Sponsorship Statement: Publication of this supplement is sponsored by the ACNP.

Individual contributor disclosures may be found within the abstracts. Part 1: All Financial Involvement with a pharmaceutical or biotechnology company, a company providing clinical assessment, scientific, or medical products or companies doing business with or proposing to do business with ACNP over past 2 years (Calendar Years 2014–Present); Part 2: Income Sources & Equity of $10,000 per year or greater (Calendar Years 2014 - Present): List those financial relationships which are listed in part one and have a value greater than $10,000 per year, OR financial holdings that are listed in part one and have a value of $10,000 or greater as of the date of disclosure; Part 3: Financial Involvement with a pharmaceutical or biotechnology company, a company providing clinical assessment, scientific, or medical products or companies doing business with or proposing to do business with ACNP which constitutes more than 5% of personal income (Calendar Years 2014 - Present); Part 4: Grants from pharmaceutical or biotechnology company, a company providing clinical assessment, scientific, or medical products directly, or indirectly through a foundation, university, or any other organization (Calendar Years 2014 - Present); Part 5: My primary employer is a pharmaceutical/biotech/medical device company.

Asterisks in the author lists indicate presenter of the abstract at the annual meeting.

W1. Association Between Neuroinflammation and Cortical Thickness in Alzheimer's Disease: A PET-MRI Study

Sina Hafizi*, Anne Wheeler, Ivonne Suridjan, Nicolaas Verhoeff, Pablo M. Rusjan, Sylvain Houle, Romina Mizrahi

Centre for Addiction and Mental Health, Toronto, Canada

Background: Increased translocator protein 18kDa (TSPO) expression is a reliable marker of (micro)glial activation and neuroinflammation in-vivo, which was shown to be increased in Alzheimer's disease (AD). Moreover, structural imaging studies on AD patients showed widespread abnormal brain structural changes, such as reduced cortical thickness and decreased cortical surface area, in several brain regions, particularly in the prefrontal cortex (i.e. dorsolateral prefrontal and medial prefrontal cortices). However, no study has assessed the link between neuroinflammation and structural brain changes in AD.

Methods: Using a novel second-generation PET radiotracer for TSPO, [18F]FEPPA, we evaluated the association between neuroinflammation and brain structural deficits in dorsolateral prefrontal and medial prefrontal cortices in AD. Sixteen patients with AD underwent [18F]FEPPA PET, using a high-resolution research tomograph (HRRT), and structural (T1) MRI using a 3-Tesla scanner. PET data analyses were carried out using the validated 2-tissue compartment model to determine the total volume of distribution (VT) of [18F]FEPPA.

Results: Results show significant inverse associations (controlling for rs6971 genotype) between neuroinflammation (VTs) and cortical thickness in the right medial prefrontal cortex [r=-.562, p =.029] and the left dorsolateral prefrontal cortex [r=-.629, p =.012](p values are not corrected for multiple comparisons). No significant associations were found with surface area.

Conclusions: These results, while preliminary, suggest links between the microglial activation/neuroinflammation and cortical thickness in the dorsolateral prefrontal cortex and medial prefrontal cortex in AD patients.

Keywords: Positron Emission Tomography Imaging, Alzheimer's Disease, Structural MRI

Disclosure: Nothing to Disclose.

W2. Impact of BDNF and Sex on Maintaining Intact Memory Function in Early Midlife

Kyoko Konishi*, Emily Jacobs, Sara Cherkerzian, Blair Scribner-Weiss, Dorene Rentz, Anne Remington, Harlyn Aizley, Anne Klibanksi, Philip De Jager, Jill Goldstein

Harvard Medical School, Brigham & Women's Hospital, Boston, Massachusetts, United States

Background: The population is rapidly aging, thus understanding how to maintain intact memory function in midlife is a major public health concern. The hippocampus (HIPP), a key region in the memory circuitry, is one of the few brain structures that continue to undergo neurogenesis with aging, albeit at a reduced rate. Sex steroid hormones and growth factors, such as brain-derived neurotrophic factor (BDNF), play a significant neuroprotective role in memory circuitry aging through mechanisms of neurogenesis, cellular survival, axonal growth, dendritic growth, and synaptic plasticity. We and others have shown in some women across the menopausal transition, there are decreased memory performance and altered brain function and connectivity in memory circuitry regions. We hypothesize that the abrupt depletion of sex steroid hormones may require a higher production of growth hormones for maintaining intact memory function in the face of reproductive aging. Here, we present new findings characterizing the impact of BDNF on memory circuitry aging across the menopausal transition.

Methods: Two hundred and twelve middle-aged adults (age range 45–55) recruited from the New England Family Study (NEFS) cohort underwent neuropsychological assessment and fMRI scanning. A multiplex immunoassay analysis and genome-wide association study (GWAS) were conducted to assess BDNF plasma levels and BDNF genotype, respectively. The sample consisted of 106 women (mean age=49.82±1.99) and 106 men (mean age=50.09±2.28). Following the STRAW-10 criteria, reproductive history and serologic evaluation (FSH, LH, estradiol, progesterone) were used to determine women's reproductive stage. Women were categorized as late reproductive (“pre-menopause”), menopausal transition (“peri-menopause”), or early post-menopausal (“post-menopause”). Participants performed a verbal working memory task (N-Back) during fMRI scanning and data were analyzed in SPM8. Under normal testing conditions, usually overall deactivation of HIPP is observed. As such, an ROI analysis was conducted with an anatomically defined mask of HIPP. Mean β weights from the HIPP were extracted for each participant as a function of working memory load (2-back>0-back) and were used for subsequent analyses. Neuropsychological assessments of verbal and associative memory were conducted with the 6-Trial Selective Reminding Test (SRT) and the Face-Name Associative Memory Exam (FNAME).

Results: The menopausal transition was associated with decreased memory performance with post-menopausal women performing worse than pre- and peri-menopausal women. In the current study, this effect was attenuated in post-menopausal women with high-BDNF plasma levels (FNAME: F=4.14, p=0.008, η2=0.12; SRT Total recall: F=1.78, p=0.16, η2=0.06). Post-menopausal women with high-BDNF plasma levels did not differ on FNAME and SRT from pre- (FNAME: p=0.39; SRT Total recall: p=0.42) and peri-menopausal women (FNAME: p=0.80; SRT Total recall: p=0.53) and performed significantly better than men (FNAME: p=0.05). In contrast, post-menopausal women with low-BDNF plasma levels performed significantly worse than pre- (FNAME: p=0.02; SRT Total recall: p=0.06) and peri-menopausal women (FNAME: p=0.006; SRT Total recall: p=0.05) and performed similar to men (FNAME: p=0.47; SRT Total recall: p=0.92) on the same tests (FNAME: F=4.96, p=0.003, η2=0.14; SRT Total recall: F=4.30, p=0.007, η2=0.12). Plasma BDNF levels also impacted functional changes in working memory circuitry observed across reproductive aging. Across the menopausal transition, regional deactivation of HIPP was attenuated. However, this relationship was only evident in women with low-BDNF plasma levels. Post-menopausal women with low-BDNF plasma levels had significantly higher activity in HIPP compared to pre- (p=0.02) and peri-menopausal women (p=0.05) and men (p=0.09). In contrast, post-menopausal women with high-BDNF plasma levels showed normal regional deactivation of HIPP on the working memory task (F=0.42, p=0.74, η2=0.01), similar to pre- (p=0.91) and peri-menopausal women (p=0.39) and men (p=0.94). Consistent findings were also observed across BDNF genotype groups. Poor memory performance was only observed in post-menopausal women carrying the BDNF met allele (F=4.02, p=0.01, η2=0.20; Pre-vs Post: p=0.03; Peri vs Post: p=0.01; Men vs Post: p=0.68), the BDNF allelic isoform associated with decreased BDNF secretion.

Conclusions: These results suggest that BDNF modulates the impact of ovarian function decline on memory circuitry in early midlife. Post-menopausal women with either high-BDNF plasma levels or BDNF val homozygous genotype did not exhibit a decline in memory performance or change in functional activity in working memory circuitry. At a human population-level, these findings demonstrate that higher BDNF levels contributed to maintaining intact memory function across the menopausal transition, thus providing evidence for potential sex-dependent avenues for attenuating risk of memory decline in early midlife.

Keywords: Aging, Memory, Hippocampus, Sex Differences, Menopause

Disclosure: Nothing to Disclose.

W3. Persistent Cancer-Related Cognitive Impairment Following Cancer Chemotherapy: Comparison to Post Menopausal Subjective Cognitive Decline

Jennifer Vega, Julie Dumas, Kimberly Albert, Paul Newhouse*

Vanderbilt University Medical Center, Nashville, Tennessee, United States

Background: Cancer-related cognitive impairment (CRCI) is commonly reported following the administration of cancer treatment. Current longitudinal studies, primarily in women with breast cancer suggest that up 35%–60% exhibit persistent cognitive complaints following completion of chemotherapy. Additionally, there is evidence that in non-cancer patients, subjective cognitive decline (SCD) may be predictive of later cognitive dysfunction and we and others have shown that SCD may begin in mid-life in women during or after the menopause transition. For example, approximately 42% of postmenopausal women reported a negative change in cognition in the Study of Women Across the Nation. We have previously shown functional connectivity differences in women who endorse menopause-associated subjective cognitive decline (maSCD), and that cancer chemotherapy alters task-related brain activity even after chemotherapy is completed. These findings suggest that cortical connectivity changes or compensation may be responsible for the symptoms of maSCD and pCRCI. Given that SCD may be a risk factor for late-life cognitive decline and/or dementia, a group of primarily post-menopausal women with pCRCI was compared to two groups of post menopausal women: women who endorse maSCD (SCD+ group) and women who do not (SCD-) to explore the similarities/differences between maSCD and pCRCI and the potential role of rapid menopause in pCRCI.

Methods: This study included data from 63 total participants from two studies: a pCRCI study and a maSCD study. A total of 24 (mean age 54) women were recruited and screened with pCRCI (breast cancer=19), ovarian cancer=1, lymphoma=3). pCRCI study participants had undergone treatment with systemic chemotherapy within the last 1-5 years and endorsed pCRCI subjective complaints. 39 healthy early post-menopausal women (mean age 56) were recruited as part of a larger study (SCD MP) examine the influence of maSCD, structural and functional brain imaging, and genetic markers on estrogen effects on cholinergic-related cognitive functioning. SCD MP study participants were required to be without menses for one year and without surgically-induced menopause). Exclusion criteria included the criteria for the pCRCI study and no use of hormone therapy during the last year, history of cancer or severe menopausal symptoms. In both studies, participants were screened to exclude individuals with evidence of clinically manifest cognitive impairment or depression. The Cognitive Complaint Index (CCI) was used be used to quantify subjective complaints. For both the studies, a CCI score was calculated as the percentage of cognitive items endorsed from a consolidated scale of 120 items. For pCRCI and SCD+ groups participants had to endorse at least 20% (0.2) of all CCI items, SCD- subjects endorsed less than 20%. The two studies shared similar cognitive testing batteries enabling comparison of the datasets. These cognitive domains included tests of arousal and attention: Critical Flicker Fusion (CFF) task, Choice Reaction Time (CRT) task). The Selective Reminding Task (SRT) was used to assess immediate and delayed verbal episodic memory.

Results: pCRCI participants had a significantly (F(2,60)=70.73, p<0.0001) higher mean CCI score (mean=0.45) compared to both SCD+ (0.28) and SCD- MP subjects (0.1). BDI scores were higher in CRCI participants (4.92) and SCD+ (5.13) compared to SCD- (1.83) (p=0.001). pCRCI participants and SCD+ both had a significantly (p<0.0001) higher mean MSC score than SCD-. pCRCI showed a significantly slower CRT Total RT (F(2,57)=3.28, p=0.045) than both MP study groups. pCRCI participants had a significantly (p<0.0001) slower CRT Recognition RT (mean =499.9ms) compared to both SCD+ (mean=424.5ms) and SCD- MP subjects (mean=424.1ms), but motor RT times were not different. There was no significant difference between groups on CFF measures of arousal between groups on SRT measures of episodic memory. There was a significant positive correlation between CCI score and CRT Median Recognition RT (r(58)=0.36, p=0.005) and CRT Total RT (r(58)=0.28, p=0.03). There was a significant positive correlation between CCI and BDI score (r(61) =0.5, p<0.0001) and between CCI and MSC score (r(61)=0.6, p<0.0001).

Conclusions: pCRCI participants endorse significantly more cognitive complaints and menopausal symptoms compared to women with and without maSCD. While mood symptoms were below clinical ranges in all groups, pCRCI and SCD+ participants scored higher than women without maSCD. Objective cognitive performance also differed between groups with significantly slower RT in pCRCI participants and greater cognitive complaints and slower performance correlated with greater menopausal and mood symptoms. Performance differences and cognitive symptoms in pCRCI may indicate long-term changes in cortical activity as a result of chemotherapy and/or vulnerability for late-life cognitive decline. These results also suggest that part of the syndrome of pCRCI may be also accounted for by the rapid onset of menopause in patients post chemotherapy in vulnerable women and the persistence of a significant level of cognitive complaints in pCRCI or maSCD may indicate that such patients are at increased risk for late-life cognitive impairment.

Keywords: Cognitive Decline, Cancer, Menopause, Memory Loss

Disclosure: Nothing to Disclose.

W4. White Matter Integrity in the Corpus Callosum is Associated With Resilience Factors in Geriatric Depression

Roza Vlasova, Beatrix Krause, Prabha Siddarth, Amber Leaver, Katherine Narr, Helen Lavretsky*

Semel Institute for Neuroscience & Human Behavior, Los Angeles, California, United States

Background: Psychological resilience can be protective in depression. Disruption in the integrity of the white matter is a risk factor for geriatric depression. Understanding the neurobiological underpinnings of resilience can provide useful biomarkers of response to antidepressant treatment. We analyzed the relationship of psychological resilience factors in relation to brain white matter integrity in a cohort of depressed geriatric subjects.

Methods: Resilience factors were first derived by the factor analysis of the Connor-Davidson Resilience Scale (CD-RISC) from the large sample of depressed elderly subjects (n=337) generating four resilience factors: (1) purposefulness, (2) self-confidence, (3) reaction to stress, and (4) spiritual beliefs. Diffusion imaging data, collected using two similar MRI acquisition protocols from 70 elderly depressed subjects (mean age=70.44, [ 6.8], 49 women) were analyzed. Diffusion data were preprocessed using regular FSL Tract-Based Spatial Statistics (TBSS) workflows. Mean fractional anisotropy (FA) values were subsequently extracted from the skeletonized FA maps using tract-based and anatomical regions-of-interest (ROIs) chosen based on their links with cognitive control and the control of mood and emotion (cingulum (CGC), posterior cingulum (CGH), genu of corpus callosum (GCC) and ROI's forming components of the dopaminergic mesolimbic reward system (anterior limb of the internal capsule (ALIC) and ventral tegmental area (VTA). The corticospinal tract (CST) was included as a control region. To examine associations with resilience factors, general linear models were used including mean FA from each ROI as dependent variables and resilience factor scores as predictors and age, sex, acquisition protocol as covariates. Using the same statistical model, post-hoc whole brain voxel-based TBSS analyses were performed to probe for more regional relationships with resilience factor scores.

Results: Results showed significant associations between the resilience factor purposefulness and white matter integrity of the GCC (p=0.008). Trend level effects were further observed for purposefulness and CGC (p=0.057) and CGH (p=0.09) mean FA. Similar regional effects where observed when conducting additional whole brain TBSS analysis. No significant results were obtained for reward system-related ROIs, the control ROI or the other factors of resilience. This was confirmed with additional whole brain TBSS analysis.

Conclusions: We demonstrated the relationship between white matter integrity in the GCC to be significantly associated with the purposefulness factor of resilience in depressed older adults. Future studies could further clarify brain mechanisms resilience to stress in depressed and non-depressed subjects.

Keywords: Depression, Resilience, Geriatric, White Matter Integrity, DTI

Disclosure: Part 1: Forest Research Institute, Grant.

W5. Segregated Localization of Dopamine D2 and Kappa Opioid Receptors in the Mesencephalic Dopamine System, Implications in Compulsive Disorders

Maria Andres*, Angélica Escobar, Rodrigo Meza, Pablo Henny

Pontificia Universidad Católica de Chile, Santiago, Chile

Background: The dopamine mesolimbic system is the major circuitry involved in the generation of goal directed behaviors. Dopamine D2 receptors (D2R) and kappa opioid receptors (KOR) are protein Gi coupled receptors highly expressed in the mesolimbic system, whose malfunction is associated to compulsive behaviors. D2R and KOR share several functions in dopamine mesencephalic neurons as regulation of dopamine release and uptake, dopamine neurons firing, among other functions. In addition, KOR activation modulates D2R function, depending on the timing and whether D2R is co-activated or not. For instance, repeated pretreatment with the KOR agonist, U69593, decreases D2R presynaptic function and D2R number in the striatum. On the other hand, repeated D2R and KOR co-activation increases D2R presynaptic function. We and others have shown that repeated KOR co-activation potentiates compulsive behavior induced by repeated activation of D2R in rats. Altogether the data indicates that KOR and D2R functionally interact within the dopamine mesencephalic system, suggesting that they could physically interact in these neurons.

Methods: Using immunofluorescence assays, we studied the localization of KOR and D2R in cultured dopamine neurons and in brain slices of mouse brain.

Results: We observed that KOR and D2R are present in similar density in the dendrites and soma of cultured dopamine neurons; however, they have a segregated localization. Interestingly, KOR immunolabelling was particularly strong in the first part of the axon, colocalizing with ankyrin, indicative that KOR is present in the initial segment of dopaminergic axons. In the adult brain, KOR and D2R are also segregated in striatal tissue. While KOR label is highly located in fibers of passage such as the striae of the striatum, corpus callosum and anterior commissure, D2R is located mostly within the striatum and nucleus accumbens, surrounding fibers of passage; D2R is also localized in fibers that are mostly different from those positives for KOR.

Conclusions: In conclusion, even though KOR and D2R serve to similar functions in the mesencephalic dopaminergic system, they are mostly segregated in these dopamine neurons.

Keywords: D2 Receptor, Kappa Opioid Receptor, Striatum

Disclosure: Nothing to Disclose.

W6. Treatment Response Prediction in Pediatric Patients With OCD Using Structural Neuroimaging Correlates: Simple Linear Regression Versus Support Vector Regression

Edoardo Vattimo, Vivian Barros, Marcelo Batistuzzo, Guaraci Requena, João Sato, Daniel Fatori, Roseli Shavitt, Euripedes Miguel, Marcelo Hoexter*

University of Sao Paulo, Sao Paulo, Brazil

Background: Magnetic resonance imaging investigations in psychiatry have produced a sizable amount of data, contributing to a better understanding of the neurobiological correlates of psychiatric illnesses. As a result, there is strong evidence of the role of the cortico-basal ganglia-thalamo-cortical (CBGTC) circuitry in the pathophysiology of obsessive-compulsive disorder (OCD). These specific brain regions have also been suggested as possible morphometric biomarkers of treatment response in adults undergoing cognitive-behavior therapy (CBT) or pharmacological treatment (antidepressants). Morphometric measurements of CBGTC structures were also shown to correlate to treatment response in children, as our group has previously demonstrated. The large amount of data produced by neuroimaging studies requires, however, improvements of statistical analysis techniques. The aims of this study were to compare the power of simple linear regression (SLR) and support vector regression (SVR) as prediction techniques of treatment response using structural neuroimaging parameters.

Methods: A total of 29 children with OCD underwent magnetic resonance imaging (MRI). Gray Matter (GM) volume and cortical thickness of CBGTC structures were selected a priori for analysis using FreeSurfer Software. After MRI acquisition, patients were enrolled in a randomized clinical trial receiving either fluoxetine (n=12) or CBT (n=17) during 14 weeks. Baseline and post-treatment symptom severity was assessed using Yale Brown Obsessive-Compulsive Scale (Y-BOCS). Morphometric data from CBGTC structures were, then, used to predict clinical response (Y-BOCS changes) in a SLR model. Predictive power of the SLR model was then compared to SVR (with default and tuned parameters) through Root-Mean-Square Error calculations.

Results: Larger GM volume of left caudate (r2= 0,175; p=0.024), right lateral orbitofrontal cortex (OFC) (r2= 0,161; p=0.031) and right medial OFC (r2= 0,143; p=0.043) correlated with percentage improvement in Y-BOCS considering both groups. Thicker right lateral OFC was also correlated with Y-BOCS improvement (r2= 0,181; p=0,021), markedly in patients that underwent CBT (r2= 0,39; p=0,007). SVR was more accurate to predict treatment response in all cases and tuned parameters failed to improve even further the technique was used only for the left caudate.

Conclusions: This investigation presents a comparison of conventional and novel techniques to analyze neuroimaging data as predictors of treatment response in pediatric patients with OCD. Our results raise the possibility that machine learning methods, particularly support vector regression, may be useful to improve the predictive power of such neuroimaging studies.

Keywords: Obsessive-Compulsive Disorder (OCD), Structural Neuroimaging, Children and Adolescents, Treatment

Disclosure: Nothing to Disclose.

