was an employee of Forest Research Institute, a subsidiary of Actavis plc, at the time of the study. Kaifeng Lu and Suresh Durgam are employees of Forest Research Institute, a subsidiary of Actavis plc. Adam Ruth is an employee of Prescott Medical Communications Group, a contractor of Forest Research Institute, a subsidiary of Actavis plc. Marc Debelle and Krisztián Nagy are employees of Gedeon Richter Plc. Joseph R. Calabrese has served on advisory boards of Forest Laboratories, Inc. and Gedeon Richter Plc.

T85. The Efficacy and Safety of LY2940094, a Selective Nociceptin Receptor Antagonist, in Patients with Major Depressive Disorder: A Randomized, Double-blind, Placebo-controlled Study

Anke Post*, Trevor Smart, Judith Krikke, Jeffrey Witkin, Michael Statnick, Catherine Harmer, Gerard Dawson, Richard Mohs

Eli Lilly and Company, Windlesham, United Kingdom

Background: Nociceptin/Orphanin FQ (nociceptin) is a neuropeptide that acts as a natural ligand on the NOP receptor, which is a G protein-coupled receptor in brain regions associated with mood disorders. In non-clinical animal models, nociceptin modulates physiological functions and behaviors related to depression, stress and anxiety, feeding, locomotor activity, body temperature, substance abuse, memory, and pain. LY2940094 ([2-[4-[(2-chloro-4,4-difluoro-spiro[5H-thieno[2,3-c]pyran-7,4'-piperidine]-1'-yl)methyl]-3-methyl-pyrazol-1-yl]-3-pyridyl]methanol) is a potent and selective antagonist of human nociceptin NOP receptors that has provided preclinical evidence for anxiolytic- and antidepressant-like effects; reduced ethanol consumption; and reduced body weight.

Methods: This was an 8-week, double-blind, placebo-controlled, proof-of-concept study that evaluated LY2940094 as a novel oral medication for the treatment of patients with major depressive disorder (MDD). Eligible patients were adults (18-65 years of age) presenting with a new episode of MDD (as defined by the DSM-IV-TR) of at least 4 weeks duration; with a GRID Hamilton Depression Rating Scale 17-item (HAMD-17) total score ≥20; a Clinical Global Impression of Severity (CGI-S) score of ≥4; and a Hospital Anxiety and Depression Scale (HADS): Depression subscale score ≥11. The primary objective was to investigate the effects of LY2940094 (40 mg/day) versus placebo in patients with MDD as assessed by the HAMD-17 total score. Secondary objectives included evaluation of efficacy as measured by the Clinical Global Impression of Improvement (CGI-I), and to evaluate the safety and tolerability of LY2940094. In addition, the effect of LY2940094 on emotional processing as an indicator of improvement in MDD was evaluated utilizing the P1vital Oxford Emotional Test Battery (ETB). If the true effect size was 0.5, the study had 90% power to show that there was an 88% or larger posterior probability that LY2940094 had a greater reduction than placebo on the HAMD-17 total score. Analyses were conducted on both the full analysis set and the per protocol set.

Results: One hundred thirty-six patients were randomized to receive either LY2940094 (N=70) or placebo (N=66). Approximately, 60% of the patients were female, the average age was 40 years, and the average disease duration was 9 years. At baseline, depression severity as assessed by the HAMD-17 total score was 25+4.5 and 25±5.0 for LY2940094 and placebo, respectively. Ninety-nine patients completed the study; 49 received LY2940094 and 50 received placebo. At study endpoint least-squares (LS) mean changes from baseline in HAMD-17 total scores in the full set analysis were -11.4 for LY2940094 and -9.8 for placebo; 1.54 difference relative to placebo, Probability (LY2940094 HAMD-17<Placebo HAMD-17)=0.83; i.e., there was an 83% probability that LY2940094 has greater reduction in HAMD-17 total scores than placebo. For the Mood item on the HAMD-17, LS mean changes from baseline were -1.7 for LY2940094 vs. -1.1 for placebo; 0.61 difference relative to placebo, Probability (LY2940094 Mood<Placebo Mood)=0.99. In the per protocol set, LS mean changes from baseline in HAMD-17 total scores were -12.2 vs. -10.1, respectively, for LY2940094 vs. placebo (2.1 difference relative to placebo, Probability (LY2940094 HAMD-17<Placebo HAMD-17)=0.89, which met the criterion of a probability ≥ 0.88). On average, patients were improved with LS mean CGI-I scores of 2.4 for those treated with LY2940094 and 2.7 for those treated with placebo; 0.25 difference relative to placebo, Probability (LY2940094 CGI-I score<Placebo CGI-I score)=0.85. For the ETB assessment, the LS mean difference in accuracy in identifying faces with positive emotions compared to negative emotions at Week 1 was 17.4 for LY2940094 and 14.5 for placebo, giving a difference of 2.89, for this comparison Probability (LY2940094>Placebo)=0.89. Rates for study discontinuation due to adverse events were 8.7% vs. 3.0%, respectively, for patients treated with LY2940094 or placebo. The most common (≥ 10%) treatment-emergent adverse events reported for LY2940094 vs. placebo were headache (23.2% vs. 18.5%, p=0.531) and nausea (10.1% vs. 1.5%, p=0.063).

Conclusions: This study provided some evidence for an antidepressant effect of a 40 mg once daily dosing of LY2940094 versus placebo based on the change from baseline to Week 8 of the HAMD-17 total score. This finding was supported by the CGI-I endpoint score. Emotional processing assessment at Week 1 revealed a positive outcome on facial recognition tasks as well as on other parameters in favor of LY2940094. LY2940094 was safe and well tolerated in this first study. Overall, these are the first human data providing evidence that the blockade of NOP receptor signaling could represent a promising strategy for the treatment of patients with MDD.

Keywords: major depression, NOP receptor antagonist, nociceptin, emotional processing.

Disclosure: This study was sponsored by Eli Lilly and Company, Indianapolis, IN USA Anke Post, Judith Krikke, Jeffrey Witkin, Michael Statnick and Richard Mohs are full-time employees of Eli Lilly & Company. Catherine Harmer and Gerard Dawson (CSO of P1Vital) were consultant to the study.

T86. Dasotraline as a Novel DAT/NET Inhibitor for the Treatment of Attention-Deficit/Hyperactivity Disorder: A Randomized, Double-Blind, Placebo-Controlled, Proof-of-concept Trial in Adults

Kenneth Koblan, Seth Hopkins, Kaushik Sarma, Fengbin Jin, Robert Goldman, Antony Loebel, Scott Kollins*

Duke University School of Medicine, Durham, North Carolina

Background: Dopamine and norepinephrine are associated with the pathophysiology of ADHD, and drugs that facilitate synaptic concentrations of dopamine and norepinephrine are clinically useful in the pharmacological management of ADHD symptoms. Here we hypothesized that the new chemical entity dasotraline, at doses maintaining steady-state inhibition of both dopamine and norepinephrine transporters, would be a novel pharmacological approach to the management of ADHD symptoms. Unique relative to currently approved ADHD drugs, dasotraline (SEP-225289) has a slow (2 to 3 days) elimination half-life in humans, and achieves steady state plasma concentration by 2 weeks of daily dosing. Dasotraline was demonstrated to occupy dopamine transporters (DAT) preferentially over serotonin transporters following single oral doses in a human PET study (DeLorenzo et al., J Nucl Med, 2011), and dasotraline inhibition of norepinephrine transporters (NET) was also anticipated based on in vitro and in vivo pharmacological measurements.

Methods: Dasotraline doses were selected to achieve and maintain, throughout the 24-hour dosing interval, steady-state plasma concentrations above 4 ng/mL, corresponding to an expected DAT occupancy level above 50%. Adults (N=341) with ADHD (DSM-IV-TR criteria) were randomized 1:1:1 to 4 mg/day, 8 mg/day, or placebo for a 4-week treatment period and followed for a 2-week discontinuation period. The primary efficacy endpoint of change from baseline in ADHD RS-IV with adult prompts total score was analyzed as a mixed model for repeated measures (MMRM). The Hochberg procedure was utilized to adjust multiple comparisons. Dasotraline concentrations were sampled weekly and modeled with a one-compartment population PK model with sequential zero-order followed by first-order absorption and dual (nonlinear and linear) elimination. Norepinephrine metabolite DHPG concentrations were modeled as a power function of the time-matched dasotraline concentrations as derived from the PK model. Population PK/PD modeling of individual patients' improvements (ADHD RS-IV with adult prompts change from baseline) used individuals' average dasotraline concentrations (Cav) in a sigmoid Emax time-course model, with a maximum improvements (Emax) as linear function of Cav.

Results: The reduction in ADHD RS-IV with adult prompts total score was superior for dasotraline 8-mg (LS mean difference=-4.18, adjusted p=0.019) and numerically better for dasotraline 4-mg (LS mean difference=-2.68, adjusted p=0.076) compared with placebo at the 4-week endpoint. Both 4-mg and 8-mg demonstrated statistically significant reductions in CGI-S scores compared with placebo (p=0.021, p=0.013, respectively, at Week 4). The most frequent adverse events reported were insomnia, decreased appetite, nausea, and dry mouth, consistent with DAT/NET pharmacology. Discontinuations due to treatment-emergent adverse events were 1.8%, 11.2% and 29.7% of subjects in the placebo, 4-mg and 8-mg treatment groups, respectively. Dasotraline concentrations matched population PK model predictions, reached steady state by 2 weeks, and indicated a mean half-life of 47 hours. Concentrations of the norepinephrine metabolite DHPG indicated central NET inhibition was achieved at both dose levels within the first days of dosing. A population PK/PD model adequately characterized dasotraline-concentration dependent improvements in ADHD RS-IV and compared favorably with observed LSMean differences. Model-based clinical trial simulations of Phase 3 trial designs were performed to select dose, duration, and sample sizes for the continued development of dasotraline as a novel treatment for ADHD.

Conclusions: Dasotraline demonstrated statistically and clinically meaningful effects in adults with ADHD, with a dose- and concentration-response relationship supporting pharmacological activity in ADHD. These results support the concept that maintaining constant, steady-state inhibition of both dopamine and norepinephrine transporters is a novel pharmacological approach to the management of ADHD symptoms.

Keywords: ADHD, DAT/NET inhibition, clinical trial.

Disclosure: Dr. Kollins has received consulting fees from Sunovion Pharmaceuticals and was a site PI (Duke) for the study being presented.

T87. D-Cycloserine in Treatment Resistant Bipolar Depression

Joshua Kantrowitz*, Batsheva Halberstam, James Gangswich, Anthony Loebel, Sharon Engel, Thomas Large, Taleen Hanania, Emer Leahy, Daniel Javitt

Columbia University/Nathan Kline Institute, New York, New York

Background: Recent studies suggest that ketamine, a N-methyl-D-aspartate glutamate-receptor (NMDAR) antagonist, may provide near-immediate relief for depression. Maintenance of ketamine treatment is limited by its transient effects, even with repeated treatments. D-cycloserine is an FDA-approved anti-tuberculosis drug, and at>500 mg/d, functions as a net NMDAR antagonist at the glycine-site. Antidepressant effects of D-cycloserine were first described in the 1950’s and have recently been confirmed in a double blind RCT of treatment resistant major-depressive disorder (d=0.91 sd vs. placebo). We report on the utility of high-dose D-cycloserine as a maintenance adjunctive to acute ketamine and ongoing FDA-approved medications, as a potential practical treatment for bipolar depression (NCT01833897). In addition, we report on a preclinical study suggesting synergistic effects with lurasidone, an antipsychotic FDA-approved for bipolar depression.

Methods: 1. Clinical study of D-cycloserine in bipolar depression: Eight subjects (mean age: 37±16 years) enrolled, and demonstrated treatment-resistance to at least three-weeks of medication FDA-approved for bipolar depression (Mean 0.2±1 change from screening MADRS of 29.0±6 over a mean 4.2±1 weeks). Subjects then received open-label acute ketamine hydrochloride (0.5 mg/kg administered over 60 minutes) followed by eight-weeks of maintenance D-cycloserine (tapered up to 1000 mg/day over three-weeks) and pyridoxine, adjunctive to continued FDA-approved treatment. 2. SmartCube® evaluation of D-cycloserine: SmartCube® is a behavioral screening platform that provides a sequence of challenges to mice and captures more than 2000 features during a testing session. These features are compared to a database of behavioral signatures obtained using a set of diverse reference compounds, including antipsychotics, anxiolytics, and antidepressants. For this study, twelve mice/treatment group received dosages of D-cycloserine (3-600 mg/kg) or lurasidone (1-10 mg/kg) alone or in combination.

Results: Clinical study: Seven subjects completed, and five responded by eight weeks. On a last observation carried forward, paired-t-test analysis, a significant, sustained response was seen (d=1.1 sd after eight-weeks), in contrast to ketamine only studies, which show a return to baseline depression levels in two-weeks. Acute ketamine response was predictive of maintenance D-cycloserine response (r=0.82). Both treatments were well tolerated, with one subject requiring a dose-reduction to 500 mg at 6 weeks for mild sedation and another with hypomanic symptoms two weeks after stopping FDA-approved treatment. Five subjects requested to continue D-cycloserine post-study, with two successfully obtaining insurance coverage. Smartcube®: Dosages of D-cycloserine 200-400 mg/kg (equivalent to 1000 mg/day in humans) had profiles suggestive of an antidepressive and anxiolytic agent, while higher dosages (600 mg/kg) were suggestive of a cognitive enhancer profile. Profiles of lower D-cycloserine dosages were not suggestive of a psychoactive profile. As expected, lurasidone alone was suggestive of antipsychotic effects. In contrast, the combination of lurasidone and D-cycloserine was suggestive of an antidepressive/anxiolytic agent that was stable across dosages.

Conclusions: The development of treatments targeted at the glutamatergic system remains novel. These findings represent the first data on D-cycloserine in bipolar depression, and suggest that it may have synergistic effects with lurasidone, an FDA-approved medication for bipolar disorder. While ketamine may be an impractical long-term treatment, these findings provide proof-of concept for further study of the combination of D-cycloserine with approved medications in BPD. Correlational analysis suggests that D-cycloserine may produce similar responses to acute ketamine. Additionally, we present evidence for a tolerable, efficacious 60-minute infusion rate for ketamine.

Keywords: NMDA, depression, bipolar, translation.

Disclosure: Dr. Kantrowitz reports having received consulting payments within the last 2 years from Otsuka Pharmaceuticals, the Healthcare Advisory Board, Health Advances, LLC and Cowen and Company. He has conducted clinical research supported by the NIMH, the Stanley Foundation, Roche-Genetech, Forum, Sunovion, Novartis and Pfizer. He owns a small number of shares of common stock in GlaxoSmithKline. Dr. Javitt reports having received consulting payments within the last 2 years from Sunovion, BMS, Lilly, and Takeda. He has received research support from Pfizer and Roche. He holds intellectual property rights for use of NMDA modulators in treatment of neuropsychiatric disorders and D-cycloserine in depression. He served on the advisory board of Promentis. Drs. Loebel, Engel and Large are employees of Sunvion, and Drs. Hanania and Leahy are employees of Psychogenics.

T88. Efficacy and Safety of Lurasidone in Older Adults with Bipolar Depression: Analysis of Two Double-blind, Placebo-controlled Studies

Martha Sajatovic*, Brent Forester, Joyce Tsai, Hans Kroger, Andrei Pikalov, Josephine Cucchiaro, Antony Loebel

Case Western Reserve University, Cleveland, Ohio

Background: The treatment of bipolar disorder in the elderly has not been well-studied. This secondary analysis of two studies of adults (age 18-75 years) with bipolar depression evaluated the efficacy and safety of lurasidone in the subgroup of older adults (age ≥55 years), either as monotherapy, or as adjunctive therapy with lithium or valproate, including safety and tolerability during a 6 month extension phase.

Methods: In both the monotherapy and adjunctive therapy studies, patients who met DSM-IV-TR criteria for bipolar I depression, with or without rapid cycling, and with a Montgomery-Asberg Depression Rating Scale (MADRS) score ≥20 were randomized to 6 weeks of once-daily, double-blind treatment with lurasidone 20-60 mg/day or 80-120 mg/day, or placebo in the monotherapy study; or lurasidone 20-120 mg/day or placebo in the adjunctive therapy study with either lithium or valproate. The primary endpoint was LS mean change from baseline to Week 6 in the MADRS total score for both studies based on a mixed model for repeated measures (MMRM) analysis. The criterion for response was ≥50% reduction in MADRS score at last observation carried forward (LOCF) endpoint. Patients who completed the 6 week, double-blind trials were enrolled in a 6 month, open-label, extension study. Long-term effects on weight and metabolic parameters were evaluated from double-blind acute study baseline to 6 month endpoint, based on an observed case analysis.

Results: The proportion of older adults was 83/485 (17.1%) in the monotherapy study, and 53/340 (15.6%) in the adjunctive therapy study. At Week 6, mean change among older adults in the MADRS was significantly greater for the lurasidone 20-60 mg (-15.4; P<0.01; effect size=0.86) and 80-120 mg group (-14.1; P<0.02; effect size=0.74) vs placebo (-7.1). At Week 6, mean change in the CGI-BP-S was significantly greater for the lurasidone 20-60 mg (-1.7; P<0.05; effect size=0.73) and 80-120 mg groups (-1.6; P<0.05; effect size=0.65) vs placebo (-0.8). At LOCF-endpoint, responder rates among older adults were higher for the lurasidone 20-60 mg (53.8%; P<0.01; NNT=3) and 80-120 mg groups (40.0%; P=0.054; NNT=4) compared with placebo (14.8%). Adjunctive therapy with lurasidone (vs placebo) in older adults was associated with numerically greater but not statistically significant improvement at Week 6 in MADRS total score (-13.9 vs. -11.1; ns; effect size=0.26), and in the CGI-BP-S score (-1.4 vs. -0.9; ns; effect size=0.43). In the adjunctive therapy study, responder rates in older adults on lurasidone vs placebo, respectively, were 46.2% vs. 37.0%; ns; NNT=11). Among older adults in the monotherapy study, discontinuation due to adverse events occurred in 7.7% of patients on lurasidone 20-60 mg, 6.7% on lurasidone 80-120 mg, and 3.7% on placebo; and adverse events with an incidence ≥5% (and greater than placebo) for combined lurasidone doses were insomnia (8.6%), constipation (7.1%), diarrhea (6.9%), somnolence (6.9%), and akathisia, fatigue, vomiting, nasopharyngitis and muscle spasms (5.2% for each). Among older adults in the adjunctive therapy studies, discontinuation due to adverse events occurred in 3.8% of patients on lurasidone, and 7.4% on placebo; and adverse events on lurasidone with an incidence ≥5% (and greater than placebo) were akathisia (19.2%), insomnia (11.5%), abnormal dreams (7.7%), and tremor (7.7%). At the end of 6 months of open-label treatment with lurasidone, minimal changes were observed in the older adult population in mean weight (-0.6 kg), or median cholesterol (+4.0 mg/dL), triglycerides (+5.0 mg/dL), and glucose (0.0 mg/dL).

Conclusions: Results of these secondary analyses suggest that lurasidone, as monotherapy, or adjunctive with lithium or valproate, was an efficacious and well-tolerated treatment for bipolar depression in older adults. The most common adverse events in older adults with bipolar depression were akathisia, insomnia, somnolence, and nausea.

Keywords: lurasidone, bi-polar depression, older adults.

Disclosure: Martha Sajatovic: Research grants:Pfizer, Merck, Ortho-McNeil Janssen, Janssen, Reuter Foundation, Woodruff Foundation, Reinberger Foundation, National Institute of Health (NIH), Centers for Disease Control and Prevention (CDC) Consultant: Bracket, Prophase, Otsuka, Pfizer, Amgen Royalties: Springer Press, Johns Hopkins University Press, Oxford Press, UpToDate, Lexicomp CME activities: American Physician’s Institute, MCM Education Brent Forester: Consultant, Sunovion Pharmaceuticals, Inc. Research grants: National Institute of Mental Health (NIMH), Rogers Family Foundation Joyce Tsai, Andrei Pikalov, Hans Kroger, Josephine Cucchiaro and Antony Loebel are full-time employees of Sunovion Pharmaceuticals, Inc.

T89. Effects of Transcranial Direct Current Stimulation (TDCS) On Cognition, Brain Connectivity and Symptoms in Schizophrenia

Robert Smith*, Stanley Colcombe, Sanela Mattiuz, Mary Youssef, Mohammed Sharif, Russel H. Tobe, Revital Amiaz, MIchael Milham, John M. Davis

New York University Medical School/Nathan Kline Institute of Psychiatric Research, New York, New York

Background: Schizophrenia (SZ) is characterized by persistent cognitive deficits. TDCS has been reported to be effective in reducing hallucinations and other symptoms in schizophrenia and proposed to have potential beneficial cognitive effects in healthy controls and schizophrenics (SZ). The current research was a double-blind sham-controlled study to investigate the effects of tDCS on cognition, cigarette smoking, and symptoms in SZ. Recent research using fMRI imaging has shown abnormalities in resting state brain connectivity networks (ICN) in schizophasia compared to controls, in the fronto-parietal control network (FPN), default mode (DN) and other brain circuits, and some studies suggest that tDCS can alter these ICN in non-psychotic controls. We also present data from a preliminary open study of active tDCS which investigated tDCS effects on resting state brain connectivity circuits (ICN) in SZ.

Methods: 36 outpatients, with SZ or schizoaffective disorder (SA) participated in double-blind study of 5 sessions of active or sham tDCS (active- tDCS for 20 min, 2 ma). 29 SZ provided evaluable data on cognitive effects. Evaluations included MATRICS battery (baseline and 1 day after 5th sessions), VIGIL CPT (session 2), PANSS, and auditory hallucinations scale (Haddock). 5 SZ patients participated in a preliminary open study of active tDCS using the same design, where resting state fMRI (R-fMRI) with 3 Tesla scanner was utilized to investigate changes in ICN between baseline and 5 session of of tDCS.

Results: One day after the 5th tDCS session active tDCS SZ showed significant improvements, compared to sham, in MATRICS battery COMPOSITE scores (P=.009), MATRICS DOMAIN scores on WORKING MEMORY (P=.002), and ATTENTION-VIGILANCE (P=.02), as well as T cores on CPT (P T=.028) and Letter-Number Span (P=.009). Some of the other MATRICS Battery Domain scores also showed trends for improvement with tDCS. The other cognitive tests, which were evaluated directly after the tDCS session, did not show significant improvement of active vs sham tDCS, although there was a trend for reaction time to improve on VIGIL CPT with tDCS. There were no significant effects of tDCS on PANSS scores or auditory hallucinations. Preliminary R-fMRI data showed 5 sessions of tDCS suggested increased ICN of the fronto-parietal control network (FPN), decreased ICN in the default–mode network (DM), and decreased interhemispheric connectivity.

Conclusions: The positive effects of tDCS on cognitive scores on MATRICS battery suggest that tDCS may be an especially useful modality for improving cognition in SZ, but this findings needs to be confirmed and replicated in additional investigations. The beneficial effects of tDCS on improving cognition may be related to increasing or decreasing connectivity in specific brain ICN networks.since some of these ICN have been previously shown to be abnormal in SZ and bipolar patients compared to controls.

Keywords: TDCS, schizophrenia, cognition, R-fmri.

Disclosure: Nothing to Disclose.

T90. Oral Aripiprazole Is an Effective Maintenance Treatment in Adolescents with Schizophrenia: A Randomized, Double-blind, Placebo-controlled Trial

Christoph U. Correll*, Eva Kohegyi, Cathy Zhao, Ross A. Baker, Robert McQuade, Phyllis Salzman, Raymond Sanchez, Margaretta Nyilas, William Carson

Hofstra North Shore-LIJ School of Medicine/Zucker Hillside Hospital, Glen Oaks, New York

Background: The primary objective of this study was to evaluate the efficacy of oral aripiprazole 10–30 mg/d as maintenance treatment in adolescents with schizophrenia. Safety and tolerability were also evaluated.

Methods: This was a phase 3, multicenter, randomized, double-blind, placebo-controlled study in adolescents aged 13–17 years with schizophrenia (DSM-IV-TR) and history of the illness for ≥6 months. After screening for eligibility, inpatients and outpatients entered a 4- to 6-week conversion phase of titration to oral aripiprazole monotherapy (10–30 mg/d) and then a 7- to 21-week stabilization phase; patients receiving oral aripiprazole monotherapy 10–30 mg/d at screening directly entered the stabilization phase. Outpatients meeting stabilization criteria (Positive and Negative Syndrome Scale [PANSS] total score ≤80; score of ≤4 on PANSS items of conceptual disorganization, suspiciousness, hallucinatory behavior, and unusual thought content; Clinical Global Impressions-Severity [CGI-S] score ≤4; no current suicidal behavior [Columbia Suicide Severity Rating Scale]; and no evidence of clinically significant aggressive or violent behavior) at 2 consecutive weekly assessments starting at week 6 were randomized (2:1) to continuation of oral aripiprazole 10–30 mg/d or to placebo for up to 52 weeks in the double-blind maintenance phase. The primary outcome, exacerbation of psychotic symptoms/impending relapse, was defined as meeting any of the following criteria: (1) CGI-Improvement (CGI-I) score of ≥5 plus an increase in the PANSS items of conceptual disorganization, hallucinatory behavior, suspiciousness, and/or unusual thought content to a score>4 with an absolute increase of ≥2 on that item (or ≥4 on the combined items) since randomization; (2) CGI-I score of 6 or 7; (3) hospitalization due to worsening of illness; (4) suicidal behavior; or (5) violent or aggressive behavior. Adverse events (AEs) were monitored throughout the study. A planned interim analysis was conducted by an Independent Data Monitoring Committee (IDMC) to assess safety and efficacy after 75% of 37 impending relapse events (28 events) were achieved.

Results: The randomized population included 146 patients (98 aripiprazole, 48 placebo). After interim analysis, the IDMC recommended continuation to the endpoint of 37 events, at which point the sponsor discontinued all patients still in the trial (n=97) and offered entry into the open-label safety extension study. Demographics were similar in both groups. Mean daily doses were 19.2 mg (aripiprazole) and 17.7 mg (placebo); treatment lasted a mean of 184.6 and 158.1 days, respectively. Time to impending relapse was significantly longer for aripiprazole- vs placebo-treated patients (hazard ratio, 0.461 [95% CI, 0.242 to 0.879]; P=0.0161). Sensitivity analysis confirmed these results. The overall relapse rate was significantly lower with aripiprazole (19.4% [n=19/98]) vs placebo (37.5% [18/48]; P=0.0181), as was the rate of all-cause discontinuation (excluding those who discontinued because of sponsor termination of the trial; aripiprazole 25.5% [n=25/98], placebo 47.9% [n=23/48]; P=0.0076). The percentage of responders and patients achieving remission did not differ between groups. In general, symptom stability was maintained with aripiprazole, indicated by lack of change in PANSS, CGI-I, and CGI-S scores, and worsened with placebo. During double-blind maintenance treatment, the rate of ≥1 AE was 65.3% for aripiprazole and 68.8% for placebo; the most common AEs were schizophrenia (10.2% aripiprazole, 27.1% placebo), psychotic disorder (9.2%, 10.4%), and insomnia (5.1%, 18.8%). AEs of special interest did not occur more frequently in aripiprazole- vs placebo-treated patients, including extrapyramidal symptoms (6.1% vs 12.5%, respectively), weight gain (8.2% vs 10.4%), somnolence (2.0% vs 2.1%), hyperlipidemia (1.0% vs 0%), hepatic parameters (0% vs 2.1% [increased bilirubin]), and suicide-related AEs (0% vs 2.1% [intentional self-injury]). There were no deaths or AEs related to neuroleptic malignant syndrome, seizures, orthostasis, glucose levels, or prolactin levels; no aminotrasferase elevations>3 × upper limit of normal; and no pregnancies. Aripiprazole was associated with lower rates of serious AEs vs placebo (3.1% vs 12.5%) and discontinuations due to AEs (20.4% vs 39.6%). Sexual maturation/growth (Tanner staging) and frequency of signs/symptoms on the New York Assessment for Adverse Cognitive Effects of Neuropsychiatric Treatment were similar between groups.

Conclusions: Results from the first placebo-controlled maintenance study in adolescents with schizophrenia are consistent with previous studies of oral aripiprazole 10–30 mg/d in the acute treatment of adolescents with schizophrenia [1,2]. These findings provide valuable information on longer-term use of oral aripiprazole in adolescents, earlier in the course of their illness, who may benefit from a medication that combines good efficacy with a comparatively low risk of AEs. This research was supported by Otsuka Pharmaceutical Development and Commercialization. 1. Findling RL, et al. J Clin Psychopharmacol. 2008;28(4):441-446. 2. Findling RL, et al. Am J Psychiatry. 2008;165(11):1432-1441.

Keywords: aripiprazole, maintenance treatment, pediatric, schizophrenia.

Disclosure: CC is an employee of The Zucker Hillside Hospital and Hofstra North Shore-LIJ School of Medicine and an Investigator for The Feinstein Institute for Medical Research. EK, CZ, RB, RM, PS, RS, MN, and WC are employees of Otsuka Pharmaceutical Development and Commercialization, Inc.

T91. Withdrawn

T92. The Use of Augmentation Strategies in Treatment Resistant Anxiety Disorders: A Systematic Review and Meta-Analysis

Michael Van Ameringen*, Beth Patterson

McMaster University, Hamilton, Canada

Background: It has been estimated that 29-44% of people will meet diagnostic criteria for an anxiety disorder at some point in their life [1]. These conditions are associated with both significant burden and costs to afflicted individuals. Standard first-line treatments for anxiety disorders include antidepressants such as selective serotonin reuptake inhibitors (SSRIs) as well as cognitive behavioural therapy (CBT). Anxiety disorders are associated with high rates of non or partial response to first-line treatment, however, with 40-60% of patients continuing to have residual (and impairing) symptoms [2]. The treatment-resistant literature in anxiety disorders is still in its infancy. Although there is a small, but growing body of randomized controlled trials (RCTs), when compared to the treatment resistant literature associated with other mental disorders (i.e. schizophrenia or bipolar disorder), the information regarding next-step treatments for anxiety is sparse. While no drug has been approved for use in treatment resistance, a variety of pharmacological and psychological strategies have been recommended in the published treatment guidelines, based on evidence from RCTs and open label trials. In cases of treatment resistance, clinicians generally adopt one of several strategies, including SSRI augmentation with other class agents or with CBT [3]. The primary objective was to examine the potential benefits of using medication augmentation strategies compared to control treatments in patients who have had a partial or non-response to initial treatment for generalized anxiety disorder(GAD), social anxiety disorder(SAD) and panic disorder(PD) by meta-analysis of a pooled dataset.

Methods: Double-blind, randomized, controlled (either placebo or active comparator) (RCT) trials of medication augmentation in treatment resistant anxiety disorders were systematically reviewed. MEDLINE, EMBASE, PsycINFO, The Cochrane Library and conference proceedings were searched to identify trials. Two rater independently reviewed studies for inclusion, quality and risk of bias. Effect estimates were calculated using random effects modeling, heterogeneity was assessed and subgroup and sensitivity analyses were completed.

Results: Six parallel RCTs (3 in GAD, 1 in SAD, 2 in PD) were included in the meta-analysis, yielding a total of 557 randomized participants. Five studies had small samples; a study of pregabalin in treatment resistant GAD had a largest sample and was given the greatest weight in the analyses. Risk of bias was assessed by both raters using the Cochrane Collaboration risk of bias tool; weighted kappa was 0.86, SE: 0.0914, 95% CI: 0.69 to 1.0. Augmentation was not associated with an increased risk of response, (CGI-I ≤ 2) versus placebo RR=1.08, 95% CI 0.94 to 1.24. Response rates for augmentation were 64% compared to 59% for controls. Very little heterogeneity was found between studies (I2=0). The relative risk reduction (RRR) and absolute risk reduction (ARR) were 8% and 5% respectively, and the number needed to treat (NNT) was 20. A small, significant effect was found in favour of augmentation for reduction in symptom severity: Standard Mean Difference (SMD) -0.32, 95% CI -.56 to -0.08. No significant differences in sub-group effects were found by drug class, however, a significant effect for benzodiazepines (SMD -0.39, 95% CI -0.73 to -0.06) over placebo and pregabalin over placebo (SMD -0.25, 95% CI -.46 to -0.04) was found for reduction in symptom severity. No significant differences were found between augmentation with medication versus placebo in functional impairment and drop-outs due to adverse events.

Conclusions: In terms of overall response (by CGI-I), pharmacological augmentation did not appear to be beneficial when compared to placebo control in the treatment of refractory anxiety disorders, although a modest effect was found in the reduction of symptom severity from baseline to endpoint. These results may be limited by small study samples, and a small number of overall studies in the analysis. However, all post-hoc sensitivity analyses of response revealed a similar, null difference in effect, which may be an indicator of the strength of the findings.

Keywords: Meta-analysis, augmentation, pharmacotherapy, anxiety disorders.

Disclosure: Dr. Van Ameringen has received grant/research support from personal fees from Canadian Foundation for Innovation (CFI), Forest Laboratories, Janssen-Ortho Inc., National Institutes of Health, Pfizer Inc., Servier, and Wyeth-Ayerst. He is on the speaker’s bureau for Biovail, GlaxoSmithKine, Janssen-Ortho Inc., Lundbeck, Pfizer Inc., and Shire. He has received consultant frees from and is on advisory boards of Astra Zeneca, Biovail, Eli Lilly, Forest Laboratories, Janssen-Ortho Inc., Labo Pharm, Lundbeck, Pfizer Inc., and Shire.

T93. Efficacy of Cariprazine vs. Placebo Across Schizophrenia Symptom Domains: Pooled Analyses from 3 Phase II/III Trials

W. Fleischhacker*, Stephen Marder, Kaifeng Lu, Dayong Li, Paul Ferguson, György Nemeth, Krisztián Nagy, Willie Earley, Suresh Durgam

Medical University Innsbruck, Innsbruck, Austria

Background: Schizophrenia is a complex disorder comprising a diverse range of symptoms including positive, negative, and mood symptoms as well as cognitive impairments. New antipsychotics that show broad efficacy across the spectrum of schizophrenia symptoms may improve the management of schizophrenia and enhance patient outcomes. Cariprazine is a potent dopamine D3 and D2 receptor partial agonist with preferential binding to D3 receptors. The efficacy, safety, and tolerability of cariprazine in the treatment of schizophrenia were supported by 3 phase II/III studies. A pooled analysis evaluated the efficacy of cariprazine across the range of schizophrenia symptoms.

Methods: Data from 3 randomized, placebo-controlled trials were pooled. Studies were generally similar in design and included 6 weeks of double-blind treatment; the primary efficacy measure in all studies was the Positive and Negative Syndrome Scale (PANSS). All cariprazine dose groups (1.5 to 9 mg/d) were combined for analyses. Efficacy across symptom domains was evaluated using PANSS-derived Marder factor groupings (anxiety/depression, disorganized thought, negative symptoms, positive symptoms, uncontrolled hostility/excitement), PANSS cognitive factor, and PANSS single items. Change from baseline in factor scores and individual PANSS item scores were evaluated based on least squares mean differences (LSMDs) between cariprazine and placebo using a mixed-effects model for repeated measure (MMRM) approach; results were not adjusted for multiple comparisons. Effect sizes were calculated for all comparisons.

Results: The pooled study population comprised 442 placebo and 1024 cariprazine patients. Cariprazine was significantly superior to placebo on all PANSS Marder factors (anxiety/depression, P=.0015; all other Marder factors, P<.0001) and the PANSS cognitive factor (P<.0001). Effects sizes for cariprazine versus placebo were as follows: Marder factors - anxiety/depression (-0.21), disorganized thought (-0.47), negative symptoms (-0.39), positive symptoms (-0.37), uncontrolled hostility/excitement (-0.34); PANSS cognitive factor (0.48). Cariprazine-treated patients showed significantly greater improvement than placebo-treated patients on 26 of the 30 PANSS single items. These included 3 of 4 items in the anxiety/depression factor (anxiety, tension, and depression; all P<.01), all 7 items in the disorganized thought factor (all P<.05), all 4 items in the uncontrolled hostility/excitement factor, 6 of 7 items in the negative symptoms factor (all items except motor retardation, P<.001), 6 of 8 items in the positive symptoms factor (all except somatic concern and grandiosity, P<.01), and all 5 items in the cognitive factor (all, P<.001). Statistical significance versus placebo was not reached for 4 PANSS items (somatic concerns, guilt Feelings, motor retardation, and grandiosity). Effects sizes for PANSS single items that showed statistical significance ranged from -0.15 (mannerisms and posturing) to -0.44 (conceptual disorganization).

Conclusions: In this pooled analyses, cariprazine-treated patients demonstrated significantly greater improvement than placebo-treated patients on all Marder factors and the PANSS cognitive factor. Cariprazine was significantly superior to placebo on almost all PANSS single items. These results suggest that cariprazine has efficacy across the broad range of symptoms associated with schizophrenia. However, improvements in symptom domains during the treatment of acute schizophrenia can be secondary to improvement in psychosis.

Keywords: cariprazine, schizophrenia, dopamine.

Disclosure: Supported by funding from Forest Laboratories, Inc., a subsidiary of Actavis plc, and Gedeon Richter Plc. Wolfgang Fleischhacker is a consultant to Gedeon Richter Plc. Stephen Marder has no conflicts of interest to disclose. Kaifeng Lu, Dayong Li, Willie Earley, and Suresh Durgam are employees of Forest Research Institute, a subsidiary of Actavis plc. Paul Ferguson is an employee of Prescott Medical Communications Group, a contractor of Forest Research Institute, a subsidiary of Actavis plc. György Nemeth and Krisztián Nagy are employees of Gedeon Richter Plc.

T94. A Pilot Study of a Novel Monoamine Triple Reuptake Inhibitor Centanafadine SR (EB-1020 SR) in the Treatment of ADHD in Adults

Timothy Hsu*, Andrew Cutler, Ann Childress, Randall Marshall, Mark Bradshaw, Frank Bymaster, Anthony McKinney, Catherine O'Brien, Stephen Hurt, Timothy Wilens

Neurovance, Inc., Cambridge, Massachusetts

Background: This pilot study was designed to evaluate EB 1020-SR (centanafadine) as a novel non-stimulant treatment option for adult attention-deficit hyperactivity disorder (ADHD). Centanafadine is a norepinephrine-preferring triple reuptake inhibitor with IC50 values for transporter reuptake inhibition of 6 nM, 38 nM, and 83 nM, for norepinephrine, dopamine and serotonin respectively.

Methods: A total of 41 adult males with well-characterized ADHD enrolled in this 4-week, single-blind study with 1-week placebo run-in. EB1020-SR was given twice daily and titrated to a target dose of 500 mg daily over 7 days. Outcomes assessed included ADHD, executive functioning, and tolerability.

Results: 37 subjects completed the trial. Centanafadine-SR produced a 21-point reduction on the ADHD Rating Scale-IV (endpoint mean score=17, p<0.0001) including significant reductions in inattentive (p<0.0001) and hyperactive impulsive symptoms (p<0.0001). Overall, 68% of subjects were considered responders using the Clinical Global Impression of Improvement (much/very much improved). Clinically and statistically significant improvements in overall and specific domains of executive function using the Behavioral Inventory of Executive Functioning were also found (overall p<0.0001). No clinically meaningful trends in adverse events, laboratory values, vital signs, or ECG parameters were noted.

Conclusions: Centanafadine-SR appears effective in treating ADHD and executive functioning deficits in adult males. The maximum dose studied appears to be well tolerated. Based on these results, randomized, controlled studies of centanafadine-SR appear warranted.

Keywords: ADHD, Centenafadine SR, triple reuptake inhibitor.

Disclosure: 1. Timothy Hsu is an employee of Neurovance Inc. 2. Andrew Cutler is a consultant to Neurovance Inc. 3. Ann Childress is a consultant to Neurovance Inc. 4. Randall D. Marshall is a former employee of Neurovance Inc. 5. Mark Bradshaw is a consultant to Neurovance Inc. 6. Frank Bymaster is an employee of Neurovance Inc. 7. Anthony McKinney is an employee of Neurovance Inc. 8. Catherine O'Brien is an employee of Neurovance Inc. 9. Stephen Hurt is a consultant to Neurovance Inc. 10. Timothy E. Wilens is a consultant to Neurovance Inc.

T95. Efficacy and Safety of Cariprazine as Adjunctive Therapy in Major Depressive Disorder: A Double-blind, Placebo-controlled Study

Maurizio Fava*, Suresh Durgam, Victor Mergel, Willie Earley, György Németh, István Laszlovszky

Massachusetts General Hospital, Boston, Massachusetts

Background: Atypical antipsychotics are often used adjunctively in patients with major depressive disorder (MDD) who do not respond to standard antidepressant treatment (ADT). While atypical antipsychotics exhibit blockade of dopamine D2 receptors, there is increasing interest in the potential role of D3 receptors, which are highly expressed in the limbic system and may be important in modulating depression-related symptoms. Cariprazine is a potent D3 and D2 receptor partial agonist that has preferential binding to D3 receptors. In addition to being evaluated for the treatment of schizophrenia and bipolar mania, cariprazine is currently being investigated as an adjunctive therapy in MDD patients. Results are presented from a Phase 2b study (NCT01469377) of adjunctive cariprazine in MDD patients with inadequate response to standard ADT.

Methods: This was a multinational, randomized, double-blind, placebo-controlled, flexible-dose study of cariprazine in adults with MDD. Patients were required to have an ongoing inadequate response to standard ADT during their current episode; inadequate response was defined as a resistance rating of ≥3 and a global confidence score of ≥3 on the Antidepressant Treatment History Form. Patients who entered this study continued to receive their current ADT and were randomized to receive 8 weeks of double-blind adjunctive treatment with placebo (n=269), cariprazine 1-2 mg/d (n=274), or cariprazine 2 4.5 mg/d (n=276). The primary efficacy outcome was change from baseline to Week 8 in Montgomery-Asberg Depression Rating Scale (MADRS) total score. The secondary efficacy outcome was mean change from baseline to Week 8 in Sheehan Disability Scale (SDS) total score. The primary and secondary outcomes were analyzed using a mixed-model for repeated measures, with P values adjusted for multiple comparisons. The percentage of patients with a treatment response, defined as ≥50% improvement from baseline in MADRS total score, was included as an additional efficacy measure. Additionally, a post hoc band-pass filter analysis was used to re-analyze the primary outcome, controlling for sites with unusually high or low placebo responses (Merlo-Pich, Clin Pharmacol Ther 2010). For the band-pass filter analysis, study sites that showed mean placebo MADRS total score improvement of ≥10 or ≤3 points were excluded. This method has previously demonstrated the ability to enhance the signal to noise ratio in antidepressant trials and may better estimate true treatment effects. Safety assessments included adverse events (AEs), clinical laboratory tests, vital signs, and electrocardiograms.

Results: For change in MADRS total score, least squares mean differences (LSMDs) between cariprazine 2-4.5 mg/d and placebo were statistically significant throughout the double-blind treatment period (Weeks 2, 4, and 6; all P<.01) and at Week 8 (LSMD, 2.2; adjusted P=.0114). In the cariprazine 1-2 mg/d group, significantly greater mean improvements relative to placebo were observed at Weeks 2 and 4 (both P<.05) but the between-group difference at Week 8 was not statistically significant (LSMD, -0.9; adjusted P=.2404). Using the post hoc band-pass filter analysis, both cariprazine groups compared with placebo showed statistically significant improvements in MADRS total score, which were notably larger than in the prospectively defined analysis (LSMDs: 1-2 mg/d, -3.2, P=.0298; 2-4.5 mg/d, -6.4, P<.0001). SDS total scores in the cariprazine groups versus placebo at Week 8, did not reach statistical significance (LSMDs: 1-2 mg/d, -1.1, adjusted P=.2404; 2-4.5 mg/d, -1.4; adjusted P=.1140). Significantly greater MADRS response rates were found at Week 8 with both cariprazine doses relative to placebo (placebo, 38%; 1.2 mg/d, 48%, P=.0223; 2 4.5 mg/d, 49%, P=.0080). Treatment-emergent AEs that were reported in ≥10% of any treatment group were akathisia (2 4.5 mg/d, 22.3%), insomnia (2 4.5 mg/d, 13.6%), nausea (2-4.5 mg/d, 12.8%), and headache (placebo, 12.8%). Mean changes in metabolic parameters (eg, lipids, glucose) were generally small and similar between groups. Mean change in body weight was<1 kg at the end of double-blind treatment in all treatment groups. Mean changes in blood pressure and pulse were small and similar between groups, and there were no postbaseline QTcB or QTcF interval values>500 msec.

Conclusions: Results from this Phase 2b study indicate that cariprazine may be an effective and generally well-tolerated adjunctive treatment in adults with MDD who have inadequate responses to standard ADTs. Adjunctive cariprazine was superior to placebo on MADRS total score improvement (2 4.5 mg/d dose group) and MADRS response (both dose groups). Results from the post hoc band-pass analysis suggest that the full extent of treatment effects in this study may have been blunted by high placebo response in certain study centers.

Keywords: cariprazine, depression, dopamine.

Disclosure: Supported by funding from Forest Laboratories, Inc., a subsidiary of Actavis plc., and Gedeon Richter Plc. Maurizio Fava has received research support from, served as an advisor/consultant to, and been a speaker for Forest Pharmaceuticals, Inc., a subsidiary of Actavis plc. Suresh Durgam, Victor Mergel, and Willie Earley are employees of Forest Research Institute, a subsidiary of Actavis plc. György Németh and István Laszlovszky are employees of Gedeon Richter Plc.

T96. Effects of NSI-189, a Neurogenic Compound, on Quantitative EEG (QEEG) in Patients with Major Depressive Disorder: QEEG Effects, Dose Response Relationships, and Clinical Outcomes

Larry Ereshefsky*, Brett English, Jack Johnstone, Karl Johe, Lev Gertsik, Maurizio Fava, Marlene Freeman, Stephen Potkin

PAREXEL International, Marina del Rey, California

Background: NSI-189, a molecule (Neuralstem Inc) being studied for the treatment of MDD, stimulates neurogenesis in human hippocampus-derived neural stem cells in vitro and in mouse hippocampus in vivo. NSI-189 has behavioral efficacy in the novelty suppressed feeding paradigm following daily 28-day oral administration. Phase 1B trials typically evaluate safety and efficacy in healthy controls and have limited pharmacodynamic assessments. The use of patients rather than healthy volunteers in early development allows the incorporation of clinical rating scales and putative biomarkers such as QEEG analyses. In previous trials with SSRI’s high frequency alpha in the left posterior temporal region has been correlated with changes in mood rating scales. QEEG may provide greater sensitivity in small trials to detect clinical efficacy signal.

Methods: Twenty-four patients with MDD were randomized in a Phase 1B, double-blind, placebo-controlled, multiple-ascending-dose study to receive NSI-189 40, 80, 120 mg or placebo daily for 28 days in an early phase unit. Outpatient follow-up visits were scheduled on Day 35 (±3), Day 42 (±3), Day 49 (±3), Day 56 (±3), Day 70 (±3) and Day 84 end-of-study (±3). Eligibility was confirmed with an independent SAFER interview (diagnosis and severity by MGH CTNI raters). In addition to safety and PK, efficacy assessments included the Montgomery-Asberg Depression Rating Scale (MADRS), the Clinician Global Impression – Improvement (CGI-I), the Symptoms of Depression Questionnaire (SDQ) and the Cognitive and Physical Functioning Questionnaire (CPFQ). Resting 20-minute QEEG measurements were obtained 6 hours post-dose on Days 14 and 28. The 19-standard International 10/20 System scalp locations plus EMG, and eye movement monitoring was obtained. Digital recordings using Cadwell Laboratories instrumentation were reviewed to identify the presence of physiological and instrumentation artifacts prior to analyses, which were manually removed by an experienced technologist. EEG epochs were submitted to power spectral analyses using Brain Vision Analyzer software. Data analyses included repeated-measures ANOVA/MANOVA, Student's t and estimation of effect size as appropriate. Given the small sample, effect sizes (d) were estimated for the clinical efficacy outcomes where all doses of drug were pooled to compare with patients on placebo. Exploratory analyses for each dose group’s efficacy measures and dose response relationships with QEEG were performed. With 6 active drug treated patients per dose group, the frequency range sampled was extended (alpha/beta) to 8-20 hz to increase power.

Results: A clinically meaningful reduction in depressive and cognitive symptoms on the MADRS (Day 28 d=0.95; Day 84=0.84), SDQ (Day 28 d=0.9; Day 84=1.10), and CPFQ (Day 28 d=0.94; Day 84=1.20) assessment scales was observed for patients on NSI-189 compared with placebo. Individual dose groups were also analyzed for efficacy measures; based on effect sizes, a greater reduction in symptoms was observed for the two lower doses (40 mg/day and 80 mg/day x 28 days) than for the highest dose (120 mg/day). These improvements appeared to persist over time during the follow-up at Day 84, although less so for the 40 mg/day dose group. The quantitative EEG analyses showed increased high frequency alpha with NSI-189 treatment (change in baseline to Day 14 and to Day 28) in contrast to lower high frequency alpha or less change with placebo. The NSI-189 effect is most prominent in the left posterior temporal (T5) (t=2.45, p=0.0247) and left parietal regions (P3) (t=3.31, p=0.004) and is similar when comparing QEEG at baseline to either Day 14 or Day 28 assessments. There are no earlier QEEG assessments while on drug to pinpoint onset of increased alpha frequency. The largest coherence differences between NSI-189 (all doses combined) and placebo were for inter-hemispheric coherence between F3-F7 and F4-F8 in the 4-6 Hz frequency range at Day 28. NSI-189 decreased inter-hemispheric coherence compared to placebo (F=7.88, p=0.01). Changes in QEEG variables were analyzed for each dose level. All were significantly different than placebo on days 14 and 28. Using a mixed model analysis of variance including placebo and active dose groups, a significant Day (Baseline, Day 14, Day 28) by Dose (40 mg/day, 80 mg/day, 120 mg/day) interaction is apparent in the left temporal region (T3) for the 8-20 Hz analyses, (F=3.16, p=.0125). Similarly Pz demonstrated a trend towards significance demonstrating the same dose effect pattern (F 2.18, p=0.065). t-tests on the individual dose groups showed no significant univariate effects. Examination of group means show similar patterns for the 10-12 Hz frequency range and the 8-20 Hz range for each dose level. The largest increases for both of these frequency ranges occurred on Day 14 for the 40 mg/day and 80 mg/day groups. The 120 mg/day group on Day 14 showed an inverted pattern at T3 and Pz with lower 10-12 Hz and 8-20 Hz as compared to baseline.

Conclusions: NSI-189 was safely tolerated as monotherapy in patients with MDD. Subjects in this Early Phase study met SAFER criteria for MDD. A significant reduction in depressive and cognitive symptoms was observed at all doses of NSI-189. NSI-189 produced qEEG changes in high-frequency alpha in the left posterior temporal and parietal regions in patients with MDD, consistent with data in MDD patients responding to SSRI antidepressants. Given the limited sample size, analysis of individual dose groups’ changes in QEEG were not significant. However, the 40 mg and 80 mg group increase high frequency alpha and 8-20Hz at Day 14 as compared to Baseline. In contrast 120 mg group decreases these frequencies as compared to Baseline values. Further analysis to assess dose, QEEG, and response relationships are warranted. NSI-189 may also exhibit pro-cognitive properties, yet decreased inter-hemispheric alpha coherence was observed. These preliminary findings support development of NSI-189 as a potential neurogenesis activator for the treatment of depression. Limitations include the small sample size, and the sequential escalating dose structure of a Phase 1 trial.

Keywords: Neurogenesis, Major Depression, Quantitative EEG, Early Phase Drug Development.

Disclosure: This study was funded by Neuralstem Inc. MGH/CTNI, QMetrx, UCI, and PAREXEL conduct numerous clinical trials funded by the pharmaceutical industry.

T97. Risperidone Long-acting Injection vs. Oral Risperidone: A Secondary Analysis of Relapse and Rehospitalization Controlling for Switching in a Pragmatic Trial

Nina R. Schooler*, Srinath Gopinath, Jeremy Weedon, Peter F. Buckley, Donald C Goff, Alexander Kopelowicz, John Lauriello, Theo Manshreck, Alan J. Mendelowitz, Del D. Miller, Daniel R. Wilson, John M. Kane

State University of New York Downstate Medical Center, Washington, District of Columbia

Background: The PROACTIVE study found no significant differences in relapse or rehospitalization between long-acting injectable risperidone(LAI-R) and oral antipsychotics, consistent with other studies. In keeping with the design of a pragmatic RCT, any oral second-generation antipsychotic could be prescribed in the oral treatment arm. Therefore, for some trial participants, randomization to LAI-R entailed both randomization to injectable route of administration and changing the specific antipsychotic being received. If these participants were randomly assigned to oral medication their medication was not changed. For participants who were already receiving oral risperidone, randomization entailed only a change in route of administration. Study of the subset of participants who entered the trial receiving oral risperidone allows examination of whether relapse and rehospitalization differ between the two routes of administration (oral vs. LAI) in those for whom there has been no change of antipsychotic medication. This provides a more precise test of the hypothesis that the LAI route of administration reduces the risk of relapse and or rehospitalization. Since changing antipsychotic medication has been reported to be associated with poorer outcomes, examination of this subset of the PROACTIVE study cohort also eliminates that source of variance in outcome.

Methods: Three hundred five subjects at eight academic centers were randomly assigned to LAI-R or oral second-generation antipsychotics. Following randomization open treatment lasted up to 30 months. All subjects had confirmed diagnoses of schizophrenia or schizoaffective disorder, were between 18 and 65 years old, were living in the community for at least 4 weeks but had been hospitalized within past 12 months and were rated at least moderately ill (CGI>4). Blinded live video assessment of symptoms by Master Raters with the Brief Psychiatric Rating Scale was completed every three months. A blinded Relapse Monitoring Board determined whether relapse and/or hospitalization for symptom exacerbation had occurred. One hundred five subjects were receiving oral risperidone at enrollment. We compared their demographic characterstics and baseline symptomatology to the 200 other study participants. Within the oral risperidone cohort, we compared the 56 randomly assigned to LAI-R to the 49 randomized to Oral (continuing risperidone) on the same characteristics. Using Cox regression analysis we evaluated time (defined by visit number 1-66) to first relapse and time to first hospitalization in the risperidone cohort, covarying number of hospitalizations and the global rating on the Scale of Functioning.

Results: Comparison of Risperidone Oral to Others. There were significant differences in the proportion of subjects receiving oral risperidone by site (Fisher’s exact test p=0.003). Those receiving oral risperidone were younger at first hospitalization (Wilcoxon rank sum p=0.015, had higher SANS Affective Flattening (p=0.037) and Asociality/Anhedonia (p=0.005). Comparison within Risperidone Cohort LAI to Oral. The only significant difference at baseline was that LAI-R subjects had poorer global Scale of Functioning than oral subjects (p=0.020). Preliminary inspection of survival curves indicates that, as seen with the full cohort of 305, there are no differences in time to relapse or time to hospitalization.

Conclusions: Comparing LAI-R to oral risperidone controls for two possible confounds in the pragmatic PROACTIVE trial. The first is that only route of administration varies between the two treatment groups. The second is that treatment with LAI does not involve changing medication. Thus, this sub-group analysis allows comparison with the studies done in the 1970s that compared oral and injectable fluphenazine. There were differences between the risperidone cohort and those who received other antipsychotics. The risperidone cohort was younger at first hospitalization and had more negative symptoms. The most interesting of these is the fact that the percentage of patients receiving risperidone at baseline varied significantly from 21% to 57%, suggesting that even among academic centers there are differences in prescribing patterns. None of these differences were reflected in treatment group differences within the risperidone cohort. Perhaps the most important result of this analysis is that that the findings are consistent with the findings for the full cohort – no significant differences in time to relapse or hospitalization in a 30 month trial.

Keywords: schizophrenia, relapse and hsopitalization, Long-acting injectable antipsychotics, randomized clinical trial.

Disclosure: Dr Schooler has received grant support from Otsuka, Neurocrine, Genentech; Advisory boards and consultation: Eli Lilly, Janssen Psychiatry, Amgen, Shire, Graham Boechk Foundation, Lundbeck International Neuroscience Foundations, Genentech/Roche, EnVivo (Forum), Sunovion. Dr. Buckley has received research support from Ameritox, Sunovion and Posit Science. Dr. Kane has been a consultant for Amgen, Alkermes, Bristol-Meyers Squibb, Eli Lilly, EnVivo Pharmaceuticals (Forum), Forest, Genentech, H. Lundbeck. Intracellular Therapeutics, Janssen Pharmaceutica, Johnson and Johnson, Merck, Novartis, Otsuka, Pierre Fabre, Proteus, Reviva, Roche and Sunovion. Dr. Kane has been on the Speaker’s Bureau for Bristol-Meyers Squibb, Eli Lilly, Genentech, Janssen and Otsuka. Dr. Kane is a Shareholder in MedAvante, Inc. Dr. Bustillo consulted for Otsuka Pharmaceuticals. Dr. Miller has received grant/research support from Hoffmann-La Roche and consulting fees from Otsuka and Hoffmann-La Roche. Drs. Gopinath, Manschreck, Mendelowitz, Weedon and Wilson have no disclosures.