W7. Fibroblast Growth Factor 2 Reverses the Effects of Anxious Phenotype on the Early Emergence of Threat Learning in Infancy

Joanna Hider, Amanda White, Julie Boulanger-Bertolus, Regina Sullivan, Huda Akil, Jacek Debiec*

University of Michigan, Ann Arbor, Michigan, United States

Background: Anxiety disorders are the most prevalent psychiatric disorders in childhood affecting around 10% of youth (Kessler et al., 2005). Maladaptive anxiety emerges early in life and disturbs child's psychosocial functioning and development. It is generally accepted that anxiety disorders are caused by a combination of inborn/hereditary and environmental factors. One of the best studied inborn risk factors for anxiety disorders in youth is anxious temperament whereas one of the best known environmental risk factors for anxiety disorders is a history of emotional trauma in childhood. Threat conditioning (TC) is the most commonly used experimental model of emotional trauma. In TC a neutral stimulus (conditioned stimulus, CS), such as neutral sound or odor is paired with an aversive unconditioned stimulus (US), typically a mild electric shock. As a result, an organism expresses threat responses such as freezing upon subsequent exposures to the CS. Previous studies show that fear conditioning in rodents is naturally attenuated until postnatal day (PN) 10 (for review: Debiec & Sullivan, 2016). However, the ontogeny of fear learning in phenotypically anxious organisms is mostly unknown. To address this issue, we used a selectively-bred anxiety-prone rat model, in which spontaneous anxiety-like behaviors emerge as early as at PN 11 (Maras et al., SFN Abstracts, 2014). We also investigated effects of the fibroblast growth factor 2 (FGF 2), which in known to have anxiolytic effects in developing rats (Graham & Richardson, 2009), on TC in anxiety-prone infant pups.

Methods: Experiment 1: Sprague-Dawley (SD) PN 4 rat pups selectively-bred for anxiety received 11 US electric shocks (0.4 mA, 1 s) to the tail, either paired with a CS peppermint odor (30 s) (Paired), or unpaired (Unpaired). Another group included pups that received 11 CS presentations (CS Only). Additional controls included wild-type SD pups matched for age and experimental conditions. OnPN 11, all pups were re-exposed to 3 CSs and their behavior was video recorded and scored for freezing. Experiment 2: SD PN 2 pups received s.c. FGF2 (Sigma-Aldrich, F0291; 20ng/g in 50μl in 0.1M PBS w/1% BSA) or vehicle (0.1M PBS w/1% BSA, s.c.) injections. On PN 4, all pups received TC training (and on PN 11 testing) as described above. Data were analyzed with ANOVA followed by post hoc test (Experiment 1) or student t-test (Experiment 2).

Results: Experiment 1: Statistical analysis revealed no significant differences in freezing behavior among the wild type animals (p>0.05), a finding consistent with previous studies using similar training parameters. However, the anxiety-prone Paired pups showed significantly higher levels of freezing as compared to the control groups (Unpaired and CS Only pups expressed comparable levels of freezing) (ANOVA: F(3,46)=16.51; p<0.0001; post hoc: p<0.05). Experiment 2: Statistical analysis revealed that pups injected with FGF 2 display significantly lower levels of freezing than vehicle controls. In other experiments, we found that early emergence of TC in anxiety-prone pups is associated with elevated levels of early expression c Fos gene expression, an effect moderated by FGF 2 neonatal administration.

Conclusions: Our results demonstrate that threat learning in phenotypically anxious pups occurs very early in life, before they spontaneously express anxious behaviors at PN 11, and before the emergence of adult-like threat conditioning in wild-type rats at PN 10. These results suggest that the early emergence of threat learning in the anxious phenotype affects early-life experiences and may contribute to the development of maladaptive threat responses, such as these seen in anxiety disorders. The fact that early postnatal administration of FGF 2 reversed the effects of anxious phenotype on TC in infancy suggests a potential role for FGF 2 in prevention of childhood anxiety disorders.

Keywords: Anxiety, Fear Conditioning, Fear Memory, Behavioral Inhibition, Amygdala

Disclosure: Nothing to Disclose.

W8. Differential Cognitive and Clinical Outcomes Following Anteromedial Subthalamic Nucleus, Ventral Capsule and Combined Deep Brain Stimulation in Obsessive Compulsive Disorder

Annemieke Apergis-Schoute*, Himanshu Tyagi, Ludwig Zrinzo, Marwan Hariz, Thomas Foltynie, Patricia Limousin, Keith Matthews, Lynne Drummond, Harith Akram, Barbara Sahakian, Trevor Robbins, Naomi Fineberg, Eileen Joyce

University of Cambridge, Cambridge, United Kingdom

Background: Deep brain stimulation (DBS) for Obsessive Compulsive Disorder (OCD) has been used in several centers worldwide to offer symptom alleviation to long term OCD sufferers who do not benefit from conventional pharmacological treatments and cognitive behavioral therapy (CBT). Both ventral capsule/ventral striatum (VC/VS) - including nucleus accumbens- and the anteromedial subthalamic nucleus (amSTN) have been targeted with DBS in severe OCD, markedly improving the lives of around 60% of these patients. We hypothesized that VC/VS and amSTN DBS may have different mechanisms of action on the well-established pathophysiological cortico-striatal-thalamo-cortical (CSTC) loops to improve OCD. Our DBS trial for OCD was the first ever within-patient approach to compare clinical and cognitive outcomes following double blind randomisation of stimulation to the VC/VS, or the amSTN, then cross-over between these targets, followed by DBS on both targets combined, then addition to conjunctive CBT. We hypothesized that 1) VC/VS stimulation would be superior to amSTN stimulation in improving mood, 2) Stimulating both areas would show greater reductions in OCD symptomatology, 3) amSTN or VC/VS stimulation would affect prefrontal efficacy, with possible effects on cognitive flexibility as measured by improved extra-dimensional set shifting (EDS), a proposed cognitive endophenotype for OCD as well as affecting inhibitory control resulting in changes in the stop signal task (SST).

Methods: Six patients with severe treatment resistant OCD with a minimum score of 32 on the Yale Brown Obsessive Compulsive Scale (YBOCS) were recruited for this study from the NHS England OCD Specialist Service. The study received approval from the Ethics Committee and tall patients had previously had pharmacological and psychological therapy, including inpatient, prior to being considered for the study.

Exclusion criteria included: substance abuse or dependence, personality disorder and psychosis. OCD medication was kept constant throughout the trial. Following neurosurgery, in which electrodes were implanted bilaterally in the amSTN and the ventral capsule (VC), the stimulation remained off for 4 weeks to allow any post-surgery effect to resolve (baseline). This was followed by 12 weeks DBS at each target site.

We tested the effects of DBS by comparing baseline, VC, amSTN, both sites (BOTH), optimized setting (OPT) and CBT stages on OCD (YBOCS), depression (Montgomery Asberg Depression Rating Scale (MADRS)), anxiety (Beck Anxiety Inventory (BAI), global assessment of function (GAF)). Our other main objective was to test cognitive performance using a selection of CANTAB tasks. For this presentation, we will discuss cognitive flexibility and impulsivity. For all stages, we used a repeated measures ANOVA with planned orthogonal (Helmert) contrasts to test if 1) DBS was effective for severe OCD; 2) DBS at both sites (BOTH) was different than amSTN and VC; 3) amSTN and VC DBS differed; 4) adjunctive CBT was better than DBS alone. All significant results are reported at P<0.05.

Results: DBS had a significant effect across all phases for all above mentioned clinical measures. Both amSTN and VC stimulation were effective in improving YBOCS scores (no significant differences between amSTN and VC), with a significant added benefit of combined stimulation (BOTH) and conjunctive CBT. A similar pattern of effects was found on anxiety (BAI) and GAF scores. Depressive symptoms (MADRS), on the other hand, showed a greater reduction following VC than amSTN stimulation, whilst also showing added benefits of combined stimulation and CBT. Cognitively, we found that ED-shift performance was only significantly improved with DBS of the amSTN compared to VC stimulation. In contrast, DBS did not show a direct benefit on impulsivity as measured by the proportion of successful stops in the SST. There was, however, a significant effect of DBS on SST mean correct GO reaction times across all conditions, showing that amSTN, VC or combined stimulation all improved reaction time.

Conclusions: This study is the first to have directly compared differential effects of amSTN versus VC DBS stimulation in OCD patients whilst testing clinical and cognitive outcomes. The results of this study suggest that both sites are equally effective in reducing OCD symptoms; stimulating VC leads to a greater improvement in mood, whilst amSTN stimulation has beneficial effects on cognitive flexibility. These results support the notion that amSTN and VC may differentially affect CSTC imbalance, and suggest that VC stimulation may, in part, be effective through improvements in mood, whilst amSTN stimulation potentially reduces compulsivity, in part, via influencing prefrontal function.

Acknowledgements: This study was conducted at Institute of Neurology and National Hospital for Neurology and neurosurgery, University College London and supported there by MRC MR/J012009/1 with additional support for the cognitive testing from a Wellcome Trust Senior Investigator Award to TW Robbins (104631/z/14/z) and a joint award from the Medical Research Council and the Wellcome Trust (G0001354).

Keywords: Obsessive Compulsive Disorder, Deep Brain Stimulation, Cognitive Flexibility, STN, VC/VS

Disclosure: Nothing to Disclose.

W9. Randomized Controlled Trial of Hydrocortisone and D-Cycloserine on Fear Extinction in PTSD

Sabra Inslicht*, Mohammed Milad, Scott P. Orr, Charles Marmar, Thomas Neylan

University of California, San Francisco, California, United States

Background: Prolonged exposure therapy, one of the most well-studied and effective treatments for posttraumatic stress disorder (PTSD) is based on principles of fear extinction. Yet, a significant proportion of PTSD patients remain symptomatic following prolonged exposure. This may be a result of deficits in fear extinction that has been found in laboratory fear conditioning studies with PTSD patients. Pharmacological agents that modulate conditioning and extinction may hold promise to enhance exposure based treatments. Cortisol is a stress hormone that modulates learning and memory and is dysregulated in PTSD. Preclinical evidence suggests that alterations in HPA axis function may affect the ability to extinguish fear. Initial animal studies indicate that Hydrocortisone facilitates fear extinction. D-Cycloserine (DCS) is an N-methyl-D-aspartate (NMDA) receptor partial agonist that has been shown to enhance fear extinction and reduce reinstatement of conditioned fear in animals and has been found to be a helpful adjunct to exposure therapy in humans with phobias and social anxiety disorder. The aim of this study was to examine whether Hydrocortisone or DCS can enhance fear extinction in individuals with PTSD. We predicted that PTSD participants receiving Hydrocortisone or DCS would have greater fear extinction, greater retention of extinction, and less reconditioning compared to those receiving placebo.

Methods: Using a double-blind placebo-controlled experimental design, 88 trauma-exposed veterans with PTSD underwent a fear conditioning task, in which they were exposed to computer-generated colored circles (CS) that were paired (CS+) or unpaired (CS-) with an aversive electrical stimulus (US). After 72 hours participants were randomly assigned to one of three groups, Hydrocortisone, DCS or placebo. The appropriate medication (25 mg. Hydrocortisone, 50 mg. DCS, or placebo) was administered 1 hour prior to the extinction task. During extinction, participants were exposed to CSs without exposure to the US. Extinction recall was tested one week later in which participants were shown the previously conditioned stimuli again with no paired aversive stimuli. This was followed by reconditioning, in which one US was subsequently administered with a CS+ stimulus. Skin conductance levels were assessed throughout each task. Square root skin conductance responses to the CS+ stimuli compared to CS- stimuli were computed and served as the dependent variable. Mixed model analyses were conducted to examine drug by stimulus effects on skin conductance scores over repeated trails.

Results: Mixed model analyses indicated greater responses to CS+ compared to CS- during fear acquisition (p<.05), indicating successful fear conditioning. Extinction training 3 days later was successful. However, there was no effect of either drug on extinction learning, extinction retention, or reconditioning compared to placebo (all p's >.05).

Conclusions: These findings suggest that Hydrocortisone and DCS do not facilitate fear extinction in comparison to placebo in PTSD. Our findings contrast limited previous work in humans that has shown the potential for Hydrocortisone or DCS to enhance extinction. The failure of these medications to produce an effect above placebo could be related to dosage or to our task. Future investigations of these drugs could consider using higher doses or a longer schedule of administration of medication. There may have been a floor effect related to our laboratory task in PTSD patients. Using trauma-relevant cues or imaginal exposure may have provided a more robust test of extinction. In conclusion, Hydrocortisone and DCS were not more effective than placebo in enhancing extinction in individuals with PTSD.

Keywords: PTSD, Fear Extinction, Hydrocortisone, d-cycloserine

Disclosure: Nothing to Disclose.

W10. Identification and Manipulation of Fear Extinction Engrams in the Hippocampus

Anthony Lacagnina, Christine Ann Denny, Michael Drew*

University of Texas at Austin, Austin, Texas, United States

Background: Fear extinction is a form of exposure therapy in which repeated presentations of a fearful stimulus in the absence of threat gradually reduce fear. Instead of erasing the original memory, extinction appears to create a parallel memory trace that inhibits or competes with the original memory. Acquisition of extinction learning is believed to involve plasticity in amygdala and prefrontal cortex, but the mechanisms controlling retrieval of extinction are not well understood. Based on recent evidence that activity of granule cell ensembles in the dentate gyrus (DG) is necessary and sufficient for recall of contextual fear memories, we assessed the role of these cells in acquisition and recall of contextual fear extinction. We used activity-dependent neural tagging to investigate whether fear acquisition and fear extinction recruit distinct neural ensembles in DG, and whether the activity of these ensembles modulates fear expression after extinction.

Methods: Activity-dependent tagging was performed using Arc-CreERT2 mice crossed with ROSA26-CAG-stopflox-Halo-eYFP mice or ROSA26-CAG-stopflox-ChR2-mCherry mice. Mice received foreground contextual fear conditioning followed by ten 5-minute sessions of context extinction. Mice were injected with tamoxifen immediately after the fear training session or the 10th session of extinction to activate Halo or ChR2 expression in an activity-dependent manner. Mice were re-exposed to the conditioning context or a neutral alternate context either 5d or 30d after activity-dependent tagging. Acquisition-tagged or extinction-tagged neurons in the DG were optogenetically activated or inhibited during the test sessions using blue or green light targeted to the dorsal DG.

Results: When fear acquisition cells were tagged, their probability of reactivation during re-exposure to the training context was reduced after extinction training. Extinction training did not, however, reduce the overall number of cells acutely activated by context re-exposure, suggesting that extinction activates an ensemble that is distinct from the fear acquisition ensemble. Silencing extinction cells in the DG increased fear during a test of extinction retrieval, but had no effect during a spontaneous recovery test one month after extinction. Conversely, silencing fear acquisition cells in the DG had no effect during a test of extinction retrieval but reduced fear during the spontaneous recovery test one month later. The behavioral effects of silencing either population were specific to the conditioned context, suggesting that the manipulations modulate expression of specific contextual memories rather than general emotional states. ChR2-mediated activation of fear extinction cells suppressed freezing in the conditioning context and enhanced exploration in an anxiogenic novel environment.

Conclusions: Our findings suggest that contextual fear acquisition and extinction memories are coded by unique neural ensembles in the DG. Activity of extinction cells is required for expression of fear extinction, whereas activity of fear acquisition cells is required for expression of spontaneous recovery. We hypothesize that expression of fear versus extinction memories is determined by competition between ensembles in the hippocampus and that the strength and durability of extinction can be enhanced by stimulating the activity of hippocampal extinction cells.

Keywords: Fear Extinction, Dentate Gyrus, Fear Conditioning, Hippocampus

Disclosure: Nothing to Disclose.

W11. Anxiety Cells in a Hippocampal-Hypothalamic Circuit

Jessica Jimenez*, Katy Su, Alexander Goldberg, Victor Luna, Gokhan Ordek, Pengcheng Zhou, Samantha Ong, Stephanie Pena, Larry Zweifel, Liam Paninski, Rene Hen, Mazen Kheirbek

Columbia University, New York, New York, United States

Background: A shared feature of a number of anxiety disorders is the overestimation of threat, leading to enhanced avoidance. Yet the mechanisms and neural circuits by which normal adaptive avoidance behaviors arise, and how these circuits become disordered in psychiatric illness remain elusive. Although the ventral hippocampus (vHPC) has been implicated in regulating anxiety and mood-related behavior, little is known about how the vHPC represents emotionally salient information and how those representations contribute to behavior. Therefore, elucidating how anxiety provoking contexts are represented within the vHPC will be critical to understanding how it may guide avoidance behaviors during conflict-based anxiety tasks.

Methods: We have used cell-type specific calcium imaging in freely behaving mice using miniaturized microscopes to visualize vHPC activity during anxiety-related behaviors. We expressed GCaMP6f in vHPC, and a gradient index lens was implanted above vCA1. We imaged the activity patterns of the same population of vCA1 neurons across multiple behavioral dimensions, including innately anxiogenic tasks (Open Field Test (OFT), and Elevated Plus Maze (EPM)), and innately rewarding tasks (novel object exploration). For imaging vCA1 neurons defined by their projection target, we used a retrograde Cre viral approach (CAV2-Cre) to selectively express GCaMP6f in vCA1 neurons projecting to the Basal Amygdala (BA) or Lateral Hypothalamus (LHA). For optogenetic manipulations, we virally expressed Arch or ChR2 opsins in vHPC and implanted a fiber optic either directly in vHPC or at vHPC terminal fields in the BA or LHA.

Results: In our imaging experiments, we found that the majority of vCA1 neurons exhibit heightened activity during exploration of innately anxiety-provoking environments, including the open arms of the EPM and center of the OFT, but not to a rewarding novel object that elicited approach. We next tested whether this vCA1 activity increase was directly modulating anxiety behavior. We expressed the inhibitory opsin Arch selectively in vCA1 pyramidal neurons and silenced their activity only during exploration bouts of the anxiogenic open arms of the EPM. We found that inhibiting activity of vCA1 anxiety cells increased exploration of the anxiogenic EPM open arms, indicating that vCA1 anxiety cell activity is necessary for avoidance behavior. We next investigated in which hippocampal circuits these anxiety cells were distributed. We found that this enrichment for anxiety cells in vCA1 was not found in mice imaged in the dorsal hippocampus, a sub-structure of HPC thought to be more specialized in spatial navigation and contextual processing than in mood-related behaviors. Moreover, imaging inhibitory interneurons in vCA1 by selectively expressing Cre-dependent GCaMP6f in vesicular GABA transporter (VGAT)-Cre mice indicated that heightened activity in anxiogenic environments was restricted to vCA1 pyramidal neurons. We next assessed whether these vCA1 anxiety cells could be segregated into projection-specific cell populations. We first determined which vCA1 limbic output streams are involved in mediating innate anxiety behavior, and employed optogenetic techniques in vCA1 downstream terminal fields in two subcortical nuclei implicated in anxiety, fear, and behavioral responses to stress, the BA and LHA. We found that modulation of vCA1-LHA terminals impacts innate anxiety and aversion, but not contextual fear conditioning, while vCA1-BA terminal modulation impacts contextual fear memory but not innate anxiety behavior. Further, retrograde tracing studies in these terminal fields revealed that vCA1-LHA and vCA1-BA projections represent largely non-overlapping cell populations. To determine how these projection-specific cell populations differentially process innate anxiety and learned fear behaviors, we next utilized retrograde labeling techniques to selectively express the calcium indicator GCaMP6f in vCA1-LHA or vCA1-BA neurons and imaged their activity in innately anxiogenic environments. We found that vCA1-LHA projecting neurons were highly enriched in anxiety cells relative to the vCA1-BA projecting population, with ~80% of all vCA1-LHA projecting neurons responding to anxiogenic environments. These results support the hypothesis that negative valence information is represented at the level of vCA1, is necessary for avoidance behavior, and is enriched in neurons projecting to the LHA.

Conclusions: Our findings provide novel insights into the representation of innately aversive information in the vHPC, and the role of pathway-specific vCA1-subcortical projections in the generation of anxiety-related behavior. The identification of a novel vCA1-LHA circuit that rapidly controls anxiety-related behavior, without impacting learned fear, may provide novel targets for the treatment of mood and anxiety disorders.

Keywords: in vivo Calcium Imaging, Ventral Hippocampus, Anxiety, Lateral Hypothalamus, Basolateral Amygdala

Disclosure: Nothing to Disclose.

W12. Temporal Dynamics of Memory-Linking

Phil (Zhe) Dong, Mimi La-Vu, Christopher Lee, Brandon Wei, Tristan Shuman, Alcino Silva, Denise Cai*

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

Background: Posttraumatic stress disorder (PTSD) is a debilitating disorder characterized by re-experiencing a traumatic event in the form of persistent and intrusive memories. Determining how memories generalize and are reactivated is an essential part of understanding and treating this disorder. Recently, we found that fear can transfer from an aversive context to a safe context if the two distinct contexts were initially encoded close in time (Cai et al., 2016). Importantly, these two experiences shared a common neural representation that led to the reactivation of a fearful association memory even in a safe environment.

Methods: Mice explored 2 different contexts either 5h, 1d, 2d or 7d apart. Negative valence was altered by the pairing of a footshock with one of the context. Mice later returned to each of the context for a recall test. We used in vivo calcium imaging with Miniscopes to record the neural activity while mice encoded and retrieved the memories of both contexts.

Results: We found a high overlap between the neural ensembles representing two neutral contexts when the two contexts were separated by hours but not days. Similarly, there was a decrease in behavioral linking across days. Adding negative valence during the encoding of one of the contexts extended the temporal window of ensemble overlap between the fearful and neutral contexts and increased the memory-linking retrospectively (i.e. the fear of the shocked context transferred to previously seen neutral contexts), but not prospectively.

Conclusions: These surprising findings indicate that there is a circuit mechanism to retrospectively link aversive memories to past experience. Decoding how and when the hippocampus binds together aversive with safe events through overlapping ensembles is critical in knowing how to diminish transferring fear to a safe context. These findings have important implications for how extreme negative valence alters the process of memory-linking and may lead to pathological PTSD symptoms.

Keywords: Fear Generalization, Memory Encoding and Retrieval, in vivo Calcium Imaging

Disclosure: Nothing to Disclose.

W13. Impairments in Cognitive and Emotional Processes in Mental Illness: Transdiagnostic Meta-Analyses of Neurocircuit Structure and Function

Lisa McTeague*, Benjamin Rosenberg, David Carreon, Julia Huemer, Ying Jiang, Christina Chick, James Lopez, Eickhoff Simon, Amit Etkin

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Recently we investigated across Axis I disorders whether a shared substrate might be revealed in regional gray matter volume (Goodkind et al., 2015). We utilized a meta-analysis of patient/control differences in 193 studies of voxel-based morphometry (VBM) across a range of diagnoses (bipolar and unipolar depression, anxiety, substance abuse/dependence, psychosis). We observed that gray matter volume was decreased in the bilateral insula and dorsal anterior cingulate across diagnoses. We followed this with a parallel transdiagnostic functional neuroimaging meta-analysis of 283 studies of cognitive functioning (McTeague et al., 2017). Across disorders abnormal activation was evident in the cognitive control or “multiple demand network” including left prefrontal cortex, as well as anterior insula, right ventrolateral prefrontal cortex, right intraparietal sulcus, and mid-cingulate/pre-supplementary motor area. Disruption was also observed in a more anterior cluster in the dorsal cingulate cortex. Notably the cognitive control network shown to be disrupted transdiagnostically, interfaces with the anterior-cingulo-insular or “salience network” demonstrated to be transdiagnostically vulnerable to gray matter reduction. In the current meta-analysis, we investigated the extent to which these transdiagnostic functional and structural network disruptions would overlap with neurocircuit disruptions in emotional processing.

Methods: Studies on emotional functioning across a range of Axis I diagnoses (bipolar and unipolar depression, anxiety, substance abuse/dependence, psychosis) and control participants were submitted to meta-analysis with Activation Likelihood Estimation. Emotion provocation tasks during functional brain imaging that reported significant whole-brain, voxel-wise group differences in stereotaxic space were included. Reported contrasts included pleasant and/or unpleasant relative to neutral or baseline. Meta-analysis utilized peak voxel coordinates for activation likelihood estimation of patterns reflecting patient hyper- or hypo-reactivity in relation to control participants.

Results: The resulting set of studies (n=290) included 5,272 patients and 5,367 control participants. Studies were predominantly passive reactivity (n=220) with functional MRI (n=275). Pooling across patient hypo- and hyper-activation and valence revealed aberrant activation in limbic and paralimbic regions including bilateral amygdala extending to parahippocampal and hippocampal gyri, thalamus, dorsal anterior cingulate, and right anterior insula extending to ventrolateral prefrontal cortex (vlPFC). Separate contrasts of patient hyper- versus hypo-activation revealed that the aberrant activation pattern was primarily attributable to hyper-activation. Only a cluster in the right vlPFC showed patient hypo-activation. Further refining contrasts by valence demonstrated that the overall aberrant activation was primarily due to patient hyper-activation during unpleasant processing. Effects during pleasant processing were only observed in patient hypo-activation in the caudate and dorsal anterior cingulate. ALE by disorder revealed that basal ganglia effects were more characteristics of psychotic disorders whereas the limbic and paralimbic dysfunction was more characteristic of non-psychotic disorders, particularly bipolar and anxiety disorders. Considering regions of concordant dysfunction in structure and function (i.e., comparison across the VBM and two functional neuroimaging meta-analyses) revealed shared impairment particularly in right anterior insula extending to vlPFC and dorsal anterior cingulate.

Conclusions: The findings from these two large-scale functional neuroimaging meta-analyses of Axis I disorders suggest that cognitive and emotional perturbations amidst psychopathology are most pronounced within networks promoting adaptive functioning and moreover, are evident across disorders (i.e., transdiagnostically). That is, cognitive disruptions are evident across both frontal-parietal and cingular-opercular networks whereas emotional reactivity aberrations are more pronounced in cingular-opercular nodes. Demonstrating an essential functional parallel to our prior structural findings, the cingular-opercular or salience network was highlighted as a particular common pathway to cognitive and emotional dyscontrol in psychopathology and potentially as a powerful common target for therapeutic intervention.

Keywords: Meta-Analysis, Research Domain Criteria (RDoC), Transdiagnostic, Cognitive/Emotional Task Performance, Functional Neuroimaging

Disclosure: Nothing to Disclose.

W14. White Matter Biomarkers of Risk for Posttraumatic Anhedonia: A Prospective Examination

Negar Fani*, Zachary Clifford, Jennifer Stevens, Sanne van Rooij, Tanja Jovanovic, Barbara Rothbaum, Kerry Ressler, Vasiliki Michopoulos

Emory University School of Medicine, Atlanta, Georgia, United States

Background: Anhedonia, or the inability to experience pleasure, emerges in some people after psychological trauma. Three symptoms of PTSD— feelings of detachment from others, diminished emotional responsiveness, and decreased interest in activities—reflect posttraumatic anhedonia (PTA). These symptoms have been linked to significant impairment in social and occupational functioning. Anhedonic features of depression have been linked to abnormal neural network connectivity, particularly within white matter pathways such as the cingulum bundle and uncinate fasciculus. However, no extant longitudinal studies have examined how abnormal connectivity of white matter pathways at the time of trauma may help give rise PTA, which was the goal of the present study.

Methods: Twenty-nine men and women aged 19-58 (Mean=34, SD=12.2) were recruited from the emergency department of a level 1 trauma center as part of an NIH-funded study of PTSD biomarkers. Participants received clinical assessment and diffusion tensor imaging (DTI) at 1 month and 6 months post-trauma. The Predicting PTSD Questionnaire (PPQ), which includes items related to history of childhood trauma, peritraumatic dissociation, family history of psychopathology, and prior drug use, was used to account for variance in PTA associated with clinical risk factors. The PTSD Symptom Scale was used to measure PTSD symptoms at 6 months post trauma; three items were summed to create a PTA index score, in accordance with prior studies. To examine structural connectivity between regions of interest implicated in anhedonia and PTSD, probabilistic fiber tracking was conducted with DTI data. We used PROBTRACKX implemented in FSL; fractional anisotropy (FA) values were extracted from cingulum bundle and uncinate fasciculus tracts to assess white matter integrity. Logistic regression was used to assess the contributions of clinical risk factors (PPQ score) and white matter integrity (cingulum bundle and uncinate fasciculus FA) at the time of trauma to the emergence of PTA symptoms (presence or absence) at 6 months post-trauma.

Results: At 6 months post-trauma, few participants exhibited PTA symptoms (Mean PTA score=1.5, SD=2.8, Median and Mode=0); 21 endorsed no PTA, 9 endorsed a PTA score of 1 or more. Logistic regression results indicated that PPQ score did not significantly contribute to the overall model (Beta=-.9, p=.1), but the integrity of the uncinate fasciculus significantly contributed (Beta=-80.8, p=.04), and the left cingulum bundle added marginally to the overall model (Beta=-37.9; p=.07).

Conclusions: These findings indicate that the integrity of the uncinate fasciculus and the left cingulum bundle at the time of trauma may directly affect vulnerability for the development of PTA symptoms following psychological trauma, even accounting for clinical risk factors. These findings replicate our earlier cross-sectional studies of white matter in PTSD and highlight the salience of these pathways for the development of anhedonic symptoms of PTSD.

Keywords: PTSD, Risk, Longitudinal Analysis, White Matter

Disclosure: Nothing to Disclose.

W15. Characterization of the Amplificatory Effect of Norepinephrine in the Acquisition of Pavlovian Threat Associations

Lorenzo Diaz-Mataix*, Walter T. Piper, Hillary C. Schiff, Clark H. Roberts, Sara A. Stark, Vincent D. Campese, Robert M. Sears, Joseph LeDoux

New York University, New York, New York, United States

Background: Norepinephrine (NE) is involved in many adaptive physiological and pathophysiological processes, being released in response to a large variety of stimuli. By acting during learning, NE strengthens Pavlovian threat memories. However, the mechanisms underlying this effect are unknown. Here, we examine the temporal dynamics underlying the NE-induced enhancement of these associative memories.

Methods: Subjects were 167 male Sprague-Dawley rats weighing 225-400g at the beginning of the experiments. All conditions and procedures followed the National Institutes of Health Guide for the Care and Use of Experimental Animals and were approved by the New York University Animal Care and Use Committee. The beta-adrenergic antagonist (±) propranolol hydrochloride was freshly prepared before use. On day one, rats were habituated to the conditioning context. Twenty-four hours later the rats were weighed and randomly assigned to be injected i.p. with either 10 mg/mL of propranolol (Prop) or its vehicle (saline), given at 1mL/kg, thirty minutes before the conditioning session. Conditioning consisted of the presentation of one or two pairings, depending on the experiment, of a 30 second tone CS (5kHz; 80dB) that co-terminated with a 1 second footshock US (0.4 or 0.6mA). Forty-eight hours later, long-term memory (LTM) was tested in a modified context by presenting 10 unreinforced CS's. All sessions were recorded for posterior analysis of freezing by at least two blind raters. Statistical significance was assessed by a two-way ANOVA followed by post hoc tests when appropriate.

Results: Long-term threat memory (LTM) is impaired by pretraining propranolol in animals conditioned with 2 CS–US pairing trials but not in animals that received 1 CS–US trial. This pattern of results suggests that the first conditioning trial induces an increase in the release of NE that strengthens the learning of the following CS–US presentation. In order to confirm this finding, a second experiment was conducted in which the rats were exposed to an unsignaled US (U-US) prior to conditioning with only one CS-US pairing. The animals that were exposed to high intensity U-US and treated with propranolol exhibited significantly less freezing than their vehicle controls during LTM. This provides further evidence for the hypothesis that the first shock during conditioning can be the trigger of the NE-induced facilitation of learning. After a footshock, the concentration of NE in the amygdala peaks 10 minutes after the shock. The increase of NE after a shock returns to basal levels 30 minutes after the shock. Accordingly, the NE facilitatory effect must be circumscribed to the temporal window in which the concentration of the neurotransmitter is increased. To test this, another experiment was conducted. 30 min after systemic propranolol or saline, rats underwent two-trial conditioning with an intertrial interval (ITI) of either 2 or 40 minutes. Propranolol only had an effect on the animals conditioned with an ITI of 2 minutes indicating that propranolol only affects memory formation when the second CS–US pairing occurs within a certain time following the first CS–US pairing.

Conclusions: The studies presented here show that the NE-induced facilitation of learning is governed by precise temporal dynamics initiated by an increase in the NE release triggered by the footshock presented as part of the first conditioning association. These results help define the conditions under which NE should and should not be expected to alter threat processing and fill an important gap in the understanding of the neural processes relevant to the pathophysiology of stress and anxiety disorders.

Keywords: Norepinephrine, Pavlovian Conditioning, PTSD, Mood and Anxiety Disorders, Translational Pharmacology

Disclosure: Nothing to Disclose.

W16. Impairments in Cognitive Flexibility Relevant to OCD and Accompanying Alterations in Cortico-Striatal Activity in SAPAP3 Knockout Mice

Elizabeth Manning, Mary Torregrossa, Susanne Ahmari*

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: Functional imaging studies have strongly implicated cortico-striatal circuit dysfunction in the pathophysiology of obsessive compulsive disorder (OCD). However, the mechanisms by which this dysfunction gives rise to OCD symptoms are unclear, with hyperactivity typically observed at baseline and during symptom provocation, and hypoactivity typically observed during cognitive testing. Studies in preclinical rodent models provide a unique opportunity to investigate this discrepancy. To date, transgenic mouse models have provided substantial insight about striatal dysfunction underlying OCD-relevant compulsive grooming. In contrast, there are no studies to date examining the neural mechanisms underlying cognitive impairment in OCD-relevant mouse models.

Methods: SAPAP3 knockout mice (KOs)–a leading transgenic OCD model– and WT littermate controls were tested in an operant reversal learning paradigm (VR2; based on Gourley et al, 2010) to assess cognitive flexibility [n=22 WT and 29 KO males with 2 cohorts combined; each cohort showed statistically significant genotype differences independently]. Cortical and striatal activation associated with training on day 1 of reversal learning was assessed via quantitative cFos expression in 12 regions of interest (ROIs). Analyses included cross-region correlation to infer network functional connectivity, and comparison of neural activity to reversal behavioral performance.

Results: SAPAP3 KOs were significantly impaired in reversal learning (p<0.001), with ~40% of mutant mice (n=13/29) failing to acquire a reversed contingency (criteria: <20 active lever presses per day across 5 days of reversal training). Reversal learning impairment was unrelated to the severity of compulsive grooming observed in SAPAP3 KOs. Impaired reversal learning was also seen in female SAPAP3 KO mice (n=9 WT, 8 KO; 4 KOs failed reversal and 4 KOs acquired the task; training day x active lever press interaction p=0.008). To rule out differences in motivation between groups, lever contingencies were returned to their original order after 5 days; SAPAP3 KOs that reversed and those that failed were indistinguishable using the old rule, suggesting that the mice that failed were still motivated to perform the task. Reversal learning-related cFos expression revealed significantly correlated activity (following Bonferroni correction p<0.05) only between striatal ROIs in wild-type controls (WT, n=11), whereas SAPAP3 KOs (n= 12) showed additional correlations between cortical ROIs and cortex and striatum ROIs. Analysis on a subset of SAPAP3 KO and WT mice that achieved reversal criteria on the first day of training suggests that this elevated correlated activity in SAPAP3 KOs may reflect compensatory network activation necessary for them to acquire reversal. Specifically, whereas no significant correlations were detected in WT mice, SAPAP3 KOs that acquired reversal showed significantly correlated cFos activation in several prelimbic prefrontal cortex (PrPFC) associated networks [PrPFC-nucleus accumbens (NAc) core, PrPFC-NAc shell and PrPFC-lateral orbitofrontal cortex (lOFC)].

Conclusions: Our studies are among the first to describe neurocognitive impairments in a transgenic OCD mouse model. These findings implicate compensatory neural activity in the PrPFC-NAc circuitry in successful reversal learning in SAPAP3 KO mice, in line with recent studies demonstrating that stronger functional connectivity in cortico-striatal circuits is associated with intact cognition in OCD patients (Vaghi et al., 2017). Ongoing studies using in vivo microscopy to measure neural activity in SAPAP3 KOs during reversal learning are directly testing this hypothesis. Our results also highlight the utility of using OCD-relevant cognitive paradigms in preclinical mouse models to gain mechanistic insight regarding the role of cortico-striatal circuit dysfunction in OCD.

Keywords: Obsessive Compulsive Disorder, SAPAP3, Reversal Learning, mPFC, Nucleus Accumbens

Disclosure: Nothing to Disclose.

W17. Structural Connectivity Predictors of CBT and SSRI Response in a Transdiagnostic Sample

Olusola Ajilore*, Julia DiGangi, Joshua Nathan, Jennifer Francis, Scott Langenecker, Alex Leow, Heide Klumpp, K. Luan Phan

University of Illinois at Chicago, Chicago, Illinois, United States

Background: There have been a number of studies that have identified white matter abnormalities across a number of psychiatric disorders. In a recent meta-analysis that examined common white matter abnormalities in emotional disorders, reduced fractional anisotropy was identified in a number of prefrontal white matter tracts (1). Very few studies have examined whether these white matter alterations. The purpose of the present study was to examine the structural connectivity correlates of treatment response to cognitive behavioral therapy (CBT) and selective serotonin reuptake inhibitors (SSRIs) in a transdiagnostic sample of patients.

Methods: A preliminary sample of participants from a larger RDoC study was selected for white matter connectometry analysis comprised of 11 healthy controls (mean age 25.8 ±11.6, 3M, 8F) and 21 patients (mean age 29.5±10.6, 6M, 15F). Patients received 12 weeks of either CBT or SSRI treatment. Among the assessments obtained with the Hamilton Rating Scale for Depression and Anxiety (HAM-D, HAM-A) and the Depression Anxiety Stress Scale (DASS). All participants had a MRI scan at the beginning and end of the 12-week period. A DTI diffusion scheme was used, and a total of 64 diffusion sampling directions with 8 b0 images were acquired on 3 Tesla GE Discovery MR750 System (Milwaukee, WI) at the UIC Center for Magnetic Resonance Research. The b-value was 1000 s/mm2. The in-plane resolution was 0.9375 mm. The slice thickness was 2.5 mm. Connectometry analysis was conducted using DSI Studio (2; http://dsi-studio.labsolver.org).

Results: Out of 21 patients sampled, 8 had a primary diagnosis of generalized anxiety disorder, 4 had a primary diagnosis of major depressive disorder, 4 with social anxiety disorder 3 with PTSD, and 1 with panic disorder (mean baseline DASS score 30±7.3). 10 participants received CBT and 11 were treated with SSRI treatment. There were no significant differences in SSRI and CBT treatment response according to HAM-D scores (p =.11) and HAM-A scores (p =.35). Overall there was a 70% response rate across both treatment groups. Connectometry analysis revealed that increased connectivity in the forceps minor over time that was significantly different in patients compared to controls and correlated with improving DASS scores.

Conclusions: In a preliminary transdiagnostic sample, there was a significant increase in connectivity in the forceps minor that also correlated with treatment response. This white matter region is increasingly emerging as a key circuit involved in treatment response across a wide range of emotional disorders.

Keywords: Diffusion Tensor Imaging, Research Domain Criteria (RDoC), CBT, SSRI, Connectome

Disclosure: Nothing to Disclose.

W18. Disruption of Prepulse Inhibition in SAPAP3 Knockout Mice is Associated With Severity of Compulsive Grooming and Dorsal Striatum D1 Receptor Binding

Elizabeth Manning*, Abigail Wang, Susanne Ahmari

University of Pittsburgh Medical Center, Western Psychiatric Institute & Clinic, Pittsburgh, Pennsylvania, United States

Background: Obsessive compulsive disorder (OCD) is a debilitating psychiatric disorder that affects 1-3% of the population. Impairments in sensorimotor gating have been demonstrated in patients, which may contribute to difficulties inhibiting intrusive thoughts and repetitive rituals. Prepulse inhibition (PPI) is a translational measure of sensorimotor gating that can be tested similarly in human subjects and preclinical models. The goal of this study was to measure PPI in SAPAP3 knockout mice (KOs), a widely used preclinical mouse model that displays OCD-relevant compulsive grooming and associated skin lesions that is sensitive to treatment with selective serotonin reuptake inhibitors (SSRIs). PPI was compared with compulsive grooming severity and measurements of dopamine and serotonin receptor and transporter binding in cortical and striatal regions that are implicated in the regulation of PPI and OCD pathophysiology.

Methods: Four cohorts of male and female SAPAP3 KOs and wild-type (WT) littermate controls were tested [Cohort 1: Across development as the compulsive grooming phenotype emerges (n=24 WT, 23 KO); Cohort 2: Naïve mice tested at an age when ~50% of KO mice show a compulsive grooming phenotype and accompanying skin lesions (n=13 WT, 18 KO); Cohort 3: Chronic treatment with SSRIs; and Cohort 4: Acute treatment with D1 receptor antagonist SCH23390 (n=16 WT, 16 KO)]. Brains were collected from mice from cohorts 2 and 3 following behavioural characterization, and autoradiography was used to investigate levels of dopamine D1 and D2 receptors, the dopamine transporter (DAT), and serotonin 5HT1B receptors.

Results: Monthly analysis from 2-8 months of age revealed progressive and heterogeneous development of compulsive grooming in SAPAP3 KOs. Compulsive grooming severity was significantly correlated with disruption of PPI (R= -0.57, p<0.001), although grouped comparisons of SAPAP3 KOs and WT mice did not reveal PPI impairments. We therefore categorically segregated SAPAP3 KOs based on skin lesion status as a proxy for grooming severity. This analysis demonstrated that SAPAP3 KOs with lesions had significantly lower PPI than both SAPAP3 KOs without lesions and WT controls; KOs without lesions and WT controls were not significantly different (lesioned KO vs WT p<0.05, lesioned vs no lesion p<0.01, WT vs no lesion p>0.05). PPI impairments in SAPAP3 KOs with skin lesions were rescued both by chronic treatment with the SSRI fluoxetine, and by acute treatment with the D1 receptor antagonist SCH-23390. Surprisingly, reduced D1 receptor binding in the dorsal striatum was significantly associated with PPI impairment in SAPAP3 KO mice (R= -0.55, p=0.017).

Conclusions: This is the first time that heterogeneity in the emergence and progression of OCD-relevant behaviors has been characterized in detail in a transgenic mouse model. Our observation that PPI is disrupted only in SAPAP3 KO mice with a severe compulsive grooming phenotype suggests that disruption of the circuits underlying sensorimotor gating may also contribute to the development of compulsive grooming. Future experiments will be necessary to further test causality and more carefully examine the relationship between these two OCD-relevant phenotypes. Pharmacological studies implicate excessive D1 receptor activation in disruption of PPI in SAPAP3 KOs. Autoradiographic binding studies demonstrating that reduced dorsal striatum D1 receptor binding is associated with impaired PPI in SAPAP3 KOs may reflect compensatory down-regulation of D1 receptors in response to hyperdopaminergic signalling. This would suggest that hyperdopaminergic signalling in the dorsal striatum plays a central role in disruption of PPI in SAPAP3 KOs. These preclinical studies may help to guide future clinical research in OCD patients exploring the relationship between sensorimotor gating deficits, compulsive behaviors, and dopaminergic system function in OCD.