T98. Efficacy of Ondansetron and Simvastatin on Cognition and Negative Symptoms in Established Schizophrenia

J.F. William Deakin*, Nusrat Husain, Alexander J. J. Parker, Mohammed O. Husain, Ajmal Kazmi, Raza ur Rahman, Mohammad M. Hamirani, Tayyaba Kiran, Nasirt Mehmood, Graham Dunn, Richard Drake, Imran B. Chaudhry

University of Manchester, Manchester, United Kingdom

Background: Despite initial excitement in the psychiatric possibilities of new 5HT3 receptor antagonists such as ondansetron in the 1980s, few if any clinical trials were carried out. The drugs are now mainly used as anti-emetics. However, a re-purposing study funded by the Stanley Medical Research Institute (SMRI) in China reported a 10-point benefit over placebo on Positive and Negative Syndrome Scale (PANSS) scores in treatment resistant schizophrenia (PMID: 16959472). We, together with the SMRI, have shown that atypical anti-inflammatory drugs such as minocycline could be of benefit in schizophrenia (PMID: 22526685). Since ondansetron has anti-inflammatory actions we designed a definitive and mechanistic study in stable but symptomatic patients. We included tests of cognitive performance. We compared ondansetron with simvastatin, identified by the SMRI as another drug with off-target anti-inflammatory actions but untested in psychosis. We tested the hypothesis that the 2 drugs might potentiate each other’s actions in improving negative symptoms of schizophrenia.

Methods: The protocol and feasibility study have been published (PMID:23782463; 25057343). The study was carried out in three centres in Pakistan supervised by PIs at the University of Manchester in collaboration with the Dow University of Health Sciences, Karachi. Patients aged 18-65 under stable treatment with a schizophrenia-related DSM IV diagnoses were recruited. They were randomly allocated to receive 8mg ondansetron daily (O) or matching placebo capsules (P) and to 40mg simvastatin (S) or matching placebo (P). Thus each patient received 2 different capsules. Treatment allocation was by a randomized permuted blocks algorithm within centres, controlled by the trial statistician in the UK via the local trial pharmacist. Power calculations resulted in an intended sample size of 54 in each of the 4 treatment groups. The primary outcome variable was the PANSS negative syndrome score rated after 3 and 6 months of treatment. Clinical Global Impression and Social and Occupational Functioning Scales were completed. Cognitive tasks included Stroop, verbal fluency and Coughlan verbal and visual list learning. Pre-planned analysis was by mixed model analysis of variance of the 3 and 6 month follow-up (within subjects) data of the intention to treat sample with factors for ondansetron, simvastatin and centre and covariation for baseline scores.

Results: The four cells of the 2x2 design over-recruited to 74-78 each. There were 7-9 drop outs from each of the 3 active treatment groups and 12 from the P+P group with none due to side effects. There were 3 deaths from natural causes and one suicide. Group mean total PANSS scores ranged from 72-78 and negative subscale scores from 17-18. The groups did not differ significantly on age or sex. Adjusted mean (95% CI) post-treatment negative PANSS scores were P+P=15.4 (14.6-16.2); O+P=13.5 (12.7-14.3); S+P=13.7 (12.9-14.4); O+S=14.1 (13.3-14.8), there being no significant effects of follow-up (3 vs 6 months). The interaction between ondansetron and simvastatin (OxS) was significant at p=.006 (F=7.7; df 1/492) reflecting the lower scores in the 3 active treatment groups than in the P+P group. The OxS interaction was also significant for total PANSS but not for positive symptoms. Calgary Depression, Clinical Global Impression and EQ50 quality life scores also showed significant OxS interactions respectively p=.001; p=.04 and p=.02. There was no evidence for positive interaction between the drugs. Ondansetron improved both Coughlan verbal and visual list learning (respective main effects p=.007 and p=.02) but there were no treatment effects on verbal fluency, Stroop or block designs. There were no group differences in side-effect ratings or withdrawals.

Conclusions: Both drugs tended to improve negative symptoms at 3 and 6 months resulting in a OxS interaction but their effects were not additive. The hypothesis that the 2 drug’s actions on immune function would synergise in improving negative symptoms was clearly not borne out. Verbal learning impairment has been associated with negative symptoms and poor social outcome in schizophrenia (PMID:8610818). The substantial main effect of ondansetron (16% improvement vs 2% on S or P) on verbal list learning suggests cognitive enhancement may contribute to benefits on symptoms and functioning. Other main effects of ondansetron occurred only in the presence of OxS interaction. There were no main effects of simvastatin. Further statistical analysis is in progress to model discontinuations and possible influences of sex and centre to formally determine relative effect sizes of the 2 drugs. Our results corroborate the earlier SMRI finding and recent reports of benefits on negative symptoms with ondansetron and granisetron (PMID:18789844). The benefit of ondansetron observed on cognition suggests it could prove useful in augmenting cognitive remediation.

Keywords: ondansetron, simvastatin, schizophrenia, negative symptoms.

Disclosure: JF William Deakin in the last 5 years has held grants from Servier, AstraZeneca and P1vital and given talks and/or advice for Servier, Johnson and Johnson, AstraZeneca and Lilley. Fees are paid as reimbursement for his time to the University of Manchester.

T99. Extinction and Change in Cognitions and Cortisol Activity in Posttraumatic Stress Disorder Treatment

Sheila Rauch*, Rebecca Sripada, Anthony King, James Abelson, Barbara Rothbaum, Israel Liberzon

University of Michigan/VA Ann Arbor Healthcare System, Ann Arbor, Michigan

Background: Mechanistic research examining the biological and psychological factors involved in effective PTSD treatment is a critical area of developing research. Understanding these mechanisms can help to refine treatments, improve efficacy and efficiency, and increase rates of response. Despite efficacy, many patients drop-out of Prolonged Exposure (PE) therapy (6) or do not remit from PTSD (7) suggesting that greater understanding of the processes involved could direct us toward better outcomes. The components of PE include exposure and emotional processing. Repeated exposures to the memory itself and cues in life (imaginal and in vivo exposure) lead to extinction/desensitization of emotional responses possibly through inhibitory learning. Emotional processing occurs as a result of these exposures and changes in cognitions, such as increasing sense of self- competence and control over negative affect and emotional difficulty (i.e., “I can handle bad things that happen”) and enhancing the experience of social support (i.e., “other people think I am a good person”) (8-10). Among the most studied biological systems involved in PTSD is the hypothalamic-pituitary-adrenal (HPA) axis with significant focus on cortisol. Patients with PTSD have consistently been shown to have HPA axis abnormalities, primarily enhanced glucocorticoid receptor sensitivity leading to enhanced negative feedback and, less consistently, hypocortisolemia (1-3) and flattened diurnal cortisol rhythms (4). There is evidence that glucocorticoids can influence the course of illness in PTSD. However, it remains unclear as to whether HPA axis dysregulation is an etiologic/pathophysiologic factor in PTSD, a biomarker of trait vulnerability, or an active agent shaping illness manifestation and perhaps recovery by promoting plasticity of specific brain circuits. Understanding cognitive and biological mechanisms of PTSD treatment can help refine treatments and increase rates of response.

Methods: Thirty-six military personnel and Veterans with PTSD were randomly assigned to receive Prolonged Exposure Therapy (PE) or Present Centered Therapy (PCT). We examined symptoms, in-session ratings of distress, trauma-related cognitions, and two indices of HPA axis function [cortisol awakening response (CAR) and cortisol response to a script-driven imagery task]. For responder status, a conservative measure of response was used to differentiate the groups for detection of biological differences. We required a 50% reduction in CAPS score from pre to posttreatment to be considered a high responder.

Results: PE resulted in significantly more symptom reduction than PCT (p=.008). In the current study, hierarchical linear modeling was used to investigate the impact of symptom change and high responder status on SUDS (level-one) nested within therapy sessions (level-two) nested within 14 patients (level-three). Symptom change (t=-2.43, df=12, p=.03) and responder status (t=-2.68, df=12, p=.02) predicted slope of SUDS across sessions, but did not predict slope of SUDS within-session, indicating that high responders demonstrated differential between- but not within-session habituation. High responders exhibited an additional 4.4 (SE=1.63) point SUDS decrease between sessions than low responders. With regard to cortisol, at pre-treatment, high responders showed a trend toward increased cortisol response at 30 minutes post trauma script when compared to low responders (p=.08). At mid-treatment, high responders showed higher cortisol levels throughout the imagery task (p’s=.03-.04). There were no differences between high and low responders at posttreatment. Thoughts of incompetence [F (1.6, 35.8)=16.8, p=.000] and a dangerous world [F (1.3, 29.9)=8.2, p=.004] significantly improved over time in high responders but showed no change in low responders. Script-associated cortisol levels prior to treatment and reductions in thoughts of incompetence accounted for 56% of the variance in reductions in PTSD with treatment.

Conclusions: Between-session reduction in reported distress, but not within session, is significantly related to treatment outcome, such that individuals who show greater habituation between treatment sessions are more likely to respond to treatment. Tracking SUDS changes across sessions can help practitioners monitor patient progress and guide decisions about treatment modifications or enhancements. In addition, both increased cortisol response to personal trauma script prior to PTSD therapy and reductions in cognitive symptoms of PTSD were significantly and uniquely related to reductions in the core symptoms of PTSD. However, contrary to our hypotheses, cortisol measures were not related to cognitive changes.

Keywords: cortisol, extinction, Posttraumatic stress disorder, Exposure therapy.

Disclosure: Nothing to Disclose.

T100. Efficacy and Safety of Adjunctive Brexpiprazole (OPC-34712) in Major Depressive Disorder: Results: of Two Pivotal Clinical Studies

Michael Thase, James Youakim, Alexander Skuban, Mary Hobart, Peter Zhang, Susan Legacy*, Robert McQuade, Margaretta Nyilas, William Carson, Raymond Sanchez

Otsuka (OAPI), Princeton, New Jersey

Background: Brexpiprazole is a serotonin-dopamine activity modulator that is a partial agonist at 5-HT1A and dopamine D2 receptors at similar potency, and an antagonist at 5-HT2A and noradrenaline alpha1B/2C receptors. The efficacy, safety and tolerability of adjunctive brexpiprazole were evaluated in patients with major depressive disorder (MDD) and inadequate response to antidepressant treatments (ADT) in two phase 3 clinical studies.

Methods: Patients with MDD and inadequate response to 1–3 ADTs were enrolled and received single-blind ADT for 8 weeks. Patients with inadequate response after this prospective phase were randomized to ADT+brexpiprazole or ADT+placebo for 6 weeks. Both studies included fixed doses (2mg [Study 1: NCT01360645]; 1mg and 3mg [Study 2: NCT01360632]). Primary efficacy endpoint was the change in MADRS total score from baseline to Week 6. Pre-specified analyses were conducted both on the efficacy population and on the final protocol population (fulfilling amended randomization criteria) for each individual study.

Results: In both studies completion rate was high (>90%) and comparable across brexpiprazole and placebo groups (Study 1: 2mg+ADT: 174/188 [92.6%], PBO+ADT: 178/191 [93.2%]; Study 2: 1mg+ADT: 216/226 [95.6%], 3mg+ADT: 210/230 [91.3%], PBO+ADT: 208/221 [94.1%]). Adjunctive brexpiprazole showed greater improvement than adjunctive placebo in MADRS total score on the efficacy population in both Study 1 (least square mean differences: 2mg+ADT [N=187]: -3.12, p=0.0001) and in Study 2 (1mg+ADT [N=225]: -1.19, p=0.0925; 3mg+ADT [N=226]: -1.52, p=0.0327). Similar results were observed for the final protocol population in both Study 1 (2mg+ADT [N=175]: −3.21, p=0.0002) and Study 2 (1mg+ADT [N=211]: -1.30, p=0.0737; 3mg+ADT [N=213]: -1.95, p=0.0079). The most frequent adverse events included akathisia (4.4%, 7.4%, 13.5%, 1.7%), weight increase (6.6%, 8.0%, 5.7%, 1.9%), tremor (4.0%, 2.1%, 5.2%, 2.2%) and somnolence (4.0%, 4.3%, 5.7%, 0.5%), in the brexpiprazole 1mg+ADT (N=226), 2mg+ADT (N=188), 3mg+ADT (N=229) and pooled placebo+ADT groups (N=411), respectively.

Conclusions: Data from adequate and well-controlled clinical studies provides evidence that brexpiprazole is efficacious as adjunctive treatment in MDD patients with an inadequate response to ADTs. All doses of adjunctive brexpiprazole were well tolerated, with notably low levels of sedating or activating side effects.

Keywords: Brexpiprazole, Major Depression.

Disclosure: All but one author is an employee of Otsuka Pharmaceutical Inc.

T101. A Multicenter, Randomized, Controlled, Phase III Trial of Fixed-dose Brexpiprazole for the Treatment of Adults with Acute Schizophrenia

John Kane, Aleksandar Skuban, James Youakim, John Ouyang, Mary Hobart, Stephanie Pfister, Steve Offord*, Robert McQuade, Margaretta Nyilas, William Carson, Raymond Sanchez

Otsuka, Princeton, New Jersey

Background: The efficacy, safety, and tolerability of brexpiprazole were evaluated in patients with acute schizophrenia.

Methods: This is a phase III, multicenter, randomized, double-blind, placebo-controlled trial (NCT01393613). Enrolled patients with acute schizophrenia were randomized to fixed doses of brexpiprazole 4mg, 2mg, 1mg, or placebo (3:3:2:3) for 6 weeks. The primary efficacy endpoint was the change in Positive and Negative Syndrome Scale (PANSS) total score from baseline to week 6 and the key secondary endpoint was the change in Clinical Global Impression – Severity Scale (CGI-S) score at week 6. Efficacy analyses were conducted using mixed model repeated measures (MMRM), including treatment, visit, site and treatment-by-visit interaction as fixed effects, and baseline score-by-visit as covariate. A gate keeping average effect method was applied to control for multiple comparisons (at an alpha level of 0.05) before proceeding with the comparisons for each group (2mg and 4mg brexpiprazole) versus placebo was allowed. The 1 mg group was included to explore the lower efficacious dose range of brexpiprazole.

Results: The gate keeper test of the average effect method met the threshold for the combined dose of brexpiprazole 4mg and 2mg compared with placebo (p=0.0093). Brexpiprazole 4mg (n=181) was superior to placebo (n=180) in change from baseline in PANSS total score at week 6 (LS mean change: -20.00 vs -13.53, p=0.0022); the change from baseline was statistically significant in favor of brexpiprazole from week 3 onwards. Brexpiprazole 2mg (n=179) showed numerical improvement over placebo at week 6 (−16.61 vs 13.53, p=0.1448). In addition, brexpiprazole 1mg (n=117) also showed numerical improvement over placebo at week 6 (p=0.1588). Results of the key secondary endpoint supported the primary results. The most common adverse events were headache (7.5%, 10.8%, 10.3%, 14.7%), insomnia (12.5%, 13.4%, 15.2%, 14.7%), and agitation (8.3%, 8.6%, 7.1%, 7.1%) in the brexpiprazole 1mg, 2mg, 4mg and placebo group, respectively; incidences of akathisia were lower in brexpiprazole treatment groups (4.2%-6.5%) vs placebo (7.1%).

Conclusions: Brexpiprazole 4mg was effective in the treatment of adults with acute schizophrenia, and was superior to placebo in the primary efficacy endpoint, as measured by the change in PANSS total score. The results of the key secondary endpoint were supportive of the primary results. All doses of brexpiprazole were well tolerated.

Keywords: brexpiprazole, schizophrenia, clinical.

Disclosure: Drs. Skuban, Youakim, Ouyang, Hobart, McQuade, Nyalis, Carson, Sanchez and Mss Pfister are employees of Otsuka Pharmaceutical Commercialization and Development, Inc., Princeton, NJ, USA Dr Offord is an employee of Otsuka America Pharmaceutical, Inc., Princeton, NJ, USA Dr Kane has received honoraria for lectures and/or consulting from Alkermes, Amgen, Bristol-Myers Squibb, Cephalon, Eisai, Boehringer Ingelheim, Eli Lilly, Intracellular Therapeutics, Janssen, Johnson and Johnson, Lundbeck, Merck, Novartis, Otsuka, Pfizer, Pierre Fabre, Proteus, Roche, Sunovion and Targacept and is a shareholder of MedAvante.

T102. Internet-based Cognitive Behavioral Therapy Effects on Symptom Severity in Major Depressive Disorder: Preliminary Results from a Randomized Controlled Trial

Scott Rauch*, Elizabeth Olson, Jennifer Buchholz, Isabelle Rosso, William Killgore, Christian Webb, Hannah Gogel

McLean Hospital, Belmont, Massachusetts

Background: Emerging evidence suggests that internet-based cognitive behavioral therapy (iCBT) is associated with symptom reduction in individuals with major depressive disorder (MDD). Researchers from the School of Psychiatry at the University of New South Wales developed an iCBT program that has shown success in treating MDD. Results of trials conducted in Australia indicate that the iCBT program, when administered in conjunction with supportive contact by technical staff, yields clinical improvement comparable to traditional face-to-face treatments rendered by doctoral therapists. This promising iCBT program has not yet been tested in the United States, and its validation among Americans with MDD is necessary before large-scale clinical applications are feasible in the United States. In this study, we tested the effectiveness of the Australian iCBT program for depression, after modifying it to be culturally relevant and making it compliant with the Health Insurance Portability and Accountability Act of 1996. We also replaced the waitlist group that had been employed in the Australian studies with a more stringent monitored attention control (MAC) group.

Methods: To date, we have recruited 47 participants (ages 18-45) who were diagnosed with MDD using the Structured Clinical Interview for DSM-IV-TR, and then randomly assigned to either the iCBT treatment condition or MAC condition. Twenty-nine participants had completed all assessments at the time of this analysis (17 iCBT; 12 MAC). Participants had no history of alcohol/substance dependence, neurologic illness, or psychotic disorder. Current psychotropic medications were also exclusionary. Participants in the treatment condition completed a ten-week intervention comprising six online lessons, printable content summaries and homework assignments. During the 10 week period, participants in both the iCBT and MAC conditions were required to log into the web-based system six times to complete a battery of self-report clinical scales, and they were contacted weekly by telephone by a non-clinician technician who provided encouragement and support. The primary outcome measure was the 17-item Hamilton Depression Rating Scale (HAMD); secondary outcome measures included the Beck Anxiety Inventory (BAI), and the Mood and Anxiety Symptom Questionnaire (MASQ).

Results: Participants in the iCBT group showed significantly greater reduction of HAMD scores than participants assigned to the MAC group (F(1,27)=7.14, p=.01). There was also a higher remission rate (HAMD score ≤7) in the iCBT group than the MAC group (59% versus 17%). BAI scores decreased significantly more in the iCBT group relative to the MAC group (F(1,26)=4.31, p<.05), as did MASQ total scores (F(1,26=6.40, p=.02). The change in MASQ scores was driven by a significant decrease in the anhedonic depression subscore (F(1,26)=8.94, p<.01).

Conclusions: These preliminary results suggest that an adapted version of an Australian iCBT program is efficacious for the reduction of depressive symptoms in individuals with MDD in the United States. They also indicate a therapeutic effect on symptoms of anxiety and anhedonia. Internet-based treatment approaches may have advantages over standard care in relation to concerns about stigma and barriers to care. Future research directions may test iCBT modules for the treatment of other psychiatric diagnoses, similarly adapted for use in the United States. Given the evolving focus on cost efficiency in American healthcare, future studies should seek to demonstrate the value (i.e., quality/cost) of internet treatment modalities in general, and this application in particular.

Keywords: depression, clinical trial, cognitive-behavioral therapy, remission.

Disclosure: Nothing to Disclose.

T103. Differential Pattern of Response to both Placebo and Antidepressants in Antidepressant Trials Using Two Different Rating Methods

Arif Khan*, James Faucett, Walter A Brown

Northwest Clinical Research Center, Bellevue, Washington

Background: In recent years there has been an upsurge in ‘enhanced interview methods’ oriented toward decreasing the magnitude of placebo response and increasing antidepressant-placebo difference in antidepressant clinical trials. We have noted that such techniques, in particular following structured (SIGMA) interviews and audio-taping interviews for outside rater feedback as compared with more traditional interview techniques may have paradoxical effects of increasing the magnitude of placebo response (Khan et al., 2014) and decreasing antidepressant-placebo differences (Khan et al., 2014). In this context, we considered the possibility that the pattern of response may be different in the trials using the traditional interview techniques compared to enhanced techniques. Using the model developed by Quitkin et al. (1984), we hypothesized that the enhanced interview method would yield an augmented pattern of delayed onset of response with true antidepressant effects appearing after the fourth week of the trial, compared to the traditional interview techniques. In order to verify this possibility we conducted the following analysis.

Methods: We reviewed data from patients randomly assigned to two trials using SIGMA rating methods with audio taping and outside review (N=41 Antidepressant, N=22 placebo) and two trials using traditional semi-structured rating methods (N=48 Antidepressant, N=37 placebo). Each of the four trials was conducted by the same pharmaceutical company, included active medication approved for use by the US FDA and each was conducted using similar trial methodology except for the rating method. Primary analyses compared the percentage symptom reduction early in the trials as well as a comparison of the response to active treatment and placebo during the final four weeks of the trial.

Results: The trials using SIGMA rating methods had a significantly faster onset of response starting with the first week of the trial compared to the trials using traditional rating methods. The percentage symptom reduction after one week was significantly larger in the SIGMA trials for patients assigned to both antidepressants (11.7% SIGMA trials, 6.5% traditional trials, t (df=87)=2.20, p=0.031) and placebo (10.8% SIGMA trials, 3.6% traditional trials, t (df=57)=2.10, p=0.046). On the other hand, the patients assigned to antidepressants in the traditional trials saw greater symptom reduction in weeks 4 through 8 of the trial (18% change from week 4 to 8) in comparison to patients assigned to antidepressants in the SIGMA trials (4% change from week 4 to 8), F (df=2)=7.00, p=0.002. The symptom reduction with placebo was similar and low in weeks 4 through 8 of the trial between traditional (2% change from week 4 to 8) and SIGMA trials (6% change from week 4 to 8), p=NS.

Conclusions: These data suggest that trials using SIGMA interview methods had a response pattern different from that seen in the trials using traditional rating methods. Interestingly, the trials using traditional interviews behaved similarly to the phenomenon observed by Quitkin et al. wherein the true antidepressant effects became apparent after the fourth week of the trial. One of the possibilities for the early onset of response in both arms of the trials using SIGMA interviews is the Hawthorne effect (Lansberger, 1958) that was described more than fifty years ago wherein a change in the experimental conditions may itself significantly influence behavior and improve outcome. This Hawthorne effect, in the circumstance of an antidepressant trial may result in greater improvement by the patients assigned to either treatment condition due to the experimental paradigm itself. Undoubtedly, the type of rating methods has a significant impact on clinical trial outcome. Whether structured, appraised interviews or traditional semi-structured interviews provide a more accurate assessment of symptom change is unclear.

Keywords: Placebo Response, Clinical Trial Rating Methods, Montgomery-Asberg Depression Rating Scale, Hawthorne effect.

Disclosure: Arif Khan, MD, Principal Investigator of over 380 clinical trials sponsored by over 85 pharmaceutical companies and 30 CROs has done no compensated consulting or speaking on their behalf, nor does he own stock in any of these or other pharmaceutical companies. Dr. Khan is not compensated for his role as author. In 2009 Dr. Khan founded Columbia Northwest Pharmaceuticals LLC, and is Medical Director of the company. Columbia Northwest Pharmaceuticals owns intellectual property rights for potential therapies for Central Nervous System disorders and other medical conditions. James Faucett is an employee of the Northwest Clinical Research Center. Walter A. Brown has nothing to disclose.

T104. Efficacy of Aripiprazole Lauroxil, a New Long-acting Injectable Atypical Antipsychotic, Across Three Geographic Regions

Marjie Hard*, Srdjan Stankovic, Robert Risinger, Anjana Bose, Yangchun Du, Jacqueline Zummo, Lisa Corey, Bernard Silverman, Elliot Ehrich

Alkermes, Inc., Waltham, Massachusetts

Background: The worldwide lifetime morbidity risk of schizophrenia is about 1% across diverse geographic, cultural, and socio-economic regions.1 Regional differences in factors such as healthcare and medication availability, medical practice, disease definition and patient characteristics may lead to different treatment response patterns both with respect to efficacy and safety.2,3 The safety and efficacy of Aripiprazole lauroxil, a novel long-acting injectable (LAI) atypical antipsychotic was demonstrated in a large, multicenter study conducted in three regions (North America, Eastern Europe and Asia).

Methods: Post-hoc analyses of efficacy response and adverse events by region were performed using data from a 12-week, double-blind, placebo (PBO) controlled study in patients experiencing acute exacerbation of schizophrenia (DSM-IV criteria) treated with monthly injections of aripiprazole lauroxil 441 mg IM, aripiprazole lauroxil 882 mg IM or matching PBO IM. The primary efficacy endpoint was the change from baseline to Day 85 on the Positive and Negative Syndrome Scale (PANSS) total score using analysis of covariance (ANCOVA) with last observation carried forward (LOCF) approach. Overall treatment response was defined as ≥30% decrease in PANSS total score or CGI-I score of 1 (very much improved) or 2 (much improved) and analyzed using logistic regression model. Incidence of frequently reported treatment-emergent adverse events (TEAEs) was also evaluated by region.

Results: A total of 622 patients were randomized and received study drug (Safety Population) at 107 sites in seven countries across three regions. The regional analyses were conducted in 307 patients from North America (aripiprazole lauroxil 882: n=102; aripiprazole lauroxil 441: n=103; PBO: n=102), 239 patients from Europe (aripiprazole lauroxil 882: n=78; aripiprazole lauroxil 441: n=81; PBO: n=80), and 76 patients from Asia (aripiprazole lauroxil 882: n=28; aripiprazole lauroxil 441: n=23; PBO: n=25). Severity of illness at baseline was within the anticipated range across regions with PANSS total scores (mean±SD) ranging from 85.9±13.2 to 94.7±9.8. Efficacy of aripiprazole lauroxil was demonstrated independently in each of the three regions. Statistically significant and clinically meaningful improvements on the change in PANSS total score at Day 85 were observed with both aripiprazole lauroxil 882 mg and aripiprazole lauroxil 441 mg in each of the regions. The PBO-adjusted differences (LS mean±SE) for aripiprazole lauroxil 882 mg were -6.9±2.4 (p=0.004) in North America, -14.5±2.9 (p<0.001) in Europe and -23.8±6.5 (p<0.001) in Asia. For aripiprazole lauroxil 441 mg, the PBO-adjusted differences were -7.1±2.4 (p=0.003) in North America, -13.9±2.9 (p<0.001) in Europe and -18.4±6.7 (p=0.008) in Asia. The OR (95%CI) for meeting response criteria was 2.5 (1.3,4.8) and 2.7 (1.4,5.0) for aripiprazole lauroxil 882 mg and 441 mg dose, respectively in North America. In Europe, the OR (95%CI) was 4.7 (2.4,9.3) and 3.1 (1.6,6.0) while in Asia the OR (95%CI) was 5.3 (1.5,18.8) and 3.3 (0.95,11.5) for aripiprazole lauroxil 882 mg and aripiprazole lauroxil 441 mg, respectively. The NNT ranged from 2.5 to 5 across the regions. In the study, the most commonly reported TEAE (>5% of patients in either of the aripiprazole lauroxil dose groups) were insomnia, akathisia, headache and anxiety. A similar but not identical TEAE profile was observed across the regions. Insomnia and akathisia met the criteria in all three regions. Headache was reported with this frequency in North American and Europe, and anxiety in Europe only.

Conclusions: Conclusions: The overall results of the study support the conclusion that aripiprazole lauroxil could represent a new treatment option for patients with schizophrenia. These analyses further establish both doses of aripiprazole lauroxil demonstrated robust efficacy with statistically significant and clinically meaningful improvements across all three regions. Observed regional response patterns are consistent with previously reported regional differences in efficacy responses and frequency in AE reporting. References: 1. World Health Organization. Schizophrenia and public health. Accessed on August 12, 2014 from http://www.who.int/mental_health/media/en/55.pdf. 2. Rothwell, PM. External validity of randomised controlled trials: "to whom do the results of this trial apply?" Lancet. 2005;365:82–93. 3. Rothwell, PM. Factors that can affect the external validity of randomised controlled trials. PLoS Clin Trials. 2006;1:e9.

Keywords: aripiprazole lauroxil, schizophrenia, long-acting injectable, efficacy.

Disclosure: The study was sponsored by Alkermes, Inc. All authors are employees of Alkermes. All authors are Alkermes shareholders.

T105. A Comparison of the CANTAB Schizophrenia Battery and the MCCB in Two Phase 2 Clinical Trials of Subjects with Stable Schizophrenia

Pradeep Nathan*, Jeff Baker, Earl Bain, George Haig

Cambridge Cognition and University of Cambridge, Cambridge, United Kingdom

Background: The MATRICS consensus cognitive battery (MCCB) was developed as a consensus approach to testing cognition in clinical trials of cognitive impairments associated with schizophrenia (CIAS). The selected tests were chosen to meet psychometric criteria, including adequate test retest reliability, minimal practice effects, and evidence of sensitivity to drugs. The battery, presented in a mostly pencil-and-paper format, contains ten tests covering seven cognitive domains known to be affected by schizophrenia. The CANTAB schizophrenia battery includes computerized neuropsychological tests presented on a touchscreen computer and selected to assess the domains covered by the MCCB. This study evaluated the relationship between the MATRICS and CANTAB schizophrenia batteries in the baseline evaluations of two phase 2 clinical trials.

Methods: Two multi-center trials in stable subjects with schizophrenia were conducted at approximately 45 sites in the United States. Both investigated the pro-cognitive effects of novel investigational compounds as add-on therapy to antipsychotics. The same design was used in both trials, with one placebo and two active dose groups (planned N=70/group; total N=420 for both trials). Eligibility criteria were consistent with MATRICS guidelines. Subjects were currently taking one or two atypical antipsychotics at the time of study entry. The CANTAB battery (excluding Verbal Recognition Memory) and the MCCB were administered for practice on the initial screening visit (screening visit 1). Baseline assessments for the two batteries were staggered in order to reduce subject burden such that the MCCB was administered on Day -1 and the CANTAB Battery administered approximately 1-2 weeks prior at Screening Visit 2. The baseline UPSA-2 was administered with the MCCB on Day -1, with no practice assessment during screening. Composite scores were calculated for both the MCCB and CANTAB. For the CANTAB a Principal Components analysis was conducted to derive a principle component that was used as the composite score on tasks that were selected based on their clinical meaningfulness based on prior experience of drug effects on these tests and effect size differences between clinical populations and healthy controls. The principal components analysis was conducted in SPSS version 21.0 using the Dimension Reduction procedure with no rotation. Components were selected based on an eigenvalue greater than one. All reported correlations are Pearson r.

Results: The correlations between MCCB and CANTAB for the MATRICS cognition domain scores varied from r=0.28 to r=0.60. The relationship of the MCCB composite with the UPSA-2 agreed with previous published results with r=0.62. The relationship of the first principal component of the CANTAB schizophrenia battery also correlated with the UPSA-2 at r=0.55. The MCCB composite correlated with the principal component of the CANTAB battery at r=0.69.

Conclusions: Using data from two large, US-based phase 2 trials of schizophrenia, we compared the cognitive scores obtained at baseline on over 400 patients on the MCCB and CANTAB cognitive batteries. We found that, while correlation of the domain scores from the two batteries varied considerably across the different cognitive domains, the summary or composite scores were well correlated. These finding may suggest that the batteries may assess different aspects cognition within some of the domains, but summary scores nevertheless can provide similar appraisals of overall cognitive performance. Furthermore both batteries were moderately correlated with functional capacity as measured by the UPSA-2.

Keywords: Cognition, Schizophrenia, CANTAB, Clinical Trials.

Disclosure: J Baker: Employee of Shire Pharmaceuticals. Former employee of Abbvie and Cambridge Cognition; holds Abbvie stock. E Bain: Employee of AbbVie; holds AbbVie stock and stock options. P. Nathan: Employee and shareholder of Cambridge Cognition. G Haig: Employee of AbbVie; holds AbbVie stock and stock options.

T106. Effects of Levomilnacipran ER on Cognition and Functioning in Patients with Major Depressive Disorder: Post Hoc Analysis of a Phase 3 Trial

Philip Harvey*, Carl Gommoll, Changzheng Chen, Alan Lipschutz, Keith Wesnes

University of Miami Miller School of Medicine, Miami, Florida

Background: Major depressive disorder (MDD) is a multidimensional disease characterized by mood symptoms (eg, depressed mood, anxiety, anhedonia, lassitude) and functional impairments. Many patients with MDD show depression-related cognitive deficits which are associated with increased functional disability. Levomilnacipran is a serotonin and norepinephrine reuptake inhibitor with greater in-vitro potency for inhibiting norepinephrine reuptake than serotonin reuptake. Levomilnacipran extended-release (ER) is FDA-approved for the treatment of MDD in adults. In a positive Phase 3 randomized, placebo-controlled trial of levomilnacipran ER 40-120 mg/day, cognition was evaluated as an additional efficacy parameter using the Cognitive Drug Research System (CDRS). This battery of computerized tests includes 2 objective measures of attention: power of attention (POA), which is the ability to focus attention and process information; and continuity of attention (COA), which is the ability to sustain attention (i.e., vigilance). In the prespecified analysis of this study, levomilnacipran ER patients compared with placebo patients showed significantly greater mean improvement in COA scores; improvement in POA scores were also observed for levomilnacipran-ER versus placebo, but the differences did not reach statistical significance. Post hoc analyses of that trial evaluated the effects of levomilnacipran ER on cognitive impairment, functional disability, and depression symptomatology, in patient subsets stratified by baseline POA and COA scores.

Methods: In this 8-week, randomized, double-blind, placebo-controlled, flexible-dose study of levomilnacipran ER (40-120 mg/d), the predefined primary and secondary efficacy outcomes were change from baseline in MADRS total score and Sheehan Disability Scale (SDS) total score, respectively. For the current post hoc analyses, patients in the study population were categorized as cognitively impaired using baseline median scores on the CDRS POA (score ≥1303), COA (score<92), or both. In the cognition subgroups (impaired and not impaired), change from baseline to end of double-blind treatment was analyzed using ANCOVA models with baseline score and treatment group for COA and POA scores, the MADRS total score, and the SDS total and subscale scores (Work/School, Social Life, Family/Home life). Missing values were imputed using the last observation carried forward method.

Results: Patients with impaired POA and/or COAhad similar baseline MADRS and SDS total scores as those without cognitive impairments. In the subgroup of patients with impaired POA at baseline, levomilnacipran ER patients showed significantly greater improvement than placebo patients in cognitive function (attention measures), as indicated by least squares mean differences (LSMDs) in score changes in POA (-111.96, P=.0164) and COA (3.43, P=.0001). In the subgroup of patients with impaired COA at baseline, the LSMD for levomilnacipran versus placebo was significant for COA (2.50, P=.0083) but not POA (-90.69, P=.0768). A total of 127 patients met the criteria for combined POA and COA impairment at baseline, and LSMDs for CDRS outcomes significantly favored levomilnacipran ER over placebo in this subgroup (POA, -155.31, P=.0130; COA, 3.47, P=.0024). In patients without baseline cognitive impairment, smaller and nonsignficant levomilnacipran ER treatment effects in POA and COA were seen. In patients with POA or COA impairment at baseline, levomilnacipran ER patients compared with placebo patients showed significantly greater improvement in SDS scores across all functional domains: POA impaired group (LSMDs: SDS total, 3.42; Work, 1.21; Social, -1.25; Family, -0.93; all P<.05); COA impaired group (LSMDs: SDS total, 3.46, Work, 1.12; Social, -1.29; Family, -1.16; all P<.05). Smaller SDS improvements were seen with levomilnacipran ER versus placebo in patients without baseline POA or COA impairment; the differences from placebo were not statistically significant. Improvement in depressive symptoms, as measured by MADRS total score, was found with levomilnacipran ER in patients with baseline POA impairment (-3.50, P=.0170) or COA impairment (-3.67, P=.0191). In patients without baseline POA or COA impairments, smaller and nonsignficant levomilnacipran ER treatment effects were seen on MADRS total score.

Conclusions: Although baseline depression symptom severity and functional impairment were similar in patients with and without POA and/or COA impairments, those with POA and/or COA impairments experienced significantly greater improvements with levomilnacipran ER compared with placebo on CDRS attention scores, SDS total and subscale scores, and MADRS total score. In comparison, improvements with levomilnacipran ER relative to placebo were smaller in patients without cognitive impairment at baseline, potentially due to higher placebo response in patients without cognitive deficits. Further research is warranted to investigate the relationship between depression symptoms, cognitive impairment, and functional disability.

Keywords: Depression, Cognition, Disability, Anridepressant.

Disclosure: Philip D. Harvey: Consultant to: Abbvie, Boeheringer-Ingelheim, Forest Labs, Forum Pharma, Genentech, Lundbeck, Otsuka America, Roche, Sanofi, Sunovion A. Lipschitz, C. Gommoll, and C. Chen are employees of Forest Research Institute, a subsidiary of Actavis, PLC. K. Wesnes was employed by Bracket at the time this trial was conducted. Bracket provided the CDR System on which the study results are based.

T107. Noradrenergic-related Depression Symptoms in Patients with Major Depressive Disorder: Post Hoc Analysis of 5 Clinical Trials of Levomilnacipran Extended-release

Pierre Blier*, Carl Gommoll, Changzheng Chen, Alan Lipschitz

University of Ottawa Institute of Mental Health Research, Ottawa, Canada

Background: Major depressive disorder (MDD), a complex illness with varied symptomatology, is one of the leading causes of disability worldwide. While some depression symptoms are associated with serotonergic dysfunction, others (eg, lassitude, decreased energy, tiredness, psychomotor retardation, and concentration difficulties) are more closely associated with deficient noradrenergic (NA) transmission (Montgomery, Neuropsychatr Dis Treat 2011) and may be particularly important in MDD-associated functional impairment. Levomilnacipran is a serotonin and norepinephrine reuptake inhibitor (SNRI) with greater in vitro potency for inhibiting norepinephrine reuptake relative to serotonin reuptake. Levomilnacipran extended-release (ER) is currently approved for the treatment of MDD in adults. In addition to improving depressive symptoms, patients treated with levomilnacipran ER compared with placebo showed significantly greater improvement in functional impairment, as measured by the Sheehan Disability Scale (SDS) (Sambunaris, Int Clin Psychopharmacol 2014). The effect of levomilnacipran on NA-related symptoms and on functional impairment in patients with low and high levels of NA-related symptoms was evaluated in a pooled post hoc analysis of 5 levomilnacipran ER studies.

Methods: Data were pooled from 5 randomized, double-blind, placebo-controlled trials (2 fixed- and 3 flexible-dose studies) in which adults with MDD received levomilnacipran ER (40-120 mg/d) or placebo for 8 or 10 weeks. The post hoc analyses included all randomized patients who had ≥1 available post-baseline Montgomery Åsberg Depression Rating Scale (MADRS) assessment and ≥1 post-baseline Hamilton Depression Rating Scale (HAMD) assessment. A cluster of NA-related symptoms was defined based on the following items: MADRS items 6 (Concentration Difficulties), 7 (Lassitude), and 8 (Inability to Feel); HAMD items 7 (Work/Activities), 8 (Retardation), and 13 (General Somatic Symptoms). The NA symptom cluster score was defined as the sum of all individual NA symptom item scores. Mean change in the NA symptom cluster score was analyzed at end of treatment using the last observation carried forward approach. Odds ratios (ORs) were calculated for ≥50% improvement from baseline in NA symptom cluster score. The subgroup of patients with “high” NA symptom cluster scores at baseline were defined as those who met all of the following criteria: score ≥4 for MADRS items 6, 7, and 8; score ≥2 for HAMD items 7 and 8; score ≥1 for HAMD item 13. Patients who did not meet these criteria comprised the “low” NA symptom cluster score subgroup. Mean changes in SDS total score and subscores (Work/School, Social Life, Family/Home Life) were analyzed in the subgroups of patients with high and low NA symptom cluster scores at baseline using a mixed-effects model for repeated measures. Effect sizes (Cohen’s d) were estimated for all mean score changes.

Results: The post hoc pooled analyses included 2538 patients (placebo, n=1017; levomilnacipran ER, n=1521). The mean baseline NA symptom scores were similar between treatment groups (placebo, 17.3; levomilnacipran ER, 17.5). At end of treatment, the improvement in the NA symptom cluster score was significantly greater with levomilnacipran ER than with placebo (least squares mean difference [LSMD]: -1.61, P<.0001), with an effect size of 0.25. Between-group differences were also significant on all of the individual MADRS and HAMD items that composed the NA symptom cluster score (all P≤.0002). The proportion of patients with ≥50% improvement in NA symptom cluster score was 43.5% and 33.5% for levomilnacipran ER and placebo, respectively (OR=1.56, 95% CI=1.32 to 1.84, P<.0001). In patients with high NA symptom cluster scores at baseline, levomilnacipran ER-treated patients showed significantly greater improvement relative to placebo-treated patients on SDS total and subscore changes (LSMD: SDS total, 3.07, P<.0001; Work, 1.09, P<.0001; Social, 0.97, P=.0002; Family, 0.84, P=.0011). In patients with low NA symptom cluster scores at baseline, levomilnacipran ER also showed significant advantage over placebo in improving SDS total and subscores (LSMD: SDS total, -1.82, P<.0001; Work, 0.62, P<.0001; Social, -0.57, P<.0001; Family, 0.54, P<.0001). Treatment effect sizes were greater in patients in the high NA symptom cluster score subgroup than in the low NA symptom cluster score subgroup (total: 0.39 vs 0.23; Work: 0.41 vs 0.20; Social: 0.35 vs 0.22; Family: 0.30 vs 0.19).

Conclusions: In this post hoc pooled analyses, levomilnacipran ER significantly improved NA-related symptoms in adults with MDD, as measured by MADRS and HAMD items, using both individual item scores and a composite NA symptom cluster score. In patients with both high and low NA symptom cluster scores at baseline, treatment with levomilnacipran ER compared with placebo resulted in significantly greater improvements in functional impairment as measured by the SDS total score and subscores. Further research is warranted to explore the relationship between NA symptoms and functional impairment.

Keywords: major depressive disorder, levomilnacipran, noradrenergic, serotonin and norepinephrine reuptake inhibitor.

Disclosure: P. Blier received a research grant and honoraria for consulting for Forest in Canada and the USA. A. Lipschitz, C. Gommoll, and C. Chen are employees of Forest Research Institute, a subsidiary of Actavis, plc.

T108. Randomized Controlled Trial of N-Acetylcysteine for Cognition and EEG Correlates in Schizophrenia

Michael Davis*, Jonathan Wynn, Katherine Weiner, Gerhard Hellemann, Michael Green, Stephen Marder

University of California, Los Angeles, Semel Institute for Neuroscience & Human Behavior, Los Angeles, California

Background: Schizophrenia is associated with significant neurocognitive deficits which predict poor functioning. Electrophysiological (EEG) biomarkers related to cognitive dysfunction in schizophrenia have been linked to abnormal NMDA glutamatergic neurotransmission. N-acetylcysteine (NAC) is an amino acid that modulates NMDA receptor function and has been previously shown to improve clinical symptoms of schizophrenia. The current pilot study evaluated whether NAC could improve cognitive function and related EEG biomarkers.

Methods: 26 adult outpatients with schizophrenia were randomized (double-blind) to receive 1200mg NAC or placebo twice daily for 8 weeks as an adjunct to antipsychotic treatment. Patients were assessed at baseline, 4 weeks, and 8 weeks on measures of clinical psychopathology (Positive and Negative Syndrome Scale (PANSS); Clinical Assessment Interview for Negative Symptoms (CAINS); Clinical Global Impression (CGI)), cognition (MATRICS Consensus Cognitive Battery (MCCB)), and side effects. Participants also underwent an EEG assessment battery examining mismatch negativity, gamma oscillation power and synchrony, and visual cortical neuroplasticity. Generalized linear mixed models for repeated measures were used to compare effects of NAC vs. placebo across the 8 weeks.

Results: 17 participants completed all study visits (8 who received NAC and 9 who received placebo). There were no significant effects of NAC vs. placebo on the PANSS total score (p=.494) or subscales; the CAINS (p=.211 on motivation/pleasure and p=.751 on expression scales); the CGI (p=.595); or the MCCB composite score (p=.830) or subdomain scores. EEG data on mismatch negativity, gamma oscillations, and the visual cortical plasticity paradigm are undergoing analysis and will be presented. The treatment was generally well-tolerated, with gastrointestinal complaints representing the most commonly reported side effect.

Conclusions: This study demonstrates the feasibility of assessing clinical symptoms and EEG biomarkers in a longitudinal clinical trial evaluating a potential treatment for schizophrenia. In our limited sample size, we found no evidence supporting NAC for improving clinical symptoms or cognitive deficits of schizophrenia. Analysis of EEG data will determine whether NAC had measurable effects on electrophysiological biomarkers related to cognition and NMDA glutamatergic function.

Keywords: N-acetylcysteine, cognition, neuroplasticity, schizophrenia.

Disclosure: Dr. Green has been a consultant to AbbVie, DSP, Forum and Roche, and he is on the scientific advisory board of Mnemosyne. He has received research funds from Amgen. Dr. Marder has been a consultant to Forum, Targacept, Lundbeck, Roche, Genentech, Boehringer-Ingelheim, and Takeda. He received research support from Synchroneuron and Genentech.

T109. Milnacipran vs. Placebo in Adult Autism Spectrum Disorder: Impact on Hyperactivity/Impulsivity Domain

Rachel Noone*, Casara Ferretti, Bonnie Taylor, Emma Racine, Eric Hollander

Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, New York

Background: Developmental dysfunction of the locus coeruleus noradrenergic locus coeruleus (NA-LC) system has been proposed as a possible cause of autism spectrum disorders (ASD). Anecdotal parent reports, clinical observations and formal studies have found that up to 40% of children with ASD experience improvement of autistic behavior during fever, which may provide clues to the pathophysiology and potential therapeutic interventions for ASD. Accordingly, we posit that the fluctuations in behavioral states observed during febrile episodes suggest the involvement of the NA-LC system, a widespread and versatile neuromodulatory system that we suggest is common to febrigenesis and the modulation of the core symptoms of ASD. Milnacipran (MLN) is predominantly a norepinephrine reuptake inhibitor that may enhance function of the NA-LC system. We hypothesize that MLN will re-activate the NA-LC system in patients with ASD in a fashion similar to that seen with fever, thus resulting in improvements in attention, irritability, repetitive behaviors and social cognition.

Methods: A 12 week randomized, double-blind, placebo-controlled trial of MLN up to 200mg/day was performed in 10 adult ASD subjects aged 19 – 41 with a DSM-IV TR diagnosis, confirmed by ADI-R and ADOS. Comorbid medical, neurological and psychiatric illnesses with the exception of ADHD and OCD were excluded. Visits occurred at two week intervals and at each visit, Conners Adults ADHD Rating Scale-Observer Version (CAARS-O), Aberrant Behavior Checklist (ABC), Clinical Global Impressions Improvement Scale (CGI-I), Yale-Brown Obsessive Compulsive Scale (YBOCS) and Diagnostic Analysis of Nonverbal Accuracy-2 (DANVA-2) were measured in each subject.

Results: Treatment with MLN versus placebo produced improvements in the impulsivity and the hyperactivity domains of Attention Deficit Hyperactivity Disorder (ADHD) symptoms, but not in the attention domain. No changes were observed on global (CGI-I), social communication (DANVA), or repetitive behavior (YBOCS) domains.

Conclusions: Milnacipran vs. placebo resulted in improvement in the hyperactivity/impulsivity associated symptom domains of ASD, which may reflect MLN’s mechanism of action on the noradrenergic system. There were no clinically significant changes in the repetitive behaviors and social cognition domains of ASD, although the study was limited by a small sample size. It is possible that MLN might be more effective in managing these symptoms in a subgroup of patients stratified for high hyperactivity/impulsivity symptoms at baseline. Further studies will address proof of mechanism of the NA-LC system using task-elicited fMRI.

Keywords: Autism Spectrum Disorder, ASD, Milnacipran, Noradrenergic System.

Disclosure: This research was funded by an investigator initiated grant from Forest Pharmaceuticals, Inc.

T110. A Pragmatic Analysis Comparing Once-monthly Paliperidone Palmitate Versus Daily Oral Antipsychotic Treatment in Patients with Schizophrenia

Larry Alphs, Carmela Benson, Cynthia Bossie, Lian Mao, H. Lynn Starr*

Ortho-McNeil Janssen Scientific Affairs, LLC, Titusville, New Jersey

Background: Clinical trials, a primary tool by which treatments are evaluated, can be broadly categorized as pragmatic or explanatory, although many study designs contain elements of both. Pragmatic trials are typically designed to evaluate the effectiveness of interventions under real practice conditions, whereas explanatory trials aim to test whether an intervention works under highly controlled and well-defined conditions. While both are important, pragmatic results can be more generalizable and relevant to routine practice settings. This analysis applies the concept of pragmatism to a study comparing active treatments for patients with schizophrenia, highlighting the importance and potential impact of pragmatic research on psychiatric practice. The Paliperidone palmitate Research In Demonstrating Effectiveness (PRIDE) study (NCT01157351) compared once-monthly injectable paliperidone palmitate (PP) to daily oral antipsychotic therapy (OA) in a study with real-world consequences as end points (eg, arrest/incarceration, hospitalization, treatment discontinuation) in a schizophrenia population inclusive of patients normally not included in trials (ie, those with a history of incarceration and coexisting substance abuse disorder). The study allowed flexibility in treatment and management decisions. PP significantly delayed time to first treatment failure versus OA, with a median difference of 190 days between groups (P=0.011). The use of more pragmatic outcomes, such as cumulative treatment failures over the entire study period, is of particular public health interest.

Methods: PRIDE, a 15-month prospective, open-label, event monitoring board (EMB)-blinded study, randomized subjects with schizophrenia to flexibly dosed PP (78–234 mg) or a daily OA assigned from a preselected group of 7 of the most commonly prescribed antipsychotics: aripiprazole, haloperidol, olanzapine, paliperidone, perphenazine, quetiapine, and risperidone. Treatment failure was defined as arrest/incarceration, psychiatric hospitalization, suicide, discontinuation due to inadequate efficacy or safety/tolerability, treatment supplementation due to inadequate efficacy, or increase in psychiatric services to prevent imminent psychiatric hospitalization. Subjects were allowed to continue study participation after a treatment failure event, regardless of whether they were maintained on their initial randomized treatment. This pragmatic analysis included all data from the intent-to-treat (ITT) population from randomization until month 15. Multiple treatment failures from the same subject were analyzed as recurrent events. The cumulative mean function (CMF) of treatment failure was a function of time, defined as the average number of treatment failures per subject in a given time interval following randomization. A proportional intensity model including a term for treatment was used to compare the CMF of any treatment failure, arrest/incarceration, or psychiatric hospitalization between the 2 treatment groups. No adjustment was made for multiplicity.

Results: The population included 226 subjects in the PP group and 218 in the OA group, of whom 41.2% and 40.4% completed 15 months of study follow-up, respectively. Risks of having multiple treatment failures, arrests/incarcerations, or psychiatric hospitalizations were all significantly lower with PP vs OA during the study period (P<0.01 for each comparison). At the end of the 15-month study, the average number of any treatment failure events was 1.09 in the PP group and 1.51 in the OA group; average numbers of arrests/incarcerations per subject were 0.72 and 1.00, respectively; and average numbers of psychiatric hospitalizations per subject were 0.17 and 0.31, respectively. Most common treatment-emergent adverse events (TEAEs; PP vs OA), were injection-site pain (18.6% vs 0%), insomnia (18.6% vs 12.4%), increased weight (13.3% vs 7.3%), akathisia (11.5% vs 7.8%), and anxiety (11.1% vs 8.3%). For PP and OA groups, respectively, TEAEs leading to treatment discontinuation occurred in 12.4% and 8.7% of subjects, and serious TEAEs occurred in 18.6% and 24.3% of subjects. One death was reported in the PP group, but it was considered by the investigator as unlikely related to study medication.

Conclusions: In this pragmatic analysis, PP had a more robust treatment effect versus OA as assessed by cumulative events of any treatment failure, arrest/incarceration, or psychiatric hospitalization over a 15-month treatment period. This information can be of value in public health and therapeutic treatment risk:benefit comparisons of these treatment options.

Keywords: paliperidone palmitate, schizophrenia, pragmatic analysis.

Disclosure: L. Alphs is an employee of Janssen Scientific Affairs, LLC, and a Johnson & Johnson stockholder. C. Benson is an employee of Janssen Scientific Affairs, LLC, and Johnson & Johnson stockholder. C. Bossie is an employee of Janssen Scientific Affairs, LLC, and Johnson & Johnson stockholder. L. Mao is an employee of Janssen Research & Development, LLC, and a Johnson & Johnson stockholder. H. L. Starr is an employee of Janssen Scientific Affairs, LLC, and a Johnson & Johnson stockholder.

T111. Effect of Paliperidone Palmitate Once-monthly in Improving and Maintaining Functioning in Subjects with Schizoaffective Disorder Using the Domains of the Personal and Social Performance Scale

Dong-Jing Fu*, Ibrahim Turkoz, R. Bruce Simonson, David Walling, Nina Schooler, J. P. Lindenmayer, Larry Alphs

Ortho-McNeil Janssen Scientific Affairs, LLC, Titusville, New Jersey

Background: Psychosocial function deficits, including poor social interaction, difficulties in maintaining relationships, and/or inadequate workplace performance, characterize schizophrenia and schizoaffective disorder (SCA). Improvement in these measures is important for achieving good long-term outcomes. A randomized, double-blind (DB), placebo-controlled clinical trial found that paliperidone palmitate once monthly (PP1M) significantly delayed relapse to psychotic, depressive, and manic symptoms; reduced the risk of relapse; and better maintained functioning in patients with SCA when given either as monotherapy or adjunctive to antidepressants and/or mood stabilizers. The present analysis evaluated the role of PP1M in improving the specific functioning domains of the Personal and Social Performance (PSP) scale during the open-label (OL) phase and maintaining these changes once patients were stabilized on PP1M during the DB relapse-prevention phase. The PSP is a validated clinician-reported instrument to assess functioning in schizophrenia.

Methods: This randomized, DB, placebo-controlled, international study (NCT01193153) included subjects who met the Structured Clinical Interview for DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition) Disorders (SCID)-confirmed diagnosis of SCA experiencing an acute exacerbation of psychotic and mood symptoms. Subjects stabilized with PP1M treatment during a 25-week OL phase could enter the 15-month DB relapse-prevention phase. Subject functioning was evaluated using the PSP, scored from 1 to 100; higher scores indicated better functioning based on evaluation of 4 domains: socially useful activities, personal/social relationships, self-care, and disturbing/aggressive behaviors. Each PSP domain was assessed on a 6-point severity scale: absent, mild, manifest, marked, severe, and very severe. Functioning was evaluated for subjects who were stabilized on PP1M during the OL phase and who entered the 15-month DB relapse-prevention phase. Comparisons of the PSP domain scores were conducted using a Cochran–Mantel–Haenszel chi-square test with modified ridit scores, and no adjustments were made for multiplicity.

Results: Only the 334 subjects who were stabilized on PP1M and who entered the DB phase were included in this analysis. At OL baseline, the percentage of subjects with absent through mild symptoms for the 4 PSP domains (ie, socially useful activities, personal and social relationships, self-care, and disturbing and aggressive behaviors) were 7.5%, 10.2%, 67.7%, and 58.4%, respectively. After OL treatment with PP1M (OL end point=DB baseline), these percentages increased to 63.5%, 76.3%, 97.6%, and 99.4%, respectively. During the DB phase, a significant difference from DB baseline to end point favoring PP1M over placebo was observed for all 4 PSP domains in the proportion of subjects with absent to mild symptoms. For socially useful activities, which include work and study, the percentage was 60.6% at DB baseline and 48.2% at end point for placebo and 66.5% and 64.0% for PP1M (P=0.004). For personal and social relationships, the percentage was 76.5% at DB baseline and 56.5% at end point for placebo and 76.2% and 70.8% for PP1M (P=0.008). For self-care, the percentage was 98.2% at DB baseline and 83.9% at end point for placebo and 97.0% and 95.0% for PP1M (P=0.001). For disturbing and aggressive behaviors, the percentage was 99.4% at DB baseline and 88.1% at end point for placebo and 99.4% and 98.1% for PP1M (P<0.001).

Conclusions: During OL treatment with PP1M the specific functioning domains of the PSP improved in subjects with an acute exacerbation of SCA. These improvements were maintained with PP1M compared to placebo during the DB relapse-prevention phase for all 4 domains: socially useful activities, personal/social relationships, self-care, and disturbing/aggressive behaviors.

Keywords: Paliperidone palmitate once-monthly, schizoaffective disorder, patient functioning.

Disclosure: D-J Fu is an employee of Janssen Scientific Affairs, LLC, and a Johnson & Johnson stockholder. I. Turkoz is an employee of Janssen Pharmaceutical Research and Development, LLC, and a Johnson & Johnson stockholder. R. B. Simonson is an employee of Janssen Pharmaceutical Research and Development, LLC, and a Johnson & Johnson shareholder. D. Walling has received grant support and/or has participated in clinical trials for Abbott, AstraZeneca, Elan, Eisai, EnVivo, Otsuka Pharmaceuticals, Janssen, Lundbeck, Eli Lilly and Company, Bristol-Myers Squibb, Merck, Pfizer, Shire, and Sunovion, and has served as consultant/speaker for Otsuka Pharmaceuticals, Janssen, ePharmaSolutions, ProPhase, and Amgen. N. Schooler has received grant support from Otsuka and Neurocrine, and has participated in advisory boards and consultations with Graham Boechk Foundation, Lundbeck International Neuroscience Foundations, Genentech/Roche, and EnVivo. J-P Lindenmayer has received grant/research support from Janssen, Eli Lilly and Company, Johnson & Johnson, Pfizer, Otsuka Pharmaceuticals, Dainippon Sumitomo, Neurocrine, Envivo and Roche; and is a consultant for Janssen. L. Alphs is an employee of Janssen Scientific Affairs, LLC, and Johnson & Johnson stockholder.