Keywords: Obsessive Compulsive Disorder, Prepulse Inhibition, Compulsive Behavior

Disclosure: Nothing to Disclose.

W19. Dissociable Meta-Analytic Networks in Post-Traumatic Stress Disorder

Meredith Reid*, Julio Yanes, Jessica Busler, Jennifer Robinson

Auburn University, Auburn University, Alabama, United States

Background: Over the last few decades, researchers have identified numerous biological correlates of post-traumatic stress disorder (PTSD). However, this enhanced understanding has not produced reliable biomarkers. Functional neuroimaging studies have provided important insight into PTSD-related brain differences; however, some findings are inconsistent. Meta-analytic tools provide a more holistic perspective, aggregating published neuroimaging results and assessing statistical convergence. Here, we sought to establish which brain regions show consistent differences among PTSD patients (across neuroimaging paradigms) by examining a corpus of studies within the Activation Likelihood Estimation (ALE) framework. To provide enhanced interpretation, we used secondary analyses to fractionate impacted regions into sub-networks and to determine which psychological processes those sub-networks may implicate.

Methods: We searched the BrainMap database for published studies that used fMRI or PET to examine functional brain differences between patients with PTSD and controls, including trauma-exposed controls without PTSD and non-trauma-exposed controls. These studies examined brain differences in the context of cognitive, affective, motor, and perceptual paradigms. Inclusion was limited to studies that reported activation increases (PTSD>Controls) and activation decreases (PTSD<Controls), encompassing the whole-brain, with results communicated as coordinates (foci) in stereotaxic space. We used GingerALE to identify regions of convergent activation across studies for the contrasts PTSD>Controls and PTSD<Controls. Next, we parsed the meta-analytic regions into sub-networks using fractional similarity network analysis (FSNA). FSNA identifies regions exhibiting co-occurring diagnosis-related changes across studies. Then, we sought to link sub-networks with psychological processes. Specifically, we performed a decoding assessment using BrainMap to reveal paradigm classifications that were linked with each sub-network.

Results: We identified 44 studies, representing data from 707 patients and 758 controls, reported as 925 foci from 113 separate statistical contrasts. PTSD patients showed increased activation in the cuneus, cingulate, amygdala, hippocampus, insula, and DLPFC. FSNA and functional decoding suggested co-occurring changes among the amygdala, hippocampus, and DLPFC (sub-network 1) associated with the paradigms Emotion Induction, Affective Pictures, and Encoding; co-occurring changes among the insula, cingulate, and occipital lobe (sub-network 2) linked with Passive Viewing, Visual Pursuit/Tracking, and Saccades; and co-occurring changes between the occipital lobe and cuneus (sub-network 3) linked with Theory of Mind. When considering activation decreases, PTSD patients showed convergent changes in the thalamus, caudate, cingulate, and prefrontal cortex. FSNA and functional decoding suggested co-occurring changes among the PCC, DMPFC, and DLPFC (sub-network 4) linked with Theory of Mind, Lexical Decisions, and Go/No-Go, as well as co-occurring changes among the caudate, ACC, VMPFC, and VLPFC (sub network 5) linked with Word Generation, Olfactory Monitoring/Discrimination, and Pitch Monitoring/Discrimination.

Conclusions: Our results indicate that PTSD is associated with region-specific changes across the brain. Specifically, sub-network 1 represents increased activations associated with inducing emotional states and processing emotional content. In addition, sub-networks 2 and 3 represent increased activations linked with monitoring visual content and subsequent evaluation. Finally, sub-networks 4 and 5 represent distributed decreased activation associated with decision-making, response inhibition, and effortful attention to one's environment. Deeper understanding of the neural systems that may be impacted by PTSD is important for establishing reliable biomarkers that may serve as targets for the development of novel therapies.

Keywords: PTSD, Functional MRI (fMRI), Meta-Analysis, Network Analysis

Disclosure: Nothing to Disclose.

W20. Ketamine Treatment During Reconsolidation of Traumatic Stress Exposure: Molecular Biological Validation of Reduced Avoidant Behavior in Marmoset Monkeys

Ingrid Philippens*, Laurijn Draaisma, Harm Krugers, Eric Vermetten

Biomedical Primate Research Centre, Rijswijk, Netherlands

Background: Memories for emotionally arousing and stressful events are retained very well in general. An extreme example occurs in posttraumatic stress disorder (PTSD) where emotional memories remain affectively over-consolidated. Several lines of evidence indicate that re-experiencing emotional memories renders these memories labile. Targeting this process of reconsolidation offers a potential window of opportunity to attenuate fearful traumatic memories and subsequent avoidant behavior.

Methods: Since the NMDA-receptor plays a role in the process of reconsolidation of arousing information, we tested the effect of ketamine, a noncompetitive NMDA receptor antagonist, on emotional memory formation and related molecular biological parameters (e.g. reflected in neurogenesis) in the adult common marmoset monkey (Callithrix jacchus). For consolidation of contextual emotional memory, an unescapable foot-shock paradigm in a passive avoidance task with two compartments (dark vs illuminated) was used. After entering the dark compartment, the monkeys (n=12) received four random foot-shocks (0.5 mA, 4 s) in a 15-minute time frame. This stressful exposure significantly increased salivary cortisol in all animals (Kruskal-Wallis test, p=0.0192). BrdU (200 mg/kg s.c.) was administered immediately after the stress exposure for measuring neurogenesis. One week later the monkeys were re-exposed to the same stressful situation. Prior to the re-exposure the monkeys were treated with ketamine (0.5 mg/kg i.m.; n=6) or vehicle (n=6). In the third week, the monkeys were again re-exposed to test their memory for the stressful situation.

Results: On behavioral analysis, the vehicle treated monkeys avoided entering the dark compartment. The ketamine treated monkeys entered the dark compartment significantly faster and more frequent (two-way ANNOVA, F(2,20)=19.18; p<0.0001). Preliminary post-mortem brain pathology showed these monkeys expressed significant less neurogenesis (BrdU+ and DCX+ cells) compared to the vehicle treated monkeys.

Conclusions: It is hypothesized that reduced neurogenesis prevented the formation of new neuronal connections and serve as a possible explanation how ketamine influenced the reconsolidation process. We think this study provided important findings for ketamine as a potential candidate to target traumatic memories and needs to be translated to patients with PTSD.

Keywords: PTSD, Non-Human Primate, Ketamine, Emotional-Memory, Neurogenesis

Disclosure: Nothing to Disclose.

W21. Stress-Induced Alterations in Functional Connectivity During Verbal Fluency

David Beversdorf*, Neetu Nair, John Hegarty, Emily Hover, Sara Hooshmand, Bradley Ferguson, Michael Tilley, Shawn Christ

University of Missouri, Columbia, Missouri, United States

Background: Stress has well-established effects on cognition. Effects of acute stressors have been observed on arousal, memory, as well as on flexible verbal problem solving tasks. Some evidence suggests that some of these effects may be adrenergically mediated, as the beta-adrenergic antagonist propranolol reverses these effects for flexible verbal problem solving, and as propranolol has widely been used off-label for the impairments associated with performance anxiety and text anxiety. While the effects of stress on resting brain networks have been explored, less is known about how stress might impact cognition during verbal tasks. To begin to determine how stress might impact language networks, we examined the effects of an established imaging-compatible stressor on connectivity during a verbal fluency task.

Methods: Thirty-five healthy individuals, age 18-24, without any history of medical or psychiatric history participated. Subjects participated in two imaging sessions. At one session, subjects were exposed to the Montreal Imaging Stress Test (MIST) while in imaging environment, derived from the Trier Social Stress Test, where participants completed challenging mathematical tasks under time pressure with a social evaluative threat. At the other session, subjects were exposed to the control task, involving mathematical tasks without time constraint or social evaluative treat. The order of the sessions was counterbalanced. During each session, participants were interrupted for 30 second epochs of letter fluency, where participants were asked to generate as many words as possible starting with a specified letter. Participants were also interrupted for 30 second blocks of rest. Images were collected on a 3T Siemens Trio Scanner at the University of Missouri Brain Imaging Center. Functional images were acquired for BOLD activation, superimposed on structural T1-weighted images. Functional MRI data were analyzed comparing fluency vs rest using FSL software Motion correction, brain extraction, Gaussian spatial smoothing (FWHM of 5 mm) and high pass temporal filtering were performed as part of preprocessing. Subjects who had relative motion greater than 2 mm along either of the x-, y- or z- axis were excluded from further analysis. The average time series from the regions of interest during task were extracted for each individual subject, including the left parietal lobe (LPL), left fusiform gyrus (LFFG), left and right inferior frontal gyrus (LIFG and RIFG), and the left middle temporal gyrus (LMTG). Functional connectivity analysis was performed, with correlation coefficients computed for each ROI pair for each subject for each session. Fisher's Z-transform was applied to the correlation coefficients and a repeated measures ANOVA was performed on the z-scores to determine differences across stress and no-stress conditions.

Results: No significant effect of stress was observed on letter fluency performance in the imaging environment. However, repeated measures ANOVA revealed a significant main effect of stress on functional connectivity across all ROI pairs (F=4.391, p=0.44), where mean functional connectivity was increased during the stress condition as compared to the control condition. Among the specific ROI pairs, increases in FC with stress were significant after correction for multiple measures for LIFG-LPL connectivity, LIFG-LFFG connectivity, and RIFG-LFFG connectivity. There was no significant effect of gender on any of these findings.

Conclusions: These findings suggest that the effect of acute psychosocial stress on verbal tasks is associated with increased functional connectivity. However, this was observed during performance of a language task that is less susceptible to impairment due to stress. Utilization of this well-established task which is less susceptible to stress effects has the advantage of avoiding a confound of task performance on the imaging results, but it leaves open the question as to the relationship between functional connectivity and task performance in the setting of stress. This will need to be explored in tasks more susceptible to the effects of stress. Additionally, the direction of the effect on connectivity is of interest, since increased connectivity was observed with stress, whereas previous literature had demonstrated that propranolol, the agent used to treat situational stress-induced anxiety, also increased functional connectivity in language areas in patient populations. Future studies should incorporate pharmacological intervention in order to better understand its effects in this setting.

Keywords: Acute Stress, Language, Functional Connectivity

Disclosure: Nothing to Disclose.

W22. Trauma Exposure and Psychopathology in an Urban Sample of Women

Vasiliki Michopoulos*, Rachel Gluck, Georgina Hartzell, Hayley Dixon, Abigail Lott, Sierra Carter, Negar Fani, Tanja Jovanovic, Ann Schwartz, Kerry Ressler, Bekh Bradley, Charles Gillespie

Emory University, Atlanta, Georgia, United States

Background: Post-traumatic stress disorder (PTSD) and major depressive disorder (MDD) are common and disabling psychiatric disorders that are associated with trauma exposure. Epidemiological studies of urban communities estimate that between 65-90% of this population have been exposed to trauma, with approximately 50% and 37% of individuals meeting diagnostic criteria for lifetime PTSD and MDD, respectively. Factors contributing to the high rates of psychopathology in low-income, urban, minority populations include high lifetime risk for experiencing multiple traumas and recurrent exposure to interpersonal violence. The impact of trauma on the mental health of minority women living in low-income, urban settings has received little research attention. This is especially relevant as substantially higher rates of PTSD and MDD are observed in women compared to men and both PTSD and MDD, particularly comorbid PTSD/MDD, are also associated with increased risk for substance abuse as well as risk and progression of stress-related medical disorders including type 2 diabetes mellitus and cardiovascular disease. To help fill the gap in the clinical research literature on the impact of trauma exposure in minority women, we examined the demographic characteristics, rates and types of trauma exposure, prevalence of PTSD and MDD, and prevalence of PTSD/MDD comorbidity within an urban, predominantly African-American population of women.

Methods: Study participants were approached while in the waiting rooms of primary care or obstetrical-gynecological clinics of Grady Memorial Hospital in Atlanta, Georgia from 2005 to 2017. Eligibility criteria included being female, at least 18 years old, not actively psychotic, and able to give informed consent. Initial screening interviews were conducted in women (n=7967) to capture information regarding trauma history, PTSD and depressive symptoms, and psychological variables with questionnaires as previously described by our group. A subset of women (n=927) self-selected to return for a structured clinical interview that included the Clinician Administered PTSD Scale (CAPS), the Structured Clinical Interview for DSM-IV (SCID), and the Mini International Neuropsychiatric Interview (MINI). All procedures were approved by the institutional review boards of Emory University School of Medicine and Grady Memorial Hospital, Atlanta, Georgia.

Results: The self-identified race of our sample was predominantly African-American (93.5%). The majority of study participants (65.3%) were unemployed at the time of initial interview and almost one-fifth of the study participants (18.6%) reported receiving disability payments. The socioeconomic status of the majority of study participants was low, with 83.2% reporting an average monthly household income of less than $2,000. 90% of the study participants reported experiencing some form of significant trauma in their lifetime, with exposure to a serious accident or injury being the most common traumatic event (44.8%) and assault without a weapon by an intimate partner as the second most endorsed type of trauma exposure (37.5%). Rates of childhood abuse were also very high in this sample, as 28.7% of study participants reported emotional abuse and 18.9% reported physical abuse in childhood. Additionally, 28.3% of study participants reported sexual contact before the age of 13, 17.3% reported sexual assault between age 14 and 17, and 14.2% reported sexual assault at age 17 or older. Current and lifetime prevalence of PTSD as assessed by the CAPS was 29.6% and 53.8%, respectively. Current and lifetime prevalence of MDD as assessed by the SCID and MINI was 22.0% and 49.9%, respectively. Comorbidity of current PTSD and MDD was 15.2% and prevalence of current PTSD without MDD was 14.3% whereas prevalence of current MDD without PTSD was 6.8%.

Conclusions: Our results show high rates of adult and childhood trauma exposure, and high rates of PTSD, MDD, and comorbid PTSD/MDD in a sample of urban, low income, women. These data extend previous reports of high trauma exposure and psychopathology in the public health care system, by focusing on women, many of whom are exposed to high levels of poverty that are more likely to suffer from stress and trauma-related psychopathology as well as stress-related medical disorders. Socioeconomic factors which influence the likelihood of exposure to neighborhood violence and limit options to escape a violent or unsafe living situation were also associated with adverse outcomes in the current study, suggesting that social determinates of health (i.e. lack of stable housing, poor relationships with law enforcement, limited education, lack of access to health care, and poor financial resources) may be important targets for intervention to reduce the adverse sequelae of trauma exposure.

Keywords: Post Traumatic Stress Disorder, Depression, Trauma Exposure, Women's Mental Health

Disclosure: Nothing to Disclose.

W23. Evidence for Circulating PACAP as a Biomarker for Anxiety Disorders in Women

Rachel Ross*, Susanne Hoeppner, Emily O'Day, Samantha Hellberg, Peter Rosencrans, Mireya Nadal, Ki Goosens, Kerry Ressler, Victor May, Naomi Simon

McLean Hospital, Harvard Medical School, Belmont, Massachusetts, United States

Background: Pituitary adenylate cyclase activating polypeptide (PACAP) is a neuropeptide and hormone released in the hypothalamus and adrenal gland thought to modulate the HPA axis. In females with PTSD, circulating PACAP levels have been correlated with symptoms. Given that patients with anxiety-spectrum disorders have maladaptive responses to stress, and PACAP's previously identified role in the stress response, we hypothesize that circulating PACAP will correlate with a range of anxiety disorder diagnoses.

Methods: Serum samples from 47 well characterized adults with generalized anxiety disorder (GAD), social anxiety disorder (SAD), panic disorder (PD), posttraumatic stress disorder (PTSD), and healthy controls (HC), recruited and formally assessed at the MGH Center for Anxiety and Traumatic Stress Disorders, were assayed for PACAP by ELISA. Then, a larger group of 211 serum samples from individuals with primary GAD (n=91), posttraumatic stress disorder (PTSD; n=8), or HC (n=112) were assayed for PACAP by the more sensitive radioimmunoassay. These data were analyzed by 3-way ANCOVA, testing diagnosis, gender, and use of psychotropic medications, with time of day of sample collection as a covariate. Of this larger cohort, 96 samples, all from individuals with primary GAD (n=88) or PTSD (n=8), were then genotyped for a PAC1R SNP associated with elevated symptoms of PTSD (the inclusion of participants with PTSD in our sample set was to use this data as a positive control).

Results: In our first set of samples (n=47) we found that PACAP in serum from males was below the level of detection for the validated PACAP ELISA, supporting the observation of sex differences in circulating PACAP levels with the use of this assay. Of the female samples (n=19), all but 3 were detectable, and these 3 were healthy controls, indicating a trend that elevated circulating PACAP may behave as a biomarker for those with anxiety-spectrum diseases. Of the 16 samples with detectable serum PACAP levels, diagnoses include GAD (n=5), SAD (n=4), and PD (n=4), PACAP serum levels were similar to those with PTSD (n=3), which were run as a positive control. When serum samples were subjected to radioimmunoassay, we excluded 4 samples whose readout on RIA was zero, and recoded all samples with RIA result was less than 9 as equal to 9, based on assay sensitivity. Samples were collected between 10am and 4pm based on when participants presented to the clinic, but there was no effect of time of blood draw on PACAP level. HAM-A14 scores (mean±SD) by diagnosis were: HAM-AHC 3.4±3.8, HAM-APTSD 22.3±6.9, and HAM-AGAD 18.0±6.6. Circulating PACAP tended to be lower in individuals diagnosed with GAD or PTSD, with lower PACAP levels inversely correlated to HAM-A score (PACAPHC=16.20ppm, PACAPANX=14.26ppm, PACAPPTSD=12.34ppm; p=0.11, and males tended to have slightly higher circulating PACAP than females (PACAPMale=16.31±6.12ppm, PACAPFemale=15.12±5.4 ppm; p=0.16, but neither of these were statistically significant. The overall outcome of the analysis of interaction of diagnosis, gender, and psychotropic medication use was non-significant (p=0.2). Ongoing genotype data will be combined with these serum PACAP levels to examine the interaction of PAC1 receptor genetic association with PACAP protein serum levels in this transdiagnostic cohort.

Conclusions: These initial data do not yet support the observation of sex differences in circulating levels of PACAP and potential sex differences in peripheral PACAP as a biomarker for anxiety, unlike with PTSD. This may relate to the biological based differences in these disorders, though mode of assay and sample handling is particularly important for the study of PACAP and greater sample sizes may also be needed. Genotyping and genetic markers may be a more stable way to differentiate biomarkers in psychiatric disease, particularly anxiety disorders.

Keywords: Anxiety Disorders, PACAP, Biomarkers, Sex Differences

Disclosure: Nothing to Disclose.

W24. Systematic Review and Meta-Analysis: Dose Response Curve of SSRIs and SNRIs in DSM-5 Anxiety Disorders

Ewgeni Jakubovski, Jessica Johnson, Kirsten Müller-Vahl, Michael Bloch*

Yale Child Study Center, New Haven, Connecticut, United States

Background: Our goal was to examine (1) the time-course of response to selective serotonin reuptake inhibitors (SSRIs) and Selective Norepinephrine Reuptake Inhibitors (SNRIs) for treating anxiety disorders and determine whether there were differences in efficacy (2) between SSRI and SNRI medication classes; (3) between individual SSRI/SNRI agents; (4) between DSM-5 anxiety disorders and (5) whether higher doses of SSRI/SNRIs were associated with an improved response.

Methods: We searched PubMed and Cochrane Central Register of Controlled Trials. We included (1) randomized placebo-controlled clinical trials of SSRI/SNRI in (2) adult patients with anxiety disorders that provided data at (3) three or more timepoints. Extracted data included weekly or bi-weekly anxiety scores for up to 12 weeks, type of medication utilized, maximum dosage of medication, dropout, side-effects, duration of trial and year of trial. We utilized linear mixed models in SAS to model time course of medication response relative to placebo. We used traditional endpoint meta-analysis in Comprehensive Meta-Analysis to examine risk of all-cause and side-effect related dropout, magnitude of improvement of anxiety symptoms and treatment response.

Results: Meta-analysis included 57 studies. A linear mixed model analysis based on weekly outcome data, suggested that a logarithmic model was the bestfit model for SNRIs whereas a linear model provided the bestfit for modeling the benefits of SSRIs compared to placebo. Higher doses of SSRIs but not SNRIs were associated with significantly greater treatment gains. Higher doses in both medication classes were associated with an increased likelihood of dropouts due to side-effects (but not all-cause dropout). The greatest benefits of SSRI/SNRIs were observed for social anxiety disorder.

Conclusions: Both SSRIs and SNRIs are effective for the treatment of anxiety disorders. Higher doses of SSRI medications within the therapeutic range are associated with greater treatment effect whereas higher doses of SNRIs are not. Higher doses of both SSRI and SNRI were associated with a significantly higher rate of dropout due to side-effects.

Keywords: Anxiety, SSRI, Antidepressants, Meta-Analysis

Disclosure: Part 1: Therapix Biosciences, Advisory Board, Part 4: Therapix Biosciences, Grant, Biohaven Pharmaceuticals, Grant.