T112. rTMS Using Summation of Electromagnetic Fields from a Two-coil Array: Efficacy for Treatment Resistant Major Depressive Disorder

Linda Carpenter*, Scott Aaronson, Gregory Clarke, Paul Holtzheimer, Clark Johnson, William McDonald, Elizabeth Stannard, M. Bret Schneider

Butler Hospital, Brown Medical School, Providence, Rhode Island

Background: Since FDA clearance of the first device for delivery of therapeutic repetitive Transcranial Magnetic Stimulation (rTMS) in 2008, rTMS has continued to grow in popularity and emerge as a standard of care for individuals with major depressive disorder (MDD) who do not benefit from, or who are unable to tolerate, antidepressant pharmacotherapy. A second generation rTMS system by Cervel Neurotech (Redwood City, CA) uses a two-coil array to generate electrical field potentials at multiple brain network locations, both deep and superficial. The positioning of coils in this device is designed to produce both a summation of the fields at relatively deep cortical sites and avoidance of non-targeted areas.

Methods: A randomized, double-blind, sham-controlled, parallel-groups clinical trial was conducted to examine the safety and efficacy of Cervel rTMS as the sole or adjunctive treatment of MDD in adult patients (n=92). Both treatment intolerant and treatment resistant participants (who failed to achieve satisfactory improvement from at least one prior antidepressant medication at or above the minimal effective dose) were enrolled at 6 US sites. Adults meeting eligibility criteria received 20 daily rTMS treatments over 4 weeks. The majority of participants got rTMS as an adjunct to stable (ineffective) pharmacotherapy. Targeted coil centers were positioned over left dorsolateral prefrontal cortex (dlPFC) and posterior dorsomedial prefrontal cortex (pdmPFC) with 10 Hz stimulation (4-second trains, 26 inter-train interval); maximum summated power for both coils was ≤ 120% of resting motor threshold. Durability of effect was measured 1 month after completion of the final treatment (#20). Serial HAMD-24 assessments were administered electronically via algorithm-based computer program, and the primary efficacy endpoint was change in HAMD-24 score from baseline to endpoint (4 weeks).

Results: Data from n=75 patients were included in the per-protocol (PP) sample. Mean HAMD-24 improvement at week 4 for the active treatment group (-15.1±9.6) was significantly greater (p=0.033; Cohen’s d=0.5) than for the sham group (-10.4±8.7). Week 4 HAMD-24 response rates were 55.3% for active versus 32.4% for sham (p=.063), and remission rates were 26.3% versus 18.9% (not significant). The proportion of patients who met response criteria at both the end of 20 treatments (week 4) and 1 month later was 41.7% for those who received active rTMS versus 19.4% who received sham (p=.072).

Conclusions: Positive results were found in the first controlled clinical trial of rTMS therapy using a new device designed to optimally direct and summate the magnetic energy fields generated by a two-coil array. Despite the modest sample size, results on the primary endpoint support antidepressant efficacy of Cervel rTMS for pharmaco-intolerant or -resistant MDD, with an effect size comparable to those reported for large trials using devices with FDA clearance. Analysis of safety data and secondary measures are currently underway.

Keywords: rTMS, depression, stimulation, treatment resistant.

Disclosure: This clinical trial was sponsored by Cervel Neurotech, Inc, the manufacturer of the rTMS device used to deliver investigational treatment. M. Bret Schneider is a founder, employee, and director of Cervel Neurotech Inc; he has stock in the company and patents related to its technologies. Elizabeth Stannard is paid consultant to Cervel Neurotech Inc. The other authors are principal investigators participating sites, employed by institutions which had a contract with Cervel Neurotech, Inc. to pay for conduct of this clinical trial. Dr. Holtzheimer has received consulting fees from St. Jude Medical Neuromodulation. Drs. Aaronson and Carpenter have received consulting fees from Neuronetics, and Drs. Aaronson, Carpenter and McDonald have received research support from Neuronetics. Drs. Aaronson and Carpenter have received research support from NeoSync. Dr. McDonald's institution (Emory University) holds a patent and receives royalties>$10,000 for the magnetic stimulator on the Neurostar rTMS device. Drs. McDonald and Holtzheimer have a contract with Oxford University Press to co-edit a book on the Clinical Guide to Transcranial Magnetic Stimulation in the Treatment of Depression.

T113. Do Implantable Cardioverter Defibrillators Contribute to New Depression Symptoms? A One Year Prospective Study

Revital Amiaz*, Elad Asher, Guy Rozen, Efrat Czerniak, Michael Glikson, Mark Weiser

Chaim Sheba Medical Center, Tel Aviv University, Ramat Gan, Israel

Background: Implantable Cardioverter Defibrillator (ICD) operations are an established form of therapy for primary and secondary prevention of sudden cardiac death. Co-morbidity of depressive symptoms and ICD implantation or ICD therapeutic shocks was found in several studies, but other studies did not find this correlation. We conducted a one year prospective study to evaluate new onset of psychopathological symptoms after ICD implantation.

Methods: We assessed 155 outpatients who were scheduled for ICD implantation and diagnosed them using the Mini International Neuropsychiatric Interview (MINI). Depression was evaluated using the HAM-D and patient’s attitude towards the ICD – using a Visual Analog Scale (VAS); 1 representing an extremely positive attitude and 5 an extremely negative one.

Results: Patients' mean age was 64 years (±12.4); 134 (86%) were men. An overall of 18 patients did not finish the one year follow up for various reasons: 10 patients died before the 3 months evaluation, two patients died before the end of the first year follow up. Five implantations were canceled, one of the patients had heart transplantation and one patient suffered from infection and his ICD was extracted before the 3 months evaluation. According to the MINI diagnosis, at baseline, three patients suffered from MDD and ten from dysthymia. Depressive symptoms (HAM-D ≥8) were found in 57 (37%) patients. Significant improvement in HAM-D mean scores were found between baseline, three months and one year after implantation (HAM_D at baseline=6.50±6.35; HAM_D after one year=4.10±5.27; F(2,100)=16.48; p<.001). Only seven (4.5%) patients experienced shocks during the one year follow-up period, but there was no significant difference in their psychiatric symptoms relative to the rest of the study group. The attitude towards the ICD device was overall positive and constant over the time-course of three months (2.51±0.99) and one year (2.19±0.98).

Conclusions: We showed that ICD significantly decreased HAM-D mean scores at three months and one year over baseline. We also showed that patients who have experienced ICD shocks did not suffer from additional depressive symptoms after implantation. In summary, the overall attitude towards the device was positive and the ICD shocks were not related to new depressive symptoms.

Keywords: ICD, Depression, ICD Shocks, HAM-D.

Disclosure: Nothing to Disclose.

T114. Randomized Controlled Trial to Evaluate the Safety and Efficacy of the Orexin Receptor Antagonist Filorexant in Patients with Painful Diabetic Neuropathy

W. J. Herring, J. Ge, S. Jackson, D. Hewitt, K. M. Connor, D. Michelson*

Merck, North Wales, Pennsylvania

Background: The orexin system plays a key role in regulating wakefulness. Orexin receptor antagonists such as filorexant have been shown to be effective for the treatment of insomnia. In addition to its role in regulating wakefulness, the orexin system has recently been associated with nociceptive processing, in line with observations of orexinergic innervation of regions responsible for modulating pain signaling. Furthermore, chronic neuropathic pain is associated with sleep disturbance, and sleep interference is commonly a secondary outcome measure in pain studies. Based on these considerations, we performed a proof-of-concept study to evaluate whether orexin receptor antagonism might be effective for treating pain in patients with painful diabetic neuropathy (PDN). We evaluated a 10 mg dose of filorexant which was previously shown to be effective for treating insomnia.

Methods: Double-blind, placebo-controlled, enriched enrollment, randomized withdrawal (EERW) study. Patients ages 18-75 with a primary diagnosis of PDN reporting moderate to severe symmetric distal extremity sensorimotor polyneuropathy for at least 6 months were eligible. The trial consisted of a 1-week screening period where eligible patients entered a 2-week single-blind active run-in period with filorexant 10 mg (every night at bedtime) followed by a 2-week double-blind treatment period in which patients were either switched to placebo or remained on filorexant (1:1 ratio). The primary efficacy endpoint was time to efficacy failure among patients who were “primary responders” to filorexant during the single-blind run-in. A primary responder was defined as a patient who achieved a ≥30% decrease in average evening pain intensity score relative to their screening period score. An efficacy failure for primary responders was defined as an occurrence of 3 consecutive days, or 4 days in a row with only one day missing and the other 3 days, satisfying the following criteria: daily evening pain intensity score ≥4 and an increase of ≥30% in daily evening pain intensity relative to the mean of the daily baseline pain measured on a numeric rating scale of 0 to 10. Other assessments included time to efficacy failure among “responders” who achieved a ≥20% decrease in average evening pain intensity score relative to their screening period score (with efficacy failure defined similarly to the primary endpoint except for an increase of ≥20% in pain intensity) and mean change from baseline in evening pain intensity averaged over the last three days of the 2-week treatment among primary responders and all responders.

Results: A total of 182 patients were enrolled into the study and all received filorexant during single blind run-in. Of these, a total of 170 patients (87 for filorexant and 83 for placebo) continued into the double-blind treatment phase and treated with study medication, of whom 65 (38.2%) were primary responders and 88 (51.8%) were responders. While trending in favor of filorexant, the comparison to placebo for the proportion of patients with efficacy failure during the double-blind treatment phase was not statistically significant among primary responders (24.3% vs. 32.1%, p-value=0.411) or all responders (34.0% vs. 43.9%, nominal p-value=0.302). Likewise, comparisons of the mean change from baseline in evening pain intensity scores averaged over the last 3 days of double-blind treatment phase among primary responders (estimated treatment difference=-0.587, p-value=0.269) or all responders did not provide evidence of efficacy (estimated treatment difference=-0.687, p-value=0.108). During the single-blind filorexant run-in period, 45/182 (24.7%) of patients reported any adverse events (AEs), 3 patients (1.6%) reported serious AEs, and 7 patients (3.8%) discontinued due to an AE. Three patients (1.6%) reported pre-specified Events of Clinical Interest (ECIs) which included excessive daytime sleepiness, sleep paralysis, and fall. The most commonly reported AEs during the single-blind run-in included somnolence (4.9%), headache (3.8%), and fatigue (3.3%). During the double-blind treatment period a higher percentage of patients in the filorexant group (N=88) versus the placebo group (N=82) reported AEs and drug-related AEs (23.9% vs. 13.4% and 8.0% vs. 3.7%, respectively). One patient in the filorexant group reported an ECI of suicidal ideation.

Conclusions: In patients with PDN, filorexant 10 mg was not demonstrated to be superior to placebo as measured by time to efficacy failure among primary responders and all responders or with regard to the mean change from baseline in evening pain intensity averaged over the last three days of the 2-week treatment among primary responders and responders. Filorexant 10 mg administered once daily in the evening was generally well tolerated for up to 4 weeks of treatment.

Keywords: orexin receptor antagonist, neuropathic pain, clinical trial.

Disclosure: All authors are full time employees of Merck & Co, Inc.

T115. Is it Safe to Conduct Antidepressant Medication Washout inTreatment-resistant Depression (TRD)?

Kyle Lapidus*, Richard Koch, Dan Iosifescu, James Murrough, Rayan Al Jurdi, Sanjay Mathew

Mount Sinai Medical Center, New York, New York

Background: Many clinical trials of novel pharmacotherapies for major depressive disorder (MDD) require patients to discontinue use of antidepressant medications prior to randomization. The medications to be discontinued are presumed to be ineffective, since the subject still meets criteria for participation in the clinical trial. The rationale for a medication-free washout period hinges on a desire to limit drug-drug interactions for a novel pharmacotherapy; washout also limits confounds for biomarker or neuroimaging studies. Medication washout is also sometimes useful in clinical practice and would serve to avoid unnecessary polypharmacy. However, drug washout may lead to withdrawal effects and worsening mood and may pose an increased safety risk for patients, including suicidality. Hence, the inclusion of a washout, particularly in outpatients, presents an important challenge and concern. Unfortunately, there is a scarcity of data regarding the effects and safety of antidepressant (AD) and more general psychotropic (PT) washout in outpatients with MDD, particularly those with treatment-resistant depression (TRD). Here we examine the washout phase of a multicenter clinical trial of intravenous ketamine to test the hypothesis that medication washout can be safely conducted in outpatients with TRD.

Methods: Seventy-three subjects with TRD in a current major depressive episode were randomized to receive IV ketamine or midazolam following a minimum one-week washout of all psychotropic medications, except for non-benzodiazepine hypnotics. Pre- and post-washout depression severity measures were analyzed from 63 subjects. Repeated-measures analyses of variance (ANOVA) were performed to investigate within-subjects differences in depression severity as measured by Montgomery-Åsberg Depression Rating Scale (MADRS) scores from pre-washout (Screening Visit) to post-washout (Randomization Visit - Baseline) and between-subjects differences among patients who underwent AD washout and/or PT washout, or no washout. An alpha level of .05 (uncorrected for multiple comparisons) was used to determine significance.

Results: Among all subjects there was no significant change in MADRS score from Screening to Randomization [F(1,62)=3.03, p=.09]. Interestingly, a large proportion of these subjects (38%) were not taking antidepressant medications at the time of screening and thus did not require washout. Demographic characteristics, including age, sex, and number of previous failed antidepressant trials did not differ between those requiring and those not requiring washout. A repeated-measures ANOVA comparing change in MADRS scores between patients who underwent AD washout (n=39) and subjects who did not showed a non-significant effect of time, F(1, 61)=3.40, p=.07, a non-significant interaction between time and group (washout vs. non-washout), F(1,61)=.46, p=.50, and a non-significant effect of group, F(1, 61)=2.18, p=.15. An ANOVA of MADRS scores and PT washout showed a non-significant effect of time, F(1, 61)=2. 86, p=.10, a non-significant interaction between time and group, F(1, 61)=.95, p=.33, and a non-significant effect of group, F(1, 61)=.29, p=.60. Subjects in the AD washout group (N=39) did not experience significant worsening in symptom severity (mean MADRS=32.9 pre-washout and 32.3 post-washout). However, one patient experienced significant worsening of mood requiring a hospitalization during lamotrigine taper.

Conclusions: In this study, we found that most subjects who underwent medication washout did not experience significant clinical deterioration. These results suggest that antidepressant washout is generally safe in patients with TRD who may be experiencing limited benefit from ongoing medication treatments. This information is significant, supporting clinicians in tapering ineffective medications and avoiding unnecessary polypharmacy. Moreover, a significant proportion of patients with TRD presenting for participation in a clinical trial were currently untreated with antidepressant medications, suggesting that they may have been already safely tapered from ineffective treatments. Future studies are indicated to replicate our findings and to optimize safety in this TRD group who may require antidepressant washout prior to participation in clinical trials.

Keywords: Antidepressant, Washout, Treatment-Resistant, Depression.

Disclosure: Kyle A. B. Lapidus received research support from NARSAD, the Brain and Behavior Research Foundation, APIRE Janssen, and Simons Foundation. He serves on the advisory board for Halo Neuro, Inc., has received devices from Medtronic, and consults for LCN Consulting, Inc. Sanjay Mathew has the following disclosures. Research support: NIMH, AstraZeneca, Janssen, Otsuka, Johnson Family Chair of Research in Psychiatry Consultant: AstraZeneca, Bristol Myers Squibb, Evotec, Genentech Use of facilities and resources at the Michael E. Debakey VA Medical Center, Houston, Texas The other authors have nothing to disclose relevant to this poster. Dr. Dennis Charney (Dean of Icahn School of Medicine at Mount Sinai), and Icahn School of Medicine at Mount Sinai have been named on a use patent on ketamine for the treatment of depression. The Icahn School of Medicine has entered into a licensing agreement for the use of ketamine as therapy for treatment-resistant depression. Dr. Charney and Icahn School of Medicine at Mount Sinai could potentially benefit if ketamine were to gain approve for the treatment of depression.

T116. ITI-007, a First-in-Class Investigational New Drug for the Treatment of Schizophrenia: Prospective Secondary Analyses from the Randomized ITI-007-005 Trial

Kimberly Vanover*, Sharon Mates, Robert Davis

Intra-Cellular Therapies, Inc., New York, New York

Background: ITI-007 is a first-in-class new molecular entity in development for the treatment of schizophrenia and other neuropsychiatric disorders. As a modulator of serotonin, dopamine, and glutamate, ITI-007 has the potential to treat a broad range of symptoms including positive and negative symptoms and improve overall social function. ITI-007 was evaluated in a randomized, double-blind, placebo- and active-controlled clinical trial for schizophrenia (Clinical Trial ITI-007-005). It was reported previously that 60 mg ITI-007 demonstrated antipsychotic efficacy in patients with an acute exacerbation of schizophrenia, as measured by a statistically significant and clinically meaningful change from baseline on the Positive and Negative Syndrome Scale (PANSS) total score after 4 weeks of treatment compared to placebo. A higher dose of ITI-007 (120 mg) did not statistically significantly separate from placebo, though the positive control, 4 mg risperidone, did separate from placebo indicating assay sensitivity. Also previously reported, the ITI-007 exhibited a favorable safety and tolerability profile. Additional data from the ITI-007-005 clinical trial were analyzed to better understand the antipsychotic efficacy of 60 mg ITI-007.

Methods: Prospectively defined data analyses on secondary endpoints were conducted on the intent-to-treat patient population: 60 mg ITI-007 (N=76), Placebo (N=80), or 4 mg risperidone (N=75) [data for 120 mg ITI-007 not shown]. Antipsychotic efficacy was measured by the PANSS Positive Symptom Factor (van der Gaag et al., 2006, Schizophr Res 85:280-287). Improvement in the PANSS Negative Symptom Factor was pre-specified as a completer analysis in a subgroup of patients exhibiting prominent negative symptoms at baseline (60 mg ITI-007, N=33; Placebo, N=29; 4 mg risperidone, N=33), as defined as having a score of 4 or higher on at least 3 PANSS Negative Symptom Subscale items, who completed 28 days of treatment.

Results: On the PANSS Positive Symptom Factor, 60 mg ITI-007 exhibited an early response with statistically significant (p<0.05) separation from placebo after the first week of treatment and maintained efficacy across 4 weeks of treatment (week 2, p<0.05; week 3, p<0.05; week 4 p<0.01). In contrast, risperidone did not significantly separate from placebo on improvement of positive symptoms until after 4 weeks of treatment (p<0.05). On the PANSS Negative Symptom Factor, 60 mg ITI-007 improved negative symptoms numerically more than risperidone or placebo in a subgroup of patients with prominent negative symptoms, whereas risperidone worsened negative symptoms compared to placebo.

Conclusions: ITI-007 at a dose of 60 mg exhibits rapid improvement of positive symptoms in patients with acutely exacerbated schizophrenia, a faster and more complete response than observed with risperidone as the positive control. Moreover, 60 mg ITI-007 improves negative symptoms, whereas risperidone worsens negative symptoms compared to placebo. Taken together with previously reported data, ITI-007 shows an early and differentiated antipsychotic response and excellent safety and tolerability profile. ITI-007 is currently in Phase 3 clinical development for the treatment of schizophrenia. Other neuropsychiatric therapeutic indications are being explored including bipolar disorder and other mood disorders and low doses of ITI-007 for the treatment of behavioral disturbances in dementia, including Alzheimer’s disease.

Keywords: PANSS, Positive Symptom Factor, Negative Symptom Factor, Antipsychotic.

Disclosure: K. Vanover and S. Mates are employees of Intra-Cellular Therapies and R. Davis is a paid consultant to Intra-Cellular Therapies.

T117. Psychophysiological and Cortisol Reactivity Predicts PTSD Treatment Outcome in Virtual Reality Exposure Therapy with D-Cycloserine

Tanja Jovanovic*, Seth Norrholm, Maryrose Gerardi, Kathryn Breazeale, Michael Davis, Erica Duncan, Kerry Ressler, Bekh Bradley, Albert Rizzo, Barbara Rothbaum

Emory University School of Medicine, Atlanta, Georgia

Background: Virtual reality (VR) methodologies have been successfully employed as an enhanced form of imaginal prolonged exposure therapy for the treatment of stressor-, trauma-, and anxiety-related disorders such as PTSD. To date, the efficacy of prolonged exposure therapies, including those that are VR-based, has been largely indexed through the use of clinician-administered or self-report measures of patient symptom severity or within-session distress. More recently, psychophysiological measures have been employed as complementary assessment tools for use in traumatized populations presenting with PTSD symptoms (Rabe et al., 2006; Griffin et al., 2012; Rhudy et al., 2010; Robison-Andrew et al.; 2014, Rothbaum et al., 2014.) The addition of psychophysiological indices, such as heart rate (HR), skin conductance (SC) and electromyography (EMG) responses to trauma-related cues, provides the potential to use objective assessments of treatment outcome and better understand the underlying biological changes that accompany successful PTSD treatment strategies. The goal of this study was to examine the predictive nature of psychophysiological and stress hormone responses during presentation of combat-related stimuli in a virtual reality environment on PTSD treatment outcome.

Methods: Patients with PTSD (N=50) resulting from OIF or OEF combat exposure underwent 6 weeks of VR exposure therapy combined with either 50 mg D-cycloserine (DCS), 0.25 mg alprazolam (ALP), or placebo (PBO), and were assessed pre- and post-treatment and at 6 months follow-up. Assessments included the Clinician Administered PTSD Scale (CAPS) and psychophysiological assessments. Psychophysiological responses including startle EMG, SC and HR were assessed while viewing combat-related VR scenes. Saliva cortisol was samples before and after VR scenes. Fifty patients completed all three visits, and were divided into three groups according to treatment condition. The study was randomized and double-blind and all assessments were conducted free of treatment medication.

Results: The study showed that startle EMG and SC responses to VR scenes decreased across timepoints, F(2,62)=6.76, p=0.006 and F(2,70)=5.13, p=0.01, respectively. HR during VR scenes did not change with treatment. In the DCS group, startle response to VR scenes prior to treatment accounted for 76% of the variance in CAPS change scores, F(1,12)=34.92, p<0.001, in that higher responses predicted greater changes in symptom severity (i.e. better outcome). The regression was not significant for either the ALP or PBO groups (both p’s>0.1). Cortisol reactivity (the difference between cortisol 15 min after VR scenes and baseline cortisol) had an opposite prediction to treatment response, in that higher reactivity was significantly associated with worse outcome. In the ALP group, cortisol reactivity prior to treatment accounted for 36% of the variance in change in CAPS scores, after controlling for baseline cortisol levels, p=0.04. In the DCS group, there was a trend for the same prediction, p=0.08.

Conclusions: The results of the present study suggest that a higher level of engagement coupled with the cognitive enhancing properties of DCS may lead to more substantial improvements in patient symptom severity. There is currently a significant void in the literature with respect to objective biomarkers that are predictive of PTSD treatment efficacy. The results of the current study suggest that the inclusion of psychophysiological assessments prior to and over the course of treatment and follow-up may prove to be an invaluable tool for monitoring therapeutic progress. This study suggests that psychophysiological responses may serve as biomarkers that predict treatment outcome, by assessing patients’ level of engagement with the exposure therapy leading to better treatment response.

Keywords: PTSD, VR exposure therapy, psychophysiology, treatment outcome prediction.

Disclosure: Nothing to Disclose.

T118. A Proof of Concept, Randomized Clinical Trial of DAR-0100A, a Dopamine-1 Receptor Agonist, for Cognitive Enhancement in Schizophrenia

Ragy Girgis*, Jared van Snellenberg, Lawrence Kegeles, Roberto Gil, Zafar Sharif, Judy Thompson, Andrew Glass, Melanie Wall, Mark Slifstein, Anissa Abi-Dargham, Jeffrey Lieberman

Columbia University, New York, New York

Background: Cognitive deficits are a core feature of schizophrenia. Evidence from preclinical and human studies suggests that cortical hypodopaminergia may contribute to cognitive deficits in schizophrenia. The purpose of this trial was to test whether stimulation of dopamine-1 receptors via a full, selective agonist of the dopamine-1 receptor (DAR-0100A) would improve cognitive deficits in schizophrenia.

Methods: We first performed a phase I, single, ascending dose trial of DAR-0100A in order to identify a maximal tolerated dose of DAR-0100A and to characterize the safety of DAR-0100A. We then randomized 49 clinically stable individuals with schizophrenia to 3 weeks of intermittent treatment with high dose (15mg), low dose DAR-0100A (0.5mg) or placebo (normal saline). fMRI BOLD imaging was used to evaluate the effects of drug administration on patterns of brain activity during performance of a working memory task. Effects on cognition were also assessed using the N-Back, MATRICS, and CogState batteries. Secondary objectives were to investigate the effects of DAR-0100A on negative symptoms.

Results: There were no observed treatment effects on either the BOLD fMRI signal during working memory tasks, on working memory domains of the MATRICS battery, nor on clinical measures. However, there were moderate to large effect size improvements in cognition as measured by the CogState Schizophrenia Battery in only the high dose group, as well as improvements in attention on the MATRICS battery in both treatment groups.

Conclusions: These results are mixed and inconclusive, but do suggest a possible pro-cognitive effect of a full, D1 receptor agonist in schizophrenia. As this drug is limited by its pharmacokinetic profile, better D1 agonists are needed to more fully test the efficacy of this mechanism for cognitive enhancement in schizophrenia and reconcile the mixed findings from this trial.

Keywords: schizophrenia, dopamine-1 receptor, clinical trial, cognitive deficits.

Disclosure: research support from Otsuka and Genentech.

T119. The "Neuroleptic Strategy Study" (NeSSy) - First vs. Second Generation Antipsychotics for the Treatment of Schizophrenia

Gerhard Gründer*, Martin Heinze, Joachim Cordes, Eckart Rüther, Jürgen Timm

RWTH Aachen University, Aachen, Germany

Background: Although the second-generation antipsychotics (SGAs) in many international guidelines are recommended as first choice over compounds from the first generation (FGAs), more recent industry-independent controlled effectiveness studies such as CATIE and CUtLASS could not unequivocally confirm the superiority of SGAs over FGAs. Furthermore, new detrimental side-effects such as weight gain attracted attention with use of some of the SGAs. In the "Neuroleptic Strategy Study" (NeSSy) we investigated with an innovative study design, whether SGAs as a group are superior to FGAs as a group.

Methods: The NeSSy study did compare treatment strategies rather than individual drugs. Although the study is randomized and double-blind, a double randomization allows for selection of a treatment for an individual patient. The randomization process for one patient consisted of two steps. In step 1 a random choice of two drug pairs was allocated to the patient, each pair consisting of a specific FGA and a specific SGA. The investigator was informed about these possibilities and decided which pair was more suitable for the specific case. In step 2 a further random decision selected either the FGA or the SGA drug from the pair determined in step 1 as the drug to be actually applied. The FGAs chosen in this study were haloperidol and flupentixol, the SGAs were aripiprazole, olanzapine and quetiapine. Thus, six different medication pairs were compared, of which two pairs were chosen in the first randomization step. 2,374 patients with schizophrenia (ICD-10) were screened in ten German psychiatric centers, 149 were randomized (FGA: 69; SGA: 80). Data of 139 patients were evaluable (FGA: 64; SGA: 75). PANSS total score baseline values were 83.6 in the FGA group and 81.4 in the SGA group. Treatment duration was 24 weeks. Primary endpoints were quality of life (as quantified with the SF-36) and clinical global impression (CGI-S). In- and exclusion criteria were selected such that a broad, naturalistic patient population could be studied. Twice during the study plasma concentrations of the administered compounds were determined. The primary efficacy analysis was dedicated to AUC values with logarithmic time scale from day 1 to week 24 integrating the effect curve with regard to its typical form of early effects flattening out for later visits. Missing values were imputed by LOCF respectively linear interpolation. Strategies were evaluated by linear contrast in ANCOVA with baseline as covariate and drugs as factors.

Results: Mean AUC values for the SF-36 were significantly higher in the SGA than in the FGA group (FGA: 79.7±17.3; SGA: 85.1±14.7). This difference is statistically significant on the 2.5% level (p=0.0112). The 5.4 point difference was considered also clinically significant as it exceeded the predefined 5.0 point threshold. CGI scores decreased in both groups (AUC FGA: 3.388; SGA: 3.256), but this difference between groups was neither statistically (p=0.3423) nor clinically significant. With regard to secondary endpoints, SGAs were statistically significantly superior to FGAs on four subscores of the SF-36: vitality (p=0.0447), emotional role function (p=0.0347), physical role functioning (p=0.0111), and physical functioning (p=0.00384). 104 of the 136 patients (76%) did not complete the whole study duration of 24 weeks. 52 patients in each group dropped out from the study prematurely, 83% in the FGA group and 71% in the SGA group (p=0.1226). Reasons for drop-out from the study were also not significantly different between groups. The increase in body mass index (BMI) was significantly more pronounced in the SGA than in the FGA group, while the increase in BMI in the SGA group was driven by olanzapine and quetiapine (olanzapine+0.8; quetiapine+0.7; aripiprazole+0.3).

Conclusions: Our study shows that quality of life (as assessed with the SF-36) was statistically and clinically significantly more improved under treatment with SGA antipsychotics compared to FGAs. CGI values, which reflect judgement by an observer, □□show the same although non-significant trend. Thus, antipsychotics from the second generation represent a significant advantage in terms of (subjective) quality of life but only a small advantage as judged by an (objective) medical observer. The evaluation of the safety analysis provides a mixed picture. On the one hand the better (although not significant) adherence with treatment with the SGAs speaks for the latter group of compounds. On the other hand, our study confirmed once again the stronger metabolic adverse effects of SGAs, which can be devastating long-term. Finally, we have shown that innovative study designs are useful for a meaningful comparison of the efficacy and tolerability of two treatment strategies. The study represents a contribution to patient-oriented assessment of opportunities and risks of the two therapeutic approaches for the treatment of schizophrenia.

Keywords: schizophrenia, antipsychotic, clinical trial.

Disclosure: Study medication was provided by the manufacturers of the applied second generation antipsychotics (AstraZeneca, Bristol-Myers Squibb, Eli Lilly.

T120. Comparative Evaluation of Quetiapine Plus Lamotrigine Versus Quetiapine Monotherapy in Bipolar Depression: A Randomized Placebo Controlled Trial (CEQUEL)

John Geddes, Chris Hinds, Jennifer Rendell, Alex Gardiner, Merryn Voysey, Mary-Jane Attenburrow, Guy Goodwin*

University of Oxford, Oxford, United Kingdom

Background: Effective short- and longer-term treatments for bipolar depression remain limited. The combination of lamotrigine plus quetiapine potentially offers enhanced treatment for patients by maximising the benefits of both drugs without inducing mood instability. CEQUEL was designed to test the hypothesis that a combination of lamotrigine with quetiapine leads to better short-term response and longer term clinical outcomes than quetiapine alone. CEQUEL also investigated the effects of adding folic acid but this report will focus on the lamotrigine: placebo comparison.

Methods: CEQUEL was a UK, multicentre, double-blind, randomized, placebo-controlled, parallel group, 2x2 factorial clinical trial with concurrent economic, biochemical and genetic analysis. Eligible patients were those with DSM-IV bipolar disorder I/II (BD1/2) who required new pharmacological treatment for an acute depressive episode, aged 16 or over with score>16 on Quick Inventory of Depressive Symptomatology. Following 14 day run-in on quetiapine, participants were randomized to added lamotrigine (200mg) or placebo. Participants were separately randomized to folic acid (0.8mg) or placebo. Allocated treatment was continued for 12 months. The primary outcome was improvement in depressive symptoms at 12 weeks from randomisation using QIDS-SR16. QIDS data was collected by participants using the True Colours SMS/email system. Secondary outcomes included improvement in depressive symptoms at 52 weeks; proportion of patients in remission (QIDS ≤ 5) at 12 and 52 weeks; time to new intervention for depressive and manic symptoms; self-harm, mortality, adverse effects and health and social care costs. Analysis was by intention-to-treat. EUdraCT No.: 2007-004513-33; REC No. 08/H0605/39; CTA:20584/0234/001-0001; ISRCTN17054996.

Results: 202 patients were randomised (101 lamotrigine; 101 placebo). The groups were balanced by minimisation for bipolar subtype; age, gender, concurrent medications, medication history and mood episodes. Diagnoses bipolar disorder type 1: bipolar 2 was approximately 3:1. The modal daily dose of quetiapine was 300mg. The mean difference in QIDS-SR16 between groups was: after 12 weeks, 1.6 points lower lamotrigine vs placebo (95% CI -3.4 to 0.23; p=0.09; after 52 weeks, 2.9 (95% -5.2 to -0.6; p=0.013). Significantly more patients treated with lamotrigine achieved remission (QIDS ≤ 5) at 12 weeks (lamotrigine 31%; placebo 16%; relative risk 2.07 [95% CI 1.1-4.0], p=0.04) and 52 weeks (36%; 13%; RR 3.7 [1.35-10.3], p=0.008).

Conclusions: Adding lamotrigine to quetiapine treatment of acute bipolar depression substantially improves outcomes by 52 weeks. The average benefit of adding lamotrigine was less pronounced at 12 weeks, although there was a significantly increased proportion of patients achieving remission which suggests that the might be a subgroup who are particularly responsive to lamotrigine. The results of CEQUEL highlight the potential benefits of combination therapy and the importance of taking a longer term view of therapy in bipolar disorder.

Keywords: Bipolar depression, Lamotrigine, Quetiapine.

Disclosure: Guy Goodwin: I: Consultancy, speaker Abbvie, AstraZeneca, Cephalon/Teva, GSK, Eli Lilly, Lundbeck (/Otsuka/Takeda), Medscape, Merck, P1Vital, Servier, Sunovion. II: Medscape, Servier. III: Servier. IV: Lundbeck, Servier.

T121. Efficacy and Safety of Adjunctive Bitopertin Versus Placebo in Patients with Sub-optimally Controlled Symptoms of Schizophrenia Treated with Antipsychotics – Results: from the Searchlyte Clinical Trial

Dragana Bugarski-Kirola*, Nakao Iwata, Snjezana Sameljak, Carol Reid, Thomas Blaettler, Jon Luca Zhu, Laurie Millar, Gang Wang, Amy Guo, Shitij Kapur

Roche, Basel, Switzerland

Background: Clinical response to antipsychotic treatment for schizophrenia is characterized by a partial remission of the acute positive psychotic symptoms.1 These residual positive symptoms have a significant effect on functioning and are associated with relapse and re-hospitalization.2 There is a strong unmet need for treatment that will improve sub-optimally controlled symptoms and can safely be given as add-on to current antipsychotics.3 N-methyl-D-aspartate (NMDA) receptor function is thought to be reduced in schizophrenia. As glycine is a necessary co-agonist at the NMDA receptor complex, inhibition of glycine reuptake can elevate extracellular levels of glycine around synaptic NMDA receptors, enhancing NMDA receptor functioning. Glycine-enhancing agents, when given as add-on treatments to stable antipsychotic-treated patients, may improve the symptoms of schizophrenia.4,5 In this series of studies, efficacy and safety of bitopertin, a glycine reuptake inhibitor thought to indirectly enhance NMDA receptor function, was evaluated in patients with schizophrenia who continued to show sub-optimally controlled psychotic symptoms despite ongoing treatment with antipsychotics.

Methods: Three (TwiLyte, NCT01235520 [NN25307]; MoonLyte, NCT01235585 [WN25306]; and NightLyte, NCT01235559 [WN25305]) Phase III, multicenter, randomized, 12-week, double-blind, parallel-group, placebo-controlled studies were conducted to evaluate the efficacy and safety of adjunctive bitopertin in stable patients treated with antipsychotics exhibiting sub-optimally controlled symptoms of schizophrenia. Key inclusion criteria were: age ≥18 years, DSM-IV-TR diagnosis of schizophrenia; baseline severity symptoms on Positive and Negative Syndrome Scale (PANSS) total ≥70 with a score ≥4 on two or more of the psychotic PANSS items; and retrospective clinical and antipsychotic treatment stability. Following the 4-week prospective stabilization period, patients were randomized 1:1:1 to bitopertin or placebo once daily for 12 weeks (doses: bitopertin 10 and 20; 5 and 10; and 10 and 20 mg, respectively). The primary efficacy endpoint was mean change from baseline in the PANSS Positive Symptom Factor Score (PSFS) at Week 12. Mean change from baseline in PANSS total score at Week 12 was a secondary endpoint. Safety and tolerability assessments included adverse event (AE) reporting, vital signs and laboratory analyses.

Results: Each study included approximately 200 patients. In all three studies, patients were randomized to one of two doses of bitopertin or placebo. In study WN25305, the 10 mg arm showed a significant improvement in the primary efficacy endpoint and separated from placebo. No other arms in any study separated from placebo. In study WN25305 alone, bitopertin 10 mg was superior to placebo on the PSFS (primary endpoint) (–5.14; SE 0.327; p=0.0028) and in PANSS total score (–14.44; SE 0.866; p=0.0217). Improvements in PSFS for the bitopertin 20 mg arm were of smaller magnitude (–4.22; SE 0.320) compared with placebo (–3.77; SE 0.319), p=0.3142. Changes in PANSS total scores were also small and non-significant. In the post-hoc pooled analysis from all three studies bitopertin 10 mg did not separate from placebo on primary endpoint. The AE profile for study WN25305 was similar to placebo with ≥1 AE reported by 35.4, 34.2 and 29.6% of patients in the bitopertin 10 mg, 20 mg and placebo arms, respectively. There were no effects on vital signs, weight or metabolic parameters, and reduction in hemoglobin was mild and dose-dependent. The incidence of serious SAEs and AEs leading to discontinuation was low and similar across all treatment arms.

Conclusions: The three well-designed and controlled studies provide a thorough examination of bitopertin safety and efficacy in this population. The AE profile of bitopertin was similar to placebo in all studies. Bitopertin 10 mg showed a significant improvement in the primary efficacy endpoint compared with placebo in the WN25305 study, but not in the other studies. It is possible that either the results of the WN25305 study are an outlier, or there is an efficacy signal in a particular population subset that may be missed in a broader examination. The overall findings of the bitopertin studies do not support the use of the drug in an unselected population of sub-optimally responsive patients. Analyses are currently underway to examine the differences between the studies and to explore whether there are subpopulations of patients that may benefit from bitopertin. References 1. Buckley PF, Shendarkar N. Curr Opin Psychiatry 2005;18(2):165–73 2. Chakos MH, et al. Psychiatr Serv 2006;57(8):1094–101 3. Bromet EJ, et al. Schizophr Bull 2005;31(3):639–49 4. Goff DC, et al. Biol Psychiatry 1999;45(4):512–4 5. Heresco-Levy U, et al. Biol Psychiatry 2005;57(6):577–85.

Keywords: Schizophrenia, Bitopertin, Treatment, N-methyl-D-aspartate.

Disclosure: Dragana Bugarski-Kirola: Employee of F. Hoffmann-La Roche Ltd. Nakao Iwata: No conflicts of interest. Snjezana Sameljak: Employee of F. Hoffmann-La Roche Ltd. Carol Reid: Employee of F. Hoffmann-La Roche Ltd. Thomas Blaettler: Employee of F. Hoffmann-La Roche Ltd. Jon Luca Zhu: Employee of F. Hoffmann-La Roche Ltd. Laurie Millar: Employee of F. Hoffmann-La Roche Ltd. Gang Wang: No conflicts of interest. Amy Guo: Employee of F. Hoffmann-La Roche Ltd. Shitij Kapur: Received honoraria for serving on the Scientific Advisory Board and as a speaker for sponsored Conference symposia for F. Hoffmann-La Roche.

T122. Evaluation of Novel Strategies for Prevention of Alzheimer's Dementia in Cognitively Normal Persons at High Risk Using Multiple Biomarker Endpoints: First Reported Findings

John Breitner*, Judes Poirier, Pierre Etienne, Jennifer Tremblay-Mercier, Marie-Elyse Lafaille-Magnan, Centre for Studies on Prevention of AD

McGill University Faculty of Medicine, Montreal, Canada

Background: Treatment trials have failed to find agents with disease-modifying effect on Alzheimer’s disease (AD) dementia. The AD development process includes several decades of brain changes that culminate in dementia onset. This pre-symptomatic phase of AD offers a time of opportunity for preventive interventions that slow the disease process, and thus defer or deter the later onset of dementia. Recently initiated prevention trials have concentrated on anti-amyloid or lifestyle/behavioral interventions (the latter being notoriously difficult to achieve in real practice.) Other interventions should be tried --- but which interventions? At present we have almost no preliminary data to suggest choice of agents. Several biomarkers are known to be present, and to change, during the course of pre-symptomatic AD. Therefore, biomarker-endpoint Phase I or Phase II trials hold promise as indicators of the likely success of specific agents in later pivotal trials. Almost certainly, no one biomarker can suffice for this purpose, since various markers change at different times in disease development. Furthermore, not all individuals with pre-symptomatic AD will show the same sequence of biomarker changes. Trials using multiple biomarker endpoints may circumvent this issue. But then whom to choose for participation in such trials? Which biomarkers? Over what period of time? And what to do about the multiple-comparisons problem that would vex this approach using conventional trials methods?.

Methods: Over the past 3 years we have assembled a cohort of persons aged 60+with a parental history of AD (age 55+if their parent’s onset was<=15 yrs older than their own current age). These persons volunteered to participate in PRevention EValuation for Experimental or Novel Treatments for AD (PREVENT-AD). PREVENT-AD participants must be medically healthy and have no current cognitive or neurological disorder. All are genotyped at APOE, BDNF, HMGR Intron M, and BChE-K. They must agree to annual or more frequent evaluations of multiple potential markers including not only tests of cognition, but MRI evaluation of cortical thickness, automated structure-specific volumetry, measures of regional and global cerebral blood flow (ASL), resting state fMRI measures of connectivity, and task fMRI. Also included are tests of olfactory identification and central auditory functioning. Serial LP’s (4 taps over 2 yrs) are optional, but have been accepted by more than 60% of participants. This last statistic exemplifies the advantages of a parental-history cohort. The problem of analytic approach to the multiple endpoints is now receiving our full attention (see Discussion).

Results: To date, more than 400 persons have been evaluated for eligibility. Some 220 have been enrolled. We recently completed recruitment of 150 of these into a placebo-controlled biomarker-endpoint trial of naproxen-sodium 220 mg b.i.d. Another 40 will be recruited for a trial of the apoE inducing drug probucol now in development. Some 75 persons in the naproxen trial have now completed their first annual follow-up visit, so we can shortly begin analyses of biomarker change over time (remaining blind to treatment assignment). Meanwhile, baseline data are of interest: Participant age (obviously related to disease development) is correlated with cortical thickness, with ASL blood flow estimates in precuneus and total brain, with hippocampal volume (R>L), with olfactory identification, with CSF tau (and P-tau) and Abeta42 and their ratio (important interaction with APOE, age-related decline occurring only in those with E4), with resting state connectivity (RSC) and, of course, with several measures of cognition. More important potentially is that numerous biomarker metrics correlate with one another after adjustment for age -- e.g., cortical thickness with tau and p-tau, with ASL (precuneus and global), with several measures of cognition, and with APOE status (interaction with age), as well as olfaction and several cognitive measures. Olfaction correlates with hippocampal volume (R>L), with tau, Abeta, and their ratio, and with ventricular volume. Source accuracy on task fMRI correlates with tau, Abeta, and their ratio. A provocative finding is a substantial difference (or in some analyses, interaction) between gender (after controlling for age) and RSC, ASL, and immediate memory, with women faring considerably worse than men.

Conclusions: AD incidence data suggest that most of the PREVENT-AD cohort are within 25 or fewer years of dementia onset. The correlation of multiple biomarkers with age, and among themselves, suggest that many of these markers represent a common underlying latent variable of progressive AD pathogenesis. The exaggerated findings here in women correspond to the known strong excess of AD dementia incidence in women vs. men starting at age 80. Is this excess caused by abrupt loss of estrogen at menopause 25 - 30 years earlier? And would the same phenomenon then appear as biomarker differences among women vs. men in their 60's and 70's? We are collecting data on peri/post-menopausal hormone therapy to investigate this hypothesis. We are also developing methods to deal with the multiple-endpoints issue using a variant of Item Response Theory (will be explained at the meeting). This method derives from educational testing theory (here, biomarkers represent questions of differing difficulty, and the latent trait is disease progression rather than ability.) IRT accommodates multiple asynchronous changes with different slopes, and avoids the multiple comparisons conundrum.

Keywords: Alzheimer, Prevention, Biomarker, Clinical trial.

Disclosure: Nothing to Disclose.

T123. Antipsychotic Re-challenge in Previous Responders

Ofer Agid, Robert Zipursky*, Cynthia Siu, Gagan Fervaha, Krysta McDonald, George Foussiasa, Gary Remington

University of Toronto, Toronto, Canada

Background: Dopamine’s proposed role in psychosis provides a starting point for our understanding of the neurobiology of relapse in schizophrenia. While perturbations in dopamine have been proposed as the final common pathway in psychosis, it is evident that relapse, like response, cannot be conceptualized as a singular process. Similarly, the relationship between response and relapse appears to be multidimensional, with patterns of response defining relapse, and trajectories of response translating to different trajectories of relapse. Relapse can be defined as either primary (i.e. idiopathic) or secondary (e.g. substance abuse, medication non-adherence). Evidence suggests that secondary relapse is more common and may be associated with a more attenuated response to antipsychotics than found in the treatment of the initial episode of psychosis. The current study sought to clarify the effect of relapse due to medication non-adherence on subsequent treatment response.

Methods: Antipsychotic-naïve individuals diagnosed with first-episode schizophrenia were treated following an algorithm that involved treatment with risperidone or olanzapine. Each trial consisted of 3 stages (low, full, or high-dose) lasting up to 4 weeks at each level and adjusted according to response and tolerability. Clinical response was defined as a Clinical Global Impression Inventory - Improvement (CGI-I) of “much” or “very much improved”, and a Brief Psychiatric Rating Scale (BPRS) Thought Disorder subscale of<6. Clinical data was collected on a monthly basis over 2 years. In each case, relapse after the first episode was due to non-adherence. The same medication and dose where response was previously noted was offered again for the second episode. Clinical data was then collected monthly over a period of 1 year. Repeated measures ANOVA was applied to compare the time course of symptom reduction between first- and second-episode treatments in schizophrenia. Growth Mixture Modeling (GMM) was used to analyze the trajectories of BPRS total score to identify latent classes corresponding to distinct patterns of response in first episode treatment, and determine if first episode treatment response patterns predicted response trajectories following relapse.

Results: A total of 132 individuals (87% male; mean age=22) met criteria for response following treatment with the first antipsychotic, subsequently relapsed due to non-adherence and went through a second trial following the same treatment. BPRS (total/core psychosis scale) improvement was significantly greater for the first episode compared to the second episode at every time point up to week 39. Three response trajectory groups for the BPRS total score were identified. The Immediate Improvement group (n=108, 82%) showed symptom reduction in the first 3 weeks (-22.7, SEM 0.60), and stabilized at that symptom severity level to week-55. The Gradual Improvement group (n=21, 16%) showed gradual improvement in BPRS total score (-33.1, SEM 1.92) over the 55-week treatment period. Finally, the Outlier Immediate Improvement group (n=3, 2%) showed mean change in BPRS total of -33.7 (SEM 0.33) over 3 weeks, stabilizing at that symptom severity level to week-55. These three response trajectories each predicted treatment response trajectories in the second episode (p<0.01), however, diminished levels of symptom improvement was evident, as noted above. That differences were not significant past the 39 week point likely reflects the impact of having fewer cases at more distant time points and a bias for poorer responders to discontinue treatment.

Conclusions: Reinitiating antipsychotic treatment for a second episode of psychosis was found to be associated with an attenuated response to antipsychotic medication. This finding raises questions about the nature of changes that may be taking place in the dopamine system in patients who have relapsed after discontinuing their antipsychotic medication. Whether this observation reflects the progression of biological changes associated schizophrenia or the impact of previous treatment or its withdrawal is not yet understood. Our results challenge the expectation that reinstituting the previous level of D2 antagonism is likely to result in a return to a remitted state. Taken together, these findings indicate that response and relapse must be viewed as multidimensional and are likely mediated by distinct mechanisms.

Keywords: Schizophrenia, antipsychotics, Response.

Disclosure: Dr. Agid has received research support from Pfizer, Inc, and Janssen-Ortho; has received consultant fees from Janssen-Ortho, Eli Lilly and Company, US, Eli Lilly Canada, Sepracor, Sunovion, and Lundbeck; and has received speaker’s fees from Janssen-Ortho, Eli Lilly and Company, Eli Lilly Canada, Novartis, Lundbeck, Sepracor, and Sunovion. Dr. Zipursky has served as a consultant to Roche, served on advisory panels for Amgen, Roche, and Sunovion, and received research contract support from Roche. Dr. Siu has served as a consultant for Pfizer, Sunovion U.S. (Dainippon Sumitomo Pharma/Sepracor), Memory Pharmaceutical (Roche Laboratories), and Wyeth (Pfizer). Mr. Gagan Fervaha BSc has no conflict of interest to disclose. Ms. Krysta McDonald has no conflict of interest to disclose. Dr Foussias has been involved in research sponsored by Medicure Inc and Neurocrine Bioscience; has received consultant fees from Roche; and has received speaker fees from Roche, Lundbeck, and Novartis. Dr. Remington has received research support from Novartis, Medicure, and Neurocrine Bioscience; has received grant support as a coinvestigator from Pfizer, Inc; has received consultant fees from Laboratorios Farmacéuticos ROVI, Synchroneuron, Novartis, and Roche; and has received speaker’s fees from Novartis.

T124. Clinical and Biomarker Effects of a Novel Vasopressin 1a Receptor Antagonist (RG7713) vs. Placebo in High Functioning Adult Autism

Eric Hollander*, Marta del Valle Rubido, Omar Khwaja, Lisa Squassante, Casara Jean Ferretti, Bonnie P. Taylor, Greg Berlin, Rachel Noone, Laura Antar, Lauren Boak, Paulo Fontoura, James McCracken, Larry Scahill, Frederick Shic, Daniel Umbricht

Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, New York

Background: Oxytocin and vasopressin play a critical role in social cognition and social signaling deficits of ASD. Experimental therapeutic interventions to enhance oxytocin or block V1a receptor signaling may modulate these domains in ASD. The study objective was to explore the impact of a novel V1a receptor antagonist RG7713 vs. placebo on core social cognition measures, exploratory biomarkers, and safety/tolerability measures in adult high functioning ASD.

Methods: High-functioning adults (M=23.4 years, range=18 to 40 years) with autism (n=19) participated in a multi-center (3-site), randomized, double-blind, placebo-controlled, cross-over study of the effects of novel vasopressin 1a receptor antagonist RG7713. Each participant was seen on two separate days (1 week apart) for dosing of a 2 hour infusion of RG7713 or placebo. Safety/tolerability, core social cognition, olfaction and AVPR1A polymorphisms measures were collected.

Results: At baseline, the Affective Speech Recognition (ASR), Reading the Mind in the Eyes (RMET) and olfaction measures correlated with measures of functioning on the Vineland and ADOS as well as IQ. The V1a antagonist showed evidence of anxiolysis. The V1a antagonist showed effects on eyetracking and RMET in the predicted direction of modest effect sizes. There were large negative effect sizes of the V1a antagonist vs. placebo on the social cognition ASR measure Lust and Fearful subscales. Olfaction and drug x phase effects influenced social cognition measures.

Conclusions: This study provides preliminary evidence of the ability of a novel V1a antagonist RG7713 to affect core symptoms of social cognition, and these findings were influenced in part by olfaction and drug x phase effects. These results should be taken as preliminary but may help to guide the development of new oral vasopressin antagonist interventions in ASD.

Keywords: Autism spectrum Disorder, clinical trial, social cognition, vasopressin.

Disclosure: This research was funded by Roche Pharmaceutical Research and Early Development, Roche Innovation Center Basel (Switzerland). Eric Hollander is also a consultant for Roche Pharmaceutical Research and Early Development, Roche Innovation Center Basel (Switzerland).

T125. Effects of Aripiprazole Once-monthly on Symptoms and Functioning of Patients with an Acute Episode of Schizophrenia Stratified by Age

W. Wolfgang Fleischhacker, Ross A. Baker, Anna Eramo*, Na Jin, Peter Hertel, Timothy Peters-Strickland, Robert McQuade, Raymond Sanchez, John Kane

H. Lundbeck, Deerfield, Illinois

Background: During an acute episode of schizophrenia, after ensuring that the patient’s psychotic symptoms are under control, it is important to consider how to achieve long-term functional preservation and management of symptoms to prevent further relapse. Preservation of functioning earlier in the course of disease is even more important, since at earlier stages of disease, patients generally have a higher level of overall functioning, and thus potentially more to lose during a relapse. In the present exploratory analysis, we evaluated whether patients aged ≤35 years showed a different pattern of response versus those aged>35 years after initiation of aripiprazole once-monthly 400 mg (AOM 400, an extended-release injectable suspension of aripiprazole), as measured by scales of symptoms and functioning.

Methods: Adults with schizophrenia who were experiencing an acute psychotic episode were randomized to 12 weeks of double-blind treatment with AOM 400 or placebo. Primary and secondary outcomes for the total population have been published [1]. In the present exploratory post hoc analyses, the within-group difference in least squares mean (LSM) change from baseline for AOM 400 vs placebo was calculated in the efficacy sample to determine treatment effect using mixed-model repeated-measures (MMRM) and analysis of covariance (ANCOVA) of observed case data. The LSM treatment effects for patients aged ≤35 years and those aged>35 years are reported at week 12 (endpoint) for the Positive and Negative Syndrome Scale (PANSS), PANSS positive and negative subscales, Clinical Global Impressions-Severity scale (CGI-S), and Personal and Social Performance scale (PSP).

Results: At baseline, symptom and functioning scores were similar in patients aged ≤35 years and those aged>35 years. Treatment effects determined by MMRM were statistically significant in favor of AOM 400 for patients in both age groups for PANSS total (≤35 years of age [n=48], –16.9 [95% CI, –27.0 to –6.7];>35 years of age [n=114], –13.3 [95% CI, –18.7 to –7.9]), PANSS positive (–4.9 [95% CI, –8.0 to –1.9]; –5.0 [–6.9 to –3.2]), and CGI-S (–0.9 [95% CI, –1.4 to –0.4]; –0.7 [95% CI, –1.0 to –0.4]). Treatment effects on PANSS negative symptoms were larger in patients ≤35 years of age (–3.5 [95% CI, –6.3 to –0.7]) than in those>35 years of age (–2.1 [95% CI, –3.5 to –0.7]), although both groups showed a statistically significant benefit. On the PSP, treatment effects in patients ≤35 years of age were more than 3-fold larger (n=25; 11.1 [95% CI, 4.8 to 17.5]; P=0.001) than those in patients>35 years of age (n=76; 3.6 [95% CI, –0.6 to 7.9]; P=0.09), for whom the benefit was not statistically significant. ANCOVA supported the results of the MMRM. No differences in adverse event profiles were observed between age groups.

Conclusions: After initiation of treatment with AOM 400, patients aged ≤35 years who were experiencing an acute episode of schizophrenia showed similar efficacy on mean change from baseline in PANSS total and positive, and CGI-S, as patients>35 years. Interestingly, larger treatment effects were observed on the PANSS negative subscale and PSP in those aged ≤35 years vs those aged>35 years, suggesting that patients earlier in the course of illness may be more sensitive to improvements in negative symptoms and social function. This research was supported by Otsuka Pharmaceutical Development and Commercialization. 1. Kane JM, Peters-Strickland T, Baker RA, et al. Aripiprazole once-monthly in the acute treatment of schizophrenia: findings from a 12-week, randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2014:in press.

Keywords: aripiprazole, functioning, schizophrenia, symptoms.

Disclosure: W. Wolfgang Fleischhacker has received research grants from Otsuka, Pfizer, Janssen, and Reckitt-Benckiser, as well as consulting honoraria from Lundbeck, Roche, Bristol-Myers Squibb, Otsuka, Janssen, Pfizer, MedAvante, Takeda, Endo, and Vanda. He has received speaker honoraria from Lundbeck, Janssen, Otsuka, Roche, and Takeda. He holds stock from MedAvante. Ross Baker, Na Jin, Timothy Peters-Strickland, Robert McQuade, and Raymond Sanchez are employees of Otsuka Pharmaceutical Development and Commercialization, Inc. Anna Eramo is an employee of Lundbeck LLC. Peter Hertel is an employee of H. Lundbeck A/S. John M. Kane has received honoraria for lectures and/or consulting from Alkermes, Amgen, Bristol-Myers Squibb, Cephalon, Eisai, Boehringer Ingelheim, Eli Lilly, Forrest, Genentech, Intracellular Therapeutics, Janssen, Johnson & Johnson, Lundbeck, Merck, Novartis, Otsuka, Pfizer, Pierre Fabre, Proteus, Reviva, Roche, Sunovion, and Targacept. He is a shareholder of MedAvante.

T126. Meditation Interventions for Treatment of PTSD in Veterans

Kelvin Lim*, Christopher Erbes, Paul Thuras, John Rodman, Scott Sponheim, Melissa Polusny

University of Minnesota, Minneapolis, Minnesota

Background: Posttraumatic stress disorder (PTSD) is a debilitating psychiatric disorder characterized by intrusive and distressing recollections of a traumatic event, avoidance of reminders of the event, and symptoms of hyper arousal, such as impaired sleep, irritability, and decreased concentration. Estimates of the rate of lifetime PTSD have ranged from 7%-12% in the US population and 19% to 30% in combat veterans. PTSD carries enormous costs to individuals, society and to the Veterans Health Administration (VHA) medical system. It is frequently chronic with high rates of psychiatric co-morbidity, is associated with substantial impairment in occupational, interpersonal, and family functioning, poorer quality of life, a greater number of physical health complaints and increased medical disease burden. While evidence-based direct-exposure psychotherapies for treatment of PTSD have been developed and disseminated within the VHA, no treatment has shown universal effectiveness and there is great concern about attenuated treatment response and elevated treatment drop out in veteran populations. Mindfulness Based Stress Reduction (MBSR) is a promising alternative that is thought to serve as a gentler form of exposure for individuals who find traditional trauma focused treatments intolerable. Although promising, MBSR has not yet been systematically evaluated as a treatment for PTSD in a clinical research trial. This study evaluated the efficacy of an 8-week MBSR intervention compared to 8 weeks of Present Centered Group Therapy (PCGT) as a group-based treatment for PTSD in veterans.

Methods: Investigators randomly assigned 116 veterans of any era or military background who suffered from PTSD to either MBSR or PCGT treatment groups. Participants were recruited from the PTSD and mental health clinics at the VHA or by response to a direct mail outreach. The study randomized participants into 9 cohorts consisting of one group each of MBSR and PCGT. Treatments were delivered by trained personnel receiving ongoing expert supervision and whose fidelity to protocol was evaluated. The MBSR course was administered following the protocol developed by Jon Kabat-Zinn with a slight modification to include an initial session on PTSD psychoeducation and treatment rationale for how mindfulness training can lead to improvements in PTSD symptoms. The PCGT intervention provided a credible and clinically relevant comparison intervention that controlled for nonspecific therapeutic factors. It consisted of nine weekly 1.5 hour group sessions including an initial session on PTSD psychoeducation and treatment rationale. Assessments were conducted at baseline, weeks 3, 6, 9 and 17. Clinical interviews were conducted by trained independent evaluators blinded to participant treatment condition. Self-report measures were collected by the study coordinator or research assistants.

Results: Of the 116 participants randomized, 99 completed therapy and follow-up and 17 were lost to attrition. All data was analyzed in this intent-to-treat model. Study groups did not differ significantly on any demographics. The two groups were equivalent on symptom severity measures of avoidance, depression, somatic symptoms and functional outcome suggesting successful randomization. However, at baseline, the MBSR group was significantly higher than the PCGT group on both the PTSD Checklist (PCL) (63.6 vs. 58.8, p=.032) and the Clinician Administered PTSD Scale (CAPS) severity scores (69.9 vs. 62.5, p=.016). An analysis of those who dropped out suggests that loss to follow-up was not due to any demographic or clinical characteristics. The drop-out was significantly higher in the MBSR group compared to the PCGT group (22.4% vs. 6.9%, suggesting that the differential demands of the two therapies may have contributed to higher dropout in the MBSR group. The results of growth curve mixed effect models found that the MBSR group showed significantly larger decreases on the PCL from baseline to the 2 month follow-up than did the PCGT group (10.2 vs. 2.9; p=.004) as well as significantly greater declines in the CAPS severity score over the same time period (22.4 vs. 12.3; p=.031). This was also reflected in significant differences in rates of decrease on each of the PCL subscales; re-experiencing (3.1 vs. 1.1; p=.012), avoidance (4.1 vs. .9; p=.015) and hyperarousal (3.0 vs. .8; p=.012). The MBSR group was also found to have significantly greater decreases on the Acceptance and Action Questionnaire-II (AAQ) compared to the PCGT group (3.6 vs. .5; p=.012). The MBSR group showed significant greater increases on the World Health Organization Quality of Life-Brief (WHO-BREF) psychological health domain (1.1 vs. -.1; p=.003), social relationships domain (2.3 vs. 1.0; p=.009) and environmental domain (.7 vs. -.1; p=.031).

Conclusions: The VHA endorses several evidence-based treatment interventions for PTSD which can be difficult for some veterans to tolerate. The current study demonstrates that MBSR can be an effective alternative treatment for veterans with PTSD.

Keywords: PTSD, meditation, mindfulness, veterans.

Disclosure: Nothing to Disclose.

T127. The Efficacy and Safety of Basimglurant as Adjunctive Therapy in Major Depression; a Randomized, Double-blind, Placebo Controlled Study

Jorge Quiroz*, Paul Tamburri, Dennis Deptula, Ludger Banken, Ulrich Beyer, Paulo Fontoura, Luca Santarelli

Hoffmann-La Roche, Inc., New York, New York

Background: Therapies targeting the glutamatergic system are known to be efficacious in the treatment of mood disorders. Antagonism of the post-synaptic mGlu5 receptor is a novel approach to indirectly modulate glutamatergic (NMDA) function and has shown efficacy in a number of preclinical behavioral models of depression. Basimglurant is a potent and selective negative allosteric modulator of the mGlu5 receptor which has been comprehensively profiled in Ph1 and Ph2a trials. The main objectives of this Ph2b trial were to evaluate the safety and efficacy of basimglurant modified release (MR) vs. placebo, as adjunctive therapy to ongoing antidepressant treatment in patients with major depressive disorder (MDD) who showed inadequate response to at least one but no more than three treatment failures within the current episode.

Methods: In this 9-week study (6-week double-blind treatment, 3-week post-treatment follow-up), adult patients with DSM-IV-TR diagnosis of MDD were randomized to basimglurant 0.5 mg/d, 1.5 mg/d, or placebo (adjunctive to ongoing SSRI or SNRI). The primary endpoint was the mean change from baseline in the Montgomery-Åsberg Depression Rating Scale Sigma total score (MADRS), as rated by the clinician at week 6. Concomitantly, patient-rated MADRS scores were also collected and analyzed. Secondary endpoints included change in the Quick Inventory of Depressive Symptomatology (QIDS-SR16), MADRS response (≥ 50% reduction in score from baseline), MADRS remission (score of ≤ 10), and Clinician and Patient Global Impression scales (CGI-I, PGI-I and CGI-S). Exploratory endpoints included change in the Quality of Life Enjoyment and Satisfaction Questionnaire Short Form (Q-LES-Q-SF) and the Sheehan Disability Scale (SDS). Due to the exploratory nature of this study, one-sided p values were estimated with no adjustment for multiplicity. Completers (observed cases) ANCOVA and ITT MMRM statistical analysis were performed.

Results: 333 male and female patients were randomized in to the study. The primary endpoint for the study (clinician-rated MADRS) was not met (p=0.127 ITT MMRM analysis); a trend for improvement was observed for basimglurant 1.5 mg vs. placebo (p=0.061 completers ANCOVA analysis) while statistical significance was reached utilizing the patient-rated MADRS (p<0.025 in both analyses). Regarding the secondary endpoints basimglurant 1.5 mg showed significant improvements vs. placebo in QIDS mean change from baseline (p=0.004 in both analyses), CGI-I mean score (p<0.039 in both analyses), PGI-I mean score (p<0.029 in both analyses). Significant improvements were also seen with in the patient-rated MADRS remission rate (p<0.024 both analyses), and to a lesser degree in the patient-rated MADRS response (p<0.1 both analyses). Lastly, significant improvements were observed in the Q-LES-Q-SF (p=0.011) and the SDS items 2-3 (p=0.047) (ITT MMRM). Basimglurant dosed at 0.5 mg showed no benefit over placebo in any of these measures. Withdrawal rates due to adverse events were 5.4%, 7.2% and 4.5% for basimglurant 0.5 mg, 1.5 mg, and placebo, respectively. The most common adverse event was dizziness (4%, 23%, and 6%), mostly transient and of mild intensity. Mania (spontaneously resolved) lead to withdrawal of 2 patients from the study in the 1.5 mg arm.

Conclusions: Adjunctive 1.5 mg/d basimglurant showed a consistent antidepressant effect across primary and secondary endpoints. Greater effects were seen in patient-rated endpoints such as the patient-rated MADRS and the QIDS, which statistically separated from placebo at several time-points including week 6, while clinician-rated MADRS only separated at earlier time-points but not at day 42. Basimglurant 0.5 mg/d was not effective compared to placebo. Study results should be considered in the context of the observed high placebo response in this trial (47% on the clinician-rated MADRS). Placebo response rates>40% have been reported as a threshold that impedes observing statistical separation for active arms in adjunctive MDD treatment trials, minimizing the possibility of detecting true antidepressant effects. In this trial, nevertheless, basimglurant 1.5mg response rates were still consistently superior to placebo. Furthermore, basimglurant was overall safe and well tolerated in combination with SSRI/SNRI. These results warrant further investigation of basimglurant in the treatment of MDD both in the adjunctive as well as monotherapy settings.

Keywords: Basimglurant, mGlu5-NAM, Antidepressant, Major Depressive Disorder.

Disclosure: All authors are employees of Hoffmann-La Roche and may posses equity in the company.

T128. Efficacy and Safety of Low-field Synchronized Transcranial Magnetic Stimulation (sTMS) for Treatment of Major Depression

Andrew Leuchter*, Ian Cook, David Feifel, John Goethe, Mustafa Husain, Linda Carpenter, Michael Thase, Andrew Krystal, Noah Philip, William Burke, Robert Howland, Yvette Sheline, Scott Aaronson, Dan Iosifescu, Johnny O'Reardon, William Gilmer, Rakesh Jain, Karl Burgoyne, Joe Massaro, Sarah Lisanby, Mark George

University of California Los Angeles, Los Angeles, California

Background: Low-field magnetic stimulation delivered at an individual’s alpha frequency (IAF) significantly affects brain function, and may have therapeutic benefit for the treatment of Major Depressive Disorder (MDD). This study aimed to test the efficacy and safety of low-field transcranial magnetic stimulation synchronized to IAF (synchronized TMS, or sTMS) for treatment of MDD.

Methods: Six-week double blind sham controlled treatment trial of a novel TMS device that used three rotating neodymium magnets to deliver sTMS treatment. IAF was determined from a single-channel EEG prior to first treatment. 120 unmedicated adult subjects with moderate MDD (mean baseline 17-item Hamilton Depression Rating Scale (HamD17) score of 21.5) completed six weeks of treatment per-protocol. Antidepressant Treatment History scores ranged from 0-6. The primary study endpoint was change in HamD17 from baseline at six weeks. Subjects who did not enter remission at this endpoint (final HamD17 score<7) were eligible to receive four weeks of treatment using a separate open-label sTMS device.

Results: Subjects who received sTMS per-protocol (N=59) had significantly greater mean decrease in HamD17 scores after six weeks than those receiving sham treatment (N=60) (-9.00 versus -6.56, p=0.033). There was significant interaction between prior history of antidepressant treatment and efficacy: subjects with history of treatment resistance or intolerance in the current episode showed greater differential improvement (-8.58 vs. -4.25, p=0.017) and higher response rates (34.2% vs. 8.3%, p=0.017) to sTMS than sham. Treatment-naïve subjects had high response rates to both active and sham treatments, and showed no difference in mean decrease in HamD17 between the treatments. Subjects enrolled in the open-label phase showed significant continued symptom improvement. No serious adverse events were attributable to sTMS.

Conclusions: These results indicate that sTMS treatment was safe and significantly more effective than sham for the treatment of MDD of at least moderate severity in the per-protocol population. Subjects who received the active treatment as planned showed significantly greater improvement compared to sham treatment during the six-week double-blind phase, and subjects who did not remit during blinded treatment showed continued improvement during four weeks of open-label sTMS. The FDA has classified sTMS as non-significant risk (NSR), consistent with the very low rate of adverse events and the absence of any serious adverse events attributable to sTMS treatment in this study. Future studies should examine longer-term benefits of sTMS treatment.

Keywords: Major Depressive Disorder (MDD), Synchronized Transcranial Magnetic Stimulation (sTMS), Quantitative Electroencephalography (qEEG), Clinical Trials.

Disclosure: Leuchter- research support from the National Institutes of Health, Wyeth Pharmaceuticals, Novartis Pharmaceuticals, Seaside Therapeutics, Genentech, Shire Pharmaceuticals, Neuronetics, Lilly, and Neurosigma; consultant to: NeoSync,, Brain Cells,, Taisho Pharmaceuticals, Lilly, and Aspect Medical Systems/Covidien; is Chief Scientific Officer of Brain Biomarker Analytics LLC (BBA); owns stock options in NeoSync, and has equity interest in BBA. Cook: grant support from Aspect Medical Systems/Covidien, Cyberonics, Lilly, High Q Foundation, John E. Fetzer Foundation, John A. Hartford Foundation, MedAvante, Merck, NARSAD, NIH, NeoSync, Neuronetics, Novartis, Pfizer, Sepracor/Sunovion, Seaside Therapeutics, and West Coast College of Biological Psychiatry; served as an advisor or consultant to Allergan, Ascend Media, Bristol-Myers Squibb, Cyberonics, Lilly, Forest Laboratories, Janssen, Neuronetics, NeuroSigma, Pfizer, Scale Venture Partners, and the U.S. Departments of Defense and of Justice; speaker for: Bristol-Myers Squibb, CME LLC, Lilly, Medical Education Speakers Network, Pfizer, Neuronetics, NeuroSigma, and Wyeth; has active biomedical device patents which are are assigned to the University of California; has been granted stock options in NeuroSigma, the licensee of some of his inventions. Feifel: research support from NIH, KAVLI Institute, Merck, Sunovion, Forest, Ferring, Retrophin, NeoSync and Brainsway; served as expert consultant for Transcept, Mylan, Actavis and Neurocrine and is inventor on patent application for the use of oxytocin assigned to the University of California, San Diego but receives no royalties. Goethe: research from AHRQ, NIH, Hartfort hospital healthcare, Burlingame endowment, Bristol-meyers Squibb, Forest, Hoffman-La Roche, Janssen, Johnson & Johnson, Neuronetics, NeoSync, Novartis, Otsuka, Pfizer, Roche, Shire, and Takeda. Mustafa: research grant support from NIH/NIMH, NIDA, NINDS, NIA, NARSD, Stanley Foundation, Cyberonics, Neuronetics (past), St. Jude Medical (ANS), MagStim (equip only), Brainsway, NeoSync, Alkmers, and Corcept. Carpenter: consultant to: Abbott/AbbVie, Corcept, Helicon, Johnson&Johnson, Taisho, Takeda-Lundbeck, Magstim, & Naurex. Resarch support: Butler Hospital and Neuronetics, Cervel Neurotech, and Neosync. Thase: consultant to: Alkermes, AstraZeneca, Bristol-Myers Squibb Company, Cerecor, Lilly, Dey Pharma, L.P., Forest Laboratories, Gerson Lehman Group, GlaxoSmithKline, Guidepoint Global, H. Lundbeck A/S, MedAvante,, Merck and Co. (formerly Schering Plough and Organon), Neuronetics, Novartis, Otsuka, Ortho-McNeil Pharmaceuticals (Johnson & Johnson; Janssen), Pamlab, L.L.C., Pfizer (formerly Wyeth Ayerst Pharmaceuticals), PGx, Shire US, Sunovion Pharmaceutical, Supernus Pharmaceuticals, Takeda, and Transcept Pharmaceuticals; resarch support: Agency for Healthcare Research and Quality, Alkermes, Lilly, Forest Pharmaceuticals, National Institute of Mental Health, Otsuka Pharmaceuticals, PharmaNeuroboost and Roche; has equity holdings from MedAvante,; and earns royalties from American Psychiatric Foundation, Guilford Publications, Herald House, and W.W. Norton & Company. Krystal: grants/research support: NIH, Teva, Sunovion, Astellas, Abbott, Neosync, Brainsway, Novartis, Janssen; consultant to Abbott, Astellas, AstraZeneca, BMS, Teva, Eisai, Eli Lilly, GlaxoSmithKline, Jazz, Janssen, Merck, Neurocrine, Novartis, Roche, Somnus, Sunovion, Somaxon, Takeda, Transcept, Vantia; inventor on a Duke University owned patent that is licensed to MECTA Corp. Phillip: research support from Neosync, Cervel Neurotech and Neuronetics. Burke: research support from NIH, NCI, Robert Wood Johnson Foundation, Department of Veteran Affairs, NeoSync, Pfizer Inc, Eli Lilly and Vanda Pharmaceuticals; DMC with Otsuka Pharmaceuetical. Howland: research support from Brain & Behavior Research Foundation, Cyberonics, Medtronic, NeoSync, NIH and Shire. Yvette I. Sheline, M.D., received research support from Neosync. Aaronson: Advisory board/consultant for Alkermes, Genomind, and Lundbeck; speakers’ bureau for Sunovion, Lundbeck/Takeda, Otsuka; research support from Dalio Family Foundation, Genomind, Stanley Medical Research Foundation. Iosifescu: research support from NIMH, AstraZeneca, Brainsway, Euthymics, Neosync, Roche and Shire; consultant for Avanir, CNS Response, Otsuka, Lundbeck, Servier and Sunovion. O’Reardon: research support from Neosync and Stanley Medical Research Institute. Gilmer: Speakers bureau for Sunovion; advisory board for AstraZeneca; and research support from NeoSync, Inc. Jain: consultant/speaker/advisory board/research support from, Lilly, Lundbeck, Otsuka, Pfizer, Shire, and Takeda; consultant/speaker/advisory board for Addrenex, Merck, Pamlab, Shionogi, and Sunovion; consultant/speaker/research support from Forest; research support from AstraZeneca. Burgoyne: none. Massaro: paid statistics analyst for NeoSync. Lisanby: research support from NIH, Stanley Medical Research Institute, Duke Institute for Brain Sciences, Coulter Foundation, Brain & Behavior Research Foundation (Formerly NARSAD), NeoSync, Brainsway, ANS/St. Jude; equipment support from MagVenture and MagStim. Dr. Lisanby is listed as co-inventor on patent applications assigned to the University but receives no royalties. George: consultant to Brainsonix, Brainsway, Cephos, Cervel/NeoStim, Lake Biosciences, Mecta, Neuronetics, NeoSync, PureTech Ventures, and RCA Holdings, LLC.; has research support from Corcept, Brainsway, Cervel/NeoStim, Mecta, and NeoSync; study equipment support from Brainsway and Neuronetics; has 4 patents or invention disclosures in his name regarding brain imaging and stimulation filed by MUSC.

T129. Once-daily Oral Aripiprazole for Treatment of Tics in Children and Adolescents with Tourette’s Disorder: A Randomized, Double-blind, Placebo-controlled Trial

Floyd R. Sallee*, Eva Kohegyi, Joan Zhao, Robert McQuade, Kevin Cox, Raymond Sanchez, Margaretta Nyilas, William Carson, Roger Kurlan

University of Cincinnati School of Medicine, Cincinnati, Ohio

Background: The precise cause of Tourette’s disorder is unknown, but disturbances in dopaminergic and/or serotonergic pathways may play a role [1]. Aripiprazole can modulate the dopamine and serotonin systems as a partial agonist at dopamine D2 and serotonin 5-hydroxytryptamine (5-HT)1A receptors and as an antagonist at serotonin 5-HT2A receptors. Results of a previous exploratory controlled study suggested that treatment with oral aripiprazole may provide benefit in children and adolescents with Tourette’s disorder [2]. The primary objective of the present study was to further evaluate the efficacy of oral aripiprazole, compared to placebo, in the suppression of tics in patients with Tourette’s disorder. Safety and tolerability were also evaluated. The indication of tics associated with Tourette’s disorder for aripiprazole is currently under review at the US Food and Drug Administration.

Methods: This was a phase 3, multicenter, randomized, double-blind, placebo-controlled clinical trial in children and adolescents aged 7–17 years with a diagnosis of Tourette’s disorder (DSM-IV-TR), documented by the Kiddie Schedule for Affective Disorders and Schizophrenia – Present and Lifetime Version including the Diagnostic Supplement 5 (Substance Abuse and Other Diseases, ie, Tic Disorders). Patients also had a total tic score of ≥20 on the Yale Global Tic Severity Scale (YGTSS) at screening and baseline (randomization), and the tic symptoms had to cause impairment of each patient’s normal routine. After a 3- to 42-day screening/washout period, eligible patients were randomized (1:1:1) to receive oral treatment with low-dose aripiprazole (5 mg/d if<50 kg, 10 mg/d if ≥50 kg), high-dose aripiprazole (10 mg/d if<50 kg, 20 mg/d if ≥50 kg), or placebo once daily for up to 8 weeks. Aripiprazole was initiated at 2 mg/d and uptitrated to achieve the randomized dose by week 3. The primary efficacy endpoint was the mean change from baseline to week 8 in the YGTSS total tic score. A key secondary efficacy endpoint was Clinical Global Impression Scale-Tourette’s Syndrome scale (CGI-TS) change score at week 8. Adverse events (AEs) were monitored throughout the study.

Results: The randomized population included 133 patients (44 low-dose aripiprazole, 45 high-dose aripiprazole, 44 placebo); most were boys (78.2%) and white (87.2%). Mean age was 11.5 years. At baseline, the mean YGTSS total tic score was 29.3–31.5 across the 3 treatment groups. The least squares mean±SE change from baseline to week 8 in YGTSS total tic score was –13.4±1.6 with low-dose aripiprazole, –16.9±1.6 with high-dose aripiprazole, and –7.1±1.6 with placebo. There were significant treatment differences between the low-dose aripiprazole and placebo groups (–6.3 [95% CI, –10.2 to –2.3]; P=0.002) and between the high-dose aripiprazole and placebo groups (–9.9 [95% CI, –13.8 to –5.9]; P<0.0001). The treatment difference between the aripiprazole and placebo groups began by week 1. At each time point, the treatment effect was numerically greater with high- vs low-dose aripiprazole. For CGI-TS change score at week 8, there were also significant treatment differences between the low-dose aripiprazole and placebo groups (–1.0 [95% CI, –1.5 to –0.5]; P=0.0001) and between the high-dose aripiprazole and placebo groups (–1.0 [95% CI, –1.5 to –0.5]; P=0.0002). These between-group differences were significant by week 1. The treatment effect of the low- and high-dose aripiprazole groups was comparable at all time points. AEs were reported in 65.9% of patients in the low-dose aripiprazole group, 75.6% in the high-dose aripiprazole group, and 40.9% in the placebo group. The most common AEs were sedation (18.2% low-dose aripiprazole, 8.9% high-dose aripiprazole, 2.3% placebo), somnolence (11.4%, 15.6%, 2.3%, respectively), fatigue (6.8%, 15.6%, 0%), headache (6.8%, 8.9%, 2.3%), increased appetite (9.1%, 6.7%, 2.3%), and nausea (6.8%, 8.9%, 2.3%). There were no serious AEs. Rates of discontinuation due to AEs were 2.3% with low-dose aripiprazole, 15.6% with high-dose aripiprazole, and 2.3% with placebo; fatigue was the only AE to result in the discontinuation of>1 patient.

Conclusions: Both high- and low-dose aripiprazole showed efficacy in the treatment of tics in patients with Tourette’s disorder. AEs were in line with other studies of aripiprazole in children and adolescents [3-7]. If approved by the US Food and Drug Administration, aripiprazole represents a new option for treatment of tics in patients with Tourette’s disorder. Funding Source: This research was supported by Otsuka Pharmaceutical Development and Commercialization. 1. Steeves TD and Fox SH. Prog Brain Res. 2008;172:495-513. 2. Yoo HK, et al. J Clin Psychiatry. 2013;74(8):e772-780. 3. Findling RL, et al. J Clin Psychiatry. 2009;70(10):1441-1451. 4. Findling RL, et al. Am J Psychiatry. 2008;165(11):1432-1441. 5. Marcus RN, et al. J Am Acad Child Adolesc Psychiatry. 2009;48(11):1110-1119. 6. Owen R, et al. Pediatrics. 2009;124(6):1533-1540. 7. Findling RL, et al. Bipolar Disord. 2013;15(2):138-149.

Keywords: aripiprazole, pediatric, tics, Tourette’s.

Disclosure: FS and RK are consultants for Otsuka Pharmaceutical Development & Commercialization, Inc.; EK, JZ, RM, KC, RS, MN, and WC are employees of Otsuka Pharmaceutical Development & Commercialization, Inc.

T130. Effects of Lurasidone on Hostility in Patients with an Acute Exacerbation of Schizophrenia: A Pooled Post Hoc Analysis of Five Short-term Studies

Leslie Citrome*, Andrei Pikalov, Michael Tocco, Jay Hsu, Antony Loebel

New York Medical College, Valhalla, New York

Background: Symptoms of hostility are present in some patients during acute exacerbations of schizophrenia and have been associated with subsequent aggressive behavior. The management of hostility/aggression in patients with schizophrenia can pose a significant treatment challenge. There are data suggesting that some second-generation antipsychotics have specific antihostility effects (ie, independent of effects on positive symptoms in general, and independent of sedative effects). Lurasidone is an atypical antipsychotic agent with demonstrated efficacy in the treatment of schizophrenia. This post hoc analysis evaluated the efficacy of lurasidone compared with placebo for reducing hostility in patients hospitalized with an acute exacerbation of schizophrenia.

Methods: Individual patient data were pooled from 5 randomized, double-blind, placebo-controlled, 6-week studies of fixed-dose, once-daily, oral lurasidone (40-160 mg/d) conducted in patients with acute schizophrenia. The primary outcome measure was the overall change from baseline to week 6 on the hostility item (P7) of the Positive and Negative Syndrome Scale (PANSS) in patients with evidence of hostility at baseline as defined by a score of ≥2 on the PANSS hostility item. Treatment group differences were analyzed using a mixed-model repeated-measures (MMRM) analysis with treatment group and time as fixed effects. The analysis was conducted with and without adjusting for the following covariates in the model: the positive symptoms of schizophrenia (calculated as the sum of the PANSS items for delusions [P1], conceptual disorganization [P2], hallucinatory behavior [P3], grandiosity [P5], suspiciousness/persecution [P6], and unusual thought content [G9]) and the presence of somnolence recorded as a spontaneously reported adverse event (defined as hypersomnia, hypersomnolence, sedation, or somnolence).

Results: A total of 1148 patients met the criteria for hostility at baseline: 775 lurasidone-treated and 373 placebo-treated patients. Lurasidone was significantly superior to placebo in reducing the PANSS hostility item throughout the 6-week study period starting at Week 1 (P=0.002) and at all subsequent study visits (P<0.001). To examine whether antihostility effects were independent of improvement in positive symptoms, the MMRM model was adjusted using PANSS positive symptoms as a time-varying covariate. After adjusting for positive symptom change, lurasidone was found to be significantly better than placebo in decreasing the PANSS hostility item starting at Week 2 (P=0.014) and at every time point measured throughout the 6-week study period (P<0.05). To control for the effect of somnolence, the analysis was repeated by also introducing the presence of somnolence as a covariate. In this analysis, lurasidone significantly reduced the PANSS hostility item relative to placebo beginning at Week 2 and at every time point measured thereafter (P<0.05), except for Week 6. The proportion of patients who demonstrated any improvement (≥1 point change) on the PANSS hostility item score from baseline to endpoint was 63.1% (489/775) for patients randomized to lurasidone and 55.0% (205/373) for patients randomized to placebo (number needed to treat=13; 95% confidence interval, 8-49).

Conclusions: The results of this post hoc analysis showed that lurasidone significantly improved symptoms of hostility compared to placebo. This was a specific antihostility effect; improvement in hostility was found to be independent of change in other positive symptoms, as well as somnolence. Sponsored by Sunovion Pharmaceuticals Inc. ClinicalTrials.gov identifiers: NCT00088634, NCT00549718, NCT00615433, and NCT00790192. One study was completed prior to the requirement to register trials.

Keywords: schizophrenia, hostility, lurasidone.

Disclosure: In the past 36 months, Dr Citrome has received research grants from and served as a consultant or on the speakers' bureau for Alexza, Alkermes, AstraZeneca, Avanir, Bristol-Myers Squibb Company, Eli Lilly, Forest, Forum, Genentech, Janssen, Jazz Pharmaceuticals, Lundbeck, Merck & Co., Inc., Medivation, Mylan, Novartis, Noven, Otsuka, Pfizer Inc, Reckitt Benckiser, Reviva, Shire, Sunovion Pharmaceuticals Inc, Takeda, Teva, and Valeant; and he also a stock shareholder in Bristol-Myers Squibb, Eli Lilly, Johnson & Johnson, Merck & Co., Inc., and Pfizer Inc. Drs Pikalov, Tocco, Hsu, and Loebel are employees of Sunovion Pharmaceuticals Inc.

T131. Lurasidone in Bipolar Disorder: Early Improvement as a Predictor of Short-term Response

Dan Iosifescu*, Joyce Tsai, Andrei Pikalov, Jay Hsu, Josephine Cucchiaro, Antony Loebel

Icahn School of Medicine at Mount Sinai, New York, New York

Background: Lurasidone has demonstrated efficacy in the treatment of bipolar depression both as monotherapy and when used adjunctively with lithium or valproate. Early clinical improvement is a potentially important predictor of subsequent response that may guide treatment selection. The aim of this analysis was to evaluate the utility of early improvement in the Montgomery-Asberg Depression Rating Scale (MADRS) and the Clinical Global Impressions Bipolar Version, Severity of Illness (CGI-BP-S) scale as predictors of response to lurasidone in patients with bipolar depression.

Methods: Patients with bipolar I depression were randomized to 6 weeks of once-daily, double-blind treatment with lurasidone in a monotherapy study with fixed-flexible doses of 20-60 mg/day (N=167) and 80-120 mg/d (N=167) or placebo (N=168); and in an adjunctive therapy study with flexible doses of 20-120 mg/d (N=183) or placebo (N=163), both adjunctive with either lithium or valproate. For both the monotherapy and adjunctive therapy studies, the relationship between early improvement and response at Week 6 was assessed using two separate criteria (MADRS improvement ≥25%; CGI-BP-S improvement ≥1-point). Early improvement was examined at Week 2. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of early improvement for the prediction of endpoint response were estimated; endpoint response was defined as ≥50% reduction from baseline in MADRS total score. Receiver operating characteristic (ROC) curves were used to evaluate the performance characteristics of early improvement criteria for the prediction of endpoint response, based on the highest area under the curve (AUC).

Results: In the monotherapy study, the proportion of patients showing early improvement at Week 2 using the MADRS≥25% criterion was 47.6%, and the proportion using the CGI-BP-S≥1 criterion was 52.0%. For prediction of endpoint response at Week 2, the MADRS≥25% criterion had 64.9% sensitivity, 78.6% specificity, 82.0% PPV, 59.7% NPV, and AUCROC=0.877; and the CGI-BP-S≥1 improvement criterion had 62.8% sensitivity, 64.3% specificity, 72.7% PPV, 53.4% NPV, and AUCROC=0.852. In the adjunctive therapy study, the proportion of patients showing early improvement at Week 2 using the MADRS≥25% criterion was 51.7%, and the proportion using the CGI-BP-S≥1 criterion was 64.3%. For prediction of endpoint response at Week 2, the MADRS≥25% criterion had 68.5% sensitivity, 78.4% specificity, 85.1% PPV, 58.0% NPV, and AUCROC=0.919; and the CGI-BP-S≥1 improvement criterion had 77.2% sensitivity, 58.8% specificity, 77.2% PPV, 58.8% NPV, and AUCROC=0.883.

Conclusions: Early improvement (by Week 2) was found to have moderate value as a predictor of clinical response at Week 6 in patients treated with lurasidone for bipolar depression. Further analyses are needed determine the utility of early improvement for clinical decision-making.

Keywords: bipolar disorder, lurasidone, predictor, early improvement.

Disclosure: Relevant to this poster, Dr. Dan Iosifescu is a consultant for Lundbeck, Otsuka, and Sunovion; he has received research support (through the Icahn School of Medicine at Mount Sinai) from Astra Zeneca, Brainsway and Roche. Drs. Joyce Tsai, Andrei Pikalov, Jay Hsu, Josephine Cucchiaro, and Antony Loebel are full time employees of Sunovion Pharmaceuticals, Inc.

T132. A Single Assessment with the Brief Adherence Rating Scale (BARS) Discriminates Responders to Long-acting Injectable Antipsychotic Treatment in Patients with Schizophrenia

Matthew Byerly*, Paul Nakonezny, T. Scott Stroup, Joseph McEvoy, Robert Hamer, Marvin Swartz, Robert Rosenheck

University of Texas Southwestern Medical Center, Dallas, Texas

Background: Antipsychotic medication non-adherence is frequent in persons with schizophrenia (i.e., approximately 50% over 1 year) and is associated with serious consequences. Long-acting injectable antipsychotics (LAIs), a potentially effective non-adherence intervention, are underutilized considering they are the first-line non-adherence intervention recommendation in all major schizophrenia treatment guidelines. A likely critical factor leading to the underutilization of LAIs is the inability of prescribers to identify non-adherence in their own patients. A prior study conducted by our group found that relative to the “gold standard” of electronically monitored adherence patients and prescribers dramatically under-estimated antipsychotic non-adherence (57% vs. 5% & 7%, respectively). A second, recent study found that 94% of schizophrenia patients with low-to-moderate (≤70%) adherence levels identified by prescription fill data had high (≥71%) prescriber-estimated adherence. The Brief Adherence Rating Scale (BARS), which was developed and validated by our group, could potentially address prescribers’ challenge in identifying non-adherence. The BARS is a brief, pencil-paper, clinician-administered adherence instrument. It consists of three questions and a visual analog scale rating (0% to 100%). The BARS requires less than 45 minutes of staff training and takes less than 5 minutes to complete each assessment. Relative to electronic monitoring, the BARS provides valid, reliable, sensitive, and specific estimates of antipsychotic medication adherence of outpatients with schizophrenia and schizoaffective disorder (Byerly, 2008). The aim of this study was to determine if, in patients with schizophrenia, a single assessment with the BARS could discriminate responders switched from oral to LAI antipsychotic treatment.

Methods: As part of the multi-site LAI ACLAIMS study (McEvoy, 2014), the BARS scale was completed at baseline in schizophrenia and schizoaffective disorder patients taking oral antipsychotics who were at risk of efficacy failure based on a history of medication noncompliance, significant substance abuse, or both. Subjects were randomized 1:1 to received LAI paliperidone palmitate or LAI haloperidol decanoate and followed for clinical outcomes. A Receiver Operating Characteristic (ROC) analysis was conducted to evaluate how well the measure of baseline BARS adherence discriminated response status (defined as at least a 20% decrease from baseline to month 3 on the PANSS total score). The Area Under the Curve (AUC) for the measure of BARS adherence was tested against a nominal area of 0.50 using the Z statistic. The ROC analysis also determined the optimal cutpoint for the measure of baseline BARS adherence (based on the highest Youden Index) in discriminating response status at month 3. Sensitivity and specificity were also determined.

Results: The evaluable sample consisted of 176 participants, which included 77% males and 53% African Americans. Average age was 43.3 years. Mean illness onset was 23.8 years, with 72% of the sample having a diagnosis of schizophrenia. Average baseline and month 3 PANSS total scores were 72 (SD=16) and 63 (SD=16), respectively. Average baseline BARS adherence was 79.5% (SD=30), with a range of 0% (n=9/176, 5%) to 100% (n=64/176, 36%). About 23% or 40 of the 176 evaluable participants had a BARS adherence rating at baseline of ≤ 66%. Responders to LAI treatment comprised 58 out of the 176 participants or 33% of the evaluable sample. The ROC analysis determined that a single BARS rating at baseline with a cutoff ≤ 66% BARS adherence best discriminated those who responded to LAI treatment at month 3 (AUC: 0.60, SE: 0.05, 95% binomial exact CI: 0.53 to 0.68, Z=2.24, p=.025), with 38% sensitivity and 85% specificity along with a PPV of 55% and NPV of 74%. Indeed, 55% of patients with a baseline BARS adherence rating of ≤ 66% responded to LAI treatment at month 3, whereas only 26% of the patients with a baseline BARS adherence rating of>66% responded (p=.001). The logistic regression revealed that patients who were ≤ 66% BARS adherent at baseline had 3.40 times the predicted odds of responding to LAI treatment at month 3 (odds ratio=3.40, 95% CI=1.64 to 7.05, Wald χ2=10.76 Chi-Square=10.76, p=0.001).

Conclusions: A single assessment with the BARS discriminated responders to LAI treatment, with nearly twice the response rate in patients identified as non-adherent vs. adherent (55% vs. 26%). The relatively low sensitivity (38%) of the single BARS assessment indicates that the test misses a considerable portion of responders who are negative on the test (i.e., responders who are “adherent” as defined by BARS of>66%). This shortcoming is mitigated in part because clinicians generally would NOT provide LAI to adherent patients. A high specificity (85%) indicates the BARS is very unlikely to “waste” resources by providing LAI to patients who will not benefit (just 15% of non-responders would receive LAI based on BARS cutpoint generated by a single BARS assessment). High specificity is what clinicians and community mental health centers would desire in the setting of prescribing LAIs. The BARS appears to offer a reasonable screening tool for those who are candidates for LAI treatment.

Keywords: Schizophrenia, Adherence, long-acting injectable, antipsychotic.

Disclosure: Drs. Byerly, McEvoy, Hamer, and Stroup report receipt of a grant from the National Institute of Mental Health (NIMH). Dr. McEvoy reports receipt of a grant from Merck, Sunovion, Roche/Genentech, GlaxoSmithKline (outside submitted work), and Psychogenics (outside submitted work); and personal fees (speaking honoraria) from Lilly, Merck, and Sunovion; and personal fees for consulting from Otsuka, Roche/Genentech, Envivo, and Alkermes. Dr. Byerly reports receipt of research support from Otsuka; a grant from Sunovion; and personal fees from Merck, Novartis, & Otsuka. Dr. Hamer reports receipt of personal fees (data safety and monitoring board) from Novartis, Roche, Protein Sciences, Alkermes, Allergan, Abbot/Abbvie, Bioline, and Columbia University, (clinical trials consulting) from Lilly, AstraZeneca, Duke University, Cenerx, and National University of Singapore/Duke, (expert witness) from Winston and Strawn, Sheppard Mullin, Rakoczy Molino Mazzochi Siwik, and Goldberg Segalla, (grant review panel) from Veterans Administration, and (mock advisory panel) from Titan, and Neurogex outside the submitted work. Dr. Swartz reports receipt of personal fees for consulting from Med-IQ outside the submitted work. Dr. Rosenheck reports receipt of personal fees (expert witness) in Jones ex rel the State of Attorney Genera of Texas in Texas v Janssen Phamaceutica et al, and (consultant) from Otsuka outside the submitted work. Dr. Stroup reports participation in CME activities funded by Genentech outside the submitted work.

T133. Treatment of Normal Older Persons with Subjective Cognitive Impairment (SCI) with Neurogenesis Enhancer, Antiamyloid Medications: Initial 2 Year Electrophysiologic Observations

Barry Reisberg*, Brittany Cerbone, Santosh Ghimire, Thet Oo, Palak Patel, George Hoover, Leslie Prichep

New York University Alzheimer's Center, New York, New York

Background: Alzheimer’s disease (AD) is now known to begin many years before dementia becomes manifest. The global deterioration scale (GDS) (Reisberg et al., Am J Psychiatry, 1982) identified 2 pre-dementia stages of eventual AD. GDS stage 3, for which the terminology mild cognitive impairment (MCI) was coined, lasting 7 years, in which there are subtle, manifest symptoms; and an earlier GDS stage 2, in which there is subjective cognitive impairment (SCI) only. Longitudinal studies have confirmed an 15 year duration for this SCI stage (Reisberg, et al., Alzheimers Dement, 2008). Therefore, SCI begins>20 years prior to the mild dementia in AD. No medications have been approved for the prevention of AD in these pre-dementia stages. We hypothesized that neurogenesis enhancer (NE) medications, such as antidepressants, might be effective in decreasing decline in the very early, SCI stage.

Methods: We are conducting a randomized, double blind, placebo controlled study of 2 antidepressants, Lexapro (escitalopram) (5 mg/day at baseline) and Effexor XR (venlafaxine extended release) (37.5 mg/day at baseline). The brand medications are used. Eligibility criteria include: healthy with SCI (GDS stage 2); 60-80 years of age; MMSE ≥ 28; no psychoactive or cognitively acting medication within 8 weeks of study entry; and no significant psychiatric or neurologic disease. Subjects are randomized to one of the 3 treatments and receive blinded medication or placebo for a 2 year period followed by a 6 month (mo.) post treatment evaluation. The primary outcome measure in this study has been change on quantitative EEGs (Q-EEGs). These measures were selected on the basis of our prior observation of: (1) continuous slowing of Q-EEG activity from no cognitive impairment (NCI), to SCI, to MCI, to successive stages of AD (Prichep, et al., Neurobiol Aging, 1994), and (2) our finding that Q-EEG slowing in SCI subjects predicted decline at a 9 year mean follow up (f/u) (Prichep, et al., Neurobiol Aging, 2006).

Results: Results are presented herein for the primary outcome measure, Q-EEG, for the initial 2 subjects to complete the double-blind protocol. For subject 1, a 66 y/o ♀ at study entry, 2 year treatment results and 6 mo. post treatment (f/u) results are presented. The following regions of interest (ROIs) were studied: (1) left (L) hippocampus, (2) right (R) hippocampus, (3) L, and (4) R, superior and transverse temporal gyri, (5) L and (6) R dorsolateral prefrontal cortex (DLPFC). Mean Z-scores for voxels in each of the ROIs were calculated based on the source localization of the scalp recorded EEG in the theta frequency band. Z-scores were computed relative to age expected normal values for the age of the subject, and expressed as probability. A Z-score of±2.54 was equivalent to p<0.01 significance. The baseline values for subject 1 showed highly significant over activation, in comparison with normative values for the subject’s age in the hippocampus (L and R) and in the superior and transverse temporal cortices (L and R), but were within normal limits for the DLPFC. At the 2 year and 6 mo. post treatment f/u, the magnitude of the EEG activity did not differ significantly from the baseline in any of the brain regions examined. Subject 2 was a 65 y/o ♀ at baseline. Her baseline values in 5 of the 6 regions examined did not differ significantly from the age related EEG norms. In the L superior and transverse temporal gyrus, there was a significant over activation at baseline in comparison with the age related normal activity level. At 2 years there was a significant decrease in EEG activity in the left hippocampus, left superior and transverse temporal gyri and in the L DLPFC. In accord with the protocol, the subjects’ blind status was broken at the 2 year f/u visit. Subject 1 received Effexor XR and subject 2 received Lexapro. Both subjects were maintained at their baseline dosage levels throughout the study. Subject 1 wished to continue with her assigned medication at the conclusion of the study. However, she did not f/u with medication treatment after the 2 year f/u point. Subject 2 reported at the 2 year visit: “No problems…with the medication…everything seems better…more acute memory...better sense of smell…recall is faster…remembers where [she] placed things…less forgetting…quicker recall, [reading]”, and that she was “sorry to be going off the medication.".

Conclusions: To our knowledge, this is the first report of the effects of NE medications in subjects with pre-MCI, SCI. In 2013, neurogenesis was conclusively shown to occur in, and be important in, humans (Kempermann, Science, 2013) and in 2014, the SSRI citalopram was demonstrated to decrease CSF amyloid (Sheline, et al., Sci Transl Med, 2014). Retrospectively, antidepressant usage has been associated with decreased brain amyloid in humans (Cirrito, et al., PNAS, 2011). Our initial data, reported herein is supportive of positive effects of the Lexapro, 5 mg, over 2 years, and the absence of changes on the Effexor XR can be viewed as positive or neutral in terms of expected age related changes. Clearly, these approaches to the remediation of SCI, on a continuum with eventual MCI and AD, are worthy of continuing investigation.

Keywords: Subjective Cognitive Impairment, Antidepressants, Prevention of Alzheimer's disease, Neurogenesis enhancers.

Disclosure: Patent filing: Prevention of mild cognitive impairment and eventual Alzheimer's disease, U.S. Provisional Patent Application Serial No. 61/592,336. Final application filed January 30, 2013.

T134. Meta-Analysis: Ketamine for the Treatment of Depressive Symptoms

Michael Bloch*, Ewgeni Jakubovski, Hope Turner

Yale Child Study Center, New Haven, Connecticut

Background: Ketamine is a high-affinity N-methyl-D-aspartate (NMDA) receptor antagonist that has demonstrated rapid antidepressant effects in clinical studies of depressed subjects. Previous meta-analysis has demonstrated that ketamine is an effective treatment for depression compared to inactive control conditions at 24-hours after infusion. The goal of the current meta-analysis is to analyze (1) the time course of ketamine response and empirically determine the time-point of maximum effect and (2) examine how moderators influenced the time-course of ketamine response.

Methods: A PubMED search was performed by 2 reviewers to identify trials that (1) examined Depressive Disorders (e.g. Major Depression, Bipolar Depression or Dysthymia), (2) studied ketamine as a treatment and (3) included a control condition. At all time-points where more than half of studies reported outcome, we conducted a meta-analysis for the effects of ketamine using Comprehensive Meta-Analysis Version 2.2. Our two primary outcomes were change in depression severity using the study-reported primary outcome scale. Standardized mean difference (SMD) was the summary measure. We also used the proc mixed procedure to conduct a meta-analysis of longitudinal time point data between 1 and 7 days following infusion in order to model ketamine response. We performed additional analyses to examine the effects of moderating variables of interest including (1) diagnostic indication; (2) concomitant medication use; (3) method of ketamine administration and (4) control (saline vs. midazolam).

Results: Eight trials involving 180 subjects studied ketamine in the treatment of depressive disorders in controlled trials. The maximum benefit of depressive symptoms from ketamine was observed Day 1 post infusion (SMD=1.24±0.13 (95% CI: 0.99-1.50), z=9.5, p<0.001, heterogeneity: χ2=5.1, df=7, p=0.64, I2=0%). The benefits of ketamine were remained significant one week after infusion (Day 7, SMD=0.43±0.13 (95% CI: 0.18-0.67), z=3.4, p=0.001, heterogeneity: χ2=9.7, df=6, p=0.14, I2=38%). Concomitant medication use was associated with decreased benefits of ketamine in the week following infusion.

Conclusions: Meta-analysis indicates that ketamine is an effective treatment for depression compared to control conditions. Subjects taking concomitant medications at the time of ketamine infusion may have more ephemeral benefit from ketamine. Further research involving meta-analysis using individual patient data is needed to establish additional predictors of acute ketamine response and more sustained benefits of treatment.

Keywords: Ketamine, Depression, Meta-analysis.

Disclosure: Nothing to Disclose.

T135. Adjunctive Armodafinil 150 mg/d in Combination with Lamotrigine, Olanzapine, or Quetiapine Maintenance Therapy for Bipolar I Depression: A Pooled Sub-group Analysis of Efficacy from Phase 3 Studies

Terence Ketter*, Jess Amchin, Ronghua Yang, Mark A. Frye

Stanford University School of Medicine, Stanford, California

Background: Olanzapine combined with fluoxetine, and lurasidone adjunctively to lithium or valproate, are currently FDA-approved treatments for depressive episodes associated with bipolar I disorder. Armodafinil (R-modafinil) is a wakefulness-promoting, low-affinity dopamine transport inhibitor currently approved in the US for the treatment of excessive sleepiness associated with narcolepsy, obstructive sleep apnea, and shift work disorder. Earlier research on modafinil and armodafinil provided a signal for potential benefit as adjunctive treatment for acute bipolar depression. Subsequently, 3 similarly designed phase 3 studies investigating adjunctive armodafinil in bipolar depression yielded varying efficacy results (only 1 with statistical significance vs placebo; 2 with a non-significant numerical advantage vs placebo). This post hoc pooled analysis evaluated the efficacy of adjunctive armodafinil vs adjunctive placebo in subsets of patients undergoing maintenance treatment with lamotrigine, olanzapine, or quetiapine.

Methods: This pooled analysis of 3 multicenter, randomized, double-blind, placebo-controlled studies investigated the efficacy of adjunctive armodafinil 150 or 200 mg/d (200 mg/d in 2 studies only) vs adjunctive placebo when added to ongoing bipolar I maintenance therapy in adults aged 18-65 years with bipolar I depression. At entry, patients were taking protocol-defined “mood stabilizers” (lithium, valproate, lamotrigine, olanzapine, risperidone, aripiprazole, ziprasidone [ziprasidone only in combination with lithium or valproate in 2 studies; only in combination with lithium, valproate, or lamotrigine in 1 study], or quetiapine [1 study only]). The primary efficacy assessment was mean change from baseline to week 8 in the 30-item Inventory of Depressive Symptomatology-Clinician-Rated (IDS-C30) total score analyzed by mixed-model repeated measures. Secondary efficacy assessments included mean change from baseline in IDS-C30, IDS-C30 response (≥50% final reduction from baseline total score) rates, and IDS-C30 remission (final IDS-C30 ≤11) rates, each assessed at weeks 1, 2, 4, 6, 7, and 8 (or early termination). Randomization to 200 mg/d (2 studies, n=62) was discontinued early, and this group was not included in the efficacy analysis. Continuous variables were analyzed using analysis of variance, and categorical efficacy variables using the Cochran-Mantel-Haenszel test. Differential efficacy is reported for armodafinil 150 mg/d vs placebo when added in subgroups taking lamotrigine, olanzapine, or quetiapine maintenance.

Results: Overall, 1,261 patients with bipolar I depression (mean age 44.0 years, 60% female, 82% white) were randomized to adjunctive armodafinil 150 mg/d (n=633) or adjunctive placebo (n=628). In the overall pooled population, least-squares mean (LSM)±standard error (SE) IDS-C30 change from baseline was significant for adjunctive armodafinil vs adjunctive placebo at week 8 (-21.2±0.59 vs -18.8±0.59, P=0.0021 ). Among the subgroups of interest, 208 took lamotrigine (armodafinil, n=101; placebo, n=107), 167 took olanzapine (armodafinil, n=80; placebo, n=87), and 67 took quetiapine (armodafinil, n=37; placebo, n=30). Baseline mean IDS-C30 scores in the adjunctive armodafinil and adjunctive placebo groups, respectively, were 43.8 and 44.7 with lamotrigine, 40.4 and 42.8 with olanzapine, and 42.1 and 42.2 with quetiapine. LSM±SE IDS-C30 change from baseline was significant for adjunctive armodafinil vs adjunctive placebo at week 8 (-20.7±2.14 vs -16.2±2.04, P=0.0096) when used with lamotrigine, but at no other time point. No statistically significant IDS-C30 changes were seen with adjunctive armodafinil vs adjunctive placebo in combination with olanzapine or quetiapine at any visit. IDS-C30 response (≥50% decrease from baseline) was statistically significant at week 1 in the olanzapine group (9% vs 2%, P=0.0178), and IDS-C30 remission (≤11) was significant at weeks 2 (4% vs 0%, P=0.0429) and 8 (38% vs 21%, P=0.0439) in the lamotrigine group, but at no other time points in any group. The overall safety population included 1,317 patients (691 taking adjunctive armodafinil 150 or 200 mg/d; 626 taking adjunctive placebo). Overall, 351 (51%) patients taking adjunctive armodafinil and 264 (42%) taking adjunctive placebo had ≥1 adverse event (AE). The most common AEs with adjunctive armodafinil were headache, nausea, diarrhea, and insomnia. Safety/tolerability findings by maintenance subgroup will be presented.

Conclusions: In a post hoc pooled analysis of 3 phase 3 studies evaluating adjunctive armodafinil vs adjunctive placebo in bipolar I depression, the addition of armodafinil 150 mg/d to lamotrigine or olanzapine resulted in significant improvements vs adjunctive placebo at a few assessment points; no superiority over placebo was noted with armodafinil added to quetiapine. Overall, adjunctive armodafinil was generally well tolerated. Findings are limited by the post hoc nature of this analysis and small subgroup sample sizes, especially for quetiapine.

Keywords: Bipolar I Depression.

Disclosure: J. Amchin and R. Yang are employees of Teva Pharmaceuticals. T. Ketter and M. Frye are consultants for Teva Pharmaceuticals. This study was sponsored by Teva Pharmaceuticals. Medical writing support was provided by John H. Simmons, MD, at Peloton Advantage, LLC, and was funded by Teva Pharmaceuticals.

T136. Baseline Blood Pressure is Associated with PTSD Symptom Response to Prazosin in Active Duty Combat Soldiers

Murray Raskind*, Elaine Peskind, Steve Millard, Eric Petrie

VA Puget Sound Health Care System, Seattle, Washington

Background: Prazosin, a CNS active alpha-1 adrenoreceptor (AR) antagonist, was demonstrated effective for combat trauma PTSD in a randomized controlled trial (RCT) in active duty soldiers returned from Iraq and Afghanistan combat deployments. However, therapeutic response was variable among participants. A possible neurobiologic explanation of prazosin efficacy is variability of responsiveness among soldiers of the central nervous system (CNS) alpha-1 AR that regulate the arousal response to CNS norepinephrine. Unfortunately, CNS alpha-1 AR responsiveness cannot be measured directly. However peripheral alpha-1 AR responsiveness can be estimated by measuring resting systolic blood pressure (BP) and the maintenance of systolic BP following postural change from supine to standing. Because peripheral and CNS noradrenergic activity are co-regulated in most situations, we hypothesize that higher systolic BP and smaller orthostatic systolic BP reduction at BASELINE would be associated with greater therapeutic response to prazosin in combat trauma PTSD.

Methods: We analyzed the effects of baseline BP parameters on PTSD outcome measures (total Clinician Administered PTSD Scale [CAPS] score, CAPS B-2 nightmare item, Clinical Global Impression of Change, Pittsburgh Sleep Quality Index) responses to prazosin using linear mixed effects models. Analyses are reported separately for participants who received prazosin (n=32) and for participants who received placebo (n=35). The model included terms for time, baseline BP and a time by baseline interaction term.

Results: Consistent with our hypotheses, there were significantly and substantially greater reductions (greater improvement) in total CAPS score with prazosin treatment (change in slope per 10mm Hg increase in baseline BP) (p=0.002) and a trend for smaller reduction (less improvement) with greater baseline orthostatic systolic drop (p=0.10). Other combinations of baseline BP parameters (supine systolic, supine and standing diastolic, and orthostatic diastolic change) and PTSD outcome measure responses to prazosin were similarly significant or demonstrated trends in the predicted direction. In contrast, the same analyses in participants treated with placebo detected no signal for a baseline BP effect on treatment response.

Conclusions: Higher baseline BP and smaller baseline BP drop are associated with substantially greater PTSD symptom improvement with prazosin treatment. These results suggest that peripheral BP, which reflects increased peripheral alpha-1 AR responsiveness, provides an indicator of increased CNS alpha-1 AR responsiveness in combat trauma PTSD. Such increased alpha-1 AR responsiveness and/or activation are the presumed target for prazosin therapeutic efficacy. Baseline BP could be a clinically useful biomarker for helping to predict the response to prazosin or other alpha-1 AR antagonist treatment outcomes in PTSD.

Keywords: prazosin, PTSD, combat, soldiers.

Disclosure: Nothing to Disclose.

T137. Impact of Atypical Antipsychotic Dose Reduction on Cognitive Function and Subjective Experiences

Hiroyoshi Takeuchi*, Takefumi Suzuki, Gary Remington, Robert Bies, Koichiro Watanabe, Masaru Mimura, Hiroyuki Uchida

University of Toronto, Toronto, Canada

Background: There have been only a few trials that evaluated the feasibility of dose reduction of atypical antipsychotics in stable patients with schizophrenia; moreover, no study has evaluated cognitive function or subjective experiences following the dose reduction. In addition, while acute-phase antipsychotic response has been attributed to 65–80% dopamine D2 receptor blockade, the degree of occupancy for relapse prevention in the maintenance treatment of schizophrenia remains unknown.

Methods: In this open-label, 28-week, randomized controlled trial, patients with schizophrenia were included if they had been receiving a stable dose of risperidone or olanzapine for ≥3 months and presented a score of ≤3 on all of the following items: P1, P2, P3, and P6 in the Positive and Negative Syndrome Scale (PANSS). They were randomly assigned to reduction or maintenance group. In the reduction group, the antipsychotic dose was reduced by 50% in 4 weeks and then maintained for 24 weeks while the dose was kept constant in the maintenance group. The following assessments were performed at baseline and endpoint: the PANSS, the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), the Subjective Well-being under Neuroleptic drug treatment Short form (SWNS), and the Drug Induced Extrapyramidal Symptoms Scale (DIEPSS). Plasma concentrations of risperidone or olanzapine at peak and trough were estimated using the population pharmacokinetic analysis and the corresponding dopamine D2 receptor occupancy levels were estimated.

Results: Sixty-one patients were enrolled. Two of 31 patients (6.5%) and 5 of 30 patients (16.7%) prematurely withdrew from the study in the reduction and maintenance groups, respectively. While no significant differences in changes in the PANSS or SWNS total scores were observed between the two groups, the reduction group showed greater improvements in the RBANS and the DIEPSS, compared to the maintenance group (changes in the total scores:+7.0±7.1 vs. -0.1±8.0, p<0.001; -0.9±1.7 vs.+0.1±1.2, p=0.010, respectively). Plasma concentrations were available in 31 patients (16 and 15 in the reduction and maintenance groups, respectively). Estimated D2 occupancy levels decreased following the dose reduction from 75.6±4.9% to 66.8±6.4% at peak and 72.3±5.7% to 62.0±6.8% at trough. In the reduction group, 10 patients (62.5%) did not demonstrate continuous D2 blockade above 65% (i.e.<65% at trough) after dose reduction; furthermore, seven patients (43.8%) did not achieve a threshold of 65% occupancy even at peak. Nonetheless, only one patient met our relapse criteria after dose reduction over six months.

Conclusions: Dose reduction of risperidone or olanzapine by half could improve cognitive function without significantly increasing the risk of worsening in clinical psychopathology or improving subjective experiences for stable patients with schizophrenia. Additionally, sustained D2 occupancy at 65% or more with antipsychotics may not be necessary for relapse prevention in the maintenance treatment of schizophrenia.

Keywords: antipsychotics, cognitive function, dose reduction, schizophrenia, subjective experiences.