W25. Resting State Innate Alarm System Neural Networks in PTSD and its Dissociative Subtype

Ruth Lanius*, Sherain Harricharan, Isadora Olive, Daniela Rabellino, Paul Frewen, Maria Densmore, Jean Theberge

University of Western Ontario, London, Canada

Background: Aberrant subconscious and conscious threat-detection mechanisms have been demonstrated in PTSD. The innate alarm system (IAS), originating in the superior colliculus (SC) and periaqueductal grey, models the neurocircuitry underlying threat processing in PTSD. Here, the PAG may work in tandem with the SC to initiate instinctual defensive strategies. The current study therefore examined resting state functional connectivity patterns with the PAG and the SC in PTSD, its dissociative subtype, and healthy controls.

Methods: PAG and SC resting-state functional connectivity was examined in PTSD, its dissociative subtype, and healthy controls using a seed-based approach via PickAtlas and SPM12.

Results: Almost no resting state PAG functional connectivity was observed in controls. However, PTSD patients exhibited widespread PAG functional connectivity with brain regions associated with emotional reactivity and defensive responses (e.g., cingulum, anterior insula, pre/post central gyrus). In terms of SC connectivity, we observed group-specific resting-state functional connectivity between the SC for both the PTSD and the dissociative subtype. We observed increased connectivity between the left SC and the right dorsolateral prefrontal cortex (DLPFC) in PTSD as compared to the dissociative subtype. The DLPFC is involved in modulation of emotional processes characterising active defensive responses in PTSD. When comparing PTSD+DS to PTSD, increased connectivity was observed between the right SC and the right temporoparietal junction (TPJ) in the dissociative subtype. The TPJ is involved in depersonalization responses characterising passive defensive responses typical of this subtype.

Conclusions: These findings point to the importance of identifying functional connectivity of the innate alarm system and related cortical structures involved in emotional reactivity and defensive posturing in PTSD and its dissociative subtype.

Keywords: PTSD, Superior Colliculus, Periaqueductal Grey (PAG), Innate Alarm System

Disclosure: Nothing to Disclose.

W26. Long-Term Efficacy of Cognitive-Behavioral Therapy for Pediatric OCD With and Without D-Cycloserine Augmentation

Aude Henin, Eric A. Storch, Sabine Wilhelm, Brent J. Small, Rachel Porth, Daniel Geller*

Massachusetts General Hospital, Boston, Massachusetts, United States

Background: Cognitive behavioral therapy (CBT) is considered a well-established and efficacious treatment for pediatric Obsessive-Compulsive Disorder (OCD). However, not all pediatric patients with OCD are treatment responders. Researchers have attempted to augment CBT with D-cycloserine (DCS), an N-methyl-D-aspartate (NMDA) partial agonist that has been shown to enhance extinction learning. Although DCS augmentation of CBT has had mixed results, it may be that the consolidation of extinction learning occurs over a longer period of time than has been examined in prior studies. Further, few studies have looked at whether treatment gains from CBT alone are maintained over the long-term. The present study examined OCD symptom severity at 1-, 3-, and 6-month intervals in treatment responders after a randomized, double-blind study of DCS augmentation of CBT. Because of the null findings in the original study, we predicted that treatment gains would be similar between patients in the DCS and placebo arms (5R01MH093402-02).

Methods: Two hundred and six children and adolescents were enrolled at either Massachusetts General Hospital (MGH) in Boston, MA or University of South Florida (USF) in St. Petersburg, FL. One hundred and forty-two patients with an OCD diagnosis per DSM-IV-TR criteria as determined by the Schedule for Affective Disorders and Schizophrenia for School-Age Children– Present and Lifetime Version (KSADS-PL) and a CY-BOCS≥16 (M=12.79, SD=2.99; range 7-17-years-old) were randomized to either DCS+CBT (n=70) or placebo+CBT (n=72; for full consort diagram, please refer to Storch et al., 2016). The current analysis consists of the one hundred and thirty-seven children considered treatment responders (CY-BOCS score≤14 and rated as “Much or Very Much Improved” on the CGI-I) at post treatment. The CY-BOCS, CGI-Severity, CGI-Improvement, COIS-P, MASC, and CDRS were administered at post-treatment, 1 month, 3 months, and 6 months after completing treatment. All procedures performed in studies involving human participants were in accordance with the ethical standards of the Institutional Review Boards at the USF and MGH. Written parental informed consent and child assent were obtained from all individual subjects participating in the study.

Results: As hypothesized, CY-BOCS and CGI-S scores declined significantly during the follow-up period, but with no effect of treatment group status. There were no significant changes in CGI-I scores over the follow-up period. Parent and child ratings of OCD symptoms were also consistent between groups, with both showing improvement over 6 months.

Conclusions: The results of the present study suggest that pediatric patients with OCD who are considered treatment responders continue to improve over the follow-up period. Further, patients treated with CBT with DCS for fear extinction learning augmentation did not improve more than those given CBT alone. These findings suggest that the benefits of exposure and response prevention based CBT continue beyond the active treatment period, irrespective of treatment group. Future research should examine outcomes over even longer post treatment intervals, as well as the extent to which patient variables may impact response to DCS augmentation of CBT.

Keywords: d-cycloserine, Cognitive Behavioral Therapy, Obsessive Compulsive Disorder, Children and Adolescents

Disclosure: Nothing to Disclose.

W27. Stress Exposure and Psychopathology Alter Methylation of the Serotonin Receptor HTR2A Gene in Preschoolers

Andrew Novick*, Stephanie Parade, Justin Parent, Ronald Seifer, Samantha Klaver, Camen Marsit, Asi Polly Gobin, Bao-Zhu Yang, Audrey Tyrka

Brown University, Providence, Rhode Island, United States

Background: Serotonin signaling pathways play a key role in brain development, stress reactivity, and mental health. Epigenetic alterations in the serotonin system may underlie the effect of early life stress on psychopathology.

Methods: The current study examined methylation of the serotonin receptor gene HTR2A in a sample of 228 children including 119 with child welfare documentation of moderate to severe maltreatment within the last 6 months. Child protection records, semi-structured interviews in the home, and parent reports were used to assess child stress exposure, psychiatric symptoms, and behavior. HTR2A genotype and methylation of HTR2A was measured at two CpG sites (-1420 and -1224) from saliva DNA.

Results: HTR2A genotype was associated with HTR2A methylation at both CpG sites (p<0.001). HTR2A genotype also moderated associations of contextual stress exposure and HTR2A methylation at site -1420 (p=0.031). Contextual stress was positively associated with -1420 methylation among A homozygotes, but negatively associated with -1420 methylation among G homozygotes. PTSD and MDD symptoms were negatively associated with methylation at -1420 (p=0.022), but positively associated with methylation at -1224 (p=0.014).

Conclusions: Results support the view that the serotonin system is sensitive to stress exposure and psychopathology, and HTR2A methylation may be a mechanism by which early adversity is biologically encoded.

Keywords: Epigenetics, 5-HT2A Receptor, Early Life Stress

Disclosure: Nothing to Disclose.

W28. Age-Normative Pathways of Striatal Connectivity Relate to ADHD Symptoms in the General Population

Anita Barber*, Majnu John, Christina Fales, Deepak Sarpal, Katherine Karlsgodt, Anil Malhotra, Todd Lencz

Feinstein Institute for Medical Research, Glen Oaks, New York, United States

Background: Striatal circuits support diverse affective, motor, and cognitive functions and integrate information across these domains (1). Striatal development is well-documented into early adulthood, influencing wide-spread subcortical, cortical, and cerebellar regions, spanning a number of brain networks (2, 3). Atypical striatal function plays a role in core deficits related to Attention-Deficit/Hyperactivity Disorder (ADHD) and aberrant maturation of striatal circuits may be critical for the course of the disorder. The current study examined the development of resting-state connections with striatal functional sub-regions in two large, cross-sectional developmental cohorts. In addition, age-normed function in developing regions was related to ADHD symptomology.

Methods: Resting state fMRI and T1-weighted scans were collected from nine sites for the Pediatric, Imaging, Neurobehavior and Genetics (PING) study (4) and one site for the Philadelphia Neuroimaging Cohort (PNC) (5). Participants were neurotypical children, ages 3–22 years.

Resting state scans were slice time corrected, motion corrected, co-registered, segmented, and normalized. Nuisance and motion artifacts were minimized using CompCor (6, 7) and regression of global mean signal and motion (both absolute and differential). Functional images were spatially-smoothed (6-mm FWHM) and temporally-filtered (bandpass 0.01–0.1 Hz). Participants with poor quality scans or Frame-wide Displacement (FD)>0.05 were removed from further analysis. After quality control filtering, 1215 participants had useable resting state data.

Full-brain connectivity was computed with six bilateral striatal seeds, placed in three caudate and three putamen functional sub-regions bilaterally (8). GLM models assessed the linear and nonlinear effects of age, while covarying for sex, FD, FD2, and site. Developing striatal connections were identified (i.e. connections with age effects significant at both a voxel-level and cluster-level threshold of p<0.001).

For the developing connections, z-transformed connectivity values were extracted and a growth-charting approach was taken. This involved plotting the normative development for the sample and creating age-deviation scores for each participant. Principal Components Analysis (PCA) was then performed on the age-normed connections to identify normative pathways that characterize similar striatal patterns of age deviation across the set of developing regions. ADHD symptoms, summed across nine items from the computerized GOASSESS battery in the PNC subsample, were then associated with the identified striatal pathways.

Results: Seventy developing striatal connections were identified which spanned a diverse set of cortical, subcortical, and cerebellar regions (Figure 1). PCA on the age deviation scores from this set of connections revealed four striatal pathways of normative development. Associations were found between the second PCA factor and summed measures of inattention (p=0.008) and overall ADHD symptoms (p=0.006).

Conclusions: Age-normative pathways of striatal connectivity encompassed a diverse set of regions, spanning several brain networks. One pathway, in particular, was related to subclinical ADHD symptoms, reflecting delayed or atypical striatal development.

Keywords: Neurodevelopmental Disorders, Striatal Cortical Circuits, ADHD, Resting State Brain Imaging

Disclosure: Nothing to Disclose.

W29. Impaired Corticostriatal Recruitment Following Chronic Amphetamine Treatment in Omega-3 Fatty Acid Deficient Rats: An in vivo Pharmacological MRI Study

Diana Lindquist, Ruth Asch, Jennifer Schurdak, Robert McNamara*

University of Cincinnati College of Medicine, Cincinnati, Ohio, United States

Background: Children and adolescents with attention deficit hyperactivity disorder (ADHD) exhibit significant deficits in omega-3 polyunsaturated fatty acids (n-3 PUFA) including docosahexaenoic acid (DHA). The therapeutic effects of psychostimulant medications are thought to be mediated in part by neuroplastic processes within corticostriatal circuits, and adjunctive n-3 PUFA treatment has been found to augment psychostimulant efficacy in ADHD. While prior rodent studies have found that brain DHA accrual during development is necessary for the maturation and resilience of mesocorticostriatal dopamine systems, its role in amphetamine (AMPH)-induced corticostriatal neuroadaptations following chronic treatment is not known. The present study investigated the impact of graded DHA levels and/or chronic AMPH treatment on AMPH-induced brain activation patterns in vivo.

Methods: From P21-P90, male rats were fed a diet with no n-3 fatty acids (Deficient, DEF, n=20), a diet fortified with preformed DHA (fish oil, FO, n=20), or a control diet fortified with alpha-linolenic acid (CON, n=20). Beginning on P40 to P60, rats (n=10/diet group) received an escalating b.i.d. dose of D-amphetamine sulfate (AMPH) beginning with 0.5 mg/kg (5 days), 1.0 mg/kg (5 d), 2.5 mg/kg (5 d), and 5 mg/kg (5 d). Controls (n=10/diet group) received an equivalent volume of 0.9% saline (1 ml/kg, b.i.d). All rats were then left undisturbed for a 30 d abstinence period. On P90 rats were anesthetized with 1.5-2.5% isoflurane in air and scanned in a 7T Bruker Biospec system. Blood oxygen level dependent (BOLD) responses to an AMPH challenge (7.5 mg/kg, i.v) were collected for 40 min post-injection. Voxel-wise statistical analyses were performed to identify group differences in regional BOLD responses to the AMPH challenge using a corrected significance threshold of p≤0.05. Following scanning postmortem forebrain DHA composition was determined by gas chromatography.

Results: Compared with controls, forebrain DHA levels were significantly lower in DEF rats (-35%, p<0.0001) and significantly greater in FO rats (+6%, p<0.0001). Rats in each diet group exhibited similar dose-dependent changes in locomotor activity following chronic AMPH treatment. The AMPH challenge during the scan induced similar increases in BOLD signal in SAL-pretreated rats from each diet group in the substantia nigra and basal forebrain structures including the nucleus accumbens, islands of Calleja, and ventral pallidum. In AMPH-pretreated CON rats, the AMPH challenge induced widespread BOLD activation in multiple regions, including the bilateral caudate putamen, thalamus, and frontoparietal and motor cortices, compared with SAL-pretreated CON rats, and a similar pattern was observed in FO rats. In contrast, regional BOLD response following the AMPH challenge was similar in AMPH- and SAL-pretreated DEF rats, and AMPH-pretreated CON and FO rats exhibited significantly greater regional BOLD signal compared with AMPH-pretreated DEF rats.

Conclusions: Repeated escalating AMPH dosing is associated with a wide-spread recruitment of corticostriatal structures in rat brain, and this response is not observed in rats subjected to dietary n-3 PUFA insufficiency during peri-adolescent development. These findings suggest that n-3 PUFA intake during peri-adolescent development promotes neuroadaptive processes within corticostriatal circuits following repeated AMPH treatment which may be relevant to the therapeutic efficacy of psychostimulants in ADHD youth.

Keywords: Amphetamine, Rat, Omega-3 Fatty Acid

Disclosure: Nothing to Disclose.

W30. Change in Adiposity, Hepatic Triglyceride and Carotid Intima Media Thickness During Behavioral Weight Loss Treatment in Antipsychotic-Treated Youth

Ginger Nicol*, Rachel Kolko, Eric Lenze, Michael Yingling, J. Philip Miller, Amanda Ricchio, Julia Schweiger, Robert Findling, Denise Wilfley, John Newcomer

Washington University School of Medicine, Saint Louis, Missouri, United States

Background: Pediatric-onset obesity is associated with type 2 diabetes risk and mortality in early adulthood. Populations at increased risk include children with mental illness. Antipsychotic medications, commonly used in childhood-onset psychiatric disorders, are known to cause weight gain and increase risk for diabetes. Behavioral weight loss (BWL) interventions are considered first-line treatments for pediatric obesity, and may have attenuated efficacy in psychiatric populations. We pilot-tested a 16-week BWL intervention - initially developed for use in non-psychiatric (NP) youth - in mentally ill antipsychotic (AP)-treated youth, employing sensitive biomarkers of cardiometabolic disease risk, including dual energy x-ray absorptiometry (DEXA)-measured body fat, hepatic triglyceride content using magnetic resonance (MR) imaging-estimated proton density fat fraction (PDFF), and carotid intima media thickness (CIMT). We hypothesized benefits of BWL treatment compared to usual care (UC).

Methods: Youth with overweight/obesity (BMI%ile of 85-94 and >95, respectively) ages 6-18 were assigned to interventions as follows: AP-treated youth were randomized 2:1 to the weekly 16-week BWL intervention or to a UC reference condition consisting of 4 once-per-month health education sessions; NP youth were frequency-matched on age and Body Mass Index (BMI) to AP-treated youth and assigned to BWL treatment. AP-treated participants had clinically reported weight gain during ≥6 months, with documented psychiatric stability prior to enrollment. Active suicidality and estimated or documented IQ <70 were exclusionary. DEXA-measured whole body total (kg) and %fat mass as well as whole body lean (kg) mass, PDFF determined by proton magnetic resonance spectroscopy with a 1.5T scanner, CIMT measured by 9-13-MHz B-mode Carotid Artery Ultrasonography, fasting plasma lipids (total cholesterol, low density lipoprotein or LDL cholesterol, high density lipoprotein or HDL cholesterol and triglyceride) and glucose were evaluated at baseline and at 16 weeks. The Aberrant Behavior Checklist (ABC) and Child Behavior Checklist (CBCL) were used to assess overall behavioral symptom severity at baseline and 16 weeks. Analyses of covariance (ANCOVA) were used to test for treatment group effect on the week 16 value of each primary outcome, using each outcome's baseline value as the covariate term. Contrasts of interest included the difference between each weekly BWL treatment group and the UC reference condition, as well as the difference between the two BWL groups.

Results: Nineteen AP-treated participants were randomized to BWL treatment (15 completed, 79%), 7 AP-treated youth were randomized to UC (6 completed, 86%), and 21 NP youth were assigned to BWL (17 completed, 80%). The mean age in the pooled sample was 13.3 years (SD=2.43), with 55% male and 72% white. AP-treated groups were diagnostically heterogeneous, including Autism Spectrum Disorder (ASD) (5), Attention-Deficit Hyperactivity Disorder (ADHD) (3), mood disorder (1), anxiety disorder (1), and psychotic disorder (1), with three meeting criteria for comorbid Disruptive Behavior Disorder (DBD) including Oppositional Defiant Disorder (ODD) or Conduct Disorder (CD). The ratio of male:female participants was higher in the AP-treated groups, reflecting the higher prevalence of ASD, ADHD and DBDs in boys. Groups were otherwise balanced on baseline characteristics. Session attendance was similar across groups; the AP-BWL group attended 10.26 (5.56) of 16 sessions, the NP group attended 11.91 (5.11) of 16 sessions, and the monthly UC group attended 3.14 (1.46) of 4 sessions (F[2,44]=0.92, p=0.41). Significant differences in endpoint DEXA total fat values, controlling for baseline (F[2,34]=4.81, p=0.01) occurred across weight loss intervention groups, explained by larger improvements in NP-BWL versus AP-treated youth both in UC (p=0.01) and BWL treatment (p=0.02). Significant differences in endpoint PDFF values, controlling for baseline (F[2,30]=3.60, p=0.04) were also observed across intervention groups, explained by larger improvements in NP versus UC youth (p=0.01). No significant differences in endpoint CIMT values were observed across groups, with no significant improvements from baseline in any group. No significant differences in endpoint BMI z-score, a surrogate clinical indicator of body composition, were observed across groups (F[2,34]=2.07, p=0.14). No significant changes in psychiatric symptoms measured by the ABC and CBCL were observed.

Conclusions: This study is the first to demonstrate tolerability and acceptability of BWL treatment in AP-treated youth with overweight/obesity, utilizing both clinical and gold-standard measures of cardiometabolic risk. The results extend observations by others that BWL can have a modest impact on weight and BMI in AP-treated youth as in NP youth, by additionally demonstrating effects of weight loss on whole body adiposity and liver fat content. However, reductions in total body and liver fat, were attenuated in AP-treated youth compared to NP youth, despite similar session attendance. Overall, the results indicate that BWL interventions can impact early measures of cardiometabolic risk in youth, and are feasibly delivered in the high-risk population of AP-treated youth. Future studies using optimized BWL interventions in this population are warranted.

Keywords: Antipsychotic-Associated Obesity, Cardiometabolic Risk, Pediatric

Disclosure: Part 1: Otsuka, Grant, Part 4: Otsuka, Grant.

W31. Maternal Adverse Childhood Experiences: Impact on Perinatal Glucocorticoid-Inflammatory Stress Response and Offspring Hypothalamic Pituitary Adrenal (HPA) Axis Development

Liisa Hantsoo*, Mary D. Sammel, Grace C. Ewing, Kathleen Morrison, Eileen Wang, Michal Elovitz, Tracy Bale, C. Neill Epperson

University of Pennsylvania, Philadelphia, Pennsylvania, United States

Background: Adverse childhood experiences (ACEs) affect hypothalamic pituitary adrenal (HPA) and inflammatory response to stress in adulthood. However, little is known about this in the context of human pregnancy, or the potential transgenerational effect on offspring stress response. We studied psychiatrically healthy pregnant women to assess impact of maternal ACE on 1) maternal glucocorticoid-inflammatory response to stress during pregnancy, 2) offspring fetal HPA axis development, 3) maternal cortisol response to stress postpartum, and 4) offspring HPA response to stress.

Methods: Healthy pregnant female participants were recruited to a longitudinal study; a subset were recruited to a related pilot study. The Structured Clinical Interview for DSM (SCID) confirmed healthy psychiatric status. Women completed the Adverse Childhood Experiences (ACE) Questionnaire. At 20-22 weeks gestation, a 3-D transabdominal ultrasound with Virtual Organ Computer-Aided Analysis (VOCAL) technology captured volumetric data of the fetal adrenal gland. At 22-34 weeks gestation, cytokine and cortisol response to a laboratory stressor (Trier Social Stress Test; TSST) was assessed in the pilot sample. Postpartum, participants and their offspring underwent a laboratory separation stressor at 6 months of age; during separation infant and maternal salivary cortisol were collected at six time points.

Results: 1) During pregnancy 17 women completed the TSST. High ACE women had a significantly accelerated decline in cortisol (p=0.001) and trended toward elevated interleukin-6 (IL-6) (p=0.1) in response to acute TSST stress compared with low ACE women. Further, higher post-stress cortisol was associated with lower gestational age at delivery in the low ACE group (p<0.001), but not in the high ACE group.

2) Maternal ACE impact on fetal adrenal volumes differed by sex, significant ACE x sex interaction (p=0.03). As expected, female offspring had significantly smaller weight adjusted adrenal volumes than males among low ACE mothers (p=0.006). However, in high ACE mothers, neonate adrenal volumes did not differ by sex, with small adrenal volumes in males compared to males with high ACE mothers (p=0.033).

3) At six months postpartum, high ACE mothers had a blunted (42% lower) salivary cortisol response during the separation stressor compared with low ACE mothers (p=0.01).

4) At six months of age, infants’ cortisol response differed by ACE and sex, interaction p= (p=0.021), such that female infants of high ACE mothers had a blunted cortisol response compared to low ACE girls (p=0.08) while cortisol response was greater in high ACE males (p=0.12) compared to low ACE males.

Conclusions: Our research suggests that not only do ACEs influence maternal glucocorticoid-inflammatory response to stress during pregnancy and postpartum, but may also have an intergenerational impact their offspring in terms of HPA axis development. Pregnant women with multiple ACEs had dysregulated glucocorticoid and inflammatory response to stress during pregnancy and postpartum, and their offspring showed altered HPA structure and function, with relatively small fetal adrenal volumes in males and a blunted cortisol response in females. These effects were not driven by affective symptoms, as this sample was psychiatrically healthy, suggesting that ACEs themselves can have a lasting effect on physiologic stress response.

Keywords: Psychoneuroimmunology, Perinatal Stress, Transgenerational, Sex Differences, Translational Research

Disclosure: Nothing to Disclose.

W32. Phase I Open Label Trial of Autologous Cord Blood for Treatment of Autism: Correlations Between Behavioral Improvement and Increased White Matter Connectivity in Children With Autism

Kimberly Carpenter*, Samantha Major, Catherine Tallman, Lyon Chen, Lauren Franz, Jessica Sun, Joanne Kurtzberg, Allen Song, Geraldine Dawson

Duke University, Durham, North Carolina, United States

Background: Autism spectrum disorders (ASD) are characterized by social communication deficits and the presence of restricted interests and repetitive behaviors. Although the exact pathophysiology of ASD remains unknown, there is evidence for increased neuroinflammation and aberrant neuronal connectivity in individuals with ASD. Thus, therapeutic interventions that impact immune modulation and regulation of neural connectivity show promise in the treatment of ASD. One such potential treatment is infusion of umbilical cord blood. Preclinical models have shown that umbilical cord blood contains effector cells that, through paracrine signaling, alter brain connectivity and also suppress inflammation. In a recent Phase I open label trial, we have reported significant improvements in children's behavior at six and twelve months following treatment with a single infusion of autologous cord blood. The current study explored whether these behavioral changes were associated with concurrent changes in neuronal connectivity.

Methods: Twenty-five children participated in a phase I, open-label trial to assess the feasibility and safety of a single infusion of autologous cord blood in 2-6 year old (mean 4.6 years) children with ASD. Children completed a battery of observational and neurophysiological assessments and parents completed a structured interview and questionnaires. Clinical outcome measures included the Vineland Adaptive Behavior Scales‐II (VABS‐II) Socialization subscale score, Expressive One‐Word Picture Vocabulary Test‐4 (EOWPVT‐4) raw score, and the Clinical Global Impression-Severity and Improvement Scales (CGI-S and CGI-I). With the exception of the CGI-I, which was completed at the 6- and 12-month follow-up visits, all assessments were completed at baseline (prior to infusion), as well as 6 months and 12 months post-infusion. Additionally, structural connectivity was measured at baseline and at 6-months in a subset of 19 children with 25-direction Diffusion Tensor Imaging (DTI). DTI data was analyzed with brain connectome analyses using deterministic tractography, the Connectome Mapping ToolKit (CMTK), and a priori defined nodes from a pediatric gray matter atlas. Connectivity was assessed for brain regions most commonly found to be atypical in previous MRI studies with children with ASD, namely, frontal, temporal, and specific subcortical (e.g. basal ganglia) brain networks. Only those pathways for which at least 80% of the sample had traceable streamlines were included in the analyses. The relationships between increased connectivity and changes in the VABS-II, EOWPVT-4, or CGI-I were assessed with Spearmen correlations. This study was approved by the Duke University School of Medicine Institutional Review Board.

Results: In order to reduce type 1 error, we identified networks for which correlations between increased connectivity and at least two of the three outcome measures were significant at p<0.05. Improvement on all 3 behavioral outcome measures was correlated with increased connectivity between the left temporal pole and the left hippocampus (VABS-II p<0.001; EOWPVT-4 p<0.05; CGI-I p<0.0001). Improvement on both the VABS-II and the EOWPVT-4 was correlated with increased connectivity between the fusiform and putamen (VABS-II p<0.05; EOWPVT-4 p=0.01) as well as between the inferior temporal gyrus and the hippocampus (both p<0.05), both in the left hemisphere. Finally, there was a significant correlation between improvement on both the VABS-II and the CGI-I and increased connectivity between the inferior temporal gyrus and superior temporal gyrus (VABS-II p<0.05, CGI-I p<0.01) and between the temporal pole and globus pallidus (VABS-II p<0.01; CGI-I p<0.05) in the left hemisphere. Improvement on the VABS-II and CGI-I was also significantly correlated with increased connectivity between the frontal pole and globus pallidus and the insula and putamen in the right hemisphere (all p<0.05).

Conclusions: Significant improvements in behavior in children with ASD were described in a Phase 1, open label study after autologous cord blood infusion. The current results suggest that these improvements in behavior are associated with increased neuronal connectivity in brain networks supporting social, communication, and language abilities.

Keywords: Autism, DTI, Anatomical Connectivity

Disclosure: Nothing to Disclose.

W33. Fractional Anisotropy in the Frontal Segment of the Uncinate Fasciculus in Adolescents at Risk for Depression

Tiffany Ho*, Akua Nimarko, Kira Oskirko, Josiah Leong, Sarah Ordaz, Ian Gotlib

Stanford University, Stanford, California, United States

Background: Major depressive disorder (MDD) is a prevalent and debilitating disorder with first episodes that typically begin during adolescence (Avenevoli et al., 2015). Current neurobiological models of adolescent MDD focus on frontolimbic disconnectivity, particularly in the uncinate fasciculus (UF). While some researchers report reduced fractional anisotropy (FA) in the UF in depressed versus nondepressed adolescents (LeWinn et al., 2014), other investigators have found increased FA (Aghajani et al., 2014). Because the UF is a long-range tract composed of three segments—temporal, insular, and frontal—that continues to mature throughout adolescence (Peltier et al., 2009; Lebel et al., 2012), developmental changes in the UF, particularly in the frontal segment, might account for these discrepant findings. In this study, we conducted diffusion-weighted imaging in a sample of young adolescents at risk for MDD and related FA of the three UF segments to concurrent severity of depression. We then sought to validate our primary findings in an independent sample of older adolescents at risk for MDD.

Methods: We report results from a total of 98 adolescent participants from two independent longitudinal studies conducted at Stanford University. All participants completed laboratory assessments at two timepoints (baseline and follow-up). Thirty-seven of these participants were initially recruited at ages 9-12 years on the basis of early life stress history, an environmental risk factor for MDD. They were assessed at baseline and at an 18-month follow-up, when they completed self-report measures of depression and an MRI scan (20 females, mean age: 13.52±1.03 years at follow-up). Given the relatively young age of this sample, as expected none of these participants met DSM criteria for MDD at either timepoint. The validation sample was 62 female adolescents who were initially recruited at ages 9-14 on the basis of familial MDD risk (29 with family history; 33 with no family history) and who did not meet DSM criteria for MDD at baseline; these participants were assessed longitudinally over a period of 7.5 years and completed self-report measures of depression and an MRI scan at follow-up (mean age: 18.89±2.49 years at follow-up). Of the 29 with a family history, 10 developed MDD, and of the 33 with no family history, 7 developed MDD. None of these participants were in episode at the time of follow-up.

All follow-up MRI scans were conducted in a 3T Discovery MR750 (GE Medical Systems, Milwaukee, WI) equipped with a 32-channel head coil (Nova Medical). Diffusion-weighted data were preprocessed using standard methods and a continuous tensor field was estimated with trilinear interpolation of the tensor elements after removing outlier directions (Ho et al., 2017). We used Automatic Fiber Quantification (AFQ; Yeatman et al., 2012) software to identify and segment each participant's left and right UF. Candidate fibers for UF were defined using deterministic tractography and then refined by comparing each fiber within the fascicle to a probabilistic fiber tract white matter atlas (Wakana et al. 2007). Each UF fiber group was represented as a 3D Gaussian with outlier streamlines omitted. FA tract profiles of the UF were computed by calculating FA along 100 equidistant nodes along the tract. The three UF segments were identified based on local maxima (defined by the first derivative) of these nodes in the tract profile (Ho et al., 2017).

Severity of depressive symptoms was assessed with the Children's Depression Inventory (CDI; Kovacs, 1992) at both baseline and follow-up. In our primary analysis of the 37 adolescents at environmental risk for MDD, we conducted linear regression models to test whether CDI scores at follow-up were predicted by FA of frontal UF at follow-up, controlling for baseline CDI score. We then sought to validate results from these primary analyses in the sample of 62 adolescents with and without familial risk of MDD.

Results: In our primary analysis, we found that higher severity of depression was significantly associated with greater FA in frontal UF above and beyond baseline depression severity (B=21.432±9.769, t36=2.194, p=0.038 for right UF; B=21.191±9.575, t33=2.213, p=0.034 for left UF); there were no significant associations between CDI and FA of the other UF segments. In our validation analysis, we found a significant interaction effect between group (familial risk with MDD; no familial risk with MDD; familial risk without MDD; no familial risk without MDD) and FA in frontal left UF on CDI scores at follow-up: those at familial risk who developed MDD showed a significant negative association between frontal left UF (but not frontal right UF) and concurrent depression severity (B=-41.832±11.454, t50=2.412, p=0.036). In contrast, those with familial risk who did not develop MDD exhibited greater FA in frontal UF.

Conclusions: Our results demonstrate that FA in frontal UF, which undergoes significant maturation during adolescence, may represent a risk marker of adolescent MDD that is sensitive to developmental stage and history of illness. Whereas reduced FA in frontal UF appears to reflect a “scar” of depression in older adolescents, consistent with previous work (LeWinn et al., 2014), greater FA in frontal UF earlier in development may represent risk to MDD but that, if remains intact, may indicate resilience to disorder. Additional longitudinal assessments are needed to definitively test this intriguing hypothesis.

Keywords: DTI, Adolescent Depression, Early Identification of Risk

Disclosure: Nothing to Disclose.

W34. Embryonic GABAergic Proliferation as a Contributing Mechanism of Sex Differences in Prenatal Stress Effects on Brain and Behavior

Hanna Stevens*, Stephanie Lussier, Jacob Michaelson, Sreya Radhakrishna, Benjamin Elser

University of Iowa Carver College of Medicine, Iowa City, Iowa, United States

Background: Prenatal stress contributes to the risk for neuropsychiatric diseases including autism spectrum disorders (ASDs), ADHD, Tourette syndrome and schizophrenia. However, the pathophysiology of these disorders and how early stress modification of brain development may contribute to this pathology is poorly understood. GABAergic neuron development is affected by prenatal stress and other prenatal disruptions and has been implicated in the pathophysiology of multiple psychiatric disorders. Here, we examined the effects of prenatal stress on GABAergic progenitor proliferation and relationships to developing striatal GABAergic populations and striatal-dependent behaviors.

Methods: We utilized a standard prenatal stress protocol with three times daily repetitive restraint stress beginning on embryonic day 12 in CD1 mouse dams bred to CD1 males heterozygous for a GAD67GFP transgene. We assessed male and female offspring in habit-learning, motor learning, and motor activity tasks as adults. We examined proliferative genes and GAD67GFP+ cellular outcomes in embryonic and postnatal brain of male and female offspring using immunocytochemistry with unbiased neurostereology, microarray, and qPCR. We also repetitively exposed pregnant dams to potential physiological mediators of prenatal stress, corticosterone and inflammatory cytokines, from embryonic day 12 onwards to understand the maternal mechanisms of these neurobiological changes.

Results: Motor activity and motor and habit learning were disrupted in males but not female prenatally-stressed offspring in adulthood. These behavioral changes correlated with increased GABAergic neuron density in male offspring in mature caudate putamen, a region dominated by inhibitory neuronal subtypes. We traced this increase in GABAergic neuron density back consistently through earlier postnatal time points in male offspring. In embryonic brain, prenatal stress resulted in increased GABAergic progenitor proliferation, expansion of ganglionic eminence proliferative zones, and larger ventral GABAergic populations in male but not female offspring. Gene expression microarray of embryonic ventral telencephalon demonstrated cell cycle pathways and specific stem cell and cell cycle genes to be upregulated in males but not females by prenatal stress. Among the twenty most significantly altered genes, sox2, pcna and ankrd17 were validated to show a significant sex by stress interaction in expression by qPCR. The set of genes significantly changed by prenatal stress interacting with sex were overrepresented in candidate gene lists associated with autism spectrum disorder and Tourette syndrome. We also found that repetitive maternal exposure to corticosterone did not increase GABAergic populations in embryonic brain but repetitive exposure to interleukin-6 did.

Conclusions: Changes in motor activity and habit learning were observed in male offspring after prenatal stress, which reflect some of the same functional disruptions seen in children with ASD and Tourette syndrome. These behavioral changes were associated with significant alterations in GABAergic populations in the caudate putamen after prenatal stress, neurobiology implicated in psychiatric pathophysiology more broadly. Prenatal stress had critical effects on cellular proliferation in embryonic ventral telencephalon, the site of embryonic GABAergic neurogenesis. These effects occurred only in male offspring, upregulating the expression of stem cell and cell cycle genes in male but not female brain. The significant overlap of prenatal stress-induced genes in male brain and genes implicated in Tourette syndrome and ASD, suggests that these processes may be involved in the male predominance of these disorders. Lastly, recapitulation of upregulated GABAergic proliferation was seen with interleukin-6 but not corticosterone, suggesting that prenatal inflammation may be a critical contributor to psychiatric disorders.

Keywords: Prenatal Stress, GABAergic Cells, Motor Behavior, Sex Differences, Caudate

Disclosure: Part 1: Oakstone Publishing, Honoraria.

W35. Probing the Adolescent Prefrontal Cortex and Hippocampus: Functional MRI Activation During Virtual Watermaze Performance and in vivo Brain GABA

Marisa Silveri*, Jennifer Sneider, Julia Cohen-Gilbert, Derek Hamilton, Anna Seraikas, Emily Oot, J. Eric Jensen, Lisa Nickerson, Sion Harris

McLean Hospital, Harvard Medical School, Belmont, Massachusetts, United States

Background: During adolescence, the frontal cortex undergoes the most substantial structural and functional changes, although significant developmental changes also occur in hippocampus. Neurodevelopmentally, dynamic integration of hippocampal and prefrontal circuitry is necessary to incorporate experience into behaviors that can be risky, but adaptive for successful developmental transitions. The inhibitory neurotransmitter gamma-amino-butyric-acid (GABA) also matures during this developmental period, providing important inhibitory control over behavior. While significant age-related increases in frontal lobe GABA have been reported, less is known about age-related changes in hippocampal GABA.

Methods: In the current study, multimodal neuroimaging methods were applied at 3 Tesla, including MEGAPRESS to acquire in vivo brain GABA, from single voxels placed in dorsal anterior cingulate cortex (ACC) and in a region with strong functional and anatomical connections to hippocampus, the medial temporal lobe (MTL). Metabolites were quantified using LCModel, normalized to creatine (Cr) and corrected for relative gray matter content in each voxel type. Multiband blood oxygen level dependent (BOLD) functional magnetic resonance imaging (fMRI) data were acquired during performance of a virtual translation of the classic Morris Water Maze task using a block design and analyzed using FSL. Participants included 33 healthy adolescents, aged 13-14 years who were alcohol and drug naïve, and who exhibited no psychiatric symptoms or conditions, recruited locally to participate in a three-year longitudinal study.

Results: Spectral data acquired in year 1 demonstrated significantly lower GABA/Cr in ACC compared to MTL (p<.0001). Significant BOLD activation (cluster-based thresholding, z=3.1, p<.05) was observed in hippocampus and PFC in the memory retrieval/motor contrast [in regions that overlapped with both MRS voxels]. In an exploratory analysis combining data from these imaging modalities, during memory retrieval, ACC GABA negatively predicted activation in ventral medial PFC, whereas both ACC and MTL GABA positively predicted thalamic activation. While GABA levels in the selected voxels of interest were not correlated, functional imaging data provide evidence of both hippocampal and prefrontal activation, that in future functional connectivity analyses may be used to probe the coupling between these regions.

Conclusions: Lower ACC GABA likely reflects a later maturation of this neurochemical in PFC, which will be examined in longitudinal analyses. Furthermore, while MTL GABA did not predict PFC activation, GABA levels across both regions predicted thalamic activity, which may be an important neuronal relay between these regions. Coupling between hippocampal and prefrontal regions, as described by the Experience-Driven Adaptive Cognitive Model, may shed light on the role of memory in risk-related behaviors that are developmentally adaptive, but that can become maladaptive, particularly as adolescents initiate alcohol and drug use.

Keywords: Adolescence, Functional MRI (fMRI), GABA MRS

Disclosure: Nothing to Disclose.

W36. Adolescent Stress Leads to Enduring Enrichment of the NF-kappa-B Pathway in the Hippocampus Without Peripheral Immune Consequences

Mandakh Bekhbat, Sydney Rowson, Sean Kelly, Gregory Tharp, Malu Tansey, Gretchen Neigh*

Virginia Commonwealth University, Richmond, Virginia, United States

Background: Chronic stress during development is a prominent risk factor for the development of mood disorders in adulthood, and causes neuroimmune alterations that contribute to the etiology of these conditions. Chronic adolescent stress (CAS) is a model that demonstrates in rats many of the immune, endocrine, and behavioral abnormalities caused by developmental stress, and recapitulates its sex-specific outcomes. We have previously shown that CAS primes the hippocampal inflammatory response in adult male, but not female, rats. However, the mechanism and origin of CAS-induced neuroinflammatory priming, as well as associated sex differences are currently undefined. Here we tested the hypothesis that CAS exaggerates induction of the pro-inflammatory NF-kappa-B pathway in adult rat hippocampus without compromising the peripheral immune response. Furthermore, we hypothesized that CAS-induced potentiation of NF-kappa-B signaling would be present in adult male rats only.

Methods: Male and female adolescent rats underwent a twelve-day CAS paradigm consisting of restraint and social defeat, or received no stress. Five weeks following the last stressor, all rats received a single, systemic injection of either a low dose of lipopolysaccharide (LPS) or vehicle to unmask possible priming effects of CAS. Total RNA from the hippocampus was used to perform RNA-Seq and enriched transcriptional pathways were identified using gene set enrichment analysis (GSEA). To assess the impact of CAS on peripheral NF-kappa-B function, we measured DNA binding activity of NF-kappa-B in the spleen and plasma concentrations of IL-6 and TNF-alpha.

Results: Following pathway analysis via GSEA, NF-kappa-B emerged as the most enriched pathway in both CAS males and females compared to non-stressed, same-sex controls following LPS. Targeted qPCR experiments further confirmed that CAS exaggerated the induction of I-kappa-B (F(1,105)=4.209, p=0.043), p65 (F(1,105)=5.262, p=0.024), and p52 (F(1,105)=8.186, p=0.005). CAS impacted neither LPS-induced NF-kappa-B activity in the spleen (F(1,57)=0.256, p>0.05) nor plasma IL-6 (F(1,58)=1.083, p>0.05) and TNF-alpha (F(1,43)=0.60, p>0.05), suggesting that the central effects of CAS on the NF-kappa-B pathway are independent of changes to the peripheral immune response. Interestingly, CAS also led to an enhanced enrichment of the glucocorticoid receptor (GR) signaling pathway in females. As GR signaling is the primary mechanism of NF-kappa-B suppression, our results suggest sex differences in the impact of CAS on NF-kappa-B regulation via an altered balance between GR and NF-kappa-B signaling.

Conclusions: Collectively, our results indicate that chronic stress experienced during adolescence leads to long-lasting changes to the hippocampal genomic profile, and that the pro-inflammatory consequences of CAS are specific to the brain. Furthermore, these results suggest that CAS alters the balance between the stress and immune pathways differentially in male and female rats.

Keywords: Chronic Stress, Sex Differences, Glucocorticoid Receptor, NFkB, Neuroinflammation

Disclosure: Nothing to Disclose.

W37. Findings From Baseline Neuroimaging Data in a Double-Blind Randomized Controlled Trial of rTMS for Executive Function Deficits for ASD

Stephanie Ameis*, Laagishan Yoganathan, Colin Hawco, Aristotle Voineskos, Rachael Lyon, Jeff Daskalakis, Daniel Blumberger, Paul Croarkin, Peter Szatmari

Centre for Addiction and Mental Health, Toronto, Canada

Background: Executive function (EF) deficits in patients with autism spectrum disorder (ASD) are ubiquitous and understudied. Repetitive transcranial magnetic stimulation (rTMS) has demonstrated promise in addressing EF deficits in adult neuropsychiatric disorders. We performed baseline structural and functional neuroimaging in the context of an ongoing randomized, double-blind, sham controlled trial of bilateral, 20 Hz, rTMS applied to the dorsolateral prefrontal cortex (DLPFC) for the treatment of EF deficits in ASD. An N-back spatial working memory task shown to reliably activate fronto-parietal networks related to working memory was used to explore whether individuals with ASD exhibited altered functional connectivity of working memory networks at baseline compared to healthy controls during task performance.