Disclosure: Dr. Takeuchi has received fellowship grants from the Canadian Institute of Health Research (CIHR), the Centre for Addiction and Mental Health (CAMH) Foundation, the Japanese Society of Clinical Neuropsychopharmacology, and Astellas Foundation for Research on Metabolic Disorders, speaker’s honoraria from Dainippon Sumitomo Pharma, Eli Lilly, GlaxoSmithKlein, Janssen Pharmaceutical, Meiji Seika Pharma, and Otsuka Pharmaceutical, and manuscript fees from Dainippon Sumitomo Pharma within the past 5 years. Dr. Suzuki has received fellowship grants from the Japanese Society of Clinical Neuropsychopharmacology, the Government of Canada Post-Doctoral Research Fellowships, Kanae Foundation, and Mochida Memorial Foundation, speaker’s honoraria from Eli Lilly, Shionogi, and Yoshitomiyakuhin, Novartis Pharma, Meiji Seika Pharma, Astellas Pharmaceutical, Janssen, and Otsuka Pharmaceutical, and manuscript fees from Dainippon Sumitomo Pharma, Elsevier Japan, Weily Japan, and Kyowa Hakko Kirin within the past 5 years. Dr. Bies has received grants from Eli Lilly through the Indiana CTSI, Merck and Company through Regenstrief, the National Institutes of Health (NIH), the National Institute of Child Health and Development (NICHD), and the National Center for Advancing Translational Sciences (NCATS) within the past 5 years. He is also a member of the scientific advisory board for Metrum Institute (an unpaid position). Dr. Remington has received research support from Novartis, Medicure, and Neurocrine Bioscience, consultant fees from Roche, and speaker’s fees from Novartis. He holds no commercial investments in any pharmaceutical company within the past 5 years. Dr. Watanabe has served on the advisory board of Dainippon Sumitomo Pharma, Eli Lilly, Janssen Pharmaceutical, Otsuka Pharmaceutical, and Tanabe Pharma, received grants from Dainippon Sumitomo Pharma, GlaxoSmithKline, Meiji Seika Pharma, Mitsubishi Tanabe Pharma, and Pfizer Japan, and speaker’s honoraria or manuscript fees from Dainippon Sumitomo Pharma, Eli Lilly, GlaxoSmithKline, Janssen Pharmaceutical, Meiji Seika Pharma, Mitsubishi Tanabe Pharma, Mochida Pharmaceutical, Otsuka Pharmaceutical, Pfizer Japan, Shionogi, and Yoshitomiyakuhin within the past 5 years. Dr. Mimura has received grants or consultant fees from Astellas Pharma, GlaxoSmithKline, Eisai, and Meiji Seika Pharma, and speaker’s honoraria from Astellas Pharma, Dainippon Sumitomo Pharma, Eli Lilly, GlaxoSmithKline, Janssen Pharmaceutical, Meiji Seika Pharma, Otsuka Pharmaceutical, Pfizer Japan, and Yoshitomiyakuhin within the past 5 years. Dr. Uchida has received grants from Astellas Pharma, Dainippon-Sumitomo Pharma, Eisai, Eli Lilly, GlaxoSmithKline, Janssen Pharmaceutical, Meiji Seika Pharma, Mochida Pharmaceutical, Otsuka Pharmaceutical, Pfizer Japan, Shionogi, and Yoshitomiyakuhin, and speaker’s honoraria from Dainippon Sumitomo Pharma, Eli Lilly, GlaxoSmithKline, Janssen Pharmaceutical, Novartis Pharma, Otsuka Pharmaceutical, Shionogi, and Yoshitomiyakuhin within the past 5 years.

T138. Effect of Lurasidone on Metabolic Parameters in Patients with Bipolar Depression

John Newcomer*, Joyce Tsai, Andrei Pikalov, Hans Kroger, Josephine Cucchiaro, Antony Loebel

Florida Atlantic University, Charles E. Schmidt College of Medicine, Boca Raton, Florida

Background: Patients with bipolar disorder are at a higher risk of cardiovascular disease and mortality than the general population. A recent meta-analysis found that individuals with bipolar disorder treated with an antipsychotic medication had a significantly higher risk of developing metabolic syndrome compared with untreated subjects (Vancampfort et al, Am J Psychiatry 2013;170:265-74). The aim of the current analysis was to evaluate the extent to which treatment with lurasidone was associated with clinically relevant shifts in key weight and metabolic parameters in patients with bipolar depression.

Methods: Data were from 3 short-term studies in patients with bipolar I depression who were randomized to 6 weeks of once-daily, double-blind, placebo-controlled treatment with lurasidone (20-120 mg/d), either as monotherapy (one study, N=499), or adjunctive therapy with lithium (Li) or valproate (VPA; two studies, combined N=694). Patients completing these three 6-week studies continued to receive 6 months of additional treatment with lurasidone 20-120 mg/d in an open-label extension study (N=494). The proportion of patients with shifts in metabolic parameters were analyzed at the 6-week primary (LOCF) and 6 month extension (LOCF) end-points, and are reported descriptively. Shifts were defined as changes in weight (≥7% change), BMI category, triglycerides (increase ≥50 mg/dL), total cholesterol (increase ≥40 mg/dL), LDL (increase ≥30 mg/dL), glucose (shift to ≥110 mg/dL), and HBA1c (shift to ≥5.7%).

Results: At Week 6 in the monotherapy study, a similar proportion of patients treated with lurasidone vs placebo, respectively, had shifts in weight (2.0% vs 0.6%), triglycerides (15.7% vs 10.1%), total cholesterol (5.5% vs 3.4%), LDL (6.4% vs 3.4%), and glucose (11.3% vs 9.5%). At Month 6 of the extension phase study, a relatively low proportion of the patients who continued on lurasidone monotherapy met criteria for shifts in triglycerides (17.7%), total cholesterol (5.6%), and LDL (10.2%). Six months of lurasidone monotherapy was associated with a small net shift to higher BMI category in patients with normal or overweight BMI at Baseline (12.5% and 4.5%, respectively); and among patients with Baseline pre-diabetic HBA1c values (≥5.7% to ≤6.4%; n=79), there was a significant net shift (24.1%) to normal HBA1c values. At Week 6 in the adjunctive therapy studies, a similar proportion of patients treated with lurasidone vs placebo, respectively, had shifts in weight (3.4% vs 0.4%), triglycerides (16.4% vs 15.1%), total cholesterol (4.7% vs 5.6%), LDL (7.7% vs 8.0%), and glucose (6.2% vs 6.5%). Six months of lurasidone adjunctive therapy was associated with a moderate net shift to higher BMI category in patients with normal or overweight BMI at Baseline (22.5% and 9.5%, respectively); and among patients with Baseline pre-diabetic HBA1c values (≥5.7% to ≤6.4%; n=95), there was a net shift (42.1%) to normal values.

Conclusions: These data add to the substantial and growing body of information regarding the metabolic safety of lurasidone. Sponsored by Sunovion Pharmaceuticals Inc.

Keywords: Schizophrenia, Antipsychotic agents, lurasidone, lipids.

Disclosure: Dr. Newcomer has received grant/research support from the National Institutes of Health; honoraria from American Physician Institute, CME Outfitters, CMEology, and American Psychiatric Association; royalties from Jones and Bartlett Publishing; and has served on the Data Safety Monitoring Boards for Bristol-Myers Squibb, Merck, and Vivus. Drs. Tsai, Pikalov, Kroger, Cucchiaro, and Loebel are employees of Sunovion Pharmaceuticals Inc.

T139. Effects of Ketamine on Suicidal Ideation in Patients with Mood and Anxiety Spectrum Disorders: A Randomized Controlled Pilot Study

Laili Soleimani*, Kaitlin Dewilde, Joanna J. Kim, Kyle Lapidus, Marc Lener, Gloria Rodriguez, Andrew Perez, Jess Brallier, Dan V. Iosifescu, Dennis Charney, James W. Murrough

Icahn School of Medicine at Mount Sinai, New York, New York

Background: Suicide is a devastating behavioral outcome of emotional despair frequently attributed to mental illness and continues to be a serious worldwide public health problem. Patients with Major Depressive Disorder (MDD), Bipolar Disorder (BD), Post-traumatic Stress Disorder (PTSD), and Borderline Personality Disorder (BPD) are at an increased risk for suicidal ideation (SI) and behavior, however very few psychiatric interventions have been found to be effective in decreasing the intensity of suicidality. We have previously shown that a single intravenous (IV) infusion of ketamine, a high-affinity glutamate N-methyl-d-aspartate (NMDA) receptor antagonist, is associated with a rapid reduction in depressive symptom severity, including SI, in patients with treatment resistant major depression (TRD). The current study was designed to investigate whether ketamine, compared to a control condition, could decrease the severity of SI across psychiatric diagnoses.

Methods: We conducted a double-blind, randomized, controlled, pilot study of ketamine in inpatients and outpatients with clinically significant SI, defined as score of ≥4 on the Montgomery-Asberg Depression Rating Scale item 10 (MADRS-10) for suicidality. Eligible participants included males or females aged 18 to 80 who were either admitted voluntarily to an inpatient psychiatric unit with acute worsening of SI or who were outpatients with chronic SI without acute intent to harm themselves. Exclusion criteria included lifetime history of schizophrenia or other primary psychotic disorder, current psychotic or manic symptoms, or current substance use disorder within 1 month of screening. Patients underwent measurement of SI and depression severity at screening and immediately prior to treatment. Patients received a single IV infusion of ketamine (0.5 mg/kg) or midazolam (0.045 mg/kg) in addition to standard of care. Midazolam, a benzodiazepine anesthetic agent, was used in this study as an active placebo in order to mitigate threats to blinding. Change in suicidality was measured using the Beck Scale for SI (BSS) as well as the MADRS-10. The primary outcome time point was 24 hours following treatment. Patients were followed in person for at least 7 days and for up to 5 weeks via phone. Change in suicidal ideation from both screen and pre-treatment time points to 24 hours was compared separately between the two treatment groups using repeated measures analysis of variance (ANOVA).

Results: Twenty-four patients (10 inpatient, 14 outpatient) were randomized to receive ketamine or midazolam in a 1:1 scheme under double-blind conditions. Primary psychiatric diagnoses included MDD (n=16), BD (n=4), PTSD (n=3) and BPD (n=1). The two treatment groups were comparable in terms of severity of the depression and SI at screening and at pre-infusion baseline. The intervention was well tolerated and no dropouts occurred during the primary 7-day follow up period. Five SAEs occurred during the study, including re-hospitalization for worsening psychiatric symptoms (n=4) and one patient who expired from causes deemed unrelated to the study intervention. The most common non-serious adverse event reported in the ketamine group was headache (58.3%). Patients who received ketamine experienced transient dissociative symptoms measured by the Clinician Administered Dissociative Scale (CADSS) (mean 17.08±19.86), which resolved within 4 hours of the treatment. SI was lower in the ketamine compared to midazolam group at 24 hours post-treatment as measured by the BSS (9.33±7.9 and 14.1±8.4, respectively) and MADRS-10 (1.75±1.86 and 3.25±1.6, respectively). For change in BSS from screen to 24 hours, there was a significant main effect of time [F(1,22)=25.4, p<0.001] and a trend-level time x treatment interaction [F(1,22)=3.02, p=0.096). For change from pre-treatment to 24 hours, the effect was similar (p=0.15). For change in MADRS-10 from screen to 24 hours, there was a significant main effect of time [F(1,22)=30.3, p<0.001] and significant time x treatment interaction [F(1,22)=8.6, p=0.008]. For change in MADRS-10 from pre-treatment to 24 hours, there was a significant main effect of time [F(1,22)=15.7, p=0.001] and trend-level time x treatment interaction [F(1,22)=4.2, p=0.053].

Conclusions: The findings of the current study provide initial support for the safety, tolerability, and preliminary efficacy of ketamine as an intervention for SI in patients who are at elevated risk for suicidal behavior across psychiatric diagnoses. If these findings are replicated and extended in rigorous controlled trials, this line of research may have robust implications for clinical practice and public health. Larger, well-controlled studies are required to further characterize the safety and efficacy of ketamine for SI.

Keywords: Suicide, Treatment, Ketamine, Depression.

Disclosure: In the past 2 years, Dr. Murrough has served on advisory boards for Janssen Research and Development and Genentech and has provided consultation services for ProPhase, LLC and Impel Neuropharma. Dr. Iosifescu has received funding (through Icahn School of Medicine at Mount Sinai) from AstraZeneca, Brainsway, Euthymics, Neosync, Roche and Shire; and consulting fees from Avanir, CNS Response, Lundbeck, Otsuka, Servier and Sunovion. Dr. Lapidus received research support from NARSAD, the Brain and Behavior Research Foundation, APIRE Janssen, and Simons Foundation. He serves on the advisory board for Halo Neuro, Inc., has received devices from Medtronic, and consults for LCN Consulting, Inc. Dr. Dennis Charney (Dean of Icahn School of Medicine at Mount Sinai), and Icahn School of Medicine at Mount Sinai have been named on a use patent on ketamine for the treatment of depression. The Icahn School of Medicine has entered into a licensing agreement for the use of ketamine as therapy for treatment-resistant depression. Dr. Charney and Icahn School of Medicine at Mount Sinai could potentially benefit if ketamine were to gain approve for the treatment of depression.

T140. Potential EEG Biomarker for ASD in Adults: Reducing Heterogeneity in ASD Trials

James McCracken*, Sandra Loo, Iman Rezazedeh, Sara Jane Webb, Gwen Frishkoff, Bryan King, Lawrence Scahill, Margaret Grabb

University of California Los Angeles, Semel Institute for Neuroscience & Human Behavior, Los Angeles, California

Background: Previous electrophysiologic research (EEG) in children and adults with Autism Spectrum Disorder (ASD) have reported abnormal oscillatory activity in subjects versus controls as demonstrated by diverse EEG spectral power and functional connectivity differences. These abnormalities have been speculated to reflect a disruption in the excitatory/inhibitory balance of glutamate and GABA. However, few efforts have been made to utilize these measures as a means to reduce subject herterogeneity for the purpose of identifying putative etiologic subgroups. In preparation for a proof of concept clinical trial of a GABA positive modulator, we sought to develop an EEG biomarker of ASD which could be applied to screen subjects for study entry, and presumably identify those with greater disruption of GABAergic signaling.

Methods: We obtained 128 channel EEG recordings from 39 healthy adults from 18 - 35 year of age and 12 high functioning adults 18 - 35 years old with ASD under resting and cognitive activation conditions. The activation paradigm made use of the FACES paradigm, comprised of the presentation of emotional faces. To determine a candidate for discriminating between Controls and ASD patients we first generated z-scores for 12 EEG measures. The z-scores are based on the means and standard deviations of the control sample rounded to 1 digit after the decimal point. We chose this lower precision to minimize overfitting and to maximize generalizability to the population. We then used Fishers Linear Discriminant Analysis to select 3 variables that fulfilled the following conditions: (1) Their linear combination is maximally discriminating between patients and controls (2) Both tasks (eyes closed and faces) are represented (3) Both coherence and power are represented Based on these criteria we chose: A) From the FACES task: Theta relative power and Alpha Coherence B) From the eyes closed task: Low Beta Coherence The optimal linear combination of those (1*Low Beta Coherence+0.5*Alpha Coherence-0.5*Theta Ratio showed good discrimination (Area under the curve of the ROC curve: .85), and a decision criterion based on a cutoff value of .6 performs well (Sensitivity: 0.8, Specificity: 0.7).

Results: The optimal linear combination of those (1*Low Beta Coherence+0.5*Alpha Coherence-0.5*Theta Ratio showed good discrimination (Area under the curve of the ROC curve: .85), and a decision criterion based on a cutoff value of .6 performs well (Sensitivity: 0.8, Specificity: 0.7).

Conclusions: We have identified a straightforward method utilizing 3 EEG variables obtained from resting and activation conditions which show acceptable performance in discriminating ASD subjects from controls with very good sensitivity and good specificity. These EEG features are hypothesized to relate in underlying differences in the regulation of oscillatory activity of large populations of neurons, which is under prominent regulation by GABA signaling. If these findings can be validated using converging methods of GABA activity, our method may represent a convenient approach to reducing subject heterogeneity in ASD and serve to better identify individuals for targeted treatments.

Keywords: electrophysiology, clinical trials, GABA.

Disclosure: AstraZeneca-drug supply; Roche-consultant, research contract.

T141. DSM-5 Dimensions of Psychosis Symptom Severity: Understanding Treatment Response in Patients with Schizophrenia

A. Kalali*, C. Siu, J Cucchiaro, A. Pikalov, R. Goldman, F. Grossman, A. Loebel

Quintiles, San Diego, California

Background: The 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) provides a "Clinician-Rated Dimensions of Psychosis Symptom Severity" measure (Section III: Emerging Measures and Models). This dimensional assessment was developed to capture the underlying structure of psychosis, including the domains of hallucinations, delusions, disorganized speech, abnormal psychomotor behavior, negative symptoms, depression, cognitive impairment and mania. The objective of this analysis was to examine these dimensions of psychosis in acutely ill patients with schizophrenia and to evaluate antipsychotic treatment response using this dimensional structure.

Methods: Data were derived from a 6-week, placebo and active-controlled trial of lurasidone in hospitalized patients with an acute exacerbation of schizophrenia (Loebel et al., 2013). The standard 7 point scale for each PANSS item (1=Absent to 7=Extreme) was mapped on to the 5 point scale for each domain of the DSM-5 dimensional assesment (0=Not present, 1=Equivocal, 2=Mild, 3=Moderate, 4=Severe). ANCOVA and the mixed effects model were applied to compare PANSS scores for each psychosis domain. Cognitive impairment and mania symptom domains were excluded from this analysis.

Results: Of the 482 patients including in this analysis, most had moderate severity symptoms for hallucinations (70%), delusions (86%), disorganized speech (67%), and negative symptoms (75%); mild (41%) to moderate (35%) abnormal psychomotor behavior, and mild (36%) to moderate (38%) depression at study baseline. Both lurasidone (160 mg/d or 80 mg/d) and quetiapine XR (600 mg/d) showed significantly greater improvement (vs. placebo) on the 6 domains of psychosis symptoms that were assessed. In patients with at least moderate severity negative symptoms at study baseline (n=360), the higher lurasidone 160 mg/d dose group had significantly greater effect size (1.09) compared to the lower lurasidone 80 mg/d dose group (0.63) and the quetiapine XR 600 mg/d group (0.83) (p<0.05). In patients with at least moderate severity depression symptoms at study baseline (n=181), the higher lurasidone dose group (1.26) and the quetiapine XR 600 mg/d group (1.0) had significantly greater effect size than the lower lurasidone dose group (0.57) (p<0.05). For the other 4 psychosis domains, effect size was numerically greater in the higher lurasidone dose group compared to the lower dose group in patients with moderate to severe symptoms.

Conclusions: Larger treatment effects were consistently observed at the higher lurasidone dose (160 mg/d) for the treatment of moderate to severe psychosis symptoms. The lower lurasidone dose (80 mg/d) was comparably effective for less severe symptoms. Dimensional assessment of symptom severity (per DSM-5) can aid understanding of treatment response across the various domains of psychotic illness.

Keywords: DSM-5, dimensions of psychosis, schizophrenia, lurasidone.

Disclosure: AK is employee of Quintiles and consultant to Sunovion Pharmaceuticals Inc., CS is consultant to Sunovion Pharmaceuticals Inc. All other authors are employees of Sunovion Pharmaceuticals Inc.

T142. Nitrous Oxide for Treatment-resistant Major Depression: A Proof of Concept Study

Peter Nagele, Andreas Duma, Michael Kopec, Marie Gebara, Alireza Parsoei, Marie Walker, Vassilis Panagopoulos, Pilar Cristancho, J. Miller, Charles Zorumski, Charles Conway*

Washington University School of Medicine, Saint Louis, Missouri

Background: Treatment-resistant depression (TRD) is a severe form of major depressive disorder (MDD) associated with lower quality of life and affects one in three patients with MDD, of an estimated 16 million US adults. TRD patients often fail to respond to multiple (>4) treatments with standard of care antidepressants, augmentation strategies, and more aggressive treatment options (e.g., electroconvulsive therapy). Therefore, new therapeutic treatment options are needed. Recent evidence demonstrates that a sub-anesthetic dose of ketamine, an established dissociative anesthetic agent, may provide rapid and sustained antidepressant effects in TRD patients. Ketamine’s antidepressant mechanism of action is not fully understood; however, it is believed to act primarily via NMDA-receptor antagonism. Nitrous oxide, a well-tolerated inhalational general anesthetic informally known as “laughing gas”, also functions as an NMDA receptor antagonist and may also prove beneficial in this difficult to treat population. Nitrous oxide may offer similar antidepressant benefits as ketamine without some of the potential adverse side effects (e.g., psychomimetic symptoms and cognitive effects).

Methods: The goal of this pilot, proof of principle, blinded, randomized placebo-controlled crossover trial was to test the hypothesis that nitrous oxide is effective as a potential rapidly acting treatment for TRD treatment. Twenty TRD patients in a University tertiary care setting received a titrated, one-hour inhalation of either up to 50% nitrous oxide/50% oxygen mix or 50% nitrogen/50% oxygen, in random order and one week apart. All participants received both treatments. The primary endpoint was change on the 21-point Hamilton Rating Scale for Depression (HRSD) within 24 hours after treatment. Secondary endpoints included change on the Quick Inventory of Depressive Symptoms Self Report (QIDS-SR). Nitrous oxide-induced inactivation of vitamin B12 was determined by plasma total homocysteine before and after treatment.

Results: Mean duration of nitrous oxide treatment was 55.6±2.5 (SD) minutes at an inspiratory nitrous oxide concentration of 44% (37 – 45%, median, IQR). Depressive symptoms improved significantly at 24 hours after receiving nitrous oxide compared to placebo (mean difference in HRSD-21 score -5.5, 95% CI -2.5 to -8.5 points, p=0.002). Four patients (20%) had treatment response (≥50% reduction of depressive symptoms on the HRSD) with nitrous oxide compared to one patient (5%) after placebo (odds ratio [OR] 4.0, 95% CI 0.45 – 35.79). Three patients (15%) experienced full remission (HRSD ≤ 7 points) after nitrous oxide treatment, compared to none after placebo (OR 3.0, 95% CI 0.31 – 28.8). In four patients the nitrous oxide treatment had to be briefly discontinued due to adverse events (compared to none during placebo). All adverse events were temporary and resolved shortly after completion of nitrous oxide administration. No serious adverse events occurred.

Conclusions: This proof of concept trial demonstrated that nitrous oxide may have rapid and marked antidepressant effects in patients with treatment-resistant depression. Subsequent studies will be required to determine optimal antidepressant dosing strategies as well as the risk/benefit ratio of nitrous oxide in a larger and more diverse population of patients with treatment-resistant major depression.

Keywords: Treatment Resistant Depression, Depression, Mood Disorders, Nitrous Oxide.

Disclosure: Nothing to Disclose.

T143. Efficacy of Vortioxetine on Cognitive Function in Patients with Major Depressive Disorder: Cognitive Test Performance Results: from a Randomized, Double-blind, Duloxetine-referenced, Placebo-controlled

Richard Keefe*, Atul Mahableshwarkar, John Zajecka, William Jacobson, Yinzhong Chen

Duke University Medical Center, Durham, North Carolina

Background: Vortioxetine is an antidepressant with multimodal activity that combines two pharmacological modes of action: direct modulation of receptor activity and inhibition of the serotonin transporter. Vortioxetine is approved for the treatment of major depressive disorder (MDD) and has demonstrated clinical efficacy in improving symptoms of depression. In addition, evidence from a recently published double-blind, placebo-controlled study demonstrated the benefits of vortioxetine on cognitive functioning in adults with MDD.1 The primary objective of the current study (NCT01564862) was to evaluate the efficacy of flexible-dose vortioxetine (10-20 mg) on cognitive function after 8 weeks of treatment in adults with MDD.2 An active reference (duloxetine 60mg) was included to demonstrate assay sensitivity.

Methods: Adults with moderate to severe MDD (18-65 yrs, MADRS>26) and self-reported cognitive dysfunction (ie, difficulty concentrating, slow thinking, and difficulty in learning new things or remembering things) were enrolled. The primary endpoint was change from baseline to Week 8 on the Digit Symbol Substitution Test (DSST). Secondary cognitive endpoints included a battery of neuropsychological tests to evaluate overall efficacy on cognitive function, including change from baseline to Week 8 on Trail Making Test A (speed of processing), Trail Making Test B (executive functioning), Stroop Test (executive functioning), Groton Maze Learning Test (visual learning and memory), Detection Task (psychomotor function), Identification Task (attention), and One-Back Task (attention, working memory). Change from Baseline to Week 8 in the MADRS total score was included to evaluate effect on depressive symptoms and a pre-specified path analysis assessed the proportion of cognitive improvement attributed to a direct treatment effect versus the proportion attributable to improvement in mood. All endpoints were analyzed using an analysis of covariance (ANCOVA) using the full analysis set. Adverse events were assessed throughout the study.

Results: A total of 602 patients were randomized (vortioxetine, n=198; placebo, n=194; duloxetine, n=210). The change from baseline to week 8 in the DSST was statistically significantly greater for vortioxetine (+4.60; p=0.019) compared with placebo (+2.85), with a standardized effect size of 0.254. Trail Making Test B (Total Time) (-9.67 sec, p<0.001) also improved with vortioxetine compared to placebo. None of the other secondary endpoints of cognitive function improved significantly with vortioxetine. Vortioxetine significantly improved depressive symptoms compared with placebo, as measured by change from baseline to Week 8 in MADRS (p<0.05). Path analysis indicated that the beneficial effect of vortioxetine on cognitive function was primarily a direct treatment effect rather than due to alleviation of depressive symptoms. Duloxetine was not significantly different from placebo on the DSST (+4.06, p=0.099) or secondary performance-based cognitive assessments, but was significant on the MADRS (p<0.001). The most common adverse events associated with vortioxetine treatment (>5%) were nausea, headache, and diarrhea.

Conclusions: MDD is often associated with cognitive dysfunction. The Diagnostic and Statistical Manual 5 lists impairment in cognition (diminished ability to think or concentrate or indecisiveness) as a criterion for the diagnosis of a major depressive episode. In addition to subjective complaints, cognitive symptoms are often seen on tests measuring executive function, processing speed, attention, learning and memory. In this study of adults with MDD and self-reported cognitive dysfunction, treatment with vortioxetine 10-20 mg/day significantly improved objectively assessed cognitive functioning as well as depressive symptoms. Path analysis suggests this was primarily a direct effect of treatment rather than a non-specific effect due to improvement in mood. These data add to the evidence from previous studies showing that vortioxetine significantly improves cognitive function in adults with MDD.

Keywords: Cognition, Depression, Functioning.

Disclosure: Received investigator-initiated research funding support from the Department of Veteran’s Affair, Feinstein Institute for Medical Research, GlaxoSmithKline, National Institute of Mental Health, Novartis, Psychogenics, Research Foundation for Mental Hygiene, Inc., Allon, AstraZeneca, and the Singapore National Medical Research Council. Received honoraria, served as a consultant, or advisory board member for Abbvie, Akebia, Amgen, Astellas, Asubio, AviNeuro/ChemRar, BiolineRx, Biogen Idec, Biomarin, BMS, Boehringer-Ingelheim, BrainCells, CHDI, Eli Lilly, FORUM, GW Pharmaceuticals, Helicon, Lundbeck, Memory Pharmaceuticals, Merck, Minerva Neurosciences, Inc., NeuroSearch, Mitsubishi, Novartis, Orion, Otsuka, Pfizer, Roche, Sanofi/Aventis, Shire, Solvay, Sunovion, Takeda, Targacept, and Wyeth. Receives royalties from the BACS testing battery, the MATRICS Battery (BACS Symbol Coding) and the Virtual Reality Functional Capacity Assessment Tool (VRFCAT). He is also a shareholder in NeuroCog Trials, Inc. and Sengenix. Source of Funding: This study was funded by H Lundbeck A/S and the Takeda Pharmaceutical Company, Ltd.

T144. Symptomatic and Functional Remission and Recovery in Lurasidone-treated Patients with Bipolar Depression: Post-hoc Analysis of a 6-week, Placebo-Controlled Trial Followed by a 6-Month Extension

Antony Loebel*, Cynthia Siu, Krithika Rajagopalan, Andrei Pikalov, Josephine Cucchiaro, Terence Ketter

Sunovion Pharmaceuticals, Fort Lee, New Jersey

Background: Symptomatic and functional recovery in bipolar depression implies sustained remission of clinical symptoms and substantial functional improvement. The objective of this post-hoc analysis was to evaluate symptomatic and functional remission and recovery in patients with bipolar depression treated with lurasidone.

Methods: Outpatients meeting DSM-IV-TR criteria for bipolar I depression, with or without rapid cycling, were randomized to 6 weeks of once-daily, double-blind treatment with either lurasidone 20-60 mg (LUR20-60), lurasidone 80-120 mg (LUR80-120) or placebo (PBO). A total of 318 intent-to-treat subjects enrolled in a 6-month, open-label, continuation study of lurasidone. Subjects initially treated with placebo were started at extension baseline with flexible once- daily doses of lurasidone 40-160 mg/d (PBO-LUR; N=107). Symptomatic and functional recovery was defined as meeting criteria for both symptomatic remission (Montgomery-Asberg Depression Rating Scale [MADRS] total score<12) and functional remission (Sheehan Disability Scale {SDS} mean score<3 and all SDS domain scores<3 representing no more than mild impairment) for at least 3 months (i.e., at weeks 19 and 32).

Results: Demographic and clinical characteristics were comparable among the treatment groups at core baseline. At week 6 (end of the 6-week core phase), a significantly higher proportion of subjects met both symptomatic and functional remission criteria in the lurasidone group (33%, N=273 pooling the LUR20-60 and LUR80-120 groups) compared to the placebo group (15%, N=143, p<0.05, NNT=6). The proportion of subjects attaining symptomatic remission at week 6 was also significantly higher in the lurasidone group (41%) compared to the placebo group (25%, p<0.01, NNT=7). Likewise, the proportion of subjects attaining functional remission at week 6 was significantly higher in the lurasidone group (48%) compared to the placebo group (31%, p<0.01, NNT=6). In the 6-month continuation study, the proportion of subjects achieving symptomatic and functional recovery, that is, symptomatic and functional remission at both week 19 (month 3 of the study) and week 32 (month 6 of the study) was 61% (85/140) and 45% (31/69), in subjects who continued treatment on lurasidone (LUR-LUR) and subjects who switched from placebo to lurasidone (PBO-to-LUR), respectively. Symptomatic recovery rates (i.e., symptomatic remission sustained for at least 3 months) were 73% (112/153) for LUR-LUR subjects and 67% (50/75) for PBO-LUR subjects at continuation study endpoint. Functional recovery rates (i.e., functional remission sustained for at least 3 months) were 65% (95/147) for LUR-LUR subjects and 57% (42/74) for PBO-LUR subjects at continuation study endpoint. Multivariate logistic modeling revealed that statistically significant predictors of symptomatic and functional recovery included: lower baseline symptom (CGI-BP overall) severity, non-white race, and taking lurasidone (rather than placebo) during the core phase.

Conclusions: Our findings, derived from a combined 6-week acute and 6-month continuation study period, support the potential for attainment of symptomatic and functional remission and recovery in patients with bipolar depression treated with lurasidone. Lurasidone was generally well-tolerated with minimal effects on weight and metabolic parameters over the treatment period.

Keywords: bipolar depression, recovery, remission, lurasidone.

Disclosure: A. Loebel, K. Rajagopalan, A. Pikalov and J. Cucchiaro are employees of Sunovion Pharmaceuticals. C. Siu is a consultant for Sunovion Pharmaceuticals Inc. and Takeda Pharmaceuticals International Inc. Dr. Ketter has received grant/research support from AstraZeneca Pharmaceuticals LP, Cephalon Inc., Eli Lilly and Company, Inc., and Pfizer Inc.; consulting fees from Astellas Pharmaceuticals, Bristol-Myers Squibb Company, Cephalon Inc., Johnson & Johnson, Merck & Co., Inc., and Sunovion Pharmaceuticals, Inc; and CME lecture honoraria (not speaker’s bureau honoraria) from Abbott Pharmaceuticals, AstraZeneca Pharmaceuticals LP, and GlaxoSmithKline. Dr. Ketter’s spouse (Nzeera Ketter, MD) is an employee of and has stock in Johnson & Johnson.

T145. Neurocognitive Effects of Ketamine in Individuals with Treatment-resistant Depression: A Randomized Controlled Trial

James Murrough*, Katherine Burdick, Andrew Perez, Jess Brallier, Lee Chang, Alexander Foulkes, Dennis Charney, Sanjay Mathew, Dan Iosifescu

Icahn School of Medicine at Mount Sinai, New York, New York

Background: The glutamate N-methyl-d-aspartate (NMDA) receptor antagonist ketamine displays rapid antidepressant effects in patients with treatment-resistant depression (TRD); however, the potential for adverse neurocognitive effects in this population has not received adequate study. We previously found in a small open-label study that slower baseline processing speed was associated with improved symptom reduction following ketamine in patients with TRD. The current study was designed to investigate the short-term neurocognitive impact of ketamine in TRD and examine baseline response predictors in the context of a randomized controlled trial.

Methods: In the current study, 62 patients (mean age=46.2±12.2) with TRD free of concomitant antidepressant medication underwent neurocognitive assessments with the MATRICS Consensus Cognitive Battery (MCCB) before and after a single intravenous infusion of ketamine (0.5 mg/kg) or midazolam (0.045 mg/kg). The randomization to ketamine:midazolam was 2:1 and participants received a follow-up assessment at 7-days post-treatment.

Results: We found no main effect of treatment condition or clinical response on cognitive performance in any domain and no significant interaction effects. Examining baseline neurocognitive predictors of response, processing speed was the single best predictor variable in a best-fit model [beta=0.43±0.19, t=2.3, p=0.027].

Conclusions: In this study ketamine was devoid of short-term adverse neurocognitive effects. Replicating our previous finding in an independent sample, slower baseline processing speed predicted a rapid antidepressant response at 24 hours following ketamine. Future studies are required to further define the neurocognitive profile of ketamine in clinical samples and to identify clinically useful response moderators.

Keywords: depression, ketamine, neurocognition, antidepressant.

Disclosure: In the past 2 years, Dr. Murrough has served on advisory boards for Janssen Research and Development and Genentech and has provided consultation services for ProPhase, LLC and Impel Neuropharma. Dr. Burdick has served on an advisory board for Dainippon Sumitomo Pharma and has participated in a continuing medical education activity sponsored by Takeda and Sunovion Pharmaceuticals. In the last three years, Dr. Iosifescu has consulted for Avanir, CNS Response, Lundbeck, Otsuka, Servier, Sunovion and he has received grant/research support through Mount Sinai School of Medicine from Astra Zeneca, Brainsway, Euthymics Bioscience Inc, Neosync, Roche and Shire. Dr. Dennis Charney (Dean of Icahn School of Medicine at Mount Sinai), and Icahn School of Medicine at Mount Sinai have been named on a use patent on ketamine for the treatment of depression. The Icahn School of Medicine has entered into a licensing agreement for the use of ketamine as therapy for treatment-resistant depression. Dr. Charney and Icahn School of Medicine at Mount Sinai could potentially benefit if ketamine were to gain approve for the treatment of depression. In the past 12 months, Dr. Mathew has received consulting fees from Bristol-Myers Squibb, Genentech, and Naurex, and research support from AstraZeneca and Janssen Pharmaceuticals. All other authors declare no conflict of interest.

T146. A Circadian Rhythm Disorder in PTSD Affects Plasma Levels of Specific Monocyte Chemokines

Clifton L Dalgard, Ofer Eidelman, Catherine Jozwik, Meera Srivastava, Roopa Biswas, Yvonne Eudy, Stephen W. Rothwell, Gregory P. Mueller, Peixiong Yuan, Wayne Drevets, Husseini K. Manji, Meena Vythlingam, Dennis S. Charney, Robert J. Ursano, David M. Jacobowitz, Harvey B. Pollard*, Omer Bonne

Uniformed Services University of the Health Sciences, Bethesda, Maryland

Background: Post-traumatic stress disorder (PTSD) is a psychiatric disease that can occur following exposure to traumatic events. In addition to many classical behavioral symptoms, a sleep disorder is also typically found in these patients. It is possible that proinflammatory cytokines (activators of immune cell division) or chemokines (attractors of immune cells) in the CSF or plasma might be biomarkers, surrogates, or drivers for the psychophysiological mechanisms connecting a history of trauma exposure to changes in behavior and medical morbidity. Based on the sleep disorder characteristics of PTSD, we hypothesized that such a dysfunctional circadian rhythm might be reflected by time-dependent concentrations, or time-dependent patterns of expression, of cytokines or chemokines in plasma from PTSD patients.

Methods: Plasma samples from 16 chronic civilian PTSD (11 females and 5 males) and 14 Healthy Controls (9 females and 5 males) were collected hourly, over a 27 hour period, in a controlled inpatient setting at NIMH/Bethesda. The patients, age- and sex-matched, were largely from the Bonne et al (2011) cohort, who were relatively free of major Depressive Disorder. We measured 17 different cytokines and chemokines in each sample using an industry-standard MesoScale multiplexed electrochemiluminescent ELISA platform. All time data were normalized to the local “zeitgeber”, the astrophysical time of sunrise in Bethesda, Maryland, on the day of data collection. All patients lived on the east coast of the United States.

Results: We found that significant discrimination between PTSD and Healthy Controls could be made over the entire circadian time period, on the basis of absolute levels of the chemokines MCP-1, MCP-4, and of the MCP-4/MCP-1 ratio (P=0.001). MCP-1 and MCP-4 are monocyte chemoattractant proteins. Receiver Operating Condition (ROC) calculations for the MCP-4/MCP-1 ratio show that the best area-under the curve (AUC) values (ca. 0.88) were found at ca. 8 hours after sunrise. The circadian pattern of chemokine expression in PTSD was also shifted relatiuve to healthy controls, by ca.+6 hours for both MCP-4 and the calculated MCP-4/MCP-1 ratio, and by ca.+4 hours for TARC (a chemokine for lymphocytes, secreted by monocytes). Gender-dependence is modest. In addition, several other chemokines, including IP10 (interferon gamma-dependent protein-10) and Eotaxin (chemokine for eosinophils) were virtually perfectly entrained when comparing PTSD and Healthy Controls.

Conclusions: The data support the hypothesis that a dysfunctional circadian rhythm in PTSD is associated with aberrations in absolute plasma concentrations of MCP-4, and of the calculated MCP-4/MCP-1 ratio. The circadian patterns of expression for the MCP-4/MCP-1 ratio, and for TARC, are also disordered in PTSD patients. It remains to be discovered whether these disease-specific differences in circadian expression for these specific immune signaling molecules are biomarkers, surrogates, or drivers for PTSD. Reference: Bonne O., et al. J.Clin.Psych. 72:1124-1128, 2011.

Keywords: PTSD, circadian rhythm, biomarker, chemokine.

Disclosure: Nothing to Disclose.

T147. Testing Sensitivity of Different Criteria for Complicated Grief

M. Katherine Shear*, Christine Mauro, Yuanjia Wang, Natalia Skritskaya, Charles Reynolds, Naomi SImon, Sidney Zisook, Barry Lebowitz, First Michael

Columbia University, New York, New York

Background: Complicated grief is a syndrome of prolonged acute grief shown to occur in bereaved populations worldwide. Symptoms can be distinguished from normal grief as well as other psychiatric conditions and CG shows a unique pattern of response to treatment. The DSM5 workgroup on Trauma and Stress Disorders supported its inclusion as a new diagnosis. However, lack of consensus about the name, the time course or specific criteria used for diagnosis resulted in its placement in section 3 of DSM-5 with a new name (Persistent Complex Bereavement Disorder PCBD) and provisional criteria that merged two proposed evidence-based criteria sets: Prolonged Grief Disorder (PGD) derived from community samples, and Complicated Grief (CG) from clinical samples. None of the proposed criteria sets have been tested in patients in whom an expert diagnosis of CG has been made. Such a procedure can help establish clinical usefulness of diagnostic criteria. We present here results of a study to test the sensitivity of PCBD, PGD and CG criteria proposals in a sample of bereaved individuals confirmed by expert clinicians to be suffering from complicated grief.

Methods: Study participants were 178 individuals assessed at one of 4 university-based psychiatric research clinics as part of an NIMH-funded treatment study of CG. Participants scored>30 on the Inventory of Complicated Grief and were confirmed by study PI’s or their delegate to have CG on clinical interview. Exclusion criteria included current substance use disorder, lifetime psychotic disorder, Bipolar I Disorder, current cognitive impairment and active suicidality. All participants in the sensitivity analysis had significant functional impairment from grief symptoms and all were bereaved at least 12 months. All participants underwent a Structured Clinical Interview for Complicated Grief (SCI-CG; Bui et al., 2014-poster), a semi-structured SCID-like instrument with items from each of the 3 proposed criteria sets scored as 1=“Absent”, 2=“Unsure or Equivocal”, 3=“Present”. Cronbach’s alpha=.76 test-retest ICC=.68 and inter-rater reliability ICC=.98 were adequate. To assess sensitivity items from a given criteria set were matched with a corresponding item from the SCI-CG and rated as present (score – 3) or absent (score=1 or 2). In the case of two or more matching SCI-CG items for a particular symptom, the symptom was considered present if any matching SCI-CG item was rated as 3. Each classification algorithm was used to determine rate of diagnosis. Sensitivity was computed as number of people diagnosed divided by number in the sample (n=178). All statistical analyses were performed using SAS 9.2.

Results: Sensitivity of the PGD criteria was 56.2% (95% CI: 48.9% – 63.5%). Yearning, a necessary symptom for this criteria set (Domain B) was endorsed by 87.6%. Additionally, 64.0% of the sample met the required five or more symptoms from the symptom list called Domain C. Individual symptom endorsements ranged from 41.0% to 76.4% within this sample. Sensitivity of the proposed CG criteria set was 99.4% (95% CI: 98.3% – 100.0%). All but one individual had at least one symptom from a list called Domain B (separation distress) and all individuals had at least two symptoms from another list called Domain C. Individual separation distress symptoms endorsement ranged in frequency from 24.7% to 87.6% and endorsement of symptoms from Domain C ranged from 24.2% to 91.6%. Sensitivity of PCBD criteria for the CG sample was 67.4% (95% CI: 60.5% – 74.3%). 94.4% of the CG sample had at least one symptom from a list labeled separation distress (Domain B), while only 70.8% had at least 6 symptoms from a second list called Domain C. Low sensitivity of this criteria set is possibly due to the threshold for Domain C being too high. Separation distress (Domain B) symptoms ranged in frequency from 65.2% to 87.6%. Endorsement of individual symptoms from Domain C ranged in frequency from 17.4% to 76.4%.

Conclusions: Complicated grief is a debilitating condition affecting an estimated 3-5% of the population. There is substantial evidence that this problem can be reliably identified and occurs following loss of a close relationship such as a child, partner, parent, other relative or close friend. CG is often chronic and unremitting without effective treatment. It is imperative that consensus be established for diagnostic criteria in order to facilitate effective treatment and foster needed research. One roadblock is lack of field testing of proposed criteria in clinical settings. Our study provides such data and indicates that both PGD and PCBD criteria have low sensitivity for established clinical cases. By contrast, the ICD 11 proposal for PGD does show adequate sensitivity. Work is underway to evaluate specificity of the three criteria sets. However, results of this sensitivity study are important to consider in building consensus for inclusion of diagnostic criteria in DSM 5.1.

Keywords: complicated grief, diagnostic criteria, DSM5, trauma and stress disorders.

Disclosure: Nothing to Disclose.

T148. Metformin Partially Reverses Olanzapine-induced Glucose Dysregulation: A Rodent Model

Margaret Hahn*, Celine Teo, Virginia Wilson, Araba Chintoh, Melanie Guenette, Zohra Ahsan, Adria Giacca, Gary Remington

Center for Addiction and Mental Health, Toronto, Canada

Background: Atypical antipsychotics (AAP) are linked to type 2 diabetes (DM2), even in absence of significant weight gain. In this regard, evidence from preclinical rodent models consistently supports pronounced direct effects on insulin sensitivity (peripheral and hepatic), following acute dosing of specific AAP agents. Similarly, several studies examining acute to sub-chronic dosing of AAP in healthy volunteers, show early changes in glucose metabolism prior to weight gain. Metformin, a first line treatment in the general population for prediabetes and DM2, has been studied as an adjunctive pharmacological intervention for AAP-related weight gain. The majority of these studies also demonstrate improvements in indirect measures of insulin sensitivity (i.e. HOMA-IR). To our knowledge there are no studies which specifically examined metformin in patients on AAPs with DM2. In this regard, the question remains whether antidiabetic drugs are efficacious to reverse direct effects of AAPs on glucose pathways, a distinguishing factor for schizophrenia populations, as compared non-mentally ill individuals.

Methods: We tested 2 metformin doses (therapeutic, or high) to prevent the previously established impairments seen following a single dose of olanzapine (OLA) (3mg/kg) in parameters of insulin sensitivity measured through gold-standard hyperinsulinemic euglycemic clamps (HIEC). Male Sprague-Dawley rats (300-325g) were treated with Metformin (Met) (400mg/kg; n=11 or 150mg/kg; n=13) or vehicle (Veh) (n=11) administered through gavage preceding an overnight (O/N) fast, followed by a second dose the morning prior to the HIEC. Comparisons between groups were made by analysis of variance (basal, hyperinsulinemic phases), with bonferroni corrections where applicable. Significance was accepted at p<0.05.

Results: In the basal period, there were no differences in insulin or glucose levels between groups. However, the high dose metformin (Met) group demonstrated significantly higher basal glucose production (GP), as compared to the low dose Met (p<0.001) or vehicle (p<0.001) groups. Treatment with the low Met dose (p=0.019), but not the high Met dose significantly improved whole body insulin sensitivity relative to vehicle, as reflected by the glucose infusion rate (GINF)(p=0.02). Controlling for elevated basal levels in hepatic glucose production (HGP), we found that only low dose Met significantly suppressed GP relative to vehicle (p<0.001). Examining percent change under insulin challenge relative to basal glucose disappearance, neither the low or high Met dose demonstrated an increase in peripheral glucose uptake relative to vehicle. To explore the unexpected increase in basal rates of glucose production in the high dose Met group, we measured basal lactate levels (prior to insulin infusion, and olanzapine treatment). We found a significant increase in the high dose Met group relative to vehicle or low dose Met treatment (p=0.003).

Conclusions: Pretreatment with the low (but not high) dose of Met, as compared to vehicle, was associated with a significantly higher total body insulin sensitivity following OLA. While Met should enhance both liver and peripheral insulin sensitivity, we show that in the context of acute OLA treatment, improvements are observed only in reduction of glucose production, with no effect on glucose utilization. A preexisting rodent study examining Met for prevention of OLA-induced glucose intolerance demonstrated only partial reversal of these perturbations (Boyda, 2013), which might be explained by our findings of a lack of Met effect on glucose disposal. This could suggest that patients who develop DM2 on AAPs require combination therapy; for example, employing a peripheral insulin sensitizer in addition to Met for optimal management. Contrary to our working hypothesis, the high dose of Met failed to improve insulin sensitivity. Lactate, a substrate for hepatic glucose production (HGP) was significantly increased in the high Met group (relative to the lower Met dose and vehicle), possibly explaining the lack of suppression of HGP with the supratherapeutic Met dose.

Keywords: Atypical antipsychtoics, diabetes, metformin, schizophrenia.

Disclosure: Nothing to Disclose.

T149. Preterm Birth: Risk Attributable to Maternal Depression and Antidepressant Pharmacotherapy

D. Jeffrey Newport*, Bettina T. Knight, Tamar L. Gur, Brett Worly, Zachary N. Stowe

University of Miami Miller School of Medicine, Miami, Florida

Background: Existing data has implicated both depression during pregnancy and antidepressant therapy during gestation as risk factors for preterm birth (PTB). However, there are scant data disentangling the effects of these frequently concomitant risk factors. The objective was to examine the association between PTB and both maternal mental illness and psychotropic exposure during pregnancy.

Methods: This is a case-cohort analysis of pregnant women participating in prospective, longitudinal observational studies of prenatal mental illness and its treatment. Included in the present analysis were women (n=841) who had delivered a live singleton with documented gestational age at delivery by the time of data sequestration, and for whom the following data had been collected: (1) prospective weekly documentation of medication exposure across gestation; (2) psychiatric diagnostic assessment and serial assessment of prenatal depressive symptoms. Severity of depression was operationalized as area under the curve (AUC) for the Hamilton Rating Scale for Depression (HRSD) within each trimester. Excluded were women with multifetal gestation or fetuses who had known congenital or chromosomal anomalies. Risk estimates were produced using multivariate logistic regression modeling. Medication and diagnosis specific post hoc confirmatory analyses of the risk estimates were conducted.

Results: Candidate predictors of PTB identified via preliminary bivariate analyses included 3rd trimester exposure to moderate[HRSD:16-20] or severe[HRSD≥21] depression, serotonin reuptake inhibitor (SRI) antidepressants, and zolpidem, in addition to other well-attested risk factors. Logistic regression demonstrated significant associations between PTB and 3rd trimester exposure to severe depression (OR=8.82 [95%: 2.62 –29.8]), moderate depression (OR=2.76 [1.15-6.64]), zolpidem (OR=3.31 [95%: 1.39-7.90]) and SRI (OR=2.31 [95%: 1.14-4.67]) therapy. Logistic regression also demonstrated significant associations between PTB and recognized risk factors including placental abruption, maternal infection, previous preterm delivery, and gestational diabetes.

Conclusions: These findings indicate PTB risk is associated with both moderate-to-severe depression and SRI exposure during late gestation; however, the magnitude of the effect of severe depression is more than three times greater than SRI exposure. These results reiterate that concerns regarding adverse effects of prenatal SRI therapy must be weighed against the risks attributable to worsening depression during pregnancy.

Keywords: depression, antidepressants, pregnancy, preterm birth.

Disclosure: Dr. Newport has received research support from Eli Lilly, GSK, Janssen, NIH, NARSAD and Wyeth, has served on speaker or advisory boards for Astra-Zeneca, Eli Lilly, GSK, Pfizer and Wyeth and has received honoraria from Astra-Zeneca, Eli Lilly, GSK, Pfizer, and Wyeth. Ms. Knight has received research support from NIH, NARSAD, Wyeth, BMS, Cyberonics, Eli Lilly, Forest, Janssen and Novartis. A family member is a GSK employee and holds GSK stock options. Dr. Stowe has received research support from NIH, GSK, Pfizer and Wyeth, has served on speaker or advisory boards for Pfizer, Eli Lilly, Wyeth, BMS, and GSK, and has received honoraria from Eli Lilly, GSK, Pfizer, and Wyeth.

T150. Metabolic Risk in Antipsychotic-treated Children During Behavioral Weight Loss Treatment

Ginger Nicol*, Michael Yingling, Vincent Huang, Julie Schweiger, John Newcomer

Washington University, Saint Louis, Missouri

Background: The US prevalence of cardiovascular risk factors like childhood-onset obesity and type 2 diabetes has increased in recent decades. Children with mental illness, especially those treated with antipsychotic medications, may have even higher risk for obesity and related cardiometabolic risk factors. Hepatic triglyceride content (HTGC) measured by 1H Magnetic Resonance Spectroscopy (MRS) and adiposity measured by dual energy X-ray absorptiometry (DEXA) are well-validated measures of metabolic risk, with limited study in the high-risk population of children with psychiatric disorders. Psychosocial stress can increase risk for obesity and related comorbidities, but have not been evaluated in context of gold standard biomarkers. This pilot study tested the relationship between adiposity and HTGC during a 16-week behavioral weight loss intervention. The relationship between biomarker outcomes and baseline measures of psychiatric symptom severity was also explored.

Methods: Participants were overweight (BMI%ile 85-94) or obese (BMI%ile>95) antipsychotic-treated youth and healthy controls ages 6-19 participating in an ongoing pilot study of a family based behavioral weight loss program. All participants underwent 16 weeks of a family-based behavioral weight loss intervention based on the Traffic Light Diet. Primary outcome measures included HTGC, measured with 1H MRS of the liver and body composition was measured via Dual Energy X-Ray Absorptiometry (DEXA), measured at baseline and 16 weeks. Psychiatric symptoms were assessed at baseline in all participants with the Child Behavior Checklist (CBCL) and the Aberrant Behavior Checklist (ABC). Spearman correlations were performed on endpoint measures of DEXA fat and HTGC; endpoint DEXA and HTGC were correlated with baseline measures of psychiatric symptoms and psychosocial stress.

Results: A total of 26 children, mean age 13.06 (SD 2.3) years participated in the study (antipsychotic treated n=18, healthy controls n=8). Groups were frequency-matched on baseline age and BMI%ile. Mean baseline BMI%ile in the pooled group was 96.7 (2.78). Following 16 weeks of weight loss treatment, DEXA total % fat and total fat were both significantly positively correlated with HTGC (Spearman’s rho [ρ]=0.61, p=0.02 and ρ=0.63, p=0.01, respectively). Baseline total psychiatric symptoms measured by the CBCL were significantly correlated with adverse change in DEXA total % fat during treatment (ρ=0.52, p=0.03) with a trend level correlation with adverse change in DEXA total fat (ρ=0.46, p=0.07). ABC irritability subscale severity was correlated with adverse change in DEXA total fat (ρ=0.54, p=0.02) with a trend correlation between irritability and adverse change in DEXA total % fat (ρ=0.47, p=0.06). Hyperactivity was correlated with adverse change in DEXA total fat (ρ=0.55, p=0.02) again with a trend level relationship with adverse change in DEXA % fat (ρ=0.46, p=0.07).

Conclusions: To our knowledge this is the first study to examine sensitive indicators of metabolic risk such as DEXA fat and HTGC in overweight or obese youth with and without antipsychotic treatment during a behavioral weight loss intervention. The results of this study indicate that the DEXA-measured adiposity at the end of the intervention is significantly associated with measured HTGC at the end of the intervention. Additionally, baseline psychiatric symptom severity measured by the CBCL total problems score and by the ABC irritability and hyperactivity subscales was associated with worse outcomes with respect to change in adiposity during the weight loss intervention. These results indicate that lower adiposity levels after the weight loss intervention are associated with lower hepatic fat content, an important risk factor for diabetes. The results suggest the need for further study of whether and how higher baseline psychiatric symptom severity may complicate the successful implementation of behavioral weight loss treatment. Support for this research was made possible by Grant Number MH 092435 from the National Institute of Mental Health, Grant Number P30DK056341 f.

Keywords: Antipsychotic, Cardiometabolic Risk, Weight loss treatment.

Disclosure: Dr. Ginger Nicol has received research funding from NIMH, the Brain & Behavior Research Foundation, and the Sidney R. Baer, Jr. Foundation. Dr. John Newcomer has received research funding from NIMH, and has been a member of data safety monitoring committees for Bristol-Myers Squibb, Merck, Vivus, Cleveland Clinic and Amgen, receives royalties from Jones & Bartlett Publishing for the development of a metabolic monitoring form, and has received honoraria for accredited CME events via American Physician Institute, CME Outfitters, CMEology, American Psychiatric Association and American Society for Clinical Pharmacology. Mr. Yingling, Mr. Huang and Ms. Schweiger have no financial disclosures to report.

T151. Acceptability of Treatments and Services for Individuals with Hoarding Behaviors

Carolyn Rodriguez*, Amanda Levinson, Sapana Patel, Kim Rottier, Jordana Zwerling, Susan Essock, Lee Shuer, Randy Frost, Blair Simpson

Columbia University/New York State Psychiatric Institute, Bronx, New York

Background: Hoarding disorder, characterized by difficulty parting with a large volume of possessions that results in distress and impairment in functioning, is a new diagnostic entity in DSM-5. Given the high prevalence (2-6%), increased public awareness, and impact on public health, treatments and services to help individuals with hoarding disorder are in high demand. Medication- and psychotherapy-based interventions have been shown to be partially effective, however, engagement and adherence to treatment remain a challenge. Better, more acceptable treatments for hoarding disorder are needed. The aim of this study was to assess acceptability of currently available treatments and services for individuals who self-report suffering from hoarding behaviors.

Methods: The authors designed a survey to elicit acceptability ratings of treatments and services available for hoarding disorder using a Likert scale from 0 (not at all acceptable) to 10 (completely acceptable). In addition, open-ended questions were used to illicit study participant’s concerns and opinions. The treatments and services presented to participants were: self-help based resources (self-help book, facilitated self-help support group, online support group) psychotherapy (individual and group cognitive behavioral therapy), medications (serotonin reuptake inhibitor, stimulants), and community services (cleaning and removal service, professional organizing service, case management service, court-appointed guardian). This survey was administered through an internet-based survey platform (surveymonkey.com) and n=265 individuals completed the survey between November 2013 and July 2014. Hoarding symptoms were evaluated by a validated self-report scale called the Saving Inventory-Revised (SI-R) and a brief survey based on the DSM-5 hoarding disorder criteria.

Results: The majority of participants (73%) had clinically meaningful hoarding symptoms by self-report (SI-R>40). Among those participants who had clinically meaningful symptoms (n=193), the four most acceptable options were individual cognitive behavioral therapy, professional organizing service, self-help book, and online support group. Moderately acceptable were group cognitive behavioral therapy, facilitated self-help support group, and case management services. The four least acceptable options were cleaning and removal services, serotonin reuptake inhibitors, stimulants, and court-appointed guardian. Respondents most often reported cost, time, and lack of access as barriers to treatment. Stigma (e.g. “I fear others would ridicule me”), ambivalence/motivation to change (e.g. “I still don’t really think hoarding is a problem”), and anticipated distress (e.g. “It would be extremely emotionally painful”) were commonly cited unacceptable aspects of seeking treatment.

Conclusions: This is the first study to assess the acceptability of currently available treatments and services in individuals who self-report clinically meaningful hoarding behaviors. In this sample, individuals rated psychotherapy, self-help based resources, and professional organizer services as most acceptable. Least acceptable were cleaning and removal services, medications, and court-appointed guardian. These findings can help guide the design of interventions that are both acceptable and effective to maximize participant engagement. Also important will be prospectively investigating what modifies acceptability and how acceptability affects treatment outcome.

Keywords: Hoarding Disorder, Treatment Acceptability, stimulants, serotonin reuptake inhibitors.

Disclosure: Nothing to Disclose.