Methods: Baseline neuroimaging was acquired on a 3T GE MR750 in n=40 participants with ASD without comorbid intellectual disability with baseline executive function impairment and n=20 healthy controls (ASD, MeanAge=23±10; Healthy controls, MeanAge=23±9). All 40 ASD participants were randomized to active or sham rTMS treatment following baseline imaging. Post-trial imaging was also undertaken in all participants. During fMRI scanning, participants performed a visuospatial N-back task consisting of a 0-back and 2-back condition. For 0-back, participants indicated the location of a stimulus in space with 1 of 4 button presses. Participants indicated the location of a stimulus in space two trials back for the 2-back condition. Task-based fMRI analyses were carried out using SPM12. The generalized form of Psychophysiological interaction (gPPI) was used to examine potential DLPFC functional connectivity differences in the 0-back versus 2-back (working memory) condition in participants with ASD versus controls. Permutation analysis was used (FSL's PALM) to determine TFCE corrected p-values.

Results: N-back task performance reliably activated fronto-parietal networks and deactivated default mode networks in participants with ASD and in healthy controls (p<0.05 TFCE corrected in both groups). In addition, a trend-level difference in task-related fronto-parietal working memory network connectivity was found in participants with ASD compared to healthy controls during N-back task performance following statistical correction. Specifically, lower connectivity between the DLPFC and the angular cortex was found in ASD versus controls in the 2-back versus 0-back condition of the N-back task (t=2.11, p<0.1, TFCE corrected).

Conclusions: In the baseline neuroimaging data of our clinical trial, we found a trend supporting altered fronto-parietal connectivity in participants with ASD during spatial working memory performance compared to healthy control participants. Connectivity between the DLPFC and the angular cortex, a region shown to be recruited during spatial working memory performance, was specifically altered in ASD. Future directions for this work include examining whether altered fronto-parietal connectivity in ASD was associated with EF task performance and whether connectivity differences in ASD may change with active versus sham rTMS treatment in the context of the broader ongoing rTMS clinical trial.

Keywords: Brain Stimulation, Functional Neuroimaging, Autism

Disclosure: Nothing to Disclose.

W38. Functional Connectivity of the Extended Social-Affective Default Network During Social Reward is Associated With Adolescent Depressive Symptoms

Gabriela Alarcón*, Erika Forbes

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: Compared to other developmental periods, adolescence is characterized by a heightened concern with approval from peers and the emergence of depression, a disorder of disrupted social functioning. The extended social-affective default (eSAD) network supports social, affective and introspective processes that are disrupted in depression, such as responses to social reward. The eSAD network is comprised of “affective” and “social” brain regions that support distinct, but overlapping functions implicated with depression. “Affective” brain regions (e.g. amygdala, ventral striatum, ventromedial prefrontal cortex) support motivation and reward, while “social” brain regions (e.g. middle temporal gyrus, precuneus) support self-processing and mentalizing. The current study examines functional connectivity of the precuneus (PC), the most well-connected brain region of the eSAD network, during social reward in a community sample of adolescents to determine associations with depression. We hypothesized stronger functional connectivity between PC and other “social” brain regions of the eSAD network, reflecting heightened concern with social approval in adolescents with higher depressive symptoms. Additionally, we hypothesized that adolescents with higher depressive symptoms would exhibit weaker functional connectivity between PC and “affective” regions of the eSAD network, indicating a dampened affective or reward response.

Methods: Forty adolescents (14-18 years old; females=27) with no history of psychiatric disorder were recruited. Youth completed a questionnaire assessing depressive symptoms (Center for Epidemiologic Studies Depression Scale) and a social reward functional magnetic resonance imaging (fMRI) task that included positive feedback (i.e. well-liked peers liked the participant in return) and neutral feedback (i.e. peers who were rated ambivalently did not provide any feedback about participant). We conducted a psychophysiological interaction analysis using bilateral PC as a seed to assess functional connectivity during Positive>Neutral feedback. A whole-brain one-way analysis of variance comparing groups with no (clinically-relevant) depressive symptoms (CES-D<16) and sub-threshold depressive symptoms (CES-D≥16) was conducted, with a voxel-wise threshold of p<0.005, uncorrected and cluster-wise threshold of p<0.05, family-wise error (FWE) corrected.

Results: Compared with the no depression group, the sub-threshold depression group had stronger functional connectivity between PC and left middle temporal gyrus (F(1, 38)=8.88, p=0.002) and right middle temporal gyrus (at a statistical trend level, F(1, 38)=8.88, p=0.06).

Conclusions: In accordance with our hypothesis, adolescents with higher depressive symptoms had more robust functional connectivity between PC and middle temporal gyrus, a node of the eSAD network that is important for social cognition, including language processing. Contrary to our hypothesis, PC functional connectivity with “affective” brain regions of the eSAD network was not different based on depressive symptoms. Together these findings may indicate that adolescents with higher depressive symptoms are more likely to utilize cognitive (i.e. semantic properties of reward) rather than more basic affective (i.e. hedonic properties of reward) processes to integrate positive social feedback from peers into their concept of self. Higher (sub-threshold) depressive symptoms may create an imbalance in the integration of the cognitive versus affective properties of a social reward into one's self-concept.

Keywords: Adolescence, Social Reward, Functional Connectivity, Extended Social-Affective Default Network, Depression

Disclosure: Nothing to Disclose.

W39. Dynamic Changes in Safety Learning and Hippocampal-Frontoamygdala Interactions to Reduce Fear During Adolescence

Dylan Gee*, Dominic Fareri, Laurel Gabard-Durnam, Christina Caldera, Bonnie Goff, Martin Monti, Tanja Jovanovic, Nim Tottenham

Yale University, New Haven, Connecticut, United States

Background: Anxiety disorders, which peak during adolescence, are characterized by difficulty discriminating between threatening and safe contexts. Translational studies in mice and humans have demonstrated dynamic changes in frontoamygdala circuitry that supports inhibitory learning and a period of diminished fear extinction during adolescence, raising the question of whether adolescents with anxiety disorders may benefit from efforts to optimize fear reduction through novel mechanisms that bypass prefrontally-mediated extinction processes. Rodent studies have shown that safety cues effectively reduce anxiety to threat and prevent the development of new fears. Because they provide a context for the conditioned stimulus (CS), safety cues may rely on hippocampal projections to frontoamygdala circuitry, which develop earlier than frontoamygdala connections, and thus be particularly useful during adolescence.

Methods: The present functional magnetic resonance imaging (fMRI) study examined the development of safety learning across childhood and adolescence (7-17 years old) using a conditioned inhibition task that was modified for use in an MRI scanner with a developmental sample.

Results: Findings revealed behavioral evidence of safety learning and the same hippocampal mechanism in typically developing children and adolescents that has previously been shown in animals, filling a significant translational gap to move basic science more rapidly to practice. Though children and adolescents both showed robust hippocampal activation to the CS paired with the safety cue, only adolescents showed increased prefrontal activation and downregulation of amygdala reactivity that was associated with safety learning (p<.05, corrected).

Conclusions: Our findings suggest that safety cues may be a powerful way to reduce fear during this unique developmental window. This study is expected to have important implications for optimizing treatments for anxiety in youth based on the biological state of the developing brain.

Keywords: Brain Development, Anxiety, Fear, Children and Adolescents, Amygdala

Disclosure: Nothing to Disclose.

W40. Looking at the Brighter Side: Functional Connectivity Biomarkers of Resilience to Adolescent Depression in Emotion Regulation Networks

Adina Fischer*, M. Catalina Camacho, Tiffany Ho, Susan Whitfield-Gabrieli, Ian Gotlib

Stanford University, San Carlos, California, United States

Background: Approximately 25% of adolescents in the United States are diagnosed with Major Depressive Disorder (MDD), a leading cause of disability and the second-leading cause of death (i.e., by suicide) in adolescence. Despite the prevalence and devastating sequelae of adolescent depression, we have little understanding about the neural mechanisms that may be protective and promote resilience during this important period of brain development. While there is a relatively large body of research examining neural aspects of adolescent depression, investigating the neural basis of resilience has the potential to improve prevention and intervention strategies for depression. Over the course of adolescence, there is significant maturation of neural networks involved in emotion regulation, including the amygdala and prefrontal cortices; moreover, this circuitry is influenced by stressful life events. Increased plasticity in emotion regulatory circuitry has been found following treatment with antidepressant medication and cognitive behavioral therapy, first-line interventions for adolescent depression. Thus, investigating the neural circuitry that underlies emotion regulation may provide insight into the protective mechanisms that promote adaptive interpretation, reappraisal of, and response to stressful life events, and into adolescents’ ability to cope successfully with the unique stressors experienced during this developmental period. This longitudinal study was designed to examine neural correlates of resilience in adolescents at familial risk for depression who remained resilient, compared to HR adolescents who experienced an episode of depression and a low-risk control group.

Methods: This study includes 65 adolescent females who were part of a larger longitudinal investigation of familial risk for depression. Participants were high-risk adolescents who did not experience depression (resilient; RES) (n=20), high-risk adolescents who did experience an episode of depression (MDD) (n=20), and low-risk control adolescents with no personal or family history of psychopathology (control; CTL) (n=25). At the time of entry into the study, participants were 9-14 years of age and had no current or lifetime history of any Axis I disorder. Participants were followed longitudinally through age 18 (i.e., for 7.61±2.42 years) and completed clinical assessments at 18-month intervals. They also completed the Life Events Survey (LES; Sarason et al., 1978) to assess the number of negative and positive life events they experienced within the past 18 months. A seed-based resting state fMRI (rsfMRI) approach with rigorous motion correction was conducted in Conn (Whitfield-Gabrieli et al., 2012) to compare connectivity profiles in the RES, MDD, and CTL groups within brain networks implicated in emotion regulation: the limbic (amygdala seed) and executive control (dorsolateral prefrontal cortex seed; DLPFC) networks. We then examined the interaction of functional connectivity with significant life events within each group.

Results: Relative to MDD, RES adolescents had greater connectivity between the amygdala and orbitofrontal cortex (OFC), and between the DLPFC and fronto-temporal regions. Relative to CTL, RES adolescents had greater connectivity between frontal regions within the executive control and limbic networks. Only within the group of RES adolescents, the strength of amygdala-OFC connectivity was positively correlated with the perceived impact of positive life events.

Conclusions: Adolescence is a neurodevelopmental period during which experience-dependent plasticity in emotion regulatory neural circuitry may not only confer vulnerability to depression but also, and perhaps more importantly, provide opportunities to enhance resilience. In this study, we attempt to elucidate the neural basis of resilience in a cohort of adolescents at familial risk for depression and to understand the influence of life events on functional connectivity that may confer resilience to adolescent-onset depression. The characteristics of brain circuitry implicated in emotion regulation and cognitive control that we identified in this study appear to be neural biomarkers of resilience. Our findings further suggest an effect of experience-dependent plasticity on brain networks, with a direct association between positive life events and amygdala-frontal connectivity in resilient adolescents. While further research is needed to examine the neural basis of resilience, the present findings begin to illuminate key network nodes that may be the focus of novel prevention and treatment paradigms. By shifting future research to a resilience-based model of mental health, and by investigating characteristics of high-risk individuals who remain resilient, we will be better positioned to understand how to optimize resilience to psychopathology by strengthening specific brain connections.

Keywords: Risk and Resilience, Major Depressive Disorder (MDD), Adolescence, Emotion Regulation, Experience Dependent Plasticity

Disclosure: Nothing to Disclose.

W41. Biomarker Discovery for Social Impairments: Translation From a Novel Monkey Model to Patients With Autism

Karen Parker*, Joseph Garner, Ozge Oztan, Erna Tarara, Jiang Li, Valentina Sclafani, Laura Del Rosso, Katie Chun, Sean Berquist, Michael Chez, Sonia Partap, Antonio Hardan, Elliott Sherr, John Capitanio

Stanford University, Stanford, California, United States

Background: Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by core social impairments. ASD remains poorly understood due to the difficulty in studying disease biology directly in patients and reliance on mouse models which lack clinically relevant, complex social cognition abilities. These limitations underscore the tremendous value in developing a primate model of social impairments for ASD biomarker discovery. Here we tested whether low-social monkeys exhibit abnormalities in biological signaling pathways previously implicated in prosocial functioning and/or neurogenetic syndromes with high penetrance for ASD. These candidate biomarkers comprised the neuropeptides oxytocin (OXT) and arginine vasopressin (AVP), and their receptors (OXTR and AVPRV1a), as well as two kinase signaling pathways, RAS-MAPK and PI3K-AKT (including the kinases ERK and AKT, and the phosphatase PTEN). We then tested whether our preclinical findings would translate to ASD patients.

Methods: This study included three cohorts: a monkey biomarker discovery cohort (N=15 low-social monkeys; N=15 high-social monkeys), a monkey replication cohort (N=15 low-social monkeys; N=15 high-social monkeys), and a small patient cohort (N=7 children with ASD; N=7 medical control children). Cerebrospinal fluid (CSF) and blood samples were collected using species-specific standard, sterile procedures. CSF and blood OXT and AVP concentrations were quantified via enzyme immunoassay, blood OXTR and AVPRV1a gene expression levels were quantified via qPCR, and blood levels of total and phosphorylated ERK, PTEN, and AKT were quantified via western blot.

Results: We first tested whether our biological dataset, considered as a whole, could accurately distinguish low-social from high-social monkeys in cohort 1. Discriminant analysis yielded a 93% correct social classification rate (LR ChiSq=26.36; p<0.0001). We then used logistic regression to identify that CSF AVP concentration (LR ChiSq=16.55; p<0.0001) and the ratios of phosphorylated-PTEN/total-PTEN (LR ChiSq=6.792; p=0.0092) and phosphorylated-AKT/total-AKT (LR ChiSq=4.064; p=0.0438) in blood strongly and additively predicted social classification. Subsequent analysis identified CSF AVP concentration as the key driver of these group differences (F1,18=9.236; P=0.0071), which we replicated in an independent monkey cohort (F1,24=8.847; P=0.0066). To test the translational utility of our monkey model, we used logistic regression to determine that human CSF AVP concentration predicted ASD diagnostic status, whereby individuals with lower CSF AVP concentrations were more likely to have been previously diagnosed with ASD (LR-ChiSq=9.233; P=0.0024). Like low-social monkeys, ASD patients showed significantly lower CSF AVP concentrations compared to control children (F1,10=11.02; P=0.0078).

Conclusions: These findings demonstrate the translational utility of developing a primate model for ASD biomarker discovery, in order to refine, and subsequently test, hypothesis-driven questions in valuable and volume-limited patient samples. Considered collectively, these data suggest that the AVP signaling pathway warrants consideration in future research endeavors that seek to develop novel laboratory-based diagnostic tools to detect, and medications to treat, ASD.

Keywords: Autism Spectrum Disorder, Vasopressin, Primate Model, CSF Biomarkers, Social Behavior

Disclosure: Nothing to Disclose.

W42. A New Vasopressin V1a Antagonist Restores Social Behavior in the Mouse CNTNAP2KO Model of Autism

Christophe Grundschober*, Audrey Genet, Michael Saxe, Patrick Schnider, Barbara Biemans

F. Hoffmann-La Roche, Ltd., Basel, Switzerland

Background: The neuropeptide vasopressin is mainly produced in the lateral hypothalamus and plays an important role in regulating aggressive and social behavior by activating the V1a vasopressin receptor. A missense mutation in the human contactin-associated protein-like 2 (CNTNAP2) gene has been identified in an Amish family, leading to cortical dysplasia–focal epilepsy (CDFE) syndrome. Out of the 9 originally reported cases, 6 have autism. CNTNAP2 is a cell adhesion molecule, important for AMPA receptor trafficking, development of dendritic arborization, and stabilization of dendritic spines. Cntnap2 knockout mice (Cntnap2 KO) demonstrate reduced social behavior and increased repetitive behavior (grooming).

In this work we investigated the effect of central V1a receptor blockade on the phenotype of the mouse Cntnap2 KO model of autism.

Methods: Cntnap2 KO mice were treated daily during 3 weeks with a novel brain penetrant V1a receptor-specific small molecule antagonist, starting at postnatal day 40. Social behavior and self-grooming was assessed during a direct social interaction test at P48. Self-grooming was measured again in a single setting at P52, and locomotor activity was assessed at P62.

Results: At postnatal day 48, chronic treatment with our V1a receptor-specific small molecule antagonist completely reversed the impairment in social behavior seen in Cntnap2 KO at 15 and 30 mg/kg i.p., which are doses predicted to block 77 to 88% of brain V1a receptors. The exaggerated grooming was also reduced at the highest dose tested (60 mg/kg i.p). None of the tested doses affected locomotor activity at P62.

Conclusions: Our data show that sub-chronic inhibition of vasopressin V1a receptors restores normal social behavior in Cntnap2 KO mice. The effect on repetitive behavior is only apparent at the highest tested dose, suggesting that grooming is either less sensitive to V1a inhibition or not directly mediated by V1a activation. Further experiments will be needed to clarify this finding.

These results suggest that V1a antagonists have the potential to improve social interaction in autism, a core symptom for which there is currently no drug treatment.

Keywords: Autism Spectrum Disorders, Vasopressin 1a Receptor Antagonist, Genetic Mouse Model, CNTNAP2

Disclosure: Part 1: F. Hoffmann-La Roche, Ltd., Employee, Part 2: F. Hoffmann-La Roche, Ltd., Employee, Part 3: F. Hoffmann-La Roche, Ltd., Employee, Part 5: F. Hoffmann-La Roche, Ltd., Employee.

W43. Alterations of Social Behaviors and Monoamine Transmission in Rodents Under Social Crowding

Yukiori Goto*, Young-A Lee, Tsukasa Obora

Kyoto University, Primate Research Institute, Aichi, Japan

Background: Social crowding has been shown to cause stress in subjects living in such environments. In addition, exceeding amount of sensory inputs and its information processing (cognitive overloads) have also been suggested to be associated with such crowding environments. It has still remained unclear how stress and cognitive overloads may impact on transmission of monoamines such as dopamine and serotonin, and associated behaviors.

Methods: To address the above issue, in this study, we investigated the effects of housing density by adjusting a number of mice housed in each cage in two different strains, C57BL/6 (C57) and DBA2 (DBA), one showing higher sensory sensitivity (C57) than the other (DBA). Social behaviors and anxiety of mice living in dispersing and crowding home cages were examined. In addition, the impacts of housing density on dopamine and serotonin concentrations in corticolimbic regions were assessed using HPLC.

Results: Although DBA mice in crowding social groups exhibited elevated stress hormones compared to those in dispersing groups, such stress hormone elevation associated with crowding was absent in C57 mice. Although C57 mice was less anxietic than DBA mice when they were housed in dispersing groups, C57 mice became more anxiety, which was comparable to that of DBA mice when they were housed in crowding groups. In contrast, DBA mice exhibited substantially more social interactions in their home cages than C57 mice. Although both DBA and C57 mice in crowding groups became less socially interactive with mates, but such social avoidance was more prominent in C57 than DBA mice. In accordance with these behavioral observations, dopamine and serotonin and their metabolites in corticolimbic regions were significantly different between mice living in crowding and dispersing groups. In particular, serotonin was different between C57 and DBA mice regardless of housing density, whereas dopamine in some regions including the prefrontal cortex was primarily modulated by housing density.

Conclusions: These results suggest that subjects exhibiting high sensory sensitivity exhibit social avoidance as adaptive behavioral strategy in socially crowding environments to mitigate stress, which may also be relevant to psychiatric disorders such as autism spectrum disorder in which both sensory hyper-response and social communication deficits are illustrated.

Keywords: Dopamine, Serotonin, Social Interaction, Prefrontal Cortex, Autism Spectrum Disorder

Disclosure: Nothing to Disclose.

W44. Development of a Novel Point of Care EEG Device for Screening of Delirium

Gen Shinozaki*, Kasra Zarei, Elijah Dahlstrom, Theodosis Chronis, Sayeh Sabbagh, Kumi Yuki, Aubrey Chan, Nicholas Bormann, Lindsey Gaul, Ellyn Cramer, Nicholas Coon, Hannah Chicchelly, Sydney Jellison, Mason Kilsares, Gabrielle Duncan, Eri Shinozaki, Matthew Karam, Nicolas Noiseux, Sangil Lee, John Cromwell

University of Iowa, Iowa City, Iowa, United States

Background: It is well known that delirium is a prevalent and dangerous condition affecting millions of elderly patients 65 years of age or older (1-3). Delirium is, however, underdiagnosed and undertreated due to lack of effective screening instruments (4, 5). The fact that delirium is often undetected among elderly patients admitted to the hospital leads to significantly higher mortality, longer length of stay, and higher post-discharge institutionalization rates (1-3). Along with these poorer outcomes for patients associated with delirium, the financial cost related to delirium is also immense. It is estimated that annual financial costs from delirium are over $150 billion in the United States alone (3). This study aims to develop a novel, noninvasive, user-friendly, accurate/objective point-of-care (POC) device with bispectral electroencephalography (“BSEEG”) for mass screening and early detection of delirium among elderly inpatients.