T152. Clinical and Pharmacogenetic Outcomes of a Double-blind Antidepressant Treatment Study

Ma-Li Wong*, Chuanhui Dong, Deborah Flores, Monika Ehrhart-Bornstein, Stefan Bornstein, Mauricio Arcos-Burgos, Julio Licinio

South Australian Heath and Medical Research Institute, Adelaide, Australia

Background: Major depressive disorder is a serious public health issue worldwide, with a lifetime prevalence of 10% to 20% in the general population. Currently, the Hispanic population is the largest ethnic minority group in the U.S. representing over 37 million people. Within this group, almost 70% are Mexican-Americans. Three recent meta-analyses studies on pharmacogenetics of antidepressants in MDD have not being able to show genome-wide significant variations. The overall conclusion of these meta-analysis studies is that the results so far do not support any major effect of any single gene variations in the pharmacogenetics of antidepressants in MDD. We compare the effectiveness of fluoxetine and desipramine treatment in a prospective double-blind pharmacogenetics study in Mexican-Americans and examine the role of whole-exome functional gene variations in their antidepressant response.

Methods: Participants: 232 Mexican-Americans who met DSM-VI diagnostic criteria for major depressive disorder (MDD) were randomly assigned to an 8-week of double-blind desipramine (50-200 mg/day) or fluoxetine (10-40 mg/day) treatment after a one-week placebo lead-in period. Outcome measures included the Hamilton Depression (HAM-D), Hamilton Anxiety and Center for Epidemiological Rating Scales, and Beck Depression Inventory. Whole exome genotyping data were obtained for 36 remitters and 29 non-responders at week 8.

Results: Our analysis showed fluoxetine treatment produced greater HAM-D score reduction, higher response/remission rates, shorter time to response/remission, and lower incidences of anticholinogeric and cardiovascular side effect events when compare to desipramine treatment. Pharmacogenetics analysis showed that a variation in Chromosome 6 achieved exome-wide significance for treatment remission (P=1.98x10-06; FDR=0.05). This variant is located in a brain methylated DNA immunoprecipitation sequencing site suggesting that it might be involved in epigenetic regulation of neuronal gene expression.

Conclusions: Compared with desipramine, fluoxetine treatment showed a more rapid reduction of HAM-D score and lower incidence of side effects in a population comprised primarily of first generation Spanish speaking only Mexican-American individuals with MDD. Our pharmacogenetics approach strongly implicates the role of functional variants in antidepressant treatment response. Further, independent studies are needed for replication and validation.

Keywords: major depression, antidepressants, pharmacogenetics, remission.

Disclosure: Nothing to Disclose.

T153. Effect Size OPRM1 A118G and Tobacco Smoking

Edward Domino*, Lisong Ni, Mika Hirasawa-Fujita

University of Michigan, Ann Arbor, Michigan

Background: The present study examines the effect size of single nucleotide polymorphism (SNP) of the mu opioid receptor OPRM1 A118G (*G) on striatal dopamine (DA) release and tobacco smoking parameters in *G American (18%) vs. Japanese (50%) smokers.

Methods: 1. Seventeen healthy American male tobacco smokers with no history of substance abuse were studied. Subjects were between the ages of 20 and 36 years old (mean 25.8±4.8) and smoked 15-40 cigarettes per day. After overnight tobacco abstinence, the subjects underwent two separate 90 minute PET scans before and after smoking. The volunteers came to the research facility about 7:30 AM and laid in the scanner with their head and body restrained for about 4 hours. Ten mL of venous blood was taken from each subject for OPRM1 genotyping. 2. Another group of 51 Americans and 55 Japanese of both sexes who met similar criteria were recruited to compare their smoking uptake parameters. The female and male smokers were of similar age, number of cigarettes smoked per day, calculated previous 24 hr dose of nicotine, 24 hr venous plasma cotinine, exhaled CO and percent carboxyhemoglobin (COHb).

Results: 1. Smoking average nicotine (avnic) compared to denicotinized (denic) cigarettes decreased [11C] raclopride binding in the striatum about 11.5% maximum. In a very small sample of 15 AA vs. 5 *G118 the latter after avnic and denic smoking had greater DA release than the former. This effect on striatal DA release suggested that *G carriers obtain a “bigger bang per smoke.” 2. When Americans and Japanese overnight tobacco abstinent smokers were compared irrespective of sex, the only statistically significant differences were lower exhaled CO and COHb in the Japanese. However, American and Japanese male smokers differed. The latter had slightly higher plasma nicotine, but lower cotinine levels, lower calculated previous 24 hr dose of nicotine, and lower exhaled CO and COHb. American and Japanese female smokers had similar tobacco uptake parameters.

Conclusions: Munafò et al., 2013 reported in a study of 598 tobacco smokers that there was a lack of association between the OPRM1 genotype and tobacco cessation using high versus low dose nicotine patches. In the present study, the *G carriers had greater DA release after avnic smoking, suggesting that *G smokers obtain greater tobacco reinforcement. However, only a very small group with a small effect size was studied. Even though the OPRM1 *G prevalence in Japanese is 2.7 times greater than that in Americans, the former had only a few different tobacco smoke uptake parameters. The present study supports the conclusion by Munafò et al. that the OPRM1 genotype has at most a modest effect on cigarette smoking.

Keywords: Effect Size, OPRM1 A118G, Tobacco Smoking, Genetics.

Disclosure: Nothing to Disclose.

T154. Impaired Fear Processes in Young People with Attention Deficit Hyperactivity Disorder Mediate Links between COMT Genotype and Aggression

Anita Thapar*, Kate Langley, Clare Northover, Kelly Main, Katya Rubia, Karen Schepman, Michael O'Donovan, Stephanie VanGoozen

Cardiff University, Cardiff, United Kingdom

Background: Aggression is a well-established adverse, potentially harmful outcome in individuals with ADHD. Genetic vulnerability is important as increased levels of aggression in ADHD index higher familial recurrence risk, higher heritability and a greater burden of common genetic risk variation. A functional, high enzyme activity variant in catechol-O-methyl transferase (Val158Met) has been observed to predict aggression/extreme antisocial behavior in individuals with ADHD across six independent studies and in a meta-analysis. COMT Val158Met has been well-researched in human and animal studies. The variant has pleiotropic effects on cognition and emotion, processes that are important in relation to aggression. Thus impaired functioning of either or both these processes represents a plausible risk mechanism that underlies the link between COMT Val158Met and aggression in those with ADHD. In a previous population cohort, we found emotion/social function assessed by questionnaire rather than executive function mediated the link with aggression in those with higher ADHD scores. We set out to more precisely delinate links in a lab-assessed patient sample.

Methods: We conducted lab-based cognitive and physiological assessments of executive functioning, empathy and fear conditioning in a sample of adolescent males with ADHD (n=194). Our aim was to test the hypothesis that emotional processes characterised at a behavioural (emotional empathy paradigm) and physiological (fear conditioning paradigm) level rather than executive function bridge the link between COMT and aggression in ADHD. Executive functioning (EF) was assessed using the Wisconsin Card Sorting Test (WCST) and Go-NoGo task. Measures of empathy (understanding and feeling others' emotions) during dynamic film clips and skin conductance responses (SCR) to aversive stimuli during fear conditioning were obtained. Aggression was defined categorically (DSM-IV Conduct Disorder) and dimensionally (aggression scores). Analyses were conducted using ANOVA, regression and mediation analyses.

Results: COMT Val158Met showed association with executive function, fear empathy and fear conditioning. Risk genotype carriers (Val/Val, Val/Met vs. Met/Met) showed more errors on the WCST and poorer inhibition on the Go/no-Go task, lower affective (feeling) not cognitive (understanding ) empathy for others' fear and reduced SCR to the conditioned aversive stimulus. Executive function did not predict aggression whereas reduced affective empathy for others' emotions (fear, sadness, happiness) and poorer fear conditioning did. Fear empathy and fear conditioning mediated links between COMTVal158Met and aggression.

Conclusions: The present study finds that in male adolescents with ADHD, fear-related processes, defined both behaviorally and physiologically, indirectly mediate links between COMT Val158Met genotype and clinically reported aggression. These findings suggest fear mechanisms as possible targets for psychological interventions to reduce aggressive outcomes in ADHD. Our findings also reveal the potential of hypothesis-driven approaches in relatively small, well-phenotyped populations for identifying neuropsychological mechanisms that mediate genetic risk effects on behavior and psychopathology.

Keywords: ADHD, fear, genetic, aggression.

Disclosure: Nothing to Disclose.

T155. On the Link between Oxytocin Signaling and Alcohol Reward: Possible Role of the CD38 rs3796863 Polymorphism in Alcohol-induced Brain Dopamine Release

Mary Lee*, Elisabeth Caparelli, Emily Oot, Melanie Schwandt, Colin Hodgkinson, David Goldman, Markus Heilig, Vijjay Ramchandani, Lorenzo Leggio

National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, Maryland

Background: CD38, a type II transmembrane protein, is involved in the regulation of social behaviors. One genetic polymorphism of CD38, rs3796863, has been shown Behavioral and neuroplastic changes occurring in the development of addiction parallel those that occur in social bonding. This has led to the hypothesis that drugs of abuse co-opt social attachment systems, shifting attachment to drugs of abuse at the expense of social affiliation. Oxytocin (OT) mediates pair bonding in animal models of social affiliation such as prairie voles, with both OT and dopamine (DA) signaling in the striatum necessary for pair bond formation. Alcohol enhances reward via DA-dependent mesolimbic reward pathways. Dopamine D2 receptor-OT receptor heteromers in the striatum facilitating D2 receptor affinity may be an underlying molecular mechanism for the OT-DA interaction that is necessary to promote motivated behaviors to modulate OT signaling. The G allele of rs3796863 results in lower levels of OT in the brain and plasma. We examined the effect of this polymorphism on DA release following intravenous (IV) alcohol administration.

Methods: We carried out a human positron emission tomography (PET) study in male social drinkers using displacement of the DA-D2 ligand [11C]-raclopride, an indirect measure of endogenous DA release (Ramchandani et al., 2011). Participants underwent two separate PET scan sessions while receiving IV infusions of alcohol or saline (placebo). In the main study, whose results have already been reported by Ramchandani et al. (2011), the role of a functional OPRM1 A118G polymorphism in the striatal DA response to alcohol was examined. In the present post-hoc study, we reanalyzed the PET results to examine the effect of the CD38 rs3796863 polymorphism, while controlling for OPRM1 genotype.

Results: The main analysis, across striatal regions of interest (anterior and posterior ventral striatum, caudate, and putamen) showed that participants homozygous for the G allele (which has been associated with lower OT levels) had a trend toward a greater DA response to alcohol than did participants with other rs3796863 genotypes. Single-region analyses indicated significantly higher DA release in the posterior ventral striatum in G allele homozygotes. Preclinical data regarding the effect of OT on alcohol and psychostimulant-induced behavioral effects will also be presented and discussed.

Conclusions: These results further highlight the role of the OT-DA interaction in reward processing areas, suggesting a potential role of oxytocin in driving reward processing in response to drugs of abuse.

Keywords: oxytocin, reward, raclopride, dopamine.

Disclosure: Nothing to Disclose.

T156. A Human-specific Isoform of AS3MT Regulated by a Human-unique Variation Explains Susceptibility to Psychiatric Illness

Ming Li*, Ran Tao, Andrew E. Jaffe, Fengyu Zhang, Danny Chen, Joel E. Kleinman, Thomas M. Hyde, Joo Heon Shin, Daniel R. Weinberger

Lieber Institute for Brain Development, Baltimore, Maryland

Background: Recent meta-analyses of genome-wide association studies for schizophrenia have reported strong associations of single nucleotide polymorphisms (SNPs) spanning 10q24.32, in an about 757.5 kb genomic region covering numerous genes. However, little is known regarding the molecular mechanisms of the risk associations.

Methods: Using RNA-Seq techniques (Hi Seq 2000, 80-120M 100 bp paired end reads) in the dorsolateral prefrontal cortex (DLPFC) of postmortem human brain tissues from healthy controls and psychiatric patients, followed by qPCR validations, and analyses of cognitive data, brain imaging, drug response as well as further functional studies.

Results: The schizophrenia risk SNPs are significantly and specially associated with AS3MT gene expression in healthy controls and patients. We further identified a previously undescribed human-specific AS3MT transcript, which lacks 102 amino acids compared with the canonical isoform. The risk allele was associated with higher isoform expression, which is consistent with the diagnostic association of increased expression in patients. We also discovered a human-unique VNTR in high LD with the GWAS-identified risk SNPs. The VNTR is likely a functional element in regulating AS3MT transcription as identified by both eQTL analyses and in vitro promoter luciferase assays. The VNTR and isoform also showed associations with DNA methylation, cognitive function, brain imaging and drug response in independent samples.

Conclusions: Our data show that the genetic risk variants in Chr.10q24.32 have potential cis-regulatory transcriptional effects on AS3MT, and more specifically on regulating splicing of a truncated transcript, which should be explored in future biological studies.

Keywords: schizophrenia, 10q24.32, AS3MT, RNA-seq.

Disclosure: Nothing to Disclose.

T157. Integrating Genetics and Epigenetics with Large-scale RNA-sequencing of Schizophrenia Brains

Panos Roussos*

Icahn School of Medicine at Mount Sinai, New York, New York

Background: The most recent Psychiatric Genomic Consortium (PGC2) schizophrenia GWAS reported>100 linkage disequilibrium independent associated loci, implying a high degree of polygenicity. However, because the majority of SNPs reside within non-coding regions of genes or in intergenic regions, it has been difficult to determine the causal genetic variants, and there is limited knowledge about the regulatory mechanisms by which they act. To better understand the pathology of neuropsychiatric disease, we formed the CommonMind Consortium (commonmind.org) to generate large-scale data (RNA-seq, ChIP-seq, DNA-seq/genotyping) from human post-mortem brain samples.

Methods: Here, we identify functional changes in gene expression and expression quantitative trait loci (eQTL) using RNA-seq (Illumina HiSeq2000 – paired end reads) of 540 samples (259 schizophrenia cases and 281 controls) from the dorsolateral prefrontal cortex. Genotypes were assayed on the Illumina Infinium HumanOmniExpressExome8 chip and were imputed to 1000 Genomes. Data were normalized via voom using clinical (gender and age of death) and technical (brain bank, post-mortem interval, RNA quality, sequencing batch) covariates. A linear model was applied to detect eQTLs, adjusting for genetic structure. A variety of publicly available, brain specific epigenomic annotations for promoters, enhancers or open chromatin was used.

Results: Preliminary differential expression analysis using linear models implemented in voom/limma identified 3% of all expressed genes as differentially expressed between cases and controls (FDR 5%). Preliminary eQTL analysis of the assayed genotypes identified 795,507 proximal eQTLs (distance<1Mb) at FDR 5% in controls and cases with schizophrenia, representing 48.1% of expressed genes. PGC2 SNPs were enriched for eQTLs [average odds ratio (OR): 11.7] and epigenomic annotations [ORpromoter: 3.1; ORenhancer: 3.3; ORopen-chromatin: 2.9]. Combined analysis of eQTL and epigenomic annotations showed a further increase in the PGC2 SNPs enrichment [ORpromoter: 12.4; ORenhancer: 18.0; ORopen-chromatin: 13.9], indicating that risk SNPs affect gene expression through allele-specific alterations in non-coding, cis-regulatory regions.

Conclusions: This large dataset will be made public in early 2015 and will include a catalogue of brain-expressed genes and isoforms, as well as eQTLs, from cases and controls. This resource will facilitate novel discoveries relating neurobiology to disease risk and advance therapies.

Keywords: gene expression, eQTL, GWAS.

Disclosure: Nothing to Disclose.

T158. Variations in the FRA10AC1 Fragile Site Are Associated with Cerebrospinal Fluid Abeta Level

Qingqin Li*, Antonio Parrado, Mahesh Samtani, Vaibhav Narayan

Johnson & Johnson Pharmaceutical, Titusville, New Jersey

Background: Cerebrospinal fluid (CSF) biomarkers are important predictors for disease pathophysiology in Alzheimer’s Disease (AD). Elevated CSF pTau181 levels are correlated with neuronal loss and predict cognitive decline and conversion to AD in subjects with mild cognitive impairment. CSF tau levels may also be a useful marker to identify genetic variants implicated not only with risk for Alzheimer’s disease but also age at onset or rate of progression.

Methods: We performed GWAS meta-analysis to identify genetic variants predictive of the baseline CSF biomaker level (Aß1-42, Tau, pTau181, Tau:Aß1-42 ratio, and pTau181:Aß1-42 ratio) in patients enrolled in the ADNI study (Cohort 1 : N=340 genotyped using the Illumina 610 platform; Cohort 2 : N=344 genotyped using Illumina Omni2.5M platform). To directly compare variants via meta-analysis across the two Illumina platforms, we imputed genotypes, based on the reference haplotypes from the1,000 Genomes Project. Both, the directly genotyped markers and imputed dosage genotypes were assessed for association in a linear regression model, correcting for APOE ɛ4 allele dosage, age, baseline clinical diagnosis group (CN, EMCI, LMCI, or AD), and principal components to account for population substructure. A conventional p-value threshold of 5x10-8 was used to declare genome-wide significance. We further tested if the most significant association signal was associated with disease susceptibility (patients with EMCI, LMCI, or AD clinical diagnosis compared to CN) and Aß positivity in both the APOE ß4 non-carriers and in all genotype groups, respectively. Patients are defined as Aß positive using either PET imaging alone (AV45>1.1 or PIB>1.5 at any time point) or a combination of PET imaging data and CSF Aßlevel and whether this region was predictive of disease progression in subjects with mild cognitive impairment. Concurrently, GWAS analyses were performed for these other endpoints to discover novel variants. Lastly, we closely examined the top marker from baseline CSF biomarker meta-analysis and several candidate variants previously reported to be associated with disease progression using graphical exploration in the LMCI and AD cohorts, respectively.

Results: The most significantly associated markers predictive of baseline CSF Aß1-42 level in the meta-analysis (directly genotyped SNV rs10509663 PRE=5.1 x 10-9, PFE=5.1 x 10-9; imputed SNV PFE=8.6 x 10-10, PRE=8.6 x 10-10 obtained the genome-wide significance threshold of P<5x10-8) are located in a 60kb interval in chromosome 10, that includes variants in the intron and variants in the putative promoter region of rare FRA10A folate-sensitive fragile site FRA10AC1 (fragile site, folic acid type, rare, fra(10)(q23.3) or fra(10)(q24.2) candidate 1). The protein encoded by this gene is a nuclear phosphoprotein of unknown function. The 5' UTR of this gene is part of a CpG island and contains a tandem CGG repeat region that normally consists of 8-14 repeats but can expand to over 200 repeats. The expanded allele becomes hypermethylated and is not transcribed; however, an expanded repeat region has not been associated with any disease phenotype. No other variants passed the genome-wide significance for the other baseline CSF biomarker measurements. We observed a nominal association (p0.03-0.05) between the intronic variant rs10509663 and Aß positivity, interestingly the association is a bit stronger (p0.001-0.005) in the Aß positivity analysis among the ß4 non-carrier subgroup. We did not observe a significant association (p<0.05) between rs10509663 and disease progression in the MCI cohort (both EMCI and LMCI but correcting for MCI stage), or among the Aß positive subset of MCI subjects. Graphical exploration of rs10509663 in late MCI and AD subgroups also showed no trend of separation by genotype status.

Conclusions: We identified a candidate genetic marker, within FRA10AC1, associated with baseline CSF Aß1-42 level that achieved genome-wide significance in our meta-analysis combining non-overlapping subjects from the ADNI study (Cohort 1 and Cohort 2 were genotyped using different platforms, thus analyzed separately). Future studies with larger sample sizes or replication samples are needed to further dissect the genetic basis of CSF biomarkers. The disease progression is a challenging problem as the patients are likely to be in a different stage of disease continuum and the deterioration is not linear over the course of disease progression.

Keywords: Amyloid, CSF, genetic, ADNI.

Disclosure: Qingqin Li, Antonio Parrado, Mahesh N. Samtani, and Vaibhav Narayan are employees of Janssen Research and Development, LLC and own Johnson and Johnson stocks.

T159. Complex Motor Sequencing as a Potential Intermediate Phenotype for Schizophrenia Genetic Studies

Dwight Dickinson*, Jesse Hochheiser, Jose Apud, Karen Berman, Daniel Weinberger, Thomas Hyde

National Institute of Mental Health, National Institues of Heatlth, Bethesda, Maryland

Background: Neurological ‘soft signs’ occur more frequently among people with schizophrenia than in the healthy population. Since Luria, the sequencing and coordination of hand movements (complex motor sequencing or CMS) has been used to assess neuromotor impairment, in particular. People with schizophrenia show highly reliable deficits on CMS. Intermediate motor and CMS impairments have been documented in first-degree relatives. CMS and other motor signs have shown association with cognition and with enduring negative symptoms and may be less confounded than other neurological soft signs by fluctuating psychotic symptomatology. Inter-rater reliability for CMS is routinely high (IRR>.80), and the tasks take minutes to complete and score. Together, these findings frame the question of whether CMS might be suitable as an intermediate phenotype for genomics analyses. We hypothesized: (1) that CMS would be significantly worse in schizophrenia than in controls, and intermediate in unaffected siblings, (2) that, across groups, CMS performance would be significantly and positively associated with cognitive performance and (3) that CMS would show substantial heritability.

Methods: Our sample consisted of 1049 participants, including 354 people with a DSM-IV diagnosis of schizophrenia or schizoaffective disorder, 383 of their unaffected siblings, and 312 non-psychiatric healthy controls from the NIMH Study of Schizophrenia Genetics (NIMH sibling study, DR Weinberger PI). Diagnoses and eligibility were determined by MD/PhD level investigators. All participants completed CMS tasks and a comprehensive cognitive battery. Symptom ratings (Positive and Negative Syndrome Scale) were completed for cases. We tested for group differences in demographic and cognitive variables, and in CMS, and using mixed model analyses and chi-square tests. A series of planned stepwise linear regressions tested the relationship of CMS to cognitive and demographic variables within each group. In the schizophrenia group only, additional regressions were used to examine the association of CMS with PANSS symptom composites, chlorpromazine equivalents (CPZE) at the time of testing, and age of onset. Variance components from the mixed model results were used to calculate intraclass correlations (ICCs), which serve as an index the familiality of CMS and cognitive variables. For full siblings, doubling the ICC yields an estimate of heritability. We also tested the association of CMS with the sodium channel gene SCN2A rs10174400, a SNP we have previously found to be associated with cognition in our schizophrenia and sibling samples.

Results: Groups differed by age, sex, and education. Age and sex were included in all analyses as covariates. The schizophrenia sample was chronic, with impaired IQ and moderate symptomatology. Mixed model analyses, accounting for the effects of age, sex and family membership, revealed a highly significant overall group effect on CMS (F[2, 869]=72.4, p=8.3E-30). The schizophrenia cases made more CMS errors than controls, siblings were intermediate, and all pairwise group differences were significant. CMS was associated with cognition within all groups, most strongly among the cases, and was associated weakly with years of education for cases and siblings. Among the schizophrenia cases, after controlling for age and sex, regression analyses showed no significant association of CMS with total PANSS symptoms or negative, positive, concrete/disorganized, distress, or excitability symptom dimensions. Variance components analysis from the mixed model results showed significant familiality for CMS (ICC=.188, 95% CI=.140 to .248, p=4.5E-05). For comparison, parallel analysis showed significant familiality for “g” (ICC=.279, 95% CI=.236 to .326, p=7.11E-08). Doubling the ICC for CMS yields an estimate of additive genetic heritability (h2=.376; compared with h2=.558 for “g”). CMS was significantly associated with SCN2A rs10174400 genotype in schizophrenia cases and unaffected siblings (F-change [1, 237]=12.1, p=6.1E-04 and F-change [1, 158]=5.7, p=.02, respectively). The associations were directionally consistent with the associations we recently reported between SCN2A genotype and general cognitive ability.

Conclusions: CMS showed (1) clear association with schizophrenia and schizophrenia risk, (2) trait-like characteristics, (3) substantial familiality/heritability (at a level similar to levels reported for individual cognitive measures), (4) strong association with general cognitive ability, which is among the most widely-used behavioral intermediate phenotypes in schizophrenia genetics studies, (5) directionally consistent association with a genetic marker, SCN2A rs10174400 genotype, previously shown to associate with cognition in this sample. The reliability and brevity of CMS assessment are additional factors supporting the use of complex motor sequencing as a schizophrenia intermediate phenotype. The utility of global phenotypes has been questioned because they may be no simpler genetically than schizophrenia. However, genes encode protein building blocks rather than differentiated brain circuits or observable behaviors. Their influence on cognition or motor performance likely develops gradually through low-level and wide-acting brain mechanisms (eg, synaptic plasticity). Analyses focused on global phenotypes may help establish priorities in mechanisms research.

Keywords: Schziophrenia, Genetics, Phenotypes, Complex motor sequencing.

Disclosure: Nothing to Disclose.

T160. Epigenetics of Social Anxiety - Multilevel Evidence for Oxytocin Receptor (OXTR) Methylation

Katharina Domschke*, Udo Dannlowski, Christiane Ziegler, David Bräuer, Stephan Stevens, Klaus-Peter Lesch, Volker Arolt, Jürgen Hoyer, Alexander Gerlach, Peter Zwanzger, Jürgen Deckert

University of Wuerzburg, Wuerzburg, Germany

Background: Social anxiety disorder (SAD) is a frequent and highly disabling disorder affecting 15 million adults, or 6.8% of the U.S. population. The neuropeptide oxytocin - produced in the hypothalamic paraventricular and supraoptic nuclei and binding to a G protein-coupled receptor (OXTR) – has been suggested as a promising prosocial and anxiolytic agent particularly with regard to social dysfunction. Also, the oxytocin receptor gene (OXTR) located on chromosome 3p25.3 has been implicated in modulating social cognition and behavior. The present study - for the first time applying a multi-level epigenetic approach - investigates the role of OXTR methylation in categorical, dimensional and intermediate neuroendocrinological/neural network phenotypes of social anxiety.

Methods: One-hundred and ten patients with social anxiety disorder (f=76; age: 33.0+1.0) and 110 age- and sex-matched healthy controls were analyzed for OXTR methylation at 12 CpG sites by direct sequencing of sodium bisulfite converted DNA extracted from whole blood. OXTR methylation was furthermore investigated regarding dimensional measures of social anxiety severity using the Social Phobia Scale (SPS) and the Social Interaction Anxiety Scale (SIAS), salivary cortisol response during the Trier Social Stress Test (TSST) and amygdala responsiveness to social phobia related verbal stimuli (fMRI).

Results: Patients with social anxiety disorder showed significantly decreased OXTR methylation as compared to healthy controls (e.g., CpG3: p<0.001). Hypomethylation at this CpG site was furthermore associated with elevated SPS (r=-0.393, p<0.001) and SIAS scores (r=-0.399; p<0.001), increased cortisol response to the TSST (p<0.02), and increased amygdala response during social phobia related word processing (right: p corr=0.001; left: p corr=0.01).

Conclusions: The present results provide evidence for a role of OXTR hypomethylation in social anxiety. Assuming decreased OXTR methylation to confer increased OXTR expression, the present finding could reflect a compensatory upregulation for pathologically reduced oxytocin levels in SAD and related traits. Given robust replication, OXTR hypomethylation could serve as a peripheral biomarker of disease risk and/or severity and might finally allow for targeted preventive interventions as well as more personalized treatment options targeting the oxytocin system.

Keywords: Epigenetics, SAD, TSST, fMRI.

Disclosure: All authors state that they have no biomedical financial interests or potential conflicts of interest related to this manuscript.

T161. The Human Brainome: Genome, Transcriptome and Proteome Interaction in Human Cortex Identifies Quinoid Dihydropteridine Reductase as a Novel Target for Alzheimer’s Disease

Amanda Myers*, Vladislav Petyuk, Manuel Ramirez, Paul Piehowski

University of Miami, Miami, Florida

Background: Our hypothesis is that changes in expression are crucial to the development of Alzheimer’s disease. Previously we have examined how DNA alleles control the downstream expression of RNA transcripts and how those relationships are changed in pathologically confirmed Alzheimer’s Disease tissue. We have now examined how proteins are incorporated into those network relationships.

Methods: SAMPLES- Our KRONOSII series was obtained from 21 National Alzheimer’s Coordinating Center (NACC) brain banks and from the Miami Brain Bank as previously described. The RUSH series includes two large, prospectively followed cohorts maintained by investigators at Rush University medical center in Chicago, IL: the Religious Orders Study (ROS) and the Memory and Aging Project (MAP). In both cohorts, samples were de-identified before receipt, and the study met human studies institutional review board and HIPPA regulations. This work is declared not human-subjects research and is IRB exempt under regulation 45 CFR 46. DATA COLLECTION- DNA samples were analyzed on the Genome-Wide Human SNP 6.0 Array (Affymetrix, Inc. Santa Clara, CA) according to the manufacturer’s protocols. cRNA was hybridized to Illumina HumanRefseq-HT-12 v2 Expression BeadChip (Illumina, San Diego, CA) using a Scigene Little Dipper® Automated Microarray Processor (Scigene, Sunnyvale, CA). Mass Spectrometry analysis was performed using an Exactive Orbitrap mass spectrometer (Thermo Scientific, San Jose, CA) outfitted with a custom electrospray ionization (ESI) interface. The accurate mass and time tag approach was used for identification and 16O/18O labeling was used for quantification. DATA ANALYSIS-Differential expression of both transcripts (n=29,686) and peptides (n=1931) was assessed using the significance analysis of microarrays platform (SAM) version 4.01. HQCUT was used to determine co-expression irrespective of SNP genotype. DATA VALIDATION- A human embryonal kidney cell line expressing a Beta-APP complementary DNA bearing a double mutation which produces 6-8 fold more Abeta than cells without the mutations was used to examine possible causative effects of targets on APP pathways. To examine neuronal phenotypes of targets the NTERA-2 cl.D1 (ATCC, Manassas, VA) cell line was employed.

Results: Cluster prediction was performed separately in cases and controls and clustering was compared across the two human brain tissue sets. For the KRONOS II case set there were 11 clusters, with 5 containing transcripts and 6 containing proteins. For the RUSH case set there were 26 clusters, with 5 containing transcripts and 21 containing mostly peptides. Comparison between each pair of predicted modules found that for the transcripts, KRONOS II cluster 2 and RUSH cluster 8 had the most overlap with 89.7% RUSH transcripts in cluster 8 found KRONOS III cluster 2 and 98.7% KRONOS II cluster 2 transcripts in RUSH cluster 8. For the peptides, KRONOS II cluster 7 and RUSH cluster 4 had the most overlap with 91% RUSH transcripts in cluster 4 found KRONOS II cluster 7 and 69.6% KRONOS II cluster 7 transcripts in RUSH cluster 4. For the KRONOS II controls there were 14 clusters, with 6 containing transcripts and 8 containing proteins. For the RUSH controls there were 13 clusters, with 5 containing transcripts and 8 containing peptides. There was high overlap with two clusters (K13-R12 and K14-R11) which was ignored given these clusters were small. For the other modules, KRONOS II cluster 1 and RUSH cluster 1 had the most overlapping transcripts and KRONOS II cluster and RUSH cluster 9 had the most overlapping peptides. Of the list of targets that were 1. overlapping between KRONOS II and RUSH sets and 2. present in both transcript and peptide clusters QDPR was picked for further follow-up. Overexpression of QDPR in the APP double mutation cell line caused a significant increase in Abeta 42/40 protein levels; however, there was no increase in APP mRNA suggesting that QDPR is acting at the protein level (data not shown). Total tau protein levels were also increased. QDPR overexpression in the NTERA-2 cl.D1 cell line gave significant decreases in the following measures: neurite total count, neurite total length, neurite average length, neurite longest length and neurite branch points.

Conclusions: We have shown the following: 1. QDPR is significantly overexpressed in AD in two human brain cohorts, 2. QDPR is a member of a transcript and protein module which has good overlap between our two human brain cohorts, 3. In the context of known AD pathology (increased levels of Abeta) QDPR can further potentiate pathogical processes via increasing both the ratio of Abeta 42/40 and total Tau levels and 4. QDPR has an effect on human neuronal phenotypes. QDPR is an interesting target for further study since it is a member of the catecholamine synthesis pathway.

Keywords: omics, Alzheimer's, catecholamine, genetics.

Disclosure: Nothing to Disclose.

T162. A Longitudinal Study in Mothers and Firstborn Children of Genetic and Environmental Influences on Externalizing and Internalizing Disorders across Development

Mary-Anne Enoch*, Harriet Kitzman, Joyce Smith, Elizabeth Anson, Colin Hodgkinson, David Goldman, David Olds

National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, Maryland

Background: The development of behavior across childhood and adolescence is known to be both genetically and environmentally influenced. The aim of this study was to determine the effects of genetic and environmental factors on the development of externalizing and internalizing disorders in a cohort of 600 firstborn children followed from pre-birth to 18 years. We hypothesized that maternal mental health and resilience, alcohol and drug use and smoking would predict child behavioral outcomes and that an early life nurse visiting intervention program would be associated with better childhood behavioral outcomes. We also hypothesized that a GABRA2 variant, previously robustly associated with the development of externalizing disorders in late adolescence would predict childhood behavior.

Methods: Data were derived from a longitudinal study of the effects of prenatal and infancy nurse home educational visiting (NHV) on health outcomes up to age 18 years in a group of urban, predominantly African American economically disadvantaged individuals in Memphis, TN. In a sub-study, a total of 600 women were consecutively recruited from an obstetric clinic with their first viable pregnancy. Of these, 200 were randomly assigned to NHV; on average 7 visits in pregnancy and 26 from birth to age 2. Mothers were assessed and scored at intake for mental health (depression, anxiety, emotional dysregulation), self-efficacy (belief in one’s own ability to complete tasks/achieve goals) and the Pearlin mastery scale (personal sense of mastery, control). Mothers reported on smoking, alcohol and drug use during pregnancy and when their children were 6, 9 and 12 years old. The Achenbach Child Behavior Checklist (CBCL) was completed by mothers at 2, 6, 12 and 18 years and by youths at ages 12 and 18 years. Outcome measures derived from the CBCL were a composite internalizing disorders (ID) score (anxious/depressed, withdrawn/depressed, somatic complaints) and a composite externalizing disorders (ED) score (rule-breaking behavior, aggressive behavior). DSM-IV alcohol and drug use disorder (AUD, DUD) diagnoses were derived from the Substance Abuse Module (SAM) completed by youth at age 18. Mothers and children were genotyped for the GABRA2 tag SNP rs279858 together with 186 ancestry informative markers. Multiple regression analyses were performed for each of the 4 time-points.

Results: Maternal use of alcohol and drugs was minimal throughout but smoking increased from 12% in pregnancy to 30% at age 12 years. ED and ID scores were strongly positively correlated at all time-points (r=0.57 – 0.76, p<0.0001). Independent predictors for both ED and ID scores at age 2 included sex, maternal mastery, maternal mental health, a maternal mental health x GABRA2 rs279858 interaction and a maternal self-efficacy x NHV interaction. In the latter case, ED and ID scores were both significantly lower in children whose mothers had high self-efficacy and had received NHV, compared with the other 3 groups. At age 6 the influence of NHV on ED and ID scores was no longer apparent. Within the whole regression models, the strongest predictors of ED and ID from 6 and upwards were ED and ID scores at younger ages. As an illustration, the ED score at age 2 was correlated (all p<0.0001) with ED at age 6 (r=0.44), age 12 (r=0.31) and age 18 (r=0.23). Maternal mental health during pregnancy continued to influence ED and ID up to 12 years after which maternal mastery re-emerged in importance. Maternal smoking predicted ED at all time-points from 2 to 18 years and ID from 6 to 18 years. The influence of GABRA2 rs279858 emerged again at age 18 with a main effect on both ED and ID. Moreover, there was an interactive effect between GABRA2 rs279858 genotype and youth-report ED at age 12 (but not ID) on AUD and DUD at age 18.

Conclusions: Our study suggests that there are long lasting effects of maternal mental health during pregnancy, resilience (belief in one’s own ability, personal sense of control) and smoking (possibly a marker for maternal ED and ID) on childhood behavior, even though the ID and ED scores were respectively clinically significant in only 10–15% of children. NHV positively influenced early behavior but only in children of resilient mothers. As has been shown in earlier studies of ED across development, environmental influences predominate at younger ages and genetic effects, such as the influence of GABRA2 rs279858 genotype, become relatively more important towards adulthood. The results of our study have implications for strategies for prevention of pathological ED and ID in adulthood.

Keywords: GABRA2, Child Behavior Checklist, addiction, externalizing disorders.

Disclosure: Nothing to Disclose.

T163. The Influence of rs1360780 on Expression of FKBP5 and Other Glucocorticoid Regulated Genes in the Context of Childhood Trauma

Seungeun Yeo*, Mary-Anne Enoch, Colin Hodgkinson, Elena Gorodetsky, Longina Akhtar, David Goldman

National Institute on Alcohol Abuse and Alcoholism, National Institues of Heatlth, Bethesda, Maryland

Background: FKBP5 is an important regulator of the glucocorticoid receptor (GR) complex. FKBP5 variation has been associated with depression, PTSD and suicidality, particularly in the context of childhood trauma (CT). The FKBP5 SNP rs1360780 risk genotypes (AA/AG) have been associated with a decrease in GR sensitivity. However the functionality of the SNP and its interaction with previous stress exposure are still incompletely understood. We have studied the effect of rs1360780 genotype on FKBP5 expression through differential allelic expression in cell lines using rs3800373 which is located in the FKBP5 3’UTR and is in strong linkage disequilibrium (r2) with rs1360780 as a reporter for the cis regulatory effects of alleles at this locus. We also investigated whether genotypes and CT influence the expression of GR responsive genes after GR stimulation to identify genotype-induced GR resistance.

Methods: A total of 29 lymphoblastoid cell lines were selected based on rs1360780 and rs3800373 genotypes and childhood trauma exposure (CTQ) from previously described samples of African American males (Enoch et al. 2010). There were 12 rs3800373 heterozygotes in whose DAE of the reporter alleles were measured. There were 4 rs1360780 AA homozygotes and 13 GG homozygotes. There were 16 cases exposed to CT and 13 controls. GR response was determined by incubating the cells with 100 nM dexamethasone (Dex) for 24 hours after which genomic DNA and RNA were extracted. To analyze differential allelic expression (DAE), RNA was reverse transcribed, cDNA PCR amplified and the PCR product directly Sanger sequenced on an Applied Biosystems 3730 and the data analyzed using Sequencing Analysis 5.1 software. Genomic DNA was amplified and directly sequenced as a control. In the sequence chromatograms the areas under the two peaks corresponding to the G and T allele of rs3800373 were measured by ImageJ1.46r. These peak areas were normalized using ratios of the peaks from sequencing heterozygous gDNA. Expression of FKBP5 and the GR response genes (TSC22D3, BIM, and RUNX2) was analyzed by quantitative RT-PCR (qPCR) using internal GAPDH control. Relative gene expression levels for each sample were calculated using the ΔCT values.

Results: At baseline there was a trend (p=0.08) for CT to lower FKBP5 expression but no effect of genotype. GR stimulation resulted in a>5-fold increase in FKBP5 transcript levels in both cases (high CT) and controls (low CT). In cases, rs1360780 genotype predicted different dexamethasone induction of FKBP5 expression, and as previously observed, carriers of the risk genotype showed a significantly lower FKBP5 induction than carriers of the protective rs1360780 genotype (p=0.006), with no genotype effect in controls. When cases and controls with risk genotypes were compared, the case group had lower average FKBP5 expression (p=0.012). In contrast, there was no difference in FKBP5 expression level between cases and controls who were carriers of the protective genotype. Therefore the combination of FKBP5 risk genotype and CT appears to be associated with low FKBP5 induction after GR stimulation. To better understand the functionality of rs1360780 and the relation between genotype, stress exposure, and cortisol receptor activation, we analyzed DAE in cases and controls who were heterozygous for both rs1360780 (A/G) and the rs3800373 (G/T) reporter locus that is in high r2 (0.88). The haplotypes including SNPs rs1360780 and rs3800373 are AG and GT (frequencies 0.39 and 0.59, respectively). The rs1360780 risk allele (A) leads to higher expression of FKBP5. Therefore the rs3800373 G allele with which it is in high LD and r2 should tend to be overexpressed in rs3800373 heterozygotes. We found that controls did have consistently higher G:T allelic expression ratios, but cases showed more variability in DAE. We then measured expression of glucocorticoid regulated genes: TSC22D3, BIM, and RUNX2. The expression of these genes was increased after GR stimulation with differing results in relation to rs1360780 genotype and CT. Controls carrying rs1360780 risk genotype showed a higher TSC22D3 response to GR stimulation than controls carrying the protective genotype (P=0.05). On the other hand, cases carrying the rs1360780 risk genotype did not affect TSC22D3 gene expression. We also found that RUNX2 and BIM expression induced by GR activation were unaffected by rs1360780 genotype or CT.

Conclusions: FKBP5 rs1360780 has previously been associated with altered induction of FKBP5 after GR activation, impairing the function of an ultra-short feedback loop. In this study, we confirmed by DAE analysis that rs1360780 predicts a cis-regulatory effect on FKBP5 expression. We also confirmed that carriers of the risk genotype who have experienced CT have lower FKBP5 induction than controls. The difference in FKBP5 DAE that we observed between cases and controls suggests that CT may disturb the cis effect on FKBP5 expression and lead to resistance in FKBP5 induction after GR activation. Together, rs1360780 and CT modulate FKBP5 transcription after GR activation. However, other GR response genes did not follow the pattern of FKBP5. Those results suggest that the effect of rs1360780 linked to CT is associated with only FKBP5 expression, but that there are other factors modulating expression of other GR responsive genes.

Keywords: FKBP5, childhood trauma, cis-regulatory effect, rs1360780.

Disclosure: Nothing to Disclose.

T164. FKBP5 Genotype and Psychopathology are Risk Factors for Emotional Eating among African Americans of Low Socioeconomic Status

Vasiliki Michopoulos*, Bekh Bradley, Kerry Ressler

Emory University, Atlanta, Georgia

Background: Exposure to chronic stress is implicated in the etiology of adverse health outcomes. Importantly, African Americans (AA) of low socioeconomic status (SES) are exposed to greater rates of trauma than the general population and have exacerbated risk for depression, posttraumatic stress disorder (PTSD), and other adverse health outcomes that are highly comorbid. For instance, depression and PTSD result from stressor exposure and are associated with increased risk for cardiovascular disease, diabetes, and obesity. However, it remains unclear which biological signals are important for the comorbid presentation of such conditions. One possible target for comorbid stress-induced psychiatric and metabolic disorders is FKBP5, a regulator of glucocorticoid sensitivity. FKBP5 is implicated in increased vulnerability to psychopathology as well as increased vulnerability to weight gain under chronic stress conditions in rodents. The rs1360780 FKBP5 polymorphism has been implicated in risk for PTSD. However, it remains unknown whether this polymorphism is also associated with increased risk for emotional eating in humans that results from stress exposure and can lead to obesity. Thus, we hypothesized that the at-risk (T) allele of the rs1360780 FKBP5 polymorphism would be associated with increased emotional eating in a highly traumatized, low SES sample.

Methods: The current study was undertaken to characterize risk factors for emotional eating in AA men and women (n=522) seeking treatment from the obstetrics and gynecology and primary care clinics at Grady Memorial Hospital in Atlanta, GA, which serve a primarily AA, low SES, inner-city population at risk for both trauma exposure and obesity. Emotional eating was determined via the Dutch Eating Behavioral Questionnaire (DEBQ). Body mass index (BMI) was determined via self-report of height and weight. Lifetime trauma history was determined by the 14-item Traumatic Events Inventory (TEI), which assesses for experiencing and witnessing traumatic events. Childhood trauma history was assessed via the Childhood Trauma Questionnaire (CTQ). The Beck Depression Inventory (BDI) was used to identify individuals with current depression and the PTSD Symptom Scale (PSS) was used to determine participants with a current PTSD diagnosis. DNA from all participants was collected with Oragene saliva kits (DNA Genotek, Ottawa, Ontario, Canada) and extracted using the Qiagen M48 automated extraction system. FKBP5 genotyping at the rs1360780 locus was quantified by a TaqMan assay as previously described by our team.

Results: Emotional eating was significantly positively correlated with total number of trauma exposures (p=0.003), depressive symptom severity (p<0.001), and PTSD symptom severity (p<0.001), but not BMI (p=0.83). Furthermore, compared to individuals with neither depression nor PTSD, emotional eating was greater in individuals with a diagnosis of depression, PTSD, or both (all p<0.05). This main effect of psychopathology was affected by gender, as emotional eating severity in AA men was higher in those with PTSD (p=0.004), whereas in AA women, emotional eating was increased in those with depression (p<0.001) while controlling for BMI, adult and childhood trauma exposure. Lastly, emotional eating severity was significantly greater in individuals carrying the T risk allele of rs1360780 compared to CC individuals (p=0.031), and remained significant after controlling for gender, childhood trauma, BMI, and current depression and PTSD symptoms (p=0.012).

Conclusions: Taken together, these data indicate that emotional eating is associated with depression and PTSD in low SES AA individuals, and that the at-risk (T) rs1360780 FKBP5 allele is associated with increased emotional eating severity. These findings suggest that engaging in comfort food eating might serve as a coping strategy similar to what has been described with alcohol and substance abuse in depression and PTSD, and that FKBP5 genotype may be an important genetic risk factor for stress-induced eating. Therapy focusing on replacing emotional eating with more positive coping strategies could serve as an approach to attenuate adverse health outcomes in this at-risk population. Additionally, the data indicate that FKBP5 might serve as a therapeutic target for treating comorbid psychiatry and metabolic disorders.

Keywords: FKBP5, Emotional Eating, PTSD, Depression.

Disclosure: Nothing to Disclose.

T165. De Novo Genomic Investigations in Tourette’s Disorder

Thomas Fernandez*, Robert King, Gary Heiman, Jay Tischfield, Matthew State

Yale University, New Haven, Connecticut

Background: Tourette’s Disorder (TD) is an often-disabling developmental neuropsychiatric syndrome, characterized by persistent motor and vocal tics. Current treatments have limited efficacy and may carry significant long-term adverse effects. The fundamental challenge in identifying novel therapeutic targets is our limited understanding of underlying biological mechanisms. While there is strong evidence for a genetic etiology of TD, progress in gene discovery has been slow relative to other psychiatric disorders, due in large part to a shortage of well-phenotyped subjects and families with biosamples available for sharing and replication of findings. The Tourette International Collaborative Genetics (TIC Genetics) study is filling this void and, to date, has collected over 1,000 affected individuals and their family members over the last 2.5 years. Here we present initial results from genome-wide studies of de novo (DN) single nucleotide (SNV) and copy number variation (CNV) in a TIC Genetics cohort of simplex parent-child trios.

Methods: Genotyping and CNV Analysis: Whole-blood DNA from 235 TD parent-child trios were genotyped using the Illumina HumanOmniExpressExome v1.2 BeadChip array. Standard quality control measures were used, including genotype call rate, cryptic relatedness, Log R variability, sample duplication, and ancestral outlier detection. CNV predictions were made using CNVision, a pipeline integrating results from three algorithms: PennCNV, QuantiSNP, and GNOSIS. Batch effects were evaluated and corrected by using principal component analysis of Log R intensity data. Only high confidence (HC) rare transmitted and DN CNV predictions were included (≥50% agreement among at least two algorithms and visualization of Log R ratios and B allele frequencies by two blinded investigators [95% PPV]) and all DN predictions were confirmed by qPCR. Whole Exome Sequencing (WES): Whole-blood DNA from 227 TD trios was enriched for exonic sequences by hybridization with NimbleGen SeqCap EZ Exome v2 capture arrays and sequenced on the Illumina HiSeq 2000. 74-bp paired-end sequence reads were aligned to the hg19 human reference genome using BWA. Reads that did not align, or aligned outside the target region, were discarded. Duplicate reads were removed and variants were predicted using SAMtools. Alignment and variant calling used an in-house pipeline script (http://futo.cs.yale.edu/mw/index.php/SEQer), incorporating BWA and SAMtools, and insuring consistency in data processing. Validation for all predicted DN SNVs via Sanger sequencing of all trio family members, with sequence readers blind to affected status, confirmed 94%.

Results: CNV: 207 trios passed quality control and showed a statistically significant increase in the burden of rare genic DN CNVs relative to 498 control trios (3.4% vs 0.8%, p=0.04). Combining these results with our prior study of CNVs in TD confirmed an increase in genic DN CNVs in affected individuals (3.1% vs 0.8%, p=0.02). Finally, we conducted pathway analysis of the 326 genes identified within rare CNVs in this cohort using IPA and found an enrichment of genes within the following canonical pathways: complement system (p=3.3 x 10-5), G-protein coupled receptor signaling (p=1.9 x 10-4), and cAMP-mediated signaling (p=2.2 x 10-4). WES: Among 227 trios, 226 DN SNVs were confirmed by Sanger sequencing, including 48 silent, 159 missense and 19 loss of function (LoF, nonsense, altering a canonical splice site, or disrupting a start/stop codon). Among 509 control trios, 465 de novo SNVs were confirmed, including 132 silent, 313 missense, and 20 LoF. In order to ensure that batch effects influencing call rates did not confound comparisons between our cases and controls, we normalized variant rates using the rate of silent variants in each cohort as the denominator. Overall, the ratio of DN LoF to silent variants was significantly greater in cases (0.40) versus controls (0.15) (OR 2.6, p=0.007). There was no significant difference detected for ratios of DN missense to silent variants between cases (3.3) and controls (2.4) (OR 1.4, p=0.09). Interestingly, when we conducted the identical analyses separately by gender we found that male TD subjects (n=181) showed a significantly greater ratio of both DN LoF (OR 4.4, p=0.003) as well as missense (OR 1.7, p=0.04) to silent variants versus controls (n=238). No differences were observed between female TD cases (n=46) versus controls (n=271). These results were not accounted for by differential sequencing quality, rates of OCD, tic severity, or parental age. Finally, we performed a pathway analysis of genes harboring 19 de novo LoF and 159 de novo missense SNVs using IPA. Enrichment was detected for the following canonical pathways: fMLP signaling in neutrophils, G-protein coupled receptor signaling, gap junction signaling, dopamine-DARPP32 feedback in cAMP signaling, and G-Protein alpha-i signaling.

Conclusions: In contrast to the only other large-scale TD genetics study, which is currently focused largely on common variants, we have and will continue to focus on the contribution of rare alleles as well as both structural and sequence variation. By genotyping and exome sequencing apparently simplex TD trios, we found an increased burden of DN CNVs and SNVs in cases vs controls. Similar to other neurodevelopmental disorders, such as autism spectrum disorders, this approach can yield a fruitful path toward gene discovery. Despite tremendous locus heterogeneity, these rare variants have been shown to converge on a smaller number of pathways and networks that can transform our understanding of disease biology and suggest new treatment targets. In addition to demonstrating, for the first time, an over-representation of DN sequence and structural variation in TD, we show that this burden is carried predominantly by male rather than female probands. Given the male predominance in TD, this discrepancy is surprising and signals a need to sequence more TD females to confirm this finding.

Keywords: exome sequencing, copy number variation, rare variation, Tourette syndrome.

Disclosure: Dr. Fernandez receives research funding from the National Institute of Mental Health, the Allison Family Foundation, and Shire.

T166. Genome-wide Transcriptome Analyses Reveal Shared and Distinct Molecular Pathways among Major Neuropsychiatric Illnesses

Michael Gandal*, Neelroop Parikshak, Jason Stein, Dan Geschwind

University of California Los Angeles, Semel Institute for Neuroscience & Human Behavior, Los Angeles, California

Background: Recent work has demonstrated significant overlap among both common and rare genetic risk factors for multiple distinct neuropsychiatric disorders. How this shared genetic architecture integrates with epigenetic and environmental perturbations to cause disease is unknown. Transcriptomic perturbations caused by disease and measured in brain tissue can be used as a functional readout to (1) quantify cross-disorder overlap in the tissue of interest, (2) discover disorder-specific processes, and (3) put forward new and high-priority biological pathways to target for potential therapeutic intervention.

Methods: To explore this, we characterized global transcriptome abnormalities in post-mortem cortical tissue from subjects with autism (ASD; n=24), schizophrenia (SZ; n=169), bipolar disorder (BD; n=104), major depression (MDD; n=83), Alzheimer’s disease (AD; n=41), and controls (n=269) across 15 studies employing gene expression microarrays. We controlled for biological (age, sex, brain region) and technical covariates (PMI, pH, RNA quality, batch effects) within and across studies. Common patterns of differential expression (DEX) were assessed by linear regression of disease-associated beta values and by the Rank Rank Hypergeometric Overlap test. Weighted Gene Coexpression Network Analysis (WGCNA) was used to identify disease-related networks of co-expressed genes.

Results: Meta-analysis of differential expression demonstrated an excess of overlapping DEX genes across ASD, SZ, and interestingly AD (all FDR-adjusted p’s<10-22). MDD was significantly associated with SZ and BD (FDR-adjusted p’s<10-92) but showed negative overlap with AD (FDR-adjusted p<10-6). BD showed an excess of overlap with all disorders except AD. WGCNA identified 5 groups of genes, or modules, associated with disease (FDR-adjusted p<0.05) in distinct combinations. Two modules (dM1 & dM2) were associated with ASD, SCZ, and BD and enriched for markers of astrocytes, fatty acid oxidation, and mitochondrial function. dM3, upregulated only in MDD, was associated with G-protein coupled receptor and glucocorticoid signaling. dM4, downregulated in ASD and SZ, was enriched for mTOR signaling and glutamateric synaptic genes. Finally, dM5, upregulated in ASD and AD, was associated with microglial activation.

Conclusions: Together, these findings provide a systems-level view of the transcriptional architecture of major neuropsychiatric illness and demonstrate pathways of molecular convergence and specificity across distinct disorders.

Keywords: gene expression, transcriptome, systems biology.

Disclosure: Nothing to Disclose.

T167. Genetics of Early Onset Bipolar Disorder

Paul Croarkin*, Joan Luby, Kelly Cercy, Jennifer Geske, Marin Veldic, Matthew Simonson, Paramjit Joshi, Karen Dineen, John Walkup, Alfredo Cuellar-Barboza, Leah Casuto, Susan McElroy, Peter Jensen, Mark Frye, Joanna Biernacka

Mayo Clinic, Rochester, Minnesota

Background: Prior genome-wide association studies have implicated several genes in bipolar disorder (BD), including voltage-dependent calcium channel, L-Type, alpha 1C subunit (CACNA1C); ankyrin-3 (ANK3); and protein odd oz/ten-m homolog 4 or teneurin-4 (ODZ4). However, it is not known if SNPs in these genes have stronger associations with early-onset BD versus adult onset. This study sought to examine the association of 8 single nucleotide polymorphisms (SNPs) in these 3 genes with early-onset BD using samples from the Treatment of Early Age Mania (TEAM) study and Mayo Clinic biobanks. The TEAM study was a multisite, randomized, eight-week trial of risperidone, lithium carbonate, and divalproex sodium in children and adolescents ages 6 to 15 with BD I presenting as a manic or mixed episode for four continuous weeks prior to enrollment. The Mayo Clinic BD Biobank has collected biologic samples and clinical data from more than 1,000 individuals (ages 18-65) with BD I, II, and schizoaffective disorder, bipolar type. The Mayo Clinic Biobank, which we refer to as the “control biobank”, has enrolled 30,000 Mayo Clinic adult patients.

Methods: Based on prior studies, we selected 8 candidate SNPs for this study: rs1006737, rs1024582, rs4765913, and rs10848632 in CACNA1C; rs1938526, rs9804190, and rs10994336 in ANK3; and rs12576775 in ODZ4. Candidate SNPs were genotyped in patients with pediatric BD from the TEAM study (n=82), adult bipolar patients from the Mayo Clinic BD Biobank (n=828), and control subjects from the Mayo Clinic control Biobank (n=832). The analyses focused on comparisons of pediatric onset patients, patients with adult onset BD, and healthy controls. For Mayo Clinic adult bipolar cases, age of first manic (n=423) and depressive episodes (n=433) were determined based on review of the Structured Clinical Interview for DSM diagnoses (SCID), and Electronic Medical Record (EMR), where available. Based on this review, a subset of the adult Mayo Clinic bipolar cases were classified as “early onset” BD (defined as an age of onset 19 or younger). Genotypic data on Mayo Clinic and TEAM subjects were analyzed to determine if candidate SNPs were associated with BD and specifically early onset BD. To reduce the chances of confounding by population stratification, analyses used a subset of Caucasian subjects. Thus, analyses compared N=69 Caucasian pediatric-onset cases from the TEAM study, N=732 Caucasian adult cases from the Mayo Bipolar Biobank (including early-onset and adult-onset cases) and N=776 Caucasian controls. Association was assessed using logistic regression with log-additive allele effects (i.e. coding genotypes as 0, 1, 2 representing the minor allele count). Haplotype analyses of SNPs in CACNA1C and ANK3 were performed. Finally, associations of BD and early-onset BD with genetic risk scores (GRSs) composed of all genotyped candidate SNPs were evaluated, using a simple count GRS (SC-GRS) and odds ratio weighted GRS (OR-GRS).

Results: In analyses comparing TEAM pediatric-onset cases with Mayo Clinic controls, we obtained nominally significant evidence of association of the minor allele of rs10848632 in CACNA1C with early-onset BD (OR=1.54, p=0.017). Haplotype analyses demonstrated no significant associations of ANK3 and CACNA1C haplotypes with BD when comparing adult cases with controls. However, comparison of pediatric-onset TEAM cases with controls revealed significant association of early-onset BD with CACNA1C rs10848632-rs1006737-rs1024582-rs4765913 haplotypes (global test p=0.013), with the T-G-G-T haplotype being associated with increased risk of early-onset BD (maximum statistic simulation p=0.0087). ANK3 haplotypes were not significantly associated with early-onset BD. Analyses of adult cases from the Mayo Clinic biobank compared to controls provided no evidence of association of the GRSs with BD (SC-GRS p=0.71, OR-GRS p=0.56). Although the odds ratio point estimates for risk score analysis of TEAM cases compared to controls were larger than in the adult case/control comparison, the associations with GRSs were also not significant in these comparisons. However, analyses comparing the Mayo early-onset cases vs. controls provided significant evidence of association of risk scores with early-onset BD (SC-GRS p=0.06, OR-GRS p=0.02). Furthermore, analyses of the combined group of early-onset cases (TEAM+Mayo) vs. controls also demonstrated association of risk scores with early-onset BD (SC-GRS p=0.02, OR-GRS p=0.01). Results were similar when the combined early-onset case group was compared with the late-onset Mayo Clinic biobank cases, demonstrating that the loci investigated appear to primarily contribute to risk of early-onset, rather than late-onset BD.

Conclusions: These preliminary analyses suggest that rs10848632 in CACNA1C may be implicated in pediatric BD. Furthermore, genetic risk score analyses demonstrated that the loci investigated appear to primarily contribute to risk of early-onset, rather than late-onset BD.

Keywords: pediatric, CACNA1C, ANK3, ODZ4.

Disclosure: Paul Croarkin has receieved research support from Pfizer for an investigator initiated study and equipment from Neuronetics for research studies.

T168. Problematic Alcohol Behavior and Related Neural Phenotypes Are Associated with an Expression QTL of the Sodium Channel and Clathrin Linker 1 (SCLT1) Gene

Kerry Ressler*, Lynn Almli, Jacquelyn Meyers, Jaemin Shin, Negar Fani, Karen Conneely, Adam Maihofer, Caroline Nievergelt, Duna Abu-Amara, Rachel Yehuda, Charles Marbar, Bekh Bradley

Emory University, Howard Hughes Medical Institute, Atlanta, Georgia

Background: Excessive alcohol consumption is a common and costly public health problem affecting 12.5% of the U.S. population1. We report results from a genome-wide association study (GWAS) of alcohol consumption and behavior in a community cohort of African-American subjects.

Methods: We used an Illumina GWAS array in a primary and secondary meta-analytic cohorts using alcohol use phenotypes as measured by the Alcohol Use Disorders Identification Test (AUDIT). We performed additional quantitative trait gene expression analyses and resting state MRI analyses to further understand the biology of the polymorphism associated with problematic drinking behavior.

Results: Results indicate a Bonferroni-significant association between total AUDIT score and rs1433375 [N=1036, p=7.76 × 10-8 (additive), p=2.61 × 10-8 (dominant)], a single nucleotide polymorphism (SNP) located 323kb upstream of Sodium Channel and Clathrin Linker 1 (SCLT1) at 4q28, which was replicated in a meta-analysis of two independent cohorts (N=1394, p=0.0004). rs1433375 was also associated with SCLT1 gene expression and with cortical-cerebellar functional connectivity measured via fMRI.

Conclusions: These convergent findings provide evidence for a pathway that may explain prior findings related to putative effects of alcohol including sodium channel regulation at the cellular level and cerebellar functioning in neural circuits, shedding light on the etiology of alcohol-related disorders. The present findings suggest that those individuals who carry a particular risk allele for this SNP may be more susceptible to problematic drinking, possibly due to alterations in connectivity in this executive control network.

Keywords: alcohol, genetics, addiction, cerebellum.

Disclosure: Nothing to Disclose.

T169. Glutamate Networks Implicate Cognitive Impairments in Schizophrenia; Genome-wide Association Studies of 52 Cognitive Phenotypes

Ryota Hashimoto*, Kazutaka Ohi, Masashi Ikeda, Hidenaga Yamamori, Yuka Yasuda, Michiko Fujimoto, Satomi Umeda-Yano, Masaki Fukunaga, Haruo Fujino, Yoshiyuki Watanabe, Masao Iwase, Hiroaki Kazui, Nakao Iwata, Daniel Weinberger, Masatoshi Takeda

Osaka University, Suita, Japan

Background: Cognitive impairments are a core feature in patients with schizophrenia. These deficits could serve as effective tools for understanding the genetic architecture of schizophrenia. This study investigated whether genetic variants associated with cognitive impairments aggregate in functional gene-networks related to the pathogenesis of schizophrenia.

Methods: Here, genome-wide association studies (GWAS) of a range of cognitive phenotypes relevant to schizophrenia were performed in 411 healthy subjects. We attempted to replicate the GWAS data using 257 patients with schizophrenia and performed a meta-analysis of the GWAS findings and the replicated results. Because gene-networks, rather than a single gene or genetic variant, may be strongly associated with the susceptibility to schizophrenia and cognitive impairments, gene-network analysis for genes in close proximity to the replicated variants was performed.

Results: We observed nominal associations between 3,054 variants and cognitive phenotypes at a threshold of P<1.0 × 10-4. Of the 3,054 variants, the associations of 191 variants were replicated in the replication samples (P<0.05). However, no variants achieved genome-wide significance in a meta-analysis (P>5.0 × 10-8). Additionally, 115 of 191 replicated SNPs have genes located within 10kb of the SNPs (60.2%). These variants were moderately associated with cognitive phenotypes which ranged from P=2.50 × 10-5 to P=9.40 × 10-8. The genes located within 10kb from the replicated SNPs were significantly grouped in terms of glutamate receptor activity (FDR q=1.39 × 10-19) and the immune system related to major histocompatibility complex class I (FDR q=4.84 × 10-8) networks.

Conclusions: Our findings demonstrate that genetic variants related to cognitive trait impairment in schizophrenia are involved in the N-methyl-D-aspartate glutamate network.

Keywords: schizophrenia, genome-wide association study, cognitive phenotypes, glutamate receptor activity.

Disclosure: Nothing to Disclose.

T170. GWAS of Suicidality in Army STARRS

Murray Stein*, Colter Mitchell, Robert Ursano, Steven Heeringa, Chia-Yen Chen, Sonia Jain, Rema Raman, Matthew Nock, Joel Gelernter, Stephan Ripke, Tianxi Cai, Ronald Kessler, Jordan Smoller, Army STARRS Biomarkers Working Group

University of California at San Diego, La Jolla, California

Background: Suicide is one of the most common causes of death among young adults, and it is of particular concern to the US military which has seen rates of suicide exceed those in the general population. Several prior studies have pointed to the possibility that there may be genetic risk factors for suicide, but very few studies have included relatively large, population-based samples and even fewer have focused on the military. Army STARRS is a study funded by the US Army and NIMH intended to discern risk factors for suicide and related mental health problems among soldiers. We will present here results from two genomewide association studies (GWAS) focused on suicidality (ideation and attempts) in Army STARRS.

Methods: In addition to mental health surveys, participants in two studies (New Soldier Study [NSS], N 8,000; and Pre-Post Deployment Survey [PPDS], N 8,000) gave blood for DNA. DNA was extracted and genotyped using an array-based system (Illumina OmniExpress+Exome), and data were QCed and imputed to 1000 Genomes using standard methods. Subjects were phenotyped based on reporting of lifetime suicidal ideation and lifetime suicide attempt(s). Case-control analyses were conducted separately, using similar models, for both studies. After clustering of respondents into ancestral groups and the derivation of principal components (PCs) for each group, logistic regression models were run separately for each group including 10 PCs per ancestral group and the phenotype of interest. Results for each ancestral group were then meta-analyzed across ancestral groups within each study (NSS and PPDS) as well as across studies. P values of genomewide significance (p<5*10-8) are reported.

Results: Several genomewide significant findings were noted for the phenotype "lifetime suicide attempt" when comparing to controls without lifetime suicide attempt or ideation. An example of a genomewide significant finding that was significant when both studies (NSS and PPDS) were meta-analyzed was the association for African Americans of lifetime suicide attempt with a SNP on Chromosome 3 (rs59653394; MAF=8%; Odds Ratio 3.4). Another strong finding was an association within the NSS for European Americans of lifetime suicide attempt with a SNP on Chromosome 8 (rs1398491; MAF=45%; Odds Ratio 2.9). A full reporting of genomewide significant findings will be provided in the poster.

Conclusions: These preliminary GWAS findings from Army STARRS suggest that numerous common genetic variants contribute to risk for suicidality. The findings further suggest that there may be some variants of relatively large effect size (OR 2-3) that contribute to risk for suicidality, in addition to multiple variants of smaller effect size that cumulatively influence risk. These findings are in need of further replication (which is in progress with an NSS Replication Sample) and of additional determination of the genes, pathways, and systems in which these variants function in order to make inferences about their population importance and the mechanisms by which they contribute to suicide risk.

Keywords: suicide, GWAS, genetics, military.

Disclosure: Nothing to Disclose.

T171. CCL2 Genotype, CSF Inflammatory Markers, and Neurocognitive Functioning in an HIV+Sample

April Thames*, Marisa Briones, Larry Magpantay, Oto Martinez-Maza, Elyse Singer, Charles Hinkin, Keith Heinzerling, Andrew Levine

University of California Los Angeles, Los Angeles, California

Background: Genetic variation in MCP-1/CCL2 has been found to be associated with HIV-disease progression and neurocognitive functioning over time (Gonzalez et al., 2002; Levine et al., 2014). CCL2 expression in cerebral spinal fluid (CSF) and plasma show strong links to neurocognitive impairment among various neurological populations. The current study examined relationships between CCL2 genotype, expression of neuroinflammatory markers in the CSF, and neurocognition among a cohort of HIV infected individuals. We expected that HIV+carriers of the CCL2 risk allele would demonstrate higher levels of CSF inflammatory markers, and that these markers would be associated with cognitive impairment and HIV disease-related variables (i.e., CD4 and viral load).

Methods: Participants consisted of 145 HIV+individuals who were enrolled in the National Neurological Tissue Consortium. Genotyping was conducted using peripheral blood mononuclear cells (PBMCs) and/or frozen tissue samples. The Autopure LS™ nucleic acid purification instrument was used to extract DNA and then samples were genotyped for CCL2. Biomarker assays were performed using multiplexed (Luminex platform) assay kits (R & D Systems, Minneapolis, MN) according to manufacturer directions, using a Bio-Plex 200 Luminex instrument and Bio-Plex analysis software (Bio-Rad, Hercules, CA). CSF expression levels of CCL2, sTNFR2 and BAFF were quantified. Global neurocognitive functioning was determined with a comprehensive battery of neuropsychological measures by which a global deficit score was created. Higher deficit score represents greater impairment.

Results: Individuals who carried the MCP-1/CCL2 risk allele demonstrated higher levels of CCL2 expression in CSF, F (1, 144)=4.16, p=.04, but there were no observed differences between genotype groups on cognitive deficit score F (1,144)=1.12, p=.20. Higher levels of CCL2 were significantly associated with viral load, r (145)=.33, p<.0001, and CD4 count, r (145)=-.259, p=.002, and cognitive impairment r (145)=.20, p=.01. Individuals with cognitive impairment also demonstrated higher levels of BAFF, F (1,144)=5.26, p=.02 and sTNFR2, F (1, 144)=5.35, p=.02 than individuals without cognitive impairment. There was a statistically significant interaction between CCL2 genotype and cognitive status on expression of sTNFR2, F (1,144)=3.130, p=.04, such that for non-carriers of CCL2 risk allele, sTNFR2 was higher among individuals with cognitive impairment. There was also a statistically significant interaction between CCL2 genotype and cognitive status on expression of BAFF, F (1,144)=3.662, p=.03, such that for non-carriers of CCL2 risk allele, BAFF was higher among individuals with cognitive impairment. However, for carriers of the CCL2 risk allele, there were no statistically significant differences in sTNFR2 or BAFF expression. To examine the additive effects of MCP-1/CCL2 genotype and CCL2 expression, four groups were created based upon genotype (risk vs. no risk) and expression of CCL2 in CSF (high vs low). The groups were as follows: No risk/Low CCL2; No risk/High CCL2; Risk/Low CCL2; Risk/High CCL2. We found that the Risk/High CCL2 had lower CD4 count, higher CD8 count, and higher viral load than the other three groups (all p’s<.05). While there were no statistically significant differences in cognitive deficit score between No Risk/High CCL2 (M=.96; SD=.76) and Risk/High CCL2 groups (M=1.13; SD=1.0), these groups demonstrated greater impairment than the Risk/Low CCL2 (M=.76; SD=.63) and No Risk/Low CCL2 groups (M=.62; SD=.53).

Conclusions: In an HIV-positive sample, we found that carriers of the MCP-1/CCL2 risk allele expressed higher levels of CCL2 in CSF, which was associated with cognitive impairment. There was no direct association between CCL2 gene and cognitive status, suggesting the importance of examining intermediate phenotypes (i.e., CSF inflammation), when examining genetic risk to cognitive impairment. While there were no significant differences in sTNFR2 and BAFF expression between the cognitive status groups among carriers of the MCP-1/CCL2 risk allele, we observed significant differences in sTNFR2, and BAFF expression between the cognitive status groups among non-carriers. Our results are generally consistent with knowledge of HIV-associated immunocompromise. High levels of CCL2 expression, which was found among carriers of the CCL2 risk allele, may disrupt CNS immunity over time, attenuating the ability for sTNFR2 and BAFF to exert its neuroprotective effects. Elevated levels of sTNFR2 and BAFF that were found among non-carriers with cognitive compromise in the absence of elevated CCL2 may demonstrate a healthy immune response to HIV in the context of CNS inflammation.

Keywords: MCP-1/CCL2, HIV, Inflammation, Cognition.

Disclosure: Nothing to Disclose.

T172. Circadian Changes of DNA Methylation and Gene Expression in Human Blood

Chunyu Liu*, Jinsong Tang, Haiyan Tang, Hua Yun Chen, Chao Chen, Yiqiao Hu, Xiaogang Chen

University of Illinois at Chicago, Chicago, Illinois

Background: Circadian rhythms are hypothesized to involve in multiple neuropsychiatric disorders, including sleep-related disorders, mood disorders, and addiction. It is well known that some gene expression oscillate in the period of 24 hours daily, as part of the circadian rhythms. But other than the few core circadian genes, much of the circadian system and its epigenetic regulation remains unknown. Genome-wide studies of gene expression have been done in mice and human postmortem brain samples. We have performed the first genome-wide study of circadian gene expression and DNA methylation in human blood in a 24-hour period.

Methods: By sampling blood at 6 time-points in 24 hours from 8 healthy male individuals, we obtained 48 samples of genomic DNA to measure DNA methylation level of 485 K CpG sites. Four of the individuals also provided RNA samples from whole blood. Illumina HT12 chip was used to measure gene expression. After removing batch effects and positional effects, filtering out not-detected and common-SNP-containing probes, we assessed DNA methylation levels at 275,337 CpG sites, and profiled gene expression levels of 15,519 genes.

Results: By fitting to a sine curve with a period length of 24 hours, we detected 408 CpG sites that have epigenome-wide significant (P<1.82e-7, after Bonferroni correction) fit to the circadian pattern of methylation variation. The CpG sites are slightly enriched around genes involving GTPase regulator activity in DAVID GO term analysis. We also detected 66 genes that have genome-wide significant (P<3.22e-6, after Bonferroni correction) fit to the circadian pattern of expression variation. Of the core circadian genes, only PER2 was among the genes with a circadian cycling expression pattern. This is similar to what has been observed in mouse and human brains. Two genes with circadian cycling expression, PLEKHO2 and PRDM1, also have nearby CpG sites showing significant cycling methylation. Their expression and methylation showed modest correlation, as well. However, although the methylation of those CpG sites do significantly fit to 24-hr circadian cycle, their patterns are noisier and less variable than those of gene expression.

Conclusions: These data suggest that DNA methylation may play a modest role in regulating circadian rhythms, and that the regulation of circadian rhythms at the molecular level may extend far beyond the few classic circadian genes.

Keywords: circadian rhythm, gene expression, DNA methylation, blood.

Disclosure: Nothing to Disclose.

T173. More Psychiatric Illness in Parents and Grandparents of US vs. European Patients with Bipolar Disorder: Relationship to Poor Prognosis Factors

Robert Post*, Lori Altshuler, Ralph Kupka, Susan McElroy, Mark Frye, Michael Rowe, Gabriele Leverich, Heinz Grunze, Trisha Suppes, Paul Keck, Willem Nolen

Bipolar Collaborative Network, Bethesda, Maryland

Background: Arguments about the diagnosis of bipolar disorder in children have deflected attention from the high incidence in the US, the reasons for this, and the need for more treatment research for this large, underserved population. Here we examine the impact of genetic/familial loading on variables associated with a poor outcome in the illness.

Methods: 968 outpatients, average age 41, with a diagnosis of bipolar disorder, 676 from 4 sites in the US and 292 from 3 sites in Germany and the Netherlands (Europe) gave informed consent and reported on family history of UP depression, bipolar, anxiety, substance abuse, and alcohol abuse disorders and history of suicide attempt in both parents, 4 grandparents, siblings, and offspring. They additionally supplied information on demographic and course of illness variables previously related to poor prognosis (PPFs).

Results: Parents, grandparents, bipolar probands, and their offspring from the US had a greater incidence of most illnesses and more complexity of psychiatric illness than those from Europe. Parental and grandparental illness was generally related to the PPFs of the US probands, including early onset of bipolar disorder, history of abuse in childhood, anxiety and substance abuse comorbidity, rapid cycling, and 20 or more prior episodes.

Conclusions: Compared to patients from Europe, those from the US reported greater amounts of psychiatric illness over 4 generations. This combined genetic and familial/environmental vulnerability appears related to early onset illness and other serious PPFs associated with difficult to treat illness. While in need of replication, these findings support a need for new public health awareness campaigns, research, and intervention paradigms, particularly in the US, in an effort to mitigate this ongoing devastating medical situation.

Keywords: Childhood Onset Bipolar Disorder, Genetics, Familial Loading, Substance Abuse.

Disclosure: Nothing to Disclose.

T174. Synaptic, Transcriptional, and Chromatin Genes Disrupted in Autism: Findings from 13,000 Exomes

Joseph Buxbaum*, For the Autism Sequencing Consortium

Mount Sinai, New York, New York

Background: The genetic architecture of autism spectrum disorder (ASD) involves the interplay of common and rare variation and their impact on hundreds of genes. During the past two years, large-scale whole-exome sequencing (WES) has proved fruitful in uncovering risk-conferring variation, especially when considering de novo variation, which is sufficiently rare that recurrent de novo mutations in a gene provide strong causal evidence.

Methods: We conducted the largest ASD WES study to date, starting with 16,098 samples from seventeen distinct sample sources and ascertained by diverse designs. Unlike earlier WES studies, we do not rely solely on counting de novo loss-of-function (LoF) variants, rather we use novel statistical methods to assess association for autosomal genes by integrating de novo, inherited and case-control LoF counts, as well as de novo missense variants predicted to be damaging.

Results: Analyses of sequence data in a cleaned sample of 3,871 autism cases and 9,937 ancestry-matched or parent controls implicate 22 autosomal genes at a false discovery rate (FDR)<0.05, and a much broader set of 107 autosomal genes strongly enriched for those likely to affect risk (FDR<0.30). These 107 genes show unusual evolutionary constraint against mutations and map to modules in unbiased networks of early neocortical developmental. Our dataset is enriched with genes targeted by two autism-associated RNA-binding proteins (FMRP and RBFOX), genes found with de novo non-synonymous mutations in schizophrenia, and genes encoding synaptic components. Amongst critical synaptic genes found mutated in our study are voltage-gated ion channels, including those involved in propagation of action potentials (e.g., the Na+channel Nav1.2 encoded by SCN2A), neuronal pacemaking, and excitability-transcription coupling (e.g., the Ca2+channel Cav1.3 encoded by CACNA1D). Our dataset is also enriched for chromatin remodeling genes, including enzymes involved in histone post-translational modifications, especially lysine methylation/demethylation, and regulators that recognize such marks and alter chromatin plasticity such as the emergent ASD gene CHD8.

Conclusions: Our study identifies a group of 107 high-confidence risk genes that incur de novo LoF mutations in over 5% of ASD subjects and expose two tightly intertwined pathways - chromatin remodeling and synaptic development – as major themes in ASD risk. With mutations found in over 5% of individuals, the genes we identified will be important for family counseling, developing model systems, and defining targets for novel therapeutics in ASD.

Keywords: whole exome sequencing, chromatin, alternative splicing, autism.

Disclosure: Nothing to Disclose.

T175. Withdrawn

T176. Drugging the Schizophrenia Genome: A Fast Track Strategy from GWAS to Clinic

Todd Lencz*, Anil Malhotra

Feinstein Institute for Medical Research, Glen Oaks, New York

Background: The Psychiatric Genomics Consortium – Schizophrenia Workgroup has recently published a genomewide association study (GWAS) which established a list of more than 100 replicable genetic loci conveying risk for the disorder (PGC-SCZ, 2014). While representing a watershed breakthrough in psychiatric research, the complexity revealed by the PGC-SCZ results will require considerable further analysis to yield clinically applicable insights. Given the potentially large number of genes implicated the PGC-SCZ, one way to prioritize the search space is to identify those genes that are known to be druggable. Recent estimates suggest that the potential for druggability may extend to 15% of the genome or more; in practice, however, the process of validating these targets with compounds that are able to transition from preclinical models to clinical trials remains a major bottleneck in the process of drug development. Consequently, it has been suggested that more rapid utility from GWAS results may be obtained by the identification of drug/target combinations which have already been validated by preclinical work and have progressed to human clinical trials and/or FDA approval. In the present study, we examined all genes implicated by PGC-SCZ GWAS hits against databases of drug targets.

Methods: Input for all analyses was derived from results of the PGC-SCZ study; a total of 341 protein-coding RefSeq genes as listed in Supplementary Table 3 of that publication were considered. The primary source of confirmed druggable targets was the curated list of 1030 genes recently published by Rask-Anderson et al. (2014). Of the 1030 druggable genes, 555 unique genes were identified which serve as targets of approved pharmaceuticals, which they refer to as “established targets.” An additional 475 genes were targeted by compounds in registered clinical trials, referred to as “novel targets.” To compare the amount of druggable genes in the PGC-SCZ list relative to druggable genes in the whole genome, we performed a binomial test examining the number of PGC-SCZ genes on the Rask-Anderson list of 1030 druggable targets. To characterize all possible approved and in-trial drugs, we supplemented the list of Rask-Anderson et al. (2014), which was based on an April 2013 data freeze, with three additional sources of information: the proprietary CiteLine database, the publicly available DrugBank database and the Drug-Gene Interaction database. Genes were only considered to be true drug targets if direct pharmacologic binding had been demonstrated; indirect effects of up- or down-regulation were not included.

Results: Of the 341 genes on the PGC list, 27 (7.92%) are drug targets on the Rask-Anderson list. This is a 56% enrichment compared to the whole genome, of which 5.06% are known drug targets (exact binomial test, p=.03, 2-tailed). Notably, this enrichment extends across both categories of approved (established) and in-trial (novel) agents. Thus, even under this very specific definition of druggability, genes nominated by the PGC-SCZ GWAS are more likely to be known targets than expected by chance. In the absence of mechanistic understanding of the role of GWAS-derived genetic variants in schizophrenia pathophysiology, this result provides some ground for optimism that one or more of these targets may yield an effective treatment. Next, we sought to characterize drugs targeting genes derived from the PGC-SCZ in terms of FDA status, mechanism of action, and primary indication, in order to identify relative potential for repurposing. We identified 20 genes that are the targets of approved drugs, and prioritized them in terms of likelihood of relevance to schizophrenia, defined by either of the following characteristics: the gene is targeted by a drug with existing neuropsychiatric indications, or it produces a protein that is constitutive of a neurotransmitter receptor, ion channel, or ion transporter. We further prioritized genes that are sole members of the linkage disequilibrium (LD) block surrounding a PGC-SCZ GWAS hit, given that many of the LD blocks in the PGC-SCZ paper contain multiple genes (up to a maximum of 26 genes at a single locus). In addition to DRD2, five genes are noteworthy insofar as they are the sole gene within an LD block and are likely to be relevant to neuropsychiatric pathophysiology: CACNA1C, CACNB2, CACNA1l, GRIN2A, and HCN1. An additional 20 genes without approved indications, but which are the target of drugs that are currently in registered clinical trials, were also identified. Three of these targets demonstrated neuropsychiatric relevance and are lone members of a GWAS-derived LD block. Intriguingly, one gene that meets both of these criteria is GRM3, encoding the metabotropic 3 glutamate receptor. Two other genes targeted by drugs in clinical trials which are notable for relevance and likelihood are ATP2A2 and PRKD1.

Conclusions: Despite the unparalleled success of GWAS as a method for rapidly advancing knowledge, questions remain about the potential clinical utility of GWAS findings. Given the perplexing diversity of findings presented by the PGC-SCZ GWAS, it may seem reasonable to extend similar skepticism to the prospect of translating these results to the clinic. However, our results demonstrate: (1) that these results are enriched for validated targets of both approved and registered pharmaceutical compounds; and (2) that, amongst these targets, there are several examples of “low-hanging fruit” – genes that are plausibly connected to schizophrenia pathophysiology, and are relatively likely to be influenced by the SNPs identified by the PGC-SCZ GWAS. Of course, one of the genes meeting criteria for “low-hanging fruit” is DRD2, which is already well-known in the context of schizophrenia treatment. In the paragraphs below, we will examine each of the other top targets in further detail.

Keywords: schizophrenia, antipsychotic, GWAS, drug target.

Disclosure: Dr. Malhotra serves on the Advisory Board of Genomind and Forum Pharmaceuticals.

T177. Neuregulin-1 Loci Recently Associated with Psychosis Onset are Associated with Increased NRG1 mRNA and Lateral Ventricle Volume

Chad Bousman*, Vanessa Cropley, Suresh Sundram, Avril Pereira, Rhoshel Lenroot, Jason Bruggemann, Elizabeth Scarr, Thomas Weickert, Andrew Zalesky, Ian Everall, Christos Pantelis, Cyndi Shannon Weickert

University of Melbourne, Carlton South, Australia

Background: DNA sequence variation in the 5-prime region of the neuregulin 1 (NRG1) gene has consistently been linked to schizophrenia, particularly among individuals of European ancestry. Recent evidence suggests the schizophrenia associated Icelandic haplotype (HapICE) region may harbor single nucleotide polymorphisms (SNPs) that could assist in differentiating high-risk individuals who will or will not transition to psychosis. Two such SNPs (rs4281084 and rs12155594) were recently shown to independently predict psychosis transition in a large ultra-high risk cohort in Australia and for every additional rs4281084-A allele and/or rs12155594-T allele (allelic load range 0 – 4) the relative risk of psychosis onset increased 1.56 (95% CI=1.20 – 2.04) (Bousman et al., Translational Psychiatry, 2013, 3, e251). Although functional annotation of several SNPs within the HapICE region has been previously undertaken, the functional effects of rs4281084 and rs12155594 are not known. The aim of this study was to utilize online bioinformatics tools as well as post-mortem brain and neuroimaging data from individuals with and without a psychotic disorder to elucidate what effect, if any, these SNPs have on NRG1 function, regulation, gene expression, and brain structure.

Methods: To functionally annotate the NRG1 rs4281084 and rs12155594 loci a three-pronged approach was employed. First, the two SNPs were analyzed in silico via a pipeline that utilized a variety of online bioinformatics resources to provide detailed information of each SNP’s genomic location, linkage disequilibrium (LD) profile, presence of regulatory elements (e.g. transcription factor binding sites, CpG islands), and/or overlap with copy number variations. In addition, using the Sanger Institute’s gene expression variation (Genvar) software each SNP’s association with in vitro gene expression was examined in lymphoblastoid cell lines (LCLs) derived from umbilical cords of 75 Geneva GenCord individuals. Second, we examined postmortem prefrontal cortex from the Victorian Brain Bank Network (BA 9) from 40 (12 non-psychiatric controls, 28 schizophrenia) via DNA and mRNA analysis conducted using the Affymetrix Exon 1.0 ST Array, to determine if the two NRG1 SNP’s or their combined allelic load had cis-regulatory effects on pan-NRG1 gene expression. Finally, structural magnetic resonance imaging (sMRI) and genotype data from 156 non-psychiatric controls and 177 individuals with a psychotic disorder (e.g. schizophrenia) from the Australian Schizophrenia Research Bank were examined to determine if the two SNPs or their combined allelic load were associated with brain volumes in regions of interest (i.e. lateral ventricles, superior frontal, middle frontal, superior temporal, or anterior cingulate) defined based on the Desikan-Killiany atlas.

Results: In silico analysis revealed that rs4281084 is located 207 base pairs upstream of the putatively functional rs6994922 (SNP8NRG243177) SNP. The rs4281084 locus is in strong LD (D’=1.0, r2=0.95-0.02) with four of the five HapICE loci as well as a putative cis-regulatory locus (rs7014762) within the HapICE region. rs121555094 is located between two microsatellites (478B14-848 and 420M9-1395) that are part of the 7-marker HapICE haplotype and sit between the Type II and Type I promoters. The mapped location of rs4281084 sits within a v-maf musculoaponeurotic fibrosarcoma oncogene homolog K (MAFK) transcription factor binding site whereas rs12155594 was not found to overlap with any known regulatory elements. However, three previously reported copy number variants as well as a long interspersed nuclear element 1 (L1) retrotransposon overlap the rs12155594 locus. Analysis of human brain mRNA showed that neither rs4281084 nor rs12155594 were associated with NRG1 gene expression individually, but combined to impact mRNA level. Individuals (both controls and people with schizophrenia) with a greater allelic load were significantly more likely to have higher pan-NRG1 expression (F3,33=3.85, padj=0.018). Similarly, structural MRI analysis revealed no associations between each SNP and brain volumes when examined individually. However, significantly larger right (37%-50%, F2,170=6.95, padj=0.01) and left (33% - 45%, F2,170=5.21, padj=0.03) lateral ventricle volumes were observed for cases with a psychotic disorder who had greater allelic load compared to those with no to low allelic load but no associations were observed in controls.

Conclusions: These results represent the first attempt to functionally annotate and biologically characterize the impact of two recently identified NRG1 loci associated with psychosis onset. They build on a growing body of research supporting the functional importance of genetic variation within the HapICE region of the NRG1 gene. In concordance with previous functional studies of other loci located in the HapICE region, our results suggest that the combined allelic load of rs4281084 and rs12155594 have potential cis-regulatory effects on NRG1 expression and are associated with increases in bilateral ventricle volume. Collectively, these results and those of others suggest that the functional effects conveyed by sequence variation within NRG1 are likely not driven by one SNP but rather, a diverse accumulation of nucleotide changes particularly in upstream regions capable of regulating gene expression.

Keywords: gene association, gene expression, imaging genetics.

Disclosure: Nothing to Disclose.

T178. Galanin-System Genes and 5-HTTLPR Are Differentially Involved in Stress Induced Anxiety and Depression and Interact with Each Other in Anxiety but Not in Lifetime or Current Depression

Gyorgy Bagdy*, Gabor Hullam, Nora Eszlari, Xenia Gonda, Ian M Anderson, Tomas G Hökfelt, J F William Deakin, Peter Antal, Gabriella Juhasz

Semmelweis University, Budapest, Hungary

Background: The gap between the progress in fundamental neuroscience and the slow advance in the treatment of neuropsychiatric disorders has led to the re-evaluation of approaches, with special focus on novel mechanisms that have preliminary clinical validation. Results of a recent study indicated that the galanin neuropeptide and its receptors play a significant role in the pathogenesis of major depression, but this effect is present only in subjects with high exposure to stress (Juhasz G et al., Brain galanin system genes interact with life stresses in depression-related phenotypes. PNAS 111:16 pp. E1666-1673, 2014). Furthermore, chronic pain induced loss of motivation requires galanin receptor-1 triggered depression of excitatory synaptic transmission in indirect pathway nucleus accumbens medium spiny neurons in mice (Schwartz N et al., Decreased motivation during chronic pain requires long-term depression in the nucleus accumbens. Science 345:6196 pp.535-542, 2014). These results are in agreement with the fact that galanin expression is strongly stimulated by stress in animals, and validate the galanin system as a target for antidepressant drug development. The similar gene-stress interaction effects of the galanin system genes (GAL, GALR1, GALR2 and GALR3) and the previously described 5-HTTLPR polymorphism in the SERT gene raised the question as to whether these genetic effects could be separated according to different depression-related phenotypes. Other questions included possible interactions between the galanin system genes and 5-HTTLPR, and interaction of these genetic effects with gender on the different phenotypes.

Methods: We studied a population cohort (NewMood) of>2,300 individuals from Manchester and Budapest. 5-HTTLPR of the serotonin transporter gene and twelve SNPs of the galanin-system genes were genotyped. Phenotypic outcome measures included self-reported lifetime depression, current depression and current anxiety using the Brief Symptoms Inventory. The list of Threatening Life Experiences (TLE) was used to measure life stress-related environmental factors in the last 12 months. Statistical analyses were performed by multivariate system-based Bayesian analysis of relevance in highly TLE exposed persons. Analysis of interactions between the galanin system genes and 5-HTTLPR, and the genes and gender were performed by using the predictor variables jointly.

Results: For lifetime depression the order of relevance for the analysed genes was GALR2>>GALR3>GAL>GALR1>5-HTTLPR with a similar pattern observed for current depression. In contrast, the contribution of genes to anxiety was very different, 5-HTTLPR>>GAL>GALR1>GALR2>GALR3. For lifetime and current depression there were no interactions between the 5-HTTLPR polymorphism and the galanin system genes. For current anxiety, epistasis between 5-HTTLPR-GALR1 and GAL-GALR1 were found. In addition, interactions with gender were found for GAL and GALR3 in depression phenotypes, and for GAL and 5-HTTLPR in anxiety. After omitting TLE from the model none of these genes were relevant for depression-related phenotypes.

Conclusions: These results show that in subjects exposed to high life stress GAL system gene variations and combinations are associated predominantly with depression, while 5-HTTLPR s allele is predominantly associated with anxiety. In addition, interactions between galanin system genes and 5-HTTLPR are present for anxiety but not for lifetime or current depression. Our results underline the importance of assessing gene-environment interactions via different depression-related phenotypes. Furthermore, non-conventional methods of analysis will help us to better understand disease aetiology and to identify putative drug targets for anxiety and depression.

Keywords: galanin, life stress, depression, anxiety.

Disclosure: The study was supported by the Sixth Framework Program of the European Union, NewMood, LSHM-CT-2004-503474, by the National Institute for Health Research Manchester Biomedical Research Centre, and by the TAMOP-4.2.1.B-09/1/KMR-2010-0001, Hungary and the Hungarian Academy of Sciences (MTA-SE Neuropsychopharmacology and Neurochemistry Research Group), by the Hungarian Brain Research Program - Grant No. KTIA_13_NAP-A-II/14 and National Development Agency (KTIA_NAP_13-1-2013-0001); the Swedish Research Council, Karolinska Institutet, an Unrestricted Bristol-Myers Squibb Neuroscience Grant, a NARSAD Distinguished Investigator Award, the Marianne and Marcus Wallenberg Foundation, the Knut and Alice Wallenberg Foundation. Gonda is recipient of the Bolyai Scholarship of the Hungarian Academy of Sciences. The sponsors funded the work but had no further role in the design or conduct of the study or in the preparation, review, or approval of the manuscript. Conflict of interest: Drs P.A., G.B., X.G., G.H., G.J., and N.E. report no conflict of interest. J.F.W.D. variously performed consultancy, speaking engagements, and research for Bristol-Myers Squibb, AstraZeneca, Eli Lilly, Schering Plough, Janssen-Cilag, and Servier (all fees are paid to the University of Manchester to reimburse them for the time taken); he has share options in P1vital. I.M.A. received an honorarium for speaking from Lundbeck and grant support from Servier and AstraZeneca. T.H. consulted with Synaptic and received an honorarium (2005).

T179. Abnormal X Chromosome Inactivation in Females with Major Psychiatric Disorders

Baohu Ji, John Kelsoe, Xianjin Zhou*

University of California at San Diego, La Jolla, California

Background: Mania, major depression and schizophrenia are severe brain disorders. No biological hallmark has been identified for any of these disorders. Our previous studies found that the DISC1-Boymaw fusion gene from the Scottish family inhibits protein translation and thereby may contribute to pathogenesis of major psychiatric disorders. Excessive protein translation has been suggested as core pathology for Fragile X Syndrome and autism. We speculate that abnormal protein translation may contribute to a wide range of mental disorders not only in rare families, but also in the general population of patients. To investigate our hypothesis, we analyzed protein translation activity in the lymphoblastoid cells of psychiatric patients. Indeed, we observed abnormally high or low protein translation activities in the patients’ cells. Surprisingly, all of the abnormal protein translation activities come from female patients. These findings prompted us to investigate functions of the X chromosome in the female patients.

Methods: Patients with either mania and psychosis or recurrent major depression were randomly collected from the general population. All patients and healthy controls are European Caucasians with age matched. Protein translation activities and RNA expression were measured in the lymphoblastoid cells. Postmortem brain RNA samples were provided by Stanley Medical Research Institute. RNA expression analysis was conducted using Q-RTPCR. β-actin RNA expression was used as an internal control for normalization.

Results: Since we found that all variation of protein translation activities in the lymphoblastoid cells comes from female patients, we examined whether X chromosome inactivation (XCI) may be dysregulated in female patients’ lymphoblastoid cells. Indeed, expression of XIST, KDM5C, and some other genes on X chromosome is significantly higher in the lymphoblastoid cells of female patients, with either mania and psychosis or recurrent major depression, than in those of female healthy controls. Abnormal XCI presents in a majority of the female patients. Preliminary studies also suggest an increased XIST expression in the postmortem brains of female patients with schizophrenia, bipolar disorder, and major depression. We suggest that XIST and KDM5C expression can potentially be used as a diagnostic hallmark for women with major psychiatric disorders in the general population.

Conclusions: XCI is a fundamental biological process that is determined during early embryogenesis in all differentiating cells of females to ensure the same gene dosage from X chromosome between males and females. XIST is the master gene in the regulation of XCI. XCI is well preserved in human female lymphoblastoid cells, which provides an excellent cell model to study the roles of XCI in human diseases. Abnormal XCI in the female patients causes an increase of gene dosage from some genes on X chromosome, which may contribute to development of psychiatric disorders since functional disomy of partial X chromosome have been suggested to cause mental retardation and other developmental abnormalities. Mutation of KDM5C gene has been reported to cause X-linked syndromic mental retardation. We propose that abnormal XCI could play a causal role in development of major psychiatric disorders in females. Correction of abnormal X chromosome inactivation may prevent and/or cure major psychiatric disorders in female patients in future.

Keywords: X chromosome inactivation, major psychiatric disorders, XIST, KDM5C.

Disclosure: Nothing to Disclose.

T180. Variation at the COMT val158met SNP Moderates Aripiprazole Effects on Drinking and Alcohol Cue-elicited Activation of the Orbitofrontal Cortex

Joseph Schacht*, Patrick Randall, Konstantin Voronin, Raymond Anton

Medical University of South Carolina, Charleston, South Carolina

Background: Aripiprazole has shown limited efficacy in reducing drinking among individuals with alcohol dependence (e.g., Anton et al., 2008, J Clin Psychopharmacol). Genetic heterogeneity may account for some previous null findings. Aripiprazole is a dopamine partial agonist, and acts to stabilize dopamine tone in the striatum and prefrontal cortex (PFC). Dopamine clearance in the PFC is achieved primarily via degradation by the enzyme catechol-O-methyltransferase (COMT). A well-characterized single nucleotide polymorphism (SNP) in COMT, val158met (rs4680), engenders a functional change in COMT activity, such that the met allele of this SNP has been associated with a three- to four-fold reduction in COMT activity, and thereby greater synaptic accumulation of dopamine in the PFC. Variation at this SNP has been reported to moderate the effects of a variety of dopaminergic medications. The current study tested the interactive effects of aripiprazole and COMT val158met variation on alcohol cue-elicited brain activation and drinking in a bar-lab paradigm.

Methods: Sixty non-treatment-seeking individuals with alcohol dependence (mean age=27, 75% male, nine drinks/drinking day, with 81% heavy drinking days in the past month) were randomized to either aripiprazole (titrated to 15 mg) or placebo for eight days. Subjects were genotyped for the COMT val158met SNP (genotype frequencies: val/val, 0.33; val/met, 0.45; met/met, 0.22) and completed an alcohol cue reactivity fMRI task on day seven and an alcohol self-administration task (bar lab) on day eight. In the bar lab, subjects were administered a standardized priming drink and were subsequently offered the choice to consume up to eight additional drinks.

Results: There were no main effects of medication or COMT val158met genotype on alcohol cue-elicited activation, but these factors interacted in their effects on cue-elicited activation of the right and medial orbitofrontal cortex (OFC), such that aripiprazole, relative to placebo, reduced OFC activation among met-allele homozygotes, but increased it among val-allele homozygotes (whole-brain results thresholded at z<1.96, p<.05, corrected for the false discovery rate). The interaction between medication and genotype on bar-lab drinking was also significant (F(2, 53)=3.35, p<.05), such that aripiprazole, relative to placebo, reduced drinking in the bar lab among met-allele homozygotes (F(1, 53)=8.98, p=.004), but was not significantly differently from placebo among heterozygotes or val-allele homozygotes. Cue-elicited activation of the OFC region for which the medication by genotype interaction was significant was positively associated with the number of drinks subjects consumed in the bar lab (r(60)=.26, p<.05), and activation of this region statistically mediated the interacting effects of medication and genotype on bar-lab drinking (i.e., inclusion of OFC activation in this model reduced the value of the interaction statistic below the threshold for significance).

Conclusions: Collectively, these data suggest a novel pharmacogenetic interaction between variation at the COMT val158met SNP and aripiprazole response in alcohol dependence, such that individuals who carry the allele of this SNP associated with lower enzyme activity (and potentially greater synaptic dopamine accumulation in the PFC) displayed reduced drinking with aripiprazole, presumably due to dopamine receptor blockade or stabilization. This effect was mediated by a reduction in alcohol cue-elicited activation of the OFC, a prefrontal region with reciprocal dopaminergic projections from the midbrain and striatum that allows drug-related cues to be associated with the effects of the drug and whose disruption results in compulsive, stereotyped behavior. Further exploration of whether COMT genetic variation may predict aripiprazole effects among individuals with alcohol dependence is warranted.

Keywords: atypical antipsychotic, catecholamines, functional neuroimaging, genetics.

Disclosure: Part I: Dr. Anton is a member of the Alcohol Clinical Trials Initiative (ACTIVE) group, which is supported by Lilly, Lundbeck, Pfizer, Abbvie, and Ethypharm. He has received grant funding from Lilly, and serves as a consultant for Lundbeck and Biotie. He also serves on Lundbeck's Scientific Advisory Board. He does not own stock in any of these companies, nor in Bristol Myers Squibb or Otsuka, the manufacturers of aripiprazole. Part II: Dr. Anton's grant funding from Lilly is greater than $10,000 per year. Part III: Nothing to disclose. Part IV: Dr. Anton has received grant funding from Lilly. Part V: Nothing to disclose.

T181. Functional Genomic Characterization of the Schizophrenia Risk SNP rs4523957 Implicating Serine Racemase

Rebecca Birnbaum*, Fengyu Zhang, Enrico D'Ambrosio, Venkata Mattay, Qiang Chen, Joo Heon Shin, Joel Kleinman, Thomas Hyde, Daniel Weinberger

Lieber Institute for Brain Development, Baltimore, Maryland

Background: Converging evidence supports the hypothesis that N-methyl-D-aspartate (NMDA)-type glutamate receptor hypofunction is involved in the pathophysiology of schizophrenia. D-serine, a co-agonist required for full activation of the NMDA receptor, is synthesized by serine racemase. Recently a SNP, rs4523957, located in an intron of serine racemase (SRR), was reported to be associated with schizophrenia in the PGC meta-analysis of 34,241 cases and 45,604 controls (Combined OR=1.07, Discovery cohort p-value=2.86x10-10) (PGC Nature 2014). The current analysis investigated the main effect of this schizophrenia risk associated SNP (and variants in linkage disequilibrium) on SRR abundance by RNA sequencing methodology. We also investigated the effect of the risk SNP (and variants in linkage disequilibrium) on neuroimaging and cognitive intermediate phenotypes.

Methods: RNA sequencing was performed on post-mortem PFC grey matter for 101 Caucasian controls (27 females, 38+/- 23 years, 74 males, 36+/-18 years) and 64 schizophrenia cases of similar ethnicity (13 females, 53+/- 18 years, 51 males, 42+/- 13 years). Samples were enriched with PCR to create a cDNA library for high throughput sequencing. Pair-end reads of cDNA sequences were aligned to the human genome reference (TopHat v2.0.4). We tested the association of SRR rs4523957 (and variants in LD) with SRR abundance using linear regression, adjusting for age, sex, and RNA Integrity Number (RIN). We further examined the effect of SRR rs4523957 (and variants in LD) on 6 factors of cognitive performance and the higher-order factor, “g” (n=500 Caucasian controls and n=378 Caucasian schizophrenia patients), using linear regression, adjusting for age and gender. We also tested the genotype effects on fMRI-based intermediate phenotypes, related to working memory, response inhibition, and episodic memory.

Results: There was no significant difference in SRR abundance between Caucasian schizophrenia cases (n=52) and controls (n=68) (Diagnosis Effect p-value=0.53). The rs4523957 risk associated allele was associated with decreased SRR abundance in Caucasian schizophrenia DLPFC (p-value=5.2x10-4). SNPs in LD with rs4523957 showed highly significant association with SRR abundance in Caucasian control DLPFC (rs3744270 p-value=1.53 × 10-10 and rs3760229 p-value=9.88x10-4). The putative 3-SNP haplotype containing the schizophrenia associated SNP and 2 SNPs in LD--rs3760229(G), rs3744270(A), and rs4523957(T)--was associated with decreased SRR abundance in Caucasian control DLPFC (p-value=5.71x10-8). In Caucasian controls (n=259), minor allele carriers for the SNP rs3744270 demonstrated an inefficient prefrontal cortical response during working memory compared to major allele homozygotes, showing greater activation in the right DLPFC (ROI encompassing right BA9 and BA46: Montreal Neurological Institute [MNI] coordinates x=54, y=0, z=33; k=51; family-wise error-corrected, p=0.05,). There was no significant genotype effect on cognition outcome for the factors tested nor on “g” in Caucasian controls or in patients with schizophrenia (rs4523957 genotype effect on “g”: n=482 Caucasian control (p=0.63), n=372 Caucasian schizophrenia (p=0.41)).

Conclusions: The current results of decreased SRR associated with a schizophrenia risk allele and with variants in LD with a schizophrenia risk allele might be viewed as consistent with previous reports of decreased D-serine concentrations associated with schizophrenia, though we did not find an illness associated reduction in SRR expression. The cis-eQTL effect suggests a putative molecular mechanism of genetic risk. Further, prefrontal cortical inefficiency in the DLPFC region during working memory implicates a neurobiological correlate for the mechanism of risk. Association analyses using SRR haplotypes rather than single risk variants are ongoing.

Keywords: Serine Racemase, Schizophrenia, RNA Sequencing.

Disclosure: Nothing to Disclose.

T182. A Genome-wide Analysis with Suicidal Behavior Severity in Bipolar Disorder

Clement Zai*, Vanessa Goncalves, Arun Tiwari, Sarah Gagliano, Georgina Hosang, Vincenzo de Luca, Sajid Shaikh, Nicole King, Qian Chen, Wei Xu, John Strauss, Gerome Breen, Cathryn Lewis, Anne Farmer, Peter McGuffin, Jo Knight, John Vincent, James Kennedy

Centre for Addiction and Mental Health, Toronto, Canada

Background: Suicide claims one million lives worldwide annually, making it a prominent public health concern. Twin and family studies support a genetic component to the risk of suicidal behavior. A number of candidate genes have been implicated in suicidal behaviour; these include the serotonin transporter (SLC6A4) and the brain-derived neurotrophic factor (BDNF) genes. More recently, genome-wide association studies (GWASs) of suicide attempt on large bipolar disorder (BD) patient samples have identified a number of potential candidate genes. GWASs of suicide behavior severity, from suicidal ideation to serious suicide attempt, have not yet been reported for BD.

Methods: We conducted a GWAS of suicide behaviour severity in three independent BD samples. We performed quantitative analyses on the the suicidality scores (from the Schedule for Clinical Assessment in Neuropsychiatry SCAN), or the suicide specifier scores (from Structured Clinical Interview for DSM-IV SCID). We conducted the analyses of suicide behavior severity in the three samples separately, and obtained an overall significance through a meta-analysis.

Results: We did not find genome-wide significant association of any tested markers in the three BD samples, but we found a number of suggestive associations on chromosomes 8 and 10 (p<1e-5). We are currently performing pathway analysis of the GWAS data.

Conclusions: Our GWAS findings suggest that likely many gene variants of small effects contribute jointly to the severity for suicidal behavior in BD. Larger independent replications are needed to confirm the findings from the GWAS presented here.

Keywords: suicide behavior severity, genome-wide association study, pathway analysis, bipolar disorder.

Disclosure: JLK: honoraria from Roche, Novartis, and Lilly. CCZ: honorium from WebMD for Medscape review. JLK & CCZ: patent application “Genetic Markers Associated with Suicide Risk and Methods of Use Thereof” submitted.

T183. In Vivo Quantitation of MicroRNAs Using MiRNA-seq in Cerebrospinal Fluid of Patients with Schizophrenia

Juan Gallego*, Kendal Van Keuren-Jensen, Harjasleen Yadav, Christopher Morell, Todd Lencz, Anil Malhotra

Hofstra North Shore-LIJ School of Medicine, Glen Oaks, New York

Background: Recent studies have linked alterations in microRNA (MiRNA) expression to schizophrenia and other psychiatric disorders. However, most of these studies have used post-mortem brain tissue or whole blood as the source of transcript. By contrast, examination of microRNAs in cerebrospinal fluid (CSF) might provide an in vivo biomarker more directly reflecting functional changes in the brain. MicroRNA sequencing (MiRNA-seq) is a technique being increasingly used to comprehensively assess miRNA expression in various tissues in patients with cancer, multiple sclerosis, HIV, etc. However, to date, there are no published papers that have studied miRNA expression using miRNA-seq in CSF of patients with schizophrenia.

Methods: Ten patients with schizophrenia-spectrum disorders and ten healthy volunteers matched to patients in age, sex, and race underwent a lumbar puncture and a blood draw. 15-25 cc of CSF and 5-10 cc of blood were obtained from each subject. RNA was extracted from CSF using miRVana PARIS kit (Invitrogen). An Illumina TruSeq Small RNA sequencing kit was used for library preparation and sequencing was done on an Illumina HiSeq2000 machine. miRNAs values were normalized using the geometric mean of all miRNAs. Differential expression analyses were conducted using DESeq2. CSF samples with less than 100.000 mapped read counts were excluded from the analysis.

Results: The mean age was 43 years (SD=8.1) in patients and 40 years (SD=8.8) in controls. Both groups had the same proportion of males (90%) and white subjects (40%). Mean total BPRS score in patients was 30.4 (SD=5.8). One out of 20 samples was excluded due to mapped read counts less than 100.000. Mean mapped read counts for the remaining 19 samples was 893, 436. No miRNAs were significantly different between patients and controls after correcting for multiple testing, likely due to the small sample size of our study. However, some miRNAs showed meaningful differences in fold change such as miR-146a-5p (FC: 0.5, p=0.17); miR-219a-5p (FC: 1.7, p=0.17), miR-423-5p (FC: 1.5, p=0.17), miR-4732-3p (FC: 1.7, p=0.17), miR-524-3p (FC: -1.0, p=0.17), and miR-132 (FC: -0.8, p=0.63). To investigate further these findings, qPCR analysis are being conducted at this time but results are not yet available.

Conclusions: Even though there were no statistically significant differences in miRNA expression between patients and controls, likely due to the small sample size, a number of miRNAs were meaningfully dysregulated based on fold change values. Therefore, the investigation of these miRNAs in CSF may help establish an illness-specific miRNA signature that could lead to a better diagnosis and treatment.

Keywords: microRNA, cerebrospinal fluid, sequencing, schizophrenia.

Disclosure: Nothing to Disclose.

T184. Age–associated Changes in Expression of GRM3 and Splice Variants in Human Prefrontal Cortex Are Related to Novel Antisense Transcripts: Relevance to Schizophrenia

Elisabetta Buonaguro*, Gianluca Ursini, Joo Heon Shin, Andrew E. Jaffe, Yankai Jia, Thomas M. Hyde, Joel E. Kleinman, Daniel R. Weinberger

Lieber Institute for Brain Development, Baltimore, Maryland

Background: Positive associations of polymorphisms in the metabotropic glutamate receptor subtype 3 (mGluR3) encoding gene GRM3 and psychosis have been reported for over ten years, including with prefrontal activity and cognitive performances in healthy subjects as well as with schizophrenia (Egan et al., PNAS 2004). The current incarnation of the PGC dataset also found GWAS significant association of GRM3 and schizophrenia (Schizophrenia Working Group of the PGC, Nature 2014). Alternatively spliced forms of GRM3 have been identified (Sartorius et al., J.Neurochem 2006), and altered splicing of the gene may be one of the mechanisms by which GRM3 contributes to schizophrenia. GRM3 is closely flanked by an uncharacterized antisense transcript (hg19:chr7:85103669-86274156) partially overlapping the first known exon of the gene, while a second antisense transcript connects a region located between exon 3-4 of GRM3 with exon 2 of KIAA1324L. The first aim of this study was to determine whether the expression of GRM3 and its alternatively spliced isoforms (GRM3delta4, GRM3delta2, GRM3delta2delta3) are related to the expression of the antisense transcripts in healthy subjects and in schizophrenia individuals. We then investigated whether the expression of the antisense and GRM3 splice variants are affected by selected risk associated genotypes. We investigated SNPs, namely rs12704290 and rs11982256 located in GRM3 and in DMTF1 respectively, recently reported to be associated with schizophrenia (Schizophrenia Working Group of the PGC, Nature 2014).We also investigated the expression of the genes KIAA1324L and DMTF1, since rs12704290 and rs11982256 map to LD blocks that include their genomic region and because of the antisense transcript bridging KIAA1324L and GRM3.

Methods: RNA sequencing data were available from the Brain Tissue Collection of the Lieber Institute for Brain Development for 125 Caucasian and African American individuals that met DSM-IV criteria for schizophrenia (83 males,age>17) and for 301 controls (203 males,fetus– 85years) from postmortem dorsolateral prefrontal cortex (DLPFC). Analyses of covariance were conducted with age, sex and RNA integrity number as basic covariates for gene expression and genotype associations. Pearson’s linear correlation coefficient was used to analyze the significance and the direction of the association between gene expression and age, and between the expression of the antisense transcript and GRM3 gene and splice variants. All data processing was performed using R statistical language.

Results: The antisense transcript showed a statistically significant correlation with age in healthy subjects (r=-0.24, P<0.0001). GRM3 expression significantly decreased with age as well (r=-0.25, P<0.0001). Furthermore, the expression of GRM3 and the antisense transcript were strongly positively correlated across the lifespan (r=0.61, P<0.0001). Similar results were obtained when considering the correlation between the three isoforms of GRM3 and the antisense transcript in the same healthy population. There were no race effects on expression. Case (125 subjects, age>17 years)–control (185 subjects, age>17 years) differences were found in the expression of the antisense transcript (P=0.01) and for the GRM3 splice variant GRM3delta4 (P=0.05). The correlation between the antisense and GRM3 gene and splice variants was statistically significant in both schizophrenia patients and healthy controls, with the latter being especially strongly associated (P<0.0001).No statistical association of the SNPs considered with the antisense and GRM3 splicing isoforms were found. DLPFC expression of KIAA1324L and DMTF1 showed no differences between schizophrenia and control subjects. No genotype association was found with the SNPs considered in this study. Intriguingly, both KIAA1324L and DMTF1 expression was positively correlated with GRM3 expression (P<0.0001).

Conclusions: Genetic and molecular mechanisms of glutamatergic dysfunction are believed to be key players in schizophrenia pathophysiology and in the emergence of the associated cognitive impairment. The main results of this study were the findings of a statistical significant correlation with age of both the antisense transcript and GRM3 expression, a positive correlation of their reciprocal expression, as well as reductions in expression of the antisense and of GRM3delta4. Expression of the antisense transcript correlated also with the expression of the three known GRM3 isoforms. Among the molecular mechanisms that could impact gene expression, antisense transcription is increasingly being found to have a specific spatial and temporal pattern and is considered an important regulator of gene expression. Further studies are required to identify the molecular mechanism of the genetic risk association at the clinical level and to explore the functional correlates of the molecular pathwaysthese expression results begin to outline, since mGluR3 is currently considered a treatment target for schizophrenia.

Keywords: GRM3, schizophrenia, psychosis, antisense transcription.

Disclosure: Nothing to Disclose.

T185. The Expression and Secretion of miR-137 in Human iPS Cell-derived Neurons

John Ryder, Kwi-Hye Kim, David Chen, Kalpana Merchant*, Hong Wang

TransThera Consulting, Zionsville, Indiana

Background: Schizophrenia is a debilitating mental disorder that affects about 1% of the population. It is believed to be a neurodevelopmental disorder with genetic etiology. Recently, the largest genome-wide association (GWA) study on over 40,000 individuals identified a strong association between schizophrenia and a common intronic variation in the gene encoding microRNA-137, rs1625579 [1]. Additionally, variants in five of the miR-137 target genes, ZNF804A, CSMD1, C10orf26, TCF4 and CACNA1c, have also shown association to schizophrenia in GWA studies [1,2], suggesting that miR-137 and its downstream targets may represent an import module for the genetic etiology of schizophrenia. Micro RNA-137 is a neuron-enriched microRNA that has been implicated to regulate neurodevelopment and adult neurogenesis in mouse model system [3-6]. However, the role of miR-137 and its targets during human neurodevelopment remains unknown. The objective of the present set of studies is to characterize: (a) the expression of miR-137 in different brain regions (b) the expression and/or secretion of miR-137 and its targets in human induced pluripotent stem cell (iPSC)-derived neurons and (c) effects of over-expression miR-137 on its candidate target genes in human iPSC-derived neurons.