Methods: Study subjects are patients who were admitted to the University of Iowa Hospitals and Clinics from January 2016 to July 2017. We focused on mainly from two cohorts, 1) general medicine patients who are 70 year or older, and 2) orthopedic hip fracture repair patients who are 60 year or older, were recruited upon admission, and assessed for the presence of delirium with Confusion Assessment Method for ICU (CAM-ICU), and Delirium Rating Scare (DRS). For confirmed cases of delirium and controls, we obtained 10 minutes EEG recording with limited channels from a hand-held EEG device twice a day during their hospital stay. We analyzed those raw EEG data by converting to digital signals with spectral density analysis and additional algorithm to calculate “EEG score” to differentiate the two groups with and without delirium.

Results: 142 subjects were recruited for the initial test cohort. Average age was 74 yo with SD 13.5, and female was 58 %. The preliminary data analysis has demonstrated that a simplified EEG can provide good-quality brain wave signals that compare well to the signals obtained by a traditional EEG. The data has also shown that preliminary analysis of the spectral density of brain waves from bispectral EEG (BSEEG) can clearly distinguish delirious patients from normal controls, as well as a delirious state from a recovered state for a single individual. With the test sets consisting of initial 142 subjects, EEG score was chosen at 1.44, and the performance metrics were as follows. Accuracy: 87.5%, Sensitivity: 80.0%, Specificity: 87.7%. AUC with ROC was 0.7 (Figure). Based on the test data set outcome, we further conducted validation analysis using an independent cohort. We used 24 subjects for validation of algorithm and the performance metrics were as follows. Accuracy: 83.3%, Sensitivity: 83.3%, Specificity: 83.3%. AUC with ROC was 0.8 with this validation data set.

Conclusions: Our data showed the feasibility of this technology and approach for mass screening and detection of delirium among elderly inpatients, who are at high risk for delirium without using traditional EEG or other conventional questionnaire style screening instruments. Further data collection to improve the performance of algorithm is required before making this strategy useful in clinical settings.

Keywords: Delirium, EEG, Screening, Early Detection

Disclosure: Part 1: Predelix Medical, LLC, Board Member, Predelix Medical, LLC, Patent.

W45. Exercise-Induced Changes in NGF are Critical for the Maintenance and Recovery of the Septohippocampal Cholinergic System

Joseph Hall*, Lisa Savage

Binghamton University, Binghamton, New York, United States

Background: Exercise has shown remarkable capabilities at rescuing and improving cognitive function in both humans as well as rodents, which has been assumed to selectively involve neurotrophins that facilitate hippocampal neurogenesis. However, our recent data demonstrated that exercise-induced increases neurotrophins, in particular nerve growth factor (NGF), restored and maintained structure and function of the cholinergic septohippocampal circuit. Using a rat model of alcohol-related pathology, pyrithiamine-induced thiamine deficiency (PTD) to mimic the amnestic disorder Korsakoff syndrome, we found that two-weeks of voluntary wheel running, followed by a two-week restoration period recovered spatial memory performance in the PTD-model (Hall & Savage 2016). More remarkable, we found that exercise remodeled the cholinergic basal forebrain by rescuing the cholinergic neuronal phenotype that displays nestin in the medial septum/diagonal band. In mature cholinergic neurons nestin is a marker of higher excitability and plasticity (Hendrickson et al, 2011; Zhu et al., 2011). In the current set of experiments, we sought to determine which neurotrophin, NGF or BDNF, was critical for the exercise-induced recovery of the cholinergic forebrain structure and function.

Methods: To determine whether NGF and/or brain derived neurotrophic factor (BDNF) was responsible for the recovery of the cholinergic system we used a microbead delivery system to sequester these two neurotrophins during exercise. Fluorescent latex microspheres were coated with either an antibody for tyrosine kinase A (Trk-A; sequesters extracellular NGF) or an antibody for Trk-B (sequesters extracellular BDNF; see Gomez-Pinilla et al, 2008). First, rats were assigned to one of two treatment conditions: (1) PTD treatment that included daily intraperitoneal injections of pyrithiamine HBr with a thiamine-deficient rat chow for two weeks as previously described (Savage, Hall & Resende, 2012); or (2) Pair-fed control condition in which rats received daily injections of thiamine to in conjunction with thiamine-deficient chow. Following recovery from treatment, all rats were implanted with a ventral hippocampal (CA1 region) cannulae to later measure acetylcholine (ACh) and dorsal bilateral hippocampal (dentate gyrus) infusions of anti-TrkA or anti-TrkB coated microspheres to inhibit the actions of exercise-induced increases in NGF or BDNF. Following surgery, rats were placed in cages with access to an active running wheel or a blocked wheel to control for context. Rats exercised freely for two-weeks and then were prescribed a two-week restoration period. We then performed in vivo microdialysis, while rats were tested on a spatial spontaneous alternation task, to measure behaviorally activated ACh levels in the hippocampus. Finally, rats were perfused and brains were stained for ChAT (choline acetyltransferase; marker for cholinergic neurons)/Nestin, and stereological assessment of the ChAT/Nestin populations and co-expression of both markers in the medial septum/diagonal band.

Results: In PTD rats, exercise led to the recovery of hippocampal acetylcholine (ACh) efflux (F[2,242]=3.63, p<0.05), a selective re-emergence of the Nestin+ cholinergic phenotype (F[1,94]=40.92, p<0.001) and subsequently, a rescue in spatial working memory (F[1,121]=8.25, p<0.01). The most critical finding was that sequestering NGF, but not BDNF, was crucial for the rescue of hippocampal ACh efflux (p<0.001), the re-emergence of the Nestin+ cholinergic phenotype (p<0.002), and the recovery of spontaneous alternation performance (F[1,15]=0.49, p>0.80) in PTD-treated rats. In PF rats, exercise did not improve hippocampal ACh efflux (all p's>0.06), but NGF was required to maintain high ACh levels, as abolishing NGF lead to a decrease in hippocampal ACh (p<0.05). Furthermore, exercise did not improve spatial working memory in PF rats (p's>0.08), although sequestering NGF and BDNF levels with the microbeads impaired performance on the spontaneous alternation task (p's<0.04).

Conclusions: These findings shed greater insight into the extensive routes that exercise has the potential to improve health outcomes. While it has been presumed that the cognitive enhancing effects of exercise were primarily driven by BDNF and improved neurogenesis, our data provide direct evidence that NGF also plays an important role in the pro-cognitive benefits of exercise. Exercise, through the actions of NGF, has the capacity to rescue degenerating cholinergic neurons with the nestin phenotype. The rescue of cholinergic neurons led to improved hippocampal ACh efflux and spatial memory. These data have significant implications for disorders affecting the cholinergic system, and these findings also suggest that exercise could be an important adjunctive therapy alongside current pharmacological treatments to augment cognitive recovery.

Keywords: Acetylcholine, Dementia, Physical Exercise, Hippocampus, Cholinergic Function

Disclosure: Nothing to Disclose.

W46. Characterization of the Serotonin 2A Receptor Inverse Agonist Nelotanserin in Cell-Based Assays and Mouse Models of Psychosis

Travis Grim*, Laura Bohn, Brante Sampey, Lawrence Friedhoff, Rosemarie Roeloffs

The Scripps Research Institute, Jupiter, Florida, United States

Background: Inverse agonists/ antagonists at the serotonin 2A receptor (5HT2AR) have proven to be highly effective, mainstay therapies for the treatment of psychosis for decades. Many atypical antipsychotics, including clozapine, have been explored for their efficacy in treating a wide range of neuropsychiatric deficits. However, atypical antipsychotics target multiple receptors, and in the case of clozapine, demonstrate activity at dopamine and 5HT2C receptors. Nelotanserin is a 5HT2A inverse agonist with 80-fold selectivity over 5HT2C that was shown to be inactive in a panel of 66 receptors and ion channels. Nelotanserin is being investigated in a Phase 2 clinical trial of subjects diagnosed with Lewy body dementia who experience frequent visual hallucinations. In the nonclinical study presented here, we evaluated the in vitro characteristics of nelotanserin and examined its effects in mouse models of 5HT2AR activation and psychosis.

Methods: The competitive properties at the 5HT2AR of nelotanserin were characterized with the DiscoveRx PathHunter β-arrestin2 enzyme fragment complementation assay in h5HT2A-Dx-U2OS cells (n=6-7). Cells were incubated with nelotanserin or clozapine for 90 minutes at 37 oC followed by addition of DiscoveRx detection reagent for 60 minutes, after which luminescence was measured.

in vivo effects were assessed in male C57Bl/6J mice by 1) inhibition of DOI (1-(2,5-dimethoxy 4-iodophenyl)-2-amino propane hydrochloride) -induced head twitch response and 2) inhibition of PCP-induced hyperlocomotion. In the DOI-induced head twitch assay, mice were administered nelotanserin or vehicle 30 minutes prior to administration of the 5HT2A receptor agonist, DOI (1 mg/kg, i.p.) and observed for head twitch response over 1 hour. PCP-induced hyperlocomotion was measured in open field activity monitor chambers. Mice were administered nelotanserin, clozapine, or vehicle (i.p.), and their activity was monitored for 30 minutes before administration of either PCP (10 mg/kg, i.p.) or saline (i.p.) followed by an additional 30 minutes of activity monitoring. All studies were performed in a blinded and randomized manner and potencies were calculated. Clozapine was used as the basis for comparison.

Results: Nelotanserin inhibited serotonin-induced 5HT2AR–β-arrestin 2 interactions, acting as a potent antagonist of the recruitment of β-arrestin 2 to the 5HT2A receptor (IC50 (SEM): 3.74 nM (±0.65)), and was found to be more potent than clozapine (IC50 (SEM): 67.8 nM (±12.1)). In mice, nelotanserin blocked DOI-induced head twitches in a dose-dependent manner (ED50 (95% CL): 2.4 mg/kg (1.3-4.4)), demonstrating potent in vivo antagonism of the 5HT2A agonist-induced behavior. Finally, in a mouse model of psychosis-like behavior, PCP-induced hyperlocomotion was effectively inhibited by nelotanserin (ED50 (95% CL): 3.0 mg/kg (1.7-5.5)). No effect of nelotanserin alone was observed on locomotor behavior following administration of saline. This effect was comparable to the results obtained with clozapine. No significant effects on head twitches or locomotor activity were observed upon treatment of nelotanserin followed by vehicle.

Conclusions: Nelotanserin antagonizes β-arrestin 2 recruitment in cell-based 5HT2A signaling assays and in mouse models blocks the effects of 5HT2A agonism (DOI-induced head twitch) and of psychosis-like behavior (PCP-induced hyperlocomotion). In all three assays, nelotanserin acts in a manner similar to clozapine, a clinically relevant atypical antipsychotic. These data suggest that nelotanserin may prove useful in the treatment of psychotic symptoms.

Keywords: Dementia-Related Psychosis, 5HT2A Receptors, D2 Dopamine and 5HT2A Serotonin Antagonists

Disclosure: Nothing to Disclose.

W47. Poster Withdrawn W48. Neuropsychiatric Symptoms and PET Markers of Alzheimer's Pathology in Cognitively-Unimpaired Individuals With Autosomal-Dominant Alzheimer's Disease

Jennifer Gatchel*, Francisco Lopera, Daniel Norton, Ana Baena, Edmarie Guzman-Velez, Aaron Schultz, Keith Johnson, Reisa Sperling, Gad Marshall, Yakeel Quiroz

Harvard Medical School, Massachusetts General Hospital, McLean Hospital, Boston, Massachusetts, United States

Background: Neuropsychiatric symptoms (NPS) often emerge in early stages of Alzheimer's disease (AD), with anxiety being one of the earliest. Despite this, the relationship between anxiety and AD pathology (e.g. amyloid and tau aggregation) is not well understood. Also of interest is the relation between anxiety in early AD and subjective cognitive decline (SCD). Studying these relationships in carriers of an AD-causing mutation (PSEN-1 E280A) from the Colombian kindred of early-onset autosomal-dominant AD (EOAD) presents the unique opportunity to isolate effects of AD pathology in NPS onset, independent from cerebrovascular disease and other age-associated co-morbidities.

Methods: We administered the Geriatric Anxiety Inventory to twenty-seven cognitively unimpaired subjects (28-56 years old) from the Colombian kindred: ten mutation carriers and seventeen age-matched non-carriers. All subjects had one parent with the PSEN-1 mutation, and thus bore a 50% risk of carrying it themselves; all were blind to their genetic status. All subjects underwent amyloid (Pittsburgh Compound B) and tau (Flortaucipir a.k.a AV-1451) PET imaging. SCD was measured using the Memory Complaint Scale-Spanish version. Spearman's correlations were used to compare anxiety, SCD, cortical amyloid, and regional tau levels in inferior temporal lobe (IT) and entorhinal cortex (EC).

Results: In PSEN-1 carriers, greater anxiety was associated with greater cortical amyloid (r=0.803, p=0.005), but not with tau (IT: r=0.049, p=0.892; EC: r=0.494, p=0.147). Greater SCD (self and informant) was also associated with greater cortical amyloid (self: r=0.735, p=0.015), but not tau (IT: r=0.374, p=0.287; EC: r=0.405, p=0.245). There were no differences between mutation carriers and non-carriers in anxiety (p=0.820) or SCD (self: p=0.538; informant: p=0.759). Among mutation carriers, greater anxiety was associated with greater SCD (self: r=0.810; p=0.005; informant: r=0.584; p=0.076).

Conclusions: Preliminary findings in this small study sample support relationships among anxiety, SCD, and amyloid among mutation carriers from the Colombian kindred of EOAD. Results suggest that anxiety and SCD are related to early changes in AD biomarkers, years before the estimated onset of mild cognitive impairment. Future longitudinal studies with larger samples and in relation to objective cognitive measures are needed to better understand anxiety and SCD in preclinical AD.

Keywords: Anxiety, Dementia, Subjective Cognitive Decline, Amyloid, Tau

Disclosure: Nothing to Disclose.

W49. Effects of Rivastigmine Transdermal Patch on Cognitive Function and Body Weight: An Observational, Retrospective Study in Japanese Patients With Alzheimer Disease

Yasuhiro Kaneda*

Iwaki Clinic, Anan, Japan

Background: Although there are presently four kinds of drugs, donepezil, galantamine, rivastigmine, and memantine, available for the treatment of Alzheimer disease in Japan, up until 2011 there was only one, donepezil. Memantine is the only N-methyl-D-aspartate receptor antagonist while the other three drugs are cholinesterase inhibitors. Rivastigmine is the only patch formulation while the other three are oral drug formulations, and, in Japan, only patch formulation is available for rivastigmine. Cholinesterase inhibitors is the first-line drugs in the treatment of mild-to-moderate Alzheimer disease, but data on the treatment by rivastigmine transdermal patch (rivastigmine patch) in Japan are still limited. This study evaluated the effects of rivastigmine patch particularly on cognitive function in Japanese patients with Alzheimer disease. Besides, weight loss can be seen due to natural course of the disease and/or gastrointestinal side effects in the treatment of the disease. There are some reports in clinical-settings that patients with Alzheimer disease treated with rivastigmine patch showed the improvement of appetite loss, and thus gained or maintained body weight. This study also evaluated the effects of rivastigmine patch on body weight in Japanese patients with Alzheimer disease.

Methods: The study patients were 87 administered rivastigmine patch between July 2012 and December 2015 at the Department of Psychiatry, Iwaki Clinic. Patients fulfilled the following inclusion criteria: diagnosis of Alzheimer disease according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria, with Functional Assessment Staging (FAST) scores 4 or 5 (mild-to-moderate), and had been treated with rivastigmine patch as the first option or switched to rivastigmine patch from any other oral cholinesterase inhibitor, and not in combination with memantine. Patients with other types of dementia were excluded from the study. Neurocognitive function was assessed using the Revised Hasegawa's Dementia Scale (HDS-R) before (baseline) and after the 6-month treatment by rivastigmine patch. The HDS-R consists of 9 simple questions, Age (0-1), Orientation in time (0-4), Orientation in place (0-2), Immediate recall of 3 words (0-3), Serial substractions of 7s (0-2), Digits backward (0-2), Delayed recall of 3 words (0-6), Immediate recall of 5 objects (0-5), and Verbal fluency (0-5), with a maximum score of 30. In addition, body weight was assessed before and after the rivastigmine patch treatment. The JMI software (Version 10.0.2) for Macintosh was used to perform the analyses. The clinical assessment scores were compared between the two assessments by repeated-measures analysis of variance (ANOVA). The level of significance was set at p<0.05. This study was exempt from requiring informed consent from individual patients because the data investigated were retrieved from databases and de-identified before data analyses.

Results: Forty-four out of 87 patients had been treated with rivastigmine patch for 6 months or more, and 43 patients dropped out of rivastigmine patch treatment (20, few visits; 3, simply stop taking rivastigmine patch; 20, switching to any other oral cholinesterase inhibitor or memantine from rivastigmine patch). The main reason for stopping or switching rivastigmine patch was skin irritation (14/23). For 31 (20 females and 11 males) out of 44 patients treated with rivastigmine patch for 6 months or more, neurocognitive function was assessed using the HDS-R before and after the treatment by rivastigmine patch. After 6 months of the treatment by rivastigmine patch, HDS-R total scores did not show significant increase nor decrease (from 18.5±3.7 to 19.1±4.4). Among 9 items, “delayed recall” scores showed significant increase (from 1.5±1.3 to 2.3±1.9, p<0.05). Regarding body weight, there was no significant change between the two assessments (from 51.3±11.1 to 51.8±10.1).

Conclusions: As suggested, rivastigmine patch has been indicated to improve “delayed recall,” and also prevent weight loss in Japanese patients with mild-to-moderate Alzheimer disease. However, careful skin care and/or great rotation of the patch application site would be necessary to avoid dropping out of treatment by rivastigmine patch.

Keywords: Rivastigmine, Patch, Alzheimer's Disease, Cognitive Function, Body Weight

Disclosure: Nothing to Disclose.

W50. Maternal Immune Activation During the Third Trimester is Associated With Neonatal Functional Connectivity of the Salience Network and Fetal to Toddler Behavior

Marisa Spann*, Catherine Monk, Dustin Scheinost, Bradley Peterson

Columbia University, New Haven, Connecticut, United States

Background: During pregnancy, maternal immune activations (MIA) arise from infection, environmental stress, and poor physical health and, in preclinical models, have proximal and long-lasting impact on offspring. Translational human studies are just beginning to consider MIA with alterations in the brain and associated behaviors. In infants, three studies found an association with MIA and head circumference, an indirect measure of the brain. Epidemiological studies have associated MIA with increased risk of psychiatric disorders. Nevertheless, there remains a paucity of human research investigating the role of MIA in altered neurodevelopment. To date, no studies have investigated the effects of MIA on functional connectivity, or the temporal correlation between synchronous fluctuations in brain activity.

Numerous preclinical models investigating MIA with rodents or non-human primates provide templates to inform human studies. MIA has been associated with altered development in a widespread and non-specific set of brain regions including the hippocampus, the prefrontal cortex, the mid-temporal lobe, the parietal lobe, the insula, and the cingulate cortex. However, behavioral deficits are more specific to emotion, inhibition, and attention regulation. In humans, these behaviors have been related to functional connectivity patterns of the salience network, a large-scale brain network anchored in the insula and dorsal anterior cingulate (dACC). Together, these converging results suggest that the insula and anterior cingulate represent good candidate regions for investigations of functional connectivity patterns associated with MIA.

Here, we test the hypothesis that higher levels of MIA as measured by maternal interleukin (IL)-6 and C-reactive protein (CRP) during the 3rd trimester will be associated with greater functional connectivity of the insula and dACC in neonates. In turn, the strength of connectivity will be associated with fetal and infant behavioral measures.

Methods: Thirty-two pregnant women, aged 14 to 19, were recruited from Columbia University Medical Center. They received routine prenatal care and had no major health problems. At 34-37 weeks of gestation, the women underwent diagnostic evaluations and blood draws. Fetal assessments included fetal heart rate (FHR) and heart rate variability (FHRV). Infant assessment included the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III) at 14 months postnatal age. IL-6 and CRP were measured using the enzyme-linked immunosorbent assay. For the neonates, resting-state functional MRI data were acquired on a GE Signa 3T scanner. Standard seed connectivity from the dACC, and the right and left insula was performed. IL-6 and CRP was correlated with connectivity while controlling for postmenstrual age (PMA) at scan and sex.

Results: All neonates were appropriate for gestational age (birthweight: 3243.9±593.8 kg, PMA at birth: 39.3±1.4 weeks) and were scanned at 42.4±1.7 weeks PMA. The majority were female (72%). Regions in which the strength of neonatal connectivity in the salience network correlated with maternal prenatal IL-6 or CRP levels included the medial prefrontal cortex (mPFC), temporoparietal junction, and basal ganglia. Additionally, maternal CRP level correlated inversely with FHRV acquired at the same gestational age. Maternal CRP and IL-6 levels correlated positively with measures of cognitive development on the BSID-III. Finally, neonatal connectivity between the dACC and mPFC correlated with both FHRV and the BSID-III cognitive score.

Conclusions: Consistent with findings from preclinical studies, we showed that MIA during the 3rd trimester likely influences the developing brain, beginning during gestation and extending to the beginning of the 2nd year of postnatal life. Using maternal IL-6 and CRP levels measured during the 3rd trimester and neonatal functional connectivity, we show for the first time that MIA is associated with individual differences in salience network connectivity to the mPFC. The network connectivity is associated with increases in fetal behavioral indices of autonomic nervous development and infant cognition. The brain regions of the salience network are critical for performing many cognitive behaviors, emotion regulation, and autonomic functions and are consistently