Methods: (1) RNA was extracted from different brain regions of adult C57/Bl6 mice and the expression levels of miR-137 and its target genes were examined by RT-PCR. (2) Human iPSCs were differentiated to neural progenitors and further into neurons. At different time points, cells and media were collected for RNA extraction. The expression levels of miR-137 and its targets were examined. (3) miR-137 was over-expressed in human iPS cell-derived neural progenitors, and the expression of potential miR137 target genes was examined.

Results: We found that miR-137 and its target genes are expressed abundantly in mouse brains, with highest expression in the hippocampus and prefrontal cortex. Upon differentiation of neural progenitors derived from human iPS cells into neurons, miR-137 level increased with time by more than 10 fold in 4 weeks. Although studies of the effect of miR-137 over-expression on schizophrenia GWA informed genes referenced above showed no changes in expression, additional studies on new target genes are underway.

Conclusions: The regionally specific expression of miR-137 in the adult mouse brain is consistent with the recently published human miR-137 expression data [7]. Additionally, we observed a pattern of miR-137 expression in human iPS-derived neurons that is indicative of its role in differentiation and maturation of neurons. Given the strong GWA findings implicating the role of miR-137 and its targets in the susceptibility to schizophrenia, the human iPS-derived neurons provide a model system to understand the role of miR-137 in neurogenesis, differentiation and maturation. This could be addressed through functional consequences of miR-137 over-expression on cellular phenotypes as well as the rescue of the same via putative target genes. References: 1. Sullivan et al., Nat Rev Genet.13:537, 2012 2. Kwon etal., Mol Psychiatry. 18:11, 2013 3. Sun et al., Nat commun. 2:529, 2011 4. Silber et al., BMC Med, 6:14, 2008 5. Szulwach et al., J Cell Biol., 189:127, 2010 6. Smart et al., Stem Cells, 28: 1060, 2010 7. Whalley et al., Neuropsychopharmacol, 37:2720, 2012.

Keywords: schizophrenia, genetics, mRNA, induced pluripotent stem cells.

Disclosure: The authors are current or former employees of Eli Lilly and Company.

T186. GWAS Derived Polygenic Risk Score is Associated with Schizophrenia only in Individuals Exposed to Obstetric Complications

Gianluca Ursini*, Stefano Marenco, Qiang Chen, Richard E. Straub, Giovanna Punzi, Daniel R. Weinberger

Lieber Institute for Brain Development, Baltimore, Maryland

Background: Genome-wide association studies (GWAS) of schizophrenia suggest that genetic risk is conferred by a large number of small effect alleles across the genome (1). To date, GWAS have not provided well-defined mechanistic hypotheses for follow-up experiments (1). Environmental factors also have a role in the pathophysiology of schizophrenia, and obstetric complications and intrauterine adversity (OCs) slightly but significantly increase risk for adult emergence of this disorder (2,3). It is reasonable to assume that such environmental factors can disrupt the normal programs of early human development and have individually varying clinical effects depending on other modifying and protecting factors, including genetic background. On the other hand, genetic risk factors could have different outcomes depending on the specific environmental context where they act. Preliminary evidence of interactions of genes and OCs has been reported (4). Here, we test whether risk profile scores (RPSs) constructed from alleles showing association with schizophrenia in recent GWAS (1) interact with OCs exposure in predicting case-control status.

Methods: We analyzed the interaction between RPSs and OCs exposure in a sample of 272 healthy subjects and 228 patients with schizophrenia from the CBDB/Lieber GWAS study (all adult, Caucasians) on whom we had both GWAS genotypes and obstetrical histories. Diagnoses were made independently by two psychiatrists/psychologists using the structured clinical interview for DSM IV Axis I disorders. All subjects were genotyped using the Illumina HumanHap550K/610Quad Bead Chips (San Diego, California). RPSs were generated as described elsewhere, using odds ratios derived from the PGC 2 datasets excluding the CBDB/LIBD dataset (1). We used different GWAS p value thresholds for selecting risk alleles (from P<0.05 to 5E-8). OCs questionnaires were completed by mothers of affected individuals and of control subjects, and were scored using the McNeil-Sjostrom Scale. Pregnancy, delivery and neonatal complications were included. We assigned to each subject an OCs score of 0/1 based on absence/presence of at least one serious OC (absence OC: severity score ≤ 3; presence OC: severity score ≥4). Statistical analyses were performed in ‘R’, using regression models, with case-control status as dependent variable, and i) RPS, ii) OCs score, iii) RPS, OC score and their interaction as predictors.

Results: We first analyzed whether RPSs predict case-control status without taking into account exposure to OCs, and as expected, we found that all the RPSs, generated using different threshold for selecting risk alleles, predict case-control status (all p<3.65e-06). OCs exposure alone did not predict case-control status. Strikingly, however, analysis of the interaction between OC exposure and the RPS obtained with the set of SNPs showing GWAS significant association with schizophrenia (p<5E-08) show that OC exposure predicts case-control status (p=0.04), while RPS does not (p>0.34); moreover OCs and RPSs significantly interact in predict case-control status (p<0.01), so that only in presence of OCs exposure is the RPS associated with schizophrenia. We did not find significant interactions (all p>0.08) between OCs and RPS generated using less restrictive thresholds.

Conclusions: Our data suggest that the RPS obtained from SNPs showing GWAS significant association with schizophrenia interact with OCs exposure in affecting risk for schizophrenia. More specifically, the RPS obtained from these SNPs predicts case-control status in our sample only in the presence of serious OCs exposure. More dramatically, our data raise the inconvenient possibility that the weak effect sizes of these SNPs, even at the GWAS level of significance, is because they only increase risk in the context of other developmental risk factors, which are not universal among patients in these large GWAS studies. The association of OC exposure with case-control status, when covarying for RPS, is consistent with other findings showing that the risk associated with OCs exposure is dependent on genetic background (2,4). Our results require replication in other samples. References: 1. Psychiatric Genomics Consortium. Nature. 2014 Jul 24;511(7510):421-7. 2. Cannon M et al. Am J Psychiatry 2002; 159:1080–92. 3. Schmidt-Kastner R, et al. Mol Psychiatry 2012; 17: 1194-205. 4. Nicodemus KK et al. Mol Psychiatry 2008; 13: 873-7.

Keywords: schizophrenia, obstetric complications, GWAS, polygenic risk score.

Disclosure: Nothing to Disclose.

T187. Whole Transcriptome Expression in Selected Layers of Orbitofrontal Cortex in Women with Major Depressive Disorder

Craig Stockmeier*, Gouri Mahajan, Nicholas Devitt, Thiru Ramaraj, Faye Schilkey, Boris Umylny, James Overholser, George Jurjus, Lesa Dieter, Grazyna Rajkowska, M. Somair Riaz

University of Mississippi Medical Center, Jackson, Mississippi

Background: In the US, the lifetime prevalence of Major Depressive Disorder (MDD) is 17 percent with women having 2X the prevalence of men. In MDD, we reported neuropathology in layers 3 and 5 of the orbitofrontal cortex (ORB) and altered gene expression and gender-specific changes in protein expression in prefrontal cortex (Rajkowska et al., 1999; Szewczyk et al,. 2009; Kang et al., 2012). To explore the underlying pathology in ORB in women with MDD, we studied whole transcriptome expression using of cytoarchitectonically well-characterized postmortem tissues from subjects that were rigorously evaluated for depression (and antidepressant drug free), laser capture microdissection, evidence-based hypothesis testing, the use of Smart-Seq for linear amplification and then whole transcriptome sequencing followed by a bioinformatics pipeline for analysis.

Methods: The study was performed according to the Declaration of Helsinki and with a protocol approved by the University Institutional Review Board. Informed written consent was obtained from the next-of-kin for all subjects for tissue donation and for administering the Structured Clinical Interview for DSM-IV. Postmortem ORB was selected from five women meeting criteria for MDD (DSM-IV) and five normal control subjects. No antidepressant medications were detected in blood. Cryostat sections (10 μm) were collected and rapidly dehydrated and cell bodies were stained using the HistoGene LCM Frozen Section Staining Kit (Life Technologies). Laser capture microdissection (LCM; Arcturus) was used to capture all tissues in layers 3 and 5 of ORB, followed by rapid RNA extraction and 2 rounds of linear amplification of the transcriptome (Smart-Seq, Eurofins Genomics). Library-prep, adaptor ligation and paired-end RNA sequencing was performed with Illumina HiSeq 2000 (NM INBRE/NCGR), followed by high-throughput bioinformatics using Bioinformatics-in-a-Box™.

Results: We demonstrate the combined use of rapid cell staining with LCM capture of neocortical layers of human brain tissue with amplification and RNAseq. A total of 614 unique transcripts were only expressed in layers 3 and 5 of orbitofrontal cortex in female subjects with MDD. We found that various non-protein coding RNA species (including miR, SnoR, scaR, siR, lncRNA & circular RNA) comprise roughly 40 percent of transcripts differentially-expressed in MDD. In the remaining 60 percent of transcripts (that code for protein), the unique expression of olfactory and taste receptor transcripts in ORB pyramidal cell layers in depression is a novel finding. Functional analysis of MDD-specific genes via Gene Ontology tools showed that the (epigenetic) chromatin-modification process is a key altered biological process. Two novel protein-coding genes (Gamma-C Crystallin and Involucrin) and their downstream effectors were expressed only in MDD.

Conclusions: We observed pathology of RNA species in women with MDD, with the largest contribution by various types of non-coding RNAs. Neocortical olfactory and taste receptors (GPCRs) may serve as novel therapeutic targets for treatment of MDD. Existing Gene Ontology is insufficiently annotated, since none of the 614 transcripts have previously been associated with the Depressive phenotype. Furthermore, none of the highly over-expressed transcripts have been annotated for orbitofrontal cortex (or frontal lobe) anatomical localization on Gene Ontology. 5P20RR016480-12; 8P20GM103451-12; P30GM103328.

Keywords: depression, women, deep sequencing, orbitofrontal.

Disclosure: Nothing to Disclose.

T188. Pharmacogenetic Associations of Antipsychotic Drug Induced Weight Gain: A Systematic Review and Meta-Analysis

Jianping Zhang*, Todd Lencz, Delbert Robinson, Wolfgang Fleischhacker, Rene Kahn, Roel Ophoff, John Kane, Anil Malhotra, Christoph Correll

Zucker Hillside Hospital, Glen Oaks, New York

Background: Weight gain is a common and serious side effect of antipsychotic drugs (APD). Individual differences in APD-induced weight gain are large and may be explained by genetic variations. Pharmacogenetic studies have investigated many genes in association with weight gain in patients undergoing antipsychotic treatment. The present study aimed to systematically examine the literature of pharmacogenetics of APD-induced weight gain.

Methods: Meta-analysis was conducted on associations of certain single nucleotide polymorphisms (SNP) with APD-induced weight gain. To be included, a study needed to report longitudinal data on body weight or Body Mass Index (BMI) change in patients taking APDs and examined their association with SNPs that were also reported by at least one other study. In addition to published studies, three more cohorts were included in the meta-analysis: (1) a pediatric sample of 222 patients who started various APDs for the first time and were followed for 3 months; (2) a first episode schizophrenia sample of 81 patients who took either risperidone or olanzapine for 4 months; and (3) another first episode schizophrenia sample of 150 patients who were treated with APDs and had weight change data at 3 months follow-up. Primary outcome was change in body weight or BMI from baseline to follow-up, and the effect size measure was computed as Hedges’ g. Secondary outcome was percentage of patients with weight gain or BMI increase ≥7% from baseline, and the effect size measure was computed as odds ratio (OR). Random effects model was used to calculate pooled effect size in each meta-analysis, and appropriate moderator analyses or subgroup analyses were conducted as necessary. Publication bias was assessed using the trim and fill procedure. Both dominant model and recessive model were used to analyze each SNP.

Results: As of 12/31/2013, the literature search produced 566 articles, and 82 met inclusion criteria and entered in subsequent analyses. 38 SNPs from 20 genes/genomic regions were meta-analyzed in association with APD-induced weight gain, with the number of studies varying from 1 to 20 and total sample sizes varying from 123 to 2,082. 11 SNPs from 8 genes were significantly associated with the primary outcome, and 4 SNPs from 2 genes were significantly associated with the secondary outcome. Combined together, 13 SNPs from 9 genes (ADRA2A, ADRB3, BDNF, DRD2, GNB3, HTR2C, INSIG2, MC4R, and SNAP25) were significantly associated with APD-induced weight gain (p values:<0.05-0.001). SNPs in ADRA2A, DRD2, HTR2C, and MC4R had relatively large effect sizes (Hedges’ g’s ranged from 0.30 to 0.46, and ORs ranged from 1.47 to 1.96). Subgroup analyses showed that samples of young patients with less prior APD exposure and samples with short follow-up periods (i.e., 1-2 months) tended to produce significant associations with large effect sizes. There was no significant difference in the phenotype-genotype association among different APDs.

Conclusions: APD-induced weight gain may be associated with certain genetic variants, especially in genes coding for target proteins that APDs bind to. It is likely that multiple genes may contribute to APD-induced weight gain. SNP-SNP interactions, gene-gene interactions, and gene-environmental interactions should be explored in a large sample to further elucidate the pharmacogenetics of APD-induced weight gain.

Keywords: pharmacogenetics, antipsychotics, weight gain, meta-analysis.

Disclosure: Dr. Zhang has received grant support from the National Institute of Mental Health and the Brain and Behavior Research Foundation. Dr. Robinson has been a consultant to Asubio and Shire, and received grant support from Bristol Meyers Squibb, Otsuka, NIH and CMS. Dr. Kane has been a consultant for Amgen, Alkermes, Bristol-Meyers Squibb, Eli Lilly, EnVivo Pharmaceuticals (Forum), Forest, Genentech, H. Lundbeck, Intracellular Therapeutics, Janssen Pharmaceutica, Johnson and Johnson, Merck, Novartis, Otsuka, Pierre Fabre, Proteus, Reviva, Roche and Sunovion; and has been on the Speaker’s Bureau for Bristol-Meyers Squibb, Eli Lilly, Genentech, Janssen and Otsuka. Dr. Kane is a Shareholder in MedAvante, Inc., and has received grant support from NIMH, CMS and Otsuka. Dr. Malhotra has been a consultant to Genomind and Forum Pharmaceutical. Dr. Correll has been a consultant and/or advisor to or has received honoraria from: Actelion, Alexza, American Academy of Child and Adolescent Psychiatry, Bristol-Myers Squibb (BMA), Cephalon, Eli Lilly, Genetech, Gerson Lehrman Group, IntraCellular Therapies, Lundbeck, Medavante, Medscape, Merck, Janssen/J&J, Otsuka, Pfizer, ProPhase, Roche, Sunovion, Takeda, Teva and Vanda.

T189. Non-replication of Association of the GADL1 rs17026688 SNP with Lithium Response in Han Chinese

Margit Burmeister*, Sheng Li, Chen Zhang, Zhiguo Wu, Haozhe Li, Lin He, Jun Li, Yiru Fang

University of Michigan, Ann Arbor, Michigan

Background: Lithium is the drug of choice to treat Bipolar I Disorder in most settings. About a third of patients usually have an excellent response, with most at least partially responding. Recently (Chen et al., NEJM 370;2, Jan 2014), a Taiwanese Bipolar 1 Consortium identified SNPs in the GADL1 gene associated with Lithium response in a sample of 294 Taiwanese, with the most highly significant p value (10-37) in rs17026688, which were replicated in another 124 patients. Heterozygotes for the T allele were>70 more likely to respond, TT homozygotes>200 more likely to respond. This SNP is monomorphic C outside of East Asia.

Methods: Here, we attempted to replicate this finding in a sample of 153 BP1 subjects who were treated with Lithium. We used the commonly used clinical global impression improvement (CGI-I) score on patients seen for the past 10 years in Shanghai, and also retrospectively applied the ALDA scale, as it was used in the original publication. Rs17026688 was genotyped by PCR+Sanger sequencing. 141 BP1 patients primarily treated with Valproate or other drugs were also genotyped.

Results: Using a CGI-I score of 1-3 as cut-off, we observed 45% of patients responded to Lithium. Genotype (48:14: 3 cases, 56:22:10 controls) or allele frequency at rs17026688 showed no significant difference between responders and non-responders. Using an ALDA response cutoff of 1-5, 36% responded, with again, no significant findings. A trend towards significance was observed for allele frequency comparison with cgi as scale, with a p=0.07. However, this was in opposite direction, with an OR of 1.7 (C/T carriers cases 0.85:0.15 vs controls 0.76:0.24). Our results are consistent with non-replications in Japanese or Japanese/Taiwanese samples (Ikeda et al., Hou et al., NEJM).

Conclusions: Our results suggest that the previously published strong association of rs17026688 with Lithium is not generalizable, although both studies used Han Chinese. We we had ample power to detect the published effect size if it existed. Both previous and our non-replication did not use the same strict criteria of adherence as to exclude>82% of the overall sample. Although inclusion required adherence, the number of patients excluded was much less than 82%. That study’s Lithium responder samples (both discovery and replication), but not the non-responders, strongly deviates from HWE (p<0.001) due excess of heterozygotes. Our Lithium treatment sample, but not the other treatment group, by contrast, shows a small deviation of HWE in the opposite direction. We hence cannot exclude that rs17026688 is related to adherence to Lithium treatment rather than response. It would be essential to know the genotypes of the 82% of the study sample that was excluded in the original study. In conclusion, the rs17026688 SNP in GADL1 is unlikely to be useful to predict Lithium response in realistic clinical samples of BP1 patients of Han Chinese origin.

Keywords: bipolar disorder, pharmacogenetics, Lithium response, Chinese.

Disclosure: Nothing to Disclose.

T190. Variations in the Chromosome 3 Region are Associated with Treatment Resistant Depression

Qingqin Li, Andrew Jadwin, Reyna Favis, Jaskaran Singh, Giacomo Salvadore, Gayle Wittenberg, Vaibhav Narayan, Gary Romano, Wayne Drevets*

Janssen Pharmaceuticals of Johnson & Johnson, Titusville, New Jersey

Background: Major depressive disorder accounts for a large and increasing global burden among psychiatric diseases. Despite the consortium collaborative efforts of MDD meta-analysis, variants predictive of disease susceptibility remain elusive, partially due to the fact that MDD is a heterogeneous disorder. Disease subtypes could be defined by treatment response status to antidepressant to increase the chance of identifying genetic predictors in a more homogenous disease population, although a subset of the treatment responders could be an artifact from placebo response and the treatment non-responders could be attributed to various reasons.

Methods: We performed a genetic association meta-analysis of treatment resistant depression from two independent cohorts of European ancestry. Cases with TRD were defined as subjects failing two trials of antidepressant treatment regimens and were drawn either from a Janssen cohort consisting of patients enrolling in antidepressant clinical studies (n=232) genotyped using Ilumina Omni5MExome or from a cohort based on STAR*D study (n=315) genotyped using either Affymetrix 500K or Affymetrix 5.0. Controls were drawn from either the cognitively normal subjects from the ADNI study (n=255) genotyped using Illumina Omni2.5M or the psychiatrically screened healthy controls from NIMH (n=584) genotyped using Affymetrix 500K, respectively. We imputed genotypes based on the reference haplotypes from the 1,000 Genomes project prior to meta-analysis to enable direct comparison of variants across the study.

Results: The most significant associated markers in the meta-analysis (directly genotyped marker P=8.51 x 10-7; imputed marker P=3.56 x 10-8 passing the conventional genome-wide significance threshold (P=5x10-8)) were located in a 50kb interval (3p24.3) in chromosome 3 with only an unannotated spliced EST reported. This lies within a linkage interval recently implicated in a linkage meta-analysis (3p25.3-3p22.1). Notably, the genetic data from each TRD sample independently supported this association, with uncorrected significance levels of P=2.37 x 10-5 for the STAR*D cohort and P=0.005 for the Janssen cohort.

Conclusions: We have identified a candidate genetic marker for TRD with an association p-value passing genome wide significance using a relatively small sample size by GWAS standard. Future studies with larger sample sizes or replication samples are needed to further dissect the genetic basis of treatment resistant depression. The identification of genetic markers associated with resistance to biogenic amine-based antidepressant drugs holds the potential to guide researchers toward unprecedented targets in the discovery of novel treatments for TRD.

Keywords: Treatment Resistant Depression, GWAS, meta-analysis, Major Depressive Disorder.

Disclosure: Drs. Li, Favis, Singh, Salvadore, Wittenberg, Narayan, Romano, Drevets, and Mr. Jadwin are employees and shareholders of Janssen Research & Development, LLC.

T191. Potential Role of LINC01268 in Completed Suicide by Violent Means

Giovanna Punzi*, Gianluca Ursini, Joo Heon Shin, Andrew Jaffe, Joel E. Kleinman, Thomas M. Hyde, Daniel R. Weinberger

Lieber Institute for Brain Development, Baltimore, Maryland

Background: Suicide is among the principal causes of death worldwide and the 10th leading cause of death in the United States for all age groups combined. Suicide may be conceived as a maladaptive stress response and an impairment of the epigenetic/brain plasticity machinery in coping with stressful events could play a crucial role in suicidal behavior. Lithium has shown efficacy against suicidal behavior, potentially implicating epigenetic mechanisms such as DNA methylation and regulation of expression of long noncoding RNAs (lncRNA). Therapeutic effects of lithium are thought to be in the context of the Wnt signaling pathway, as a result of the inhibition of GSK-3β and the consequent stabilization of β-catenin. Interestingly, lithium has been shown to down-regulate MARCKS, a gene itself involved in neuroplasticity and associated with suicide in several clinical studies.

Methods: In an effort to focus on the impulsiveness/aggressiveness component of suicidal behavior, which is often associated with violent suicide, we analyzed post-mortem dorsolateral prefrontal cortex (DLPFC) RNA-seq data from 106 patients with schizophrenia (SCZ; 63 Caucasians, 79 Males, 45.7±13.8 years), consisting of both non-violent suicides (NV, N=13) and violent suicides (VS, N=16), plus non-suicides (NS, N=77). PolyA enriched RNA was extracted and then purified and enriched with PCR to create a final cDNA library for high throughput sequencing using the Illumina HiSeq2000. Genomic DNA was extracted from the same samples and, after bisulfite treatment, DNA methylation was assessed according to the manufacturer’s instructions with the Infinium HumanMethylation450 BeadChip. To ascertain that gene expression would not be secondary to acute exposure to a prescription or drugs of abuse, we compared the blood toxicology results (positive vs negative) for the main self-poisoning substances (all p>0.2). One-way ANOVA and Χ2 test for differences were used to compare demographic data across manner of death. ANCOVA with age, sex, race and RIN as covariates, was performed to analyze the relationship between gene expression and manner of death. Pearson Correlation was employed to analyze the relationship between genes expression and between gene expression and methylation. Finally, an estimation of the relative neuronal composition of the tissue based on the methylation profile was added as a covariate in analysis of the methylation data.

Results: Manner of death was significantly associated with MARCKS expression (p<0.05), specifically MARCKS expression was greater in VS than in both NS and NV (post hoc with Fisher’s LSD: VS>NS, p<0.00005; VS>NV, p<0.00001; NS vs NV, p=0.08). Expression of an uncharacterized antisense lincRNA (LINC01268) that closely flanks MARCKS was strongly related to violent suicide (p=0.00003; post hoc with Fisher’s LSD: VS>NS, p<0.00001; VS>NV, p<0.00001; NS vs NV, p>0.4), and also to MARCKS expression. Adding LINC01268 as a covariate, MARCKS failed to show an association with manner of death (p>0.5), while the lincRNA remained strongly associated after adding MARCKS as a covariate (p=0.001). Finally, DNA methylation analysis (Illumina 450K chip) revealed that methylation of a site within a MARCKS CpG island was also associated with manner of death, in that VS had lower methylation than NS, with NV being intermediate (p=0.04). Methylation of this site was also associated with expression of the lincRNA, but not of MARCKS, consistent with a link between intragenic methylation and expression of a noncoding antisense transcript. We then examined a list of 296 genes showing a significant (p-corr.<9.8E-10) association with the expression of the lincRNA (LINC01268) in the DLPFC from a broader sample of adults (N=367; including normal controls, subjects with bipolar disorder and depression, again all Caucasians, and with RIN≥7) using DAVID and GOrilla algorithms both in a single-list mode and by comparing with customized (from the same sample) background genes. P values corrected for multiple testing with the Benjamini method show that genes corresponding to gene ontology (GO) terms and pathways related to the immune system response and transmembrane signaling receptor activity were significantly implicated.

Conclusions: We found that epigenetic changes in the MARCKS locus are associated with suicide and likely represent two facets of a mechanism related to expression of LINC01268. GO analysis on the lincRNA are consistent with the role of GSK-3 in modulating immune pathways and endorse the noncoding RNA as a regulatory molecule within the cell as well as in cell-to-cell communication.

Keywords: suicide, noncoding RNA, MARCKS, violence.

Disclosure: Nothing to Disclose.

T192. Retrotransposon-mediated Neuronal Gene Disruption in Schizophrenia and Cocaine Addiction

Wade Berrettini*, Glenn Doyle, Chang-Gyu Hahn, Deborah Mash

University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania

Background: LINE 1 (L1) retrotransposons (RTPs) are 6 kb sequences (probably of ancient retroviral origin) comprising 17% of the human genome (Beck et al, 2010). They have a promoter, 5’ and 3’UTRs, and two open reading frames which encode a reverse transcriptase (Mathias et al, 1991) and an endonuclease (Feng et al, 1996). The reverse transcriptase and the endonuclease allow replication and insertion elsewhere into the host genome (reviewed in Richardson et al, 2014), as first proposed by Barbara McClintock (1956). Kazazian et al (1988) first demonstrated that a germline L1 RTP gene insertion can cause human disease. Data suggested that neuronal embryogenesis is accompanied by activation of L1 RTPs. While most of these somatic de novo L1 RTP insertions will have little effect on neuronal function (perhaps because they occur in gene deserts or in large introns or in genes not required for that cell’s function), some may interfere with normal neuronal activity because they have disrupted a gene needed by that neuron for normal function. If one or more functional L1 RTPs insertions occur in a neuronal precursor, all the daughter neurons that derive from that neuronal precursor will also carry the L1 insertion, perhaps leading to a dysfunctional population of neurons destined to increase risk for a CNS disorder. Bundo et al (2014) reported on L1 RTP insertions in neuronal DNA from three persons with schizophrenia (SZ), in a study of PFC neurons. The L1 RTP disrupted genes were often those implicated in SZGWAS reports. Gene Ontology pathway analysis of this small sample yielded a significant over-representation of L1 RTPs in genes for the category, synapse (uncorrected p=3 x 10-9). There are two reports of increase L1 mRNA in neuronal cells (Okudaira et al, 2014) or mouse brain (Maze et al, 2011) after exposure to cocaine. This background prompted us to study L1 RTPs in post-mortem brain of persons with SZ and in post-mortem brain of persons with cocaine addiction.

Methods: We studied post-mortem dorsolateral prefrontal cortex (DLPFC) from 26 elderly individuals (mean age=79+/- 9; 12 women; all European-American) with SZ and 26 matched (by age, sex and ethnicity) controls (no history of psychiatric illness, cause of death non-neurological). SZ patients had been examined by a Penn psychiatrist years before death to establish the diagnosis (Arnold et al., 1995). We also studied 30 medial prefrontal cortex from persons who died of cocaine overdose and who had histories of cocaine addiction. These were matched with 30 accident victim controls. De novo L1 RTP events were studied by the following steps. Differential centrifugation was used to isolate nuclei from the tissues. Nuclei were labeled with a fluorescently-tagged NeuN antibody and separated into NeuN+(neuronal origin) and NeuN- (glial origin) aliquots by fluorescence-assisted cell sorting (FACS). DNA samples of the NeuN+nuclei were subjected to a degenerate PCR procedure to amplify L1Hs sequences. The resulting amplicons were used to construct libraries for high-throughput sequencing, using the Illumina HiSeq 2000. The L1 Seq program was used for alignment and quality control measures (Ewing and Kazazian, 2010). Only those reads with both L1 sequence and genome-specific sequence were aligned. Quality control metrics for each L1 included mapscore>0.5 (a measure of the strength of alignment to a unique sequence in the reference genome) and at least two independent (maxunique>2) partially overlapping reads for a de novo L1 insertion. Further analysis was limited to de novo (not in the reference genome) L1 insertions into neuronally-expressed genes. A de novo L1 RTP was considered to be in a gene if it was located within 500 bp of the start of transcription or the polyA signal in the 3’UTR of a neuronally-expressed gene. The de novo L1 RTP events in neuronally-expressed genes were analyzed using the David program, which employs BioCarta, Gene Ontology and KEGG pathway analysis (http://david.abcc.ncifcrf.gov).

Results: Results for pathway analyses are shown below. None of the significant SZ or CA pathways were significant for the control groups. In a Biocarta pathway analysis, Synaptic Proteins pathway was significant (Bonferroni corrected p=0.002). In a Gene Ontology pathway analysis, the related pathway, Synapse Part, was significant (Bonferroni corrected p=0.0008). There were 51 genes from the SZ Gene Ontology synapse part with putative de novo L1 RTPs, including NCAM1, NRXN3, ANK3, GRIN1, CASK, DLG3 (AKA SAP102), SPTBN1, GUK1, DLG2 (AKA PSD93), ADD1, DLG1 (AKA SAP97). Many of these genes have been implicated in SZ genetics and/or SZ pathophysiology. For CA there were 7 significant overlapping Gene Ontology kinase/phosphorylation pathways, with a total of 177 genes having putative de novo L1 RTPs (Bonferroni corrected p=10-6 to 10-7). Of the kinases identified by the cocaine addiction pathway analyses, ALK, RET, SRC, GRK6, ILK, PKG, PI3K and MEK1 have all been implicated in the action of cocaine in animal model studies. There is no possibility to determine whether these de novo L1 RTPs arose early in development and increase risk for cocaine addiction or SZ, or whether these de novo L1 RTPs arose as a result of years of disease process. In the latter, the discovery that these disease processes disrupt neuronal DNA would be paradigm shifting. In the former case, the work would create a novel sets of genes to be studied for their roles in the mechanisms whereby these diseases arise. These results must be considered preliminary until they are confirmed in independent samples. It is encouraging that Bundo et al (2014) identified in pathway analysis the Gene Ontology category of ‘synapse’ in their report on 3 SZ brain samples.

Conclusions: These genes from the SZ BioCarta synaptic proteins category had putative de novo L1 RTPs: NCAM1, NRXN3, ANK3, GRIN1, CASK, DLG3 (AKA SAP102), SPTBN1, GUK1, DLG2 (AKA PSD93), ADD1, DLG1 (AKA SAP97). Many of these genes have been implicated in SZ genetics and/or SZ pathophysiology. Of the kinases identified by the cocaine addiction pathway analyses, ALK, RET, SRC and MEK1 have all been implicated in the action of cocaine in animal model studies. There is no possibility to determine whether these de novo L1 RTPs arose early in development and increase risk for cocaine addiction or SZ, or whether these de novo L1 RTPs arose as a result of years of disease process. In the latter, the discovery that these disease processes disrupt neuronal DNA would be paradigm shifting. In the former case, the work would create a novel sets of genes to be studied for their roles in the mechanisms whereby these diseases arise. These results must be considered preliminary until they are confirmed in independent samples. It is encouraging that Bundo et al (2014) identified in pathway analysis the Gene Ontology category of ‘synapse’ in their report on 3 SZ brain samples.

Keywords: retrotransposon, schizophrenia, cocaine addiction, pathway analysis.

Disclosure: Nothing to Disclose.

T193. The Functional Serotonin 1a Receptor Promoter Polymorphism, rs6295, is Associated with Psychiatric Illness and Differences in Transcription

Zoe Donaldson*, Brice le Francois, Tabia Santos, Maura Boldrini, Frances Champagne, Victoria Arango, Craig Stockmeier, Hanga Galfalvy, Paul Albert, Kerry Ressler, Rene Hen

Columbia University/New York State Psychiatric Institute, New York, New York

Background: Both preclinical and clinical data suggest a role for serotonin 1a receptors (5-HT1A) in mood and anxiety disorders. In particular, relatively small changes in receptor levels have been shown to impact behavior in animal models, suggesting that naturally-occurring mechanisms that impact 5-HT1A receptor levels in humans represent a potentially important source of disease risk. One such putative mechanism is a G/C single nucleotide polymorphism (SNP) in the serotonin 1a receptor gene promoter, known as rs6295. The G-allele of this SNP has previously been linked with increased rates of depression and suicide, as well as decreased antidepressant responsiveness. in vitro studies also suggest that rs6295 may have functional effects on the expression of the serotonin 1a receptor gene (HTR1A) through altered binding of a number of transcription factors, including Deaf1/NUDR and the development-specific factors, Hes1 and Hes5. However, to date, this has not been investigated in vivo. Thus we undertook dual lines of research to test the hypothesis that rs6295 represents a common genetic variant that may have a functional impact on mental health through alterations in gene expression.

Methods: In order to further explore the relationship between rs6295, mental illness, and gene expression, we performed dual epidemiological and biological studies. First, we genotyped a cohort of 1412 individuals to examine the relationship between rs6295 and five psychiatric outcomes: current post-traumatic stress disorder (PTSD), current depression, history of substance or alcohol abuse, history of psychiatric hospitalization, and history of suicide attempts. We also examined potential gene-environment interactions involving rs6295 and exposure to either childhood or adult trauma. In conjunction, we investigated the potential impact of rs6295 on HTR1A expression in post-mortem human brain tissue. Specifically, we used relative allelic expression assays, which measure the relative amount of mRNA derived from the C and G-allele in the same individual. This novel approach overcomes many of the limitations of working with post-mortem human samples by controlling for inherent variability attributable to differences in post-mortem tissue handling. Using this approach, we assessed the region-specific, developmental, and disease state-specific impact of rs6295 on transcription.

Results: In our large, highly traumatized cohort, we found that the rs6295G allele is associated with increased risk for substance abuse (odds ratio (OR)=1.367, p=0.021), psychiatric hospitalization (OR=1.688, p=0.001), and suicide attempts (OR=1.440, p=0.022). Overall, exposure to either childhood or non-childhood trauma resulted in increased risk for all psychiatric outcomes, and we found a significant interaction between rs6295 and childhood trauma in modulating psychiatric hospitalizations (OR=2.536, p=0.022). Specifically, GG individuals exhibited maximal rates of psychiatric hospitalization if exposed to moderate levels of early life trauma while C-allele carriers with moderate levels of trauma were indistinguishable from those experiencing mild or no stress. Moreover, in order to link these epidemiological findings with a putative biological mechanism, we also investigated HTR1A transcription in post-mortem brain tissue. Using relative allelic expression assays, we found more mRNA produced from the C versus the G allele of rs6295 in the prefrontal cortex (t(15)=6.092, p<0.001), but not in the midbrain of non-psychiatric control subjects (t(4)=2.881, p=0.135). Further, in the fetal cortex, rs6295C was associated with increased relative expression as early as gestational week 18 in humans (t(4)=6.24, p=0.003). Finally, we found that the G:C allelic expression ratio was significantly neutralized in the prefrontal cortex of subjects with Major Depressive Disorder (MDD) both with and without suicide as compared to controls, indicating that normal patterns of transcription may be disrupted in MDD/suicide (t(33)=3.235, p=0.003). These data together indicate that rs6295 has region-specific and developmental impacts on transcription and support its role as a potential risk factor for mental illness.

Conclusions: Our data provide a putative biological mechanism underlying the association between rs6295, trauma, and mental illness. Specifically, the increased incidence of psychiatric outcomes in GG individuals, as well as their increased sensitivity to childhood trauma, suggests that the G-allele decreases the threshold for the harmful effects of stress. Moreover, our results show that the rs6295G-allele corresponds with decreased HTR1A transcription in the prefrontal cortex during both gestational development and adulthood. Such decreases in 5-HT1A receptor levels may alter top-down processing of stressful experiences by limbic circuits, leading to increased risk for psychiatric outcomes. To our knowledge, this is the first demonstration of developmental and region-specific effects of a common polymorphism on gene expression in the human brain, and its potential relevance is further strengthened by our well-powered epidemiological study linking the G-allele with psychiatric outcomes. These represent an important advance towards understanding the so-called “missing heritability” of psychiatric illness.

Keywords: serotonin 1a receptor, rs6295, gene expression, post-mortem.

Disclosure: Nothing to Disclose.

T194. SLC2A1 Polymorphism Decreases Incidence of Depression and PTSD after Trauma

Gretchen Neigh*, Tanja Jovanovic, Alicia Smith, Lynn Almli, Charles Gillespie, Varun Kilaru, Constance Harrell, Kerry Ressler

Emory University School of Medicine, Atlanta, Georgia

Background: Manifestation of Major Depressive Disorder (MDD) and Post-Traumatic Stress Disorder (PTSD) can have devastating consequences. Alterations in cerebral metabolic activity have been demonstrated in both disorders, and these changes in metabolic activity are generally attributed to disease-related changes in glutamate release from neurons which thereby precipitate concomitant changes in regional glucose transport. Counter to this traditional dogma, it is possible that a primary change in facilitated glucose transport, imposed by a genetic polymorphism in glucose transporter subtype 1 (GLUT1), also known as solute carrier family 2 (facilitated glucose transporter) member 1 (SLC2A1), subsequently alters neuronal activity following trauma. GLUT1 is a rate-limiting step regulating transport and metabolism of glucose in the periphery and brain. Nonpathological genetic variations in metabolism are well characterized and these variations have been shown to differentially impact disease progression in the case of several somatic conditions. For instance, a polymorphism in GLUT1, which leads to lower gene expression, impacts cancer progression and diabetic nephropathy; however, the polymorphism does not increase the risk of developing either cancer or diabetes. These data indicate that while the GLUT1 polymorphism is not in and of itself a pathogenic genetic variation, the polymorphism can alter the physiological sequela following a primary challenge. We hypothesized that genetic variants in GLUT1 would be differentially associated with psychiatric risk and resilience following trauma exposure.

Methods: The Grady Trauma Project (GTP) has collected DNA samples and trauma history evaluations on over 5,000 participants from a high risk, highly traumatized urban population. Using genome-wide data imputed with HapMap reference samples, we evaluated a single nucleotide polymorphism (SNP) in the promoter region of GLUT1 (rs710218, quality score=0.98). In order to address the hypothesis that decreased GLUT1 expression due to a SNP in the promoter region of the GLUT1 gene would confer resilience against the development of MDD and PTSD following trauma exposure, detailed trauma interviews and Beck Depression Inventory assessments were completed as part of the Grady Trauma Project (GTP). Subjects with two or more traumas were used in the analysis.

Results: Assessment of mRNA demonstrated that this SNP associated with GLUT1 gene expression in the blood; the A allele associated with increased gene expression (p=0.01; N=307 who have both genotype and mRNA analysis, covarying for sex, age, race, income, education). These data suggest that within our dataset, rs710218 serves as an expression-linked quantitative trait locus (eQTL), and that the SNP is functionally significant. Further, we found an association of rs710218 with GTP subjects self-report measures of MDD (N=561; p=0.01) and PTSD symptoms (N=575; p=0.008) in traumatized subjects.

Conclusions: These data indicate that in high-risk subjects, the TT genotype, associated with less GLUT1 mRNA and with resilience in subjects who had exposures to multiple types of childhood trauma. Assessment of genetically-mediated metabolic factors may provide innovative insight into a potential metabolic resilience factor against trauma-induced mental health impairments. Appreciation for the role of metabolic factors in the manifestation of behavioral disorders will provide a new direction of consideration for novel therapeutic options.

Keywords: depression, PTSD, metabolism, resilience.

Disclosure: Nothing to Disclose.

T195. Exposure to Adversity in Pre-school Aged Children, Glucocorticoid Receptor Gene Methylation and Behavioral Outcomes

Kathryn Ridout*, Stephanie Parade, Ronald Seifer, Carmen Marsit, Corina Lesseur, David Armstrong, Nicole Eslinger, Melissa McWilliams, Noah Philip, Brittney Josefson, Audrey Tyrka

Brown University, Butler Hospital, Providence, Rhode Island

Background: Epigenetic modifications to the genome are a key mechanism involved in the biological encoding of experience. Animal studies and a growing body of literature in humans have shown that early adversity is linked to methylation of the gene for the glucocorticoid receptor (NR3C1) which is a key regulator of the hypothalamic-pituitary-adrenal (HPA) axis as well as a broad range of physiological systems including metabolic and immune function. No prior work has considered the contribution of methylation of NR3C1 to emerging behavior problems and psychopathology in childhood. The current study examined links between adversity, promoter methylation of NR3C1, and behavior problems in preschoolers.

Methods: One-hundred and eighty-four families, including n=74 with child welfare documentation of moderate-severe maltreatment in the past six months, participated in the gene methylation study and n=171 families (n=71 with documentation moderate-severe maltreatment in the past six months) participated in the behavioral problems assessment. Children ranged in age from 3 to 5 years and were racially and ethnically diverse. Structured record review and home interviews were used to assess a history of maltreatment, other traumas, and contextual life stressors, and a composite variable assessed the number exposures to these adversities. Child internalizing and externalizing behavior problems were determined by parental report during structured interviews. Salivary samples were obtained from the children for gene methylation testing. Methylation of regions 1D, 1F, and 1H of the NR3C1 promoter was measured via sodium bisulfite pyrosequencing.

Results: The composite measure of adversity was positively correlated with methylation at exons 1D and 1F in the promoter of NR3C1. Individual stress measures were significantly associated with a several CpG sites in these regions. Methylation of the NR3C1 promoter at exons 1D and 1F was positively associated with internalizing (r=.17, p<.05 and r=.23, p<.01 respectively), but not externalizing, behavior problems. Furthermore, this NR3C1 methylation significantly mediated the association between measures of early adversity and internalizing behavior problems.

Conclusions: These results show that early adversity is correlated with NR3C1 promoter methylation and that methylation of NR3C1 is associated with internalizing behaviors in young children. Methylation of the NR3C1 promoter may be a mechanism by which adverse exposures in young children influence the bio-behavioral outcomes associated with early childhood adversity.

Keywords: pre-school children, glucocorticoid receptor gene, methylation, internalizing behavior.

Disclosure: Nothing to Disclose.

T196. Pharmacoepigenetics of Insulin Resistance in Bipolar Disorder

Kyle Burghardt*, Jacyln Goodrich, Dana Dolinoy, Vicki Ellingrod

Wayne State University, Detroit, Michigan

Background: The rates of insulin resistance, diabetes and cardiovascular disease (CVD) are elevated in bipolar disorder and contribute to a more complicated course of psychiatric illness and shorter life expectancies. The second generation antipsychotics (SGAs) are known to increase the risk of insulin resistance which is an early and unifying feature of the metabolic syndrome and CVD. The mechanism behind this adverse effect may be related to abnormalities in the folate cycle. A key endpoint to the folate cycle is the production of methyl donors for DNA methylation. This study aimed to assess the effect of SGA use and insulin resistance on DNA methylation using two different DNA methylation interrogation techniques.

Methods: Bipolar subjects on either SGAs or mood stabilizer monotherapy were assessed for insulin resistance using the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) along with other demographic and clinical characteristics. DNA samples were collected at the time of intake and global methylation levels were assessed using the LUminometric Methylation Assay on a pyrosequencer. Multivariate linear regression was used to investigate the effect of insulin resistance and SGA use on global DNA methylation. 96 SGA-treated bipolar I subjects were used for a follow-up epigenome-wide association study (EWAS) using the Illumina 450K beadchip. The top differentially methylated probes were identified using R statistical software packages based on an insulin resistance cutoff of HOMA-IR>4.65 and adjusting for the covariates of age, gender, hip-to-waist ratio, smoking status and SGA type. The SGAs were divided into three groups based on their risk for metabolic side effects: (1) olanzapine and clozapine, (2) quetiapine, risperidone and paliperidone and (3) ziprasidone and aripiprazole. EWAS p-values were corrected using the False Discovery Rate method.

Results: A total of 115 bipolar I subjects were included in the global methylation analysis. The average age was 43.0±12.1 years, with 67% female and 81% Caucasian. Seventy-three percent were on SGAs and the rest were on mood stabilizer monotherapy. The overall average % global methylation was 77.0±3.26. A multivariate linear regression was conducted to test the association between SGA use and insulin resistance with global methylation while adjusting for smoking status, race, folate level and hip-to-waist circumference ratio. This regression was significant (F(7,86)=3.44, p=0.0029) due to the effect of both HOMA-IR (p=0.0072), smoking (p=0.0320) and the interaction between hip-to-waist ratio and SGA use (p=0.0167). For the model, an increased HOMA-IR, smoking and SGA use (at a given hip-to-waist circumference) was associated with lower global methylation. Global methylation was not influenced by any other demographic or clinical variables. For the EWAS, the average age was 44.5±11.4, 62.5% were female and 87% were Caucasian. A total of 4 CpG probes were found to be differentially methylated after correction for multiple testing using FDR. The top hit was found in the Protein Tyrosine Phosphatase Receptor Type N Polypeptide 2 (PTPRN2) which was hypomethylated in subjects with insulin resistance (HOMA-IR>4.65).

Conclusions: This is the first study to show a relationship between SGA use, insulin resistance and global DNA methylation followed by an EWAS to identify differentially methylated genes for future study. The top EWAS hit was in the PTP2RN gene which is linked to the production of cell signaling molecules. PTP2RN is closely related to PTPRN/IA-2beta which has been linked to autoantibody production in insulin dependent diabetes mellitus. Targeted disruption of the IA-2beta gene has been linked to impaired glucose tolerance and insulin secretion in preclinical models. This work may lead to identification of the folate-based mechanisms underlying the metabolic side effects of SGAs. Future work is needed to identify site-specific methylation differences based on gold-standard assessments of insulin resistance (e.g., hyperinsulinemic euglycemic clamp). Furthermore, analyzing and comparing tissue-specific methylation differences based on insulin resistance will identify mechanisms while retaining clinical translatability.

Keywords: bipolar, insulin, methylation, antipsychotic.

Disclosure: Nothing to Disclose.

T197. Genetic Ancestry Informative Markers (AIMS) and Smoking Cessation Treatment Response in African American Smokers: An Analysis of a Randomized Controlled Trial

Andrea King*, Adam Bress, Coady Wing, Rick Kittles

University of Chicago, Chicago, Illinois

Background: Nearly 20% of adults in the U.S. smoke cigarettes and 443,000 Americans die of smoking every year. Several pharmacotherapies and behavioral therapies increase smoking cessation quit rates. However, effects vary across ethnic groups, and there are disparities in outcomes, particularly for African American (AA) smokers. We conducted a preliminary pharmacogenetic study to determine if genetic background plays a role in smoking cessation differences between self-identified AA and European Americans (EA). Genetic ancestry, as measured by ancestry informative markers, was examined in the context of racial differences in smoking cessation pharmacotherapy.

Methods: Data were culled from a randomized clinical trial examining the efficacy of the opioid receptor antagonist naltrexone versus placebo. AA smokers were oversampled in this Chicago-based sample (35%, 110/315). There were 100 autosomal DNA ancestry informative markers genotyped using the SEQUENOM iPLEX MassArray platform. Individual ancestry was estimated for all participants using STRUCTURE. Logistic regression models examined 4-week quit rates for naltrexone vs. placebo condition by racial group (95 AA vs. 136 EA participants with sufficient DNA) and within AAs of high and low West African (WA) ancestry. In AAs, the median WA ancestry was 80% (range 52-99%), so we defined subjects above the median as high WA ancestry and those below median as low WA ancestry. The models adjusted for baseline clinical and demographic characteristics and confidence intervals and p-values were estimated using the bootstrap. All significant results remained after controlling for covariates including age, sex, education, nicotine dependence scores, and socio-economic status.

Results: Among the EAs, naltrexone significantly improved quit rates at four weeks (61% vs. 44% placebo, a 17% quit rate increase, p=0.03) but this was not the case in AAs (43% vs. 32% placebo, p=0.44). Examining AAs by WA ancestry revealed a significant medication by WA ancestry interaction (p=.03). Those AA with low WA ancestry exhibited significantly higher quit rates with naltrexone compared with placebo (60% vs. 27%, a 33% increase, p=.04). In contrast, for those AAs with high WA ancestry, quit rates were directionally lower in the naltrexone vs. placebo group (26 vs. 39% placebo, 13% decrease, p=0.33). The OPRM1 A118G risk allele was only evident in 6% of AA, precluding measurement of its effects on outcomes.

Conclusions: This study represents the first to our knowledge to examine genetic ancestry, as measured by West African ancestry informative markers, in the context of racial differences in smoking cessation pharmacotherapy. The results are consistent with two other studies examining drug response based on genetic ancestry in Hispanic asthmatics and in EA and AA patients receiving anti-depressant treatment. Naltrexone remains an experimental medication for smoking cessation with mixed results in predominantly EA smokers. This investigation elucidates that AA smokers were less responsive that EU smokers, and that genetic factors may partially explain this differential response to naltrexone. Self-identified AAs of low WA ancestry exhibited better treatment response to naltrexone, more than doubling their quit rates with placebo, but there was no advantage of this treatment in AAs of high WA ancestry. Further examination of genetic ancestry using admixture mapping methods, as well as investigating other genetic variants and biomarkers closely linked with ancestry that may affect pharmacokinetics and pharmacodynamics, may help elucidate the functional significance of these findings. Overall, stratifying on self-reported race and examining heterogeneity by genetic ancestry allows important information about potential biological differences that may contribute to the heterogeneity observed across self-reported racial groups.

Keywords: genetic ancestry, nicotine dependence, smoking cessation, African American.

Disclosure: Nothing to Disclose.

T198. Intersecting Large-scale Genetic Studies of Schizophrenia with Drug Target Information to Inform Drug Design and Repurposing

Douglas Ruderfer*, Alexander Charney, Ben Readhead, Swedish Schizophrenia Sequencing, PGC Schizophrenia Working, Shaun Purcell, Joel Dudley, Pamela Sklar

Mount Sinai School of Medicine, New York, New York

Background: Most drugs for psychiatric disease have variable efficacy and tolerability. For example, antipsychotics are generally effective at treating the psychosis, not the cognitive or negative symptoms of schizophrenia (SCZ) and often produce severe side effects. The understanding of the therapeutic mechanism of action for these drugs remains incomplete. Recent large-scale genetic studies have begun to identify both specific loci and functional gene sets that are contributing to SCZ risk. The most recent PGC SCZ GWAS has identified genome-wide significant association to over 100 loci including common variants in some of the main therapeutic targets of antipsychotics like DRD2. We hypothesize that genes implicated by genetic studies represent plausible therapeutic targets and that genetic information can be used to inform drug design and repurposing.

Methods: We downloaded drug classifications from the Anatomical Therapeutic Chemical Classification System (ATC). Drug targets were collected from multiple databases of both referenced (DrugBank, TTD, etc.) and predicted targets of drugs. In particular, we utilized a proprietary method to identify novel drug targets based on ligand similarity (SeaChange). Enrichment testing was performed using Inrich for GWAS regions and SMP for rare coding variants.

Results: First, we sought to identify if the gene targets of any particular class of drugs showed enrichment for genes associated with SCZ. For each ATC level 3 drug class we created a single set of gene targets by combining all referenced and predicted targets for the drugs in that class. The gene targets from each drug class were tested for overlap with genome-wide significant regions from SCZ GWAS and enrichment of singleton case loss of function variants in a dataset of over 5,000 individuals. Of the 175 drug classes tested, the most significant enrichment was seen in the set of 63 drugs labeled antipsychotics (p=0.0002), of which there were significant independent results for both the GWAS data (p=0.0026) and the sequencing data (p=0.007). Significant enrichment was present in the 44 referenced drug targets of antipsychotics but also in the 284 novel predicted targets. We further aim to prioritize gene targets based on genetic data and other information including brain expression and to utilize network based methods to predict plausible new therapeutic targets.

Conclusions: Our results suggest that the effectiveness of antipsychotics is not limited to their actions on dopamine and serotonin receptors but likely the result of targeting numerous other proteins, many of which are off-target. We show that of the 175 classes of drugs tested antipsychotics are most significantly implicated by the combination of common and rare genetic variation. These results suggest efforts to develop a “cleaner” antipsychotic may be worth pursuing and that genetic data may help identify which targets are most important. In addition, the inclusion of genetic data from other disorders related to drug side effect profiles (e.g. metabolic disorder for antipsychotics) may point to targets contributing to these side effects.

Keywords: Schizophrenia, genetics, drugs.

Disclosure: Nothing to Disclose.

T199. Polygenic Correlates of Psychotic Disorder across Neurobiological Taxonomy: Potential Opportunity for Precision Medicine

Pranav Nanda*, Jaya Padmanabhan, Neeraj Tandon, Ian Mathew, Gualberto Ruano, Andreas Windemuth, Brett Clementz, Godfrey Pearlson, John Sweeney, Carol Tamminga, Matcheri Keshavan

Columbia College of Physicians and Surgeons, New York, New York

Background: The complex nature and imprecise diagnostic boundaries of psychotic disorders have obfuscated their genetic architecture. Although over 1,000 studies have been published investigating associations between proposed risk genes and schizophrenia alone, results have largely been inconsistent. The failure of monogenic approaches to produce robust results has suggested the value of polygenic analysis, which represents overall genetic loading for disease and thereby enables exploration of genotype-phenotype correlations in complex non-Mendelian disorders, such as psychotic disorders. Furthermore, the diagnostic taxonomy used to classify psychotic disorders, which is based on clinical phenomenology, does not capture etiological differences between patients. Rather, schizophrenia, schizoaffective disorder, and psychotic bipolar disorders exhibit similar characteristics, including crosscutting symptom profiles, overlapping diagnoses within family lineages, and common putative susceptibility genes. One approach to address this issue is to use neurobiological measures to derive sub-groups of psychosis in a data-driven manner agnostic to DSM diagnosis, as Bipolar-Schizophrenia Network for Intermediate Phenotypes (B-SNIP) collaborators have done using cognitive and sensorimotor biomarkers (Clementz et al., under review). A reformulated taxonomy based on biomarkers may more successfully differentiate psychotic disorders’ etiologies, and may thereby aid in the tailoring of therapies to patients’ disease pathogeneses. In this study, we examined the polygenic underpinnings of neurobiologically defined psychotic disorder patient subgroups in order to genetically validate and characterize them.

Methods: Cognitive and sensorimotor biomarkers (EEG, saccades, and cognitive testing) were used to group 344 probands (131 schizophrenic (SZ), 76 schizoaffective (SZA), and 137 psychotic bipolar (PBP)) into three Biotype classes (92 BT1, 102 BT2, 150 BT3), according to the protocol of Clementz et al. (under review). Along with 118 controls, these subjects were genotyped for over 1.1 million target single nucleotide polymorphisms (SNPs). SNPs from an independent discovery sample (Schizophrenia Psychiatric Genome-Wide Association Consortium (PGC) stage 1) were screened for inclusion in analysis by identifying the SNP set with highest polygenic association with probandicity. Candidate SNP sets were generated by incrementally adding SNPs in descending order of their PGC-1 association with schizophrenia. All SNPs in the screened SNP set with maximal probandicity association were individually associated with case/control status for each Biotype and diagnostic group. The screened SNP set was converted to a set of genes using the SCAN database and then was clustered using functional annotation in DAVID version 6.7. Significant clusters and the entire screened SNP set were tested by logistic regression for polygenic association with Biotypes and diagnoses, correcting for the number of clusters and patient groups by Benjamini-Hochberg adjustment. A total of 10,000 permutations of the size of significantly associated clusters were taken from the screened SNP set and also tested by logistic regression for polygenic association with Biotypes and diagnoses, correcting for the number of permutations and patient groups by Benjamini-Hochberg adjustment. Functional annotation clustering was performed on significantly associated permutations. Population stratification was controlled in statistical analyses using multidimensional scaling (MDS) in PLINK version 1.07.

Results: A screened set of 8,646 SNPs was identified as conferring maximal polygenic risk of disease. It exhibited significant polygenic association with SZ, SZA, PBP, BT1, and BT3 (p<0.05) and trending polygenic association with BT2 (p<0.1). One SNP (rs894243: PRDM12) was found to be significantly associated with BT1 (p<5.8x10-6, OR=4.90). The screened SNP set was found to comprise eight significant clusters (p<0.0001): ion binding, ion channel activity, neuron projection development, voltage-gated channel activity, cell motility and migration, regulation of neuron differentiation, regulation of neuron projection development, and phosphorus metabolism. One of these functional clusters (ion binding), consisting of 895 SNPs, was significantly associated with BT1 (p<0.001). Twelve permutations of size 895 SNPs were significantly associated with BT1 after multiple comparisons correction (p<8.3x10-7). Eleven of these permutations were characterized by either ion binding or ion channel activity functional clusters. No significant monogenic, polygenic cluster, or polygenic permutation associations were observed for any other Biotype or diagnosis.

Conclusions: Our results indicate that psychotic disorders are characterized by genetic risk spanning large and functionally heterogeneous sets of genes. Genetic risk appeared to differentiate Biotype but not diagnostic categories, corroborating genetic validation for the biotyping method of patient classification. Compared to the other biotype groups and even compared to all diagnostic groups, BT1 exhibited substantially greater monogenic association, particularly with PRDM12, and substantially greater polygenic association, particularly with risk genes in functional clusters involving ion channel binding and activity. The marked degree of risk conferred by ion channel related genes raises the possibility of ion channels as therapeutic targets in BT1 patients, who represent an objectively and neurobiologically identifiable subset of psychotic disorder patients.

Keywords: Psychotic Disorders, Schizophrenia, Polygenic Risk Score.

Disclosure: Nothing to Disclose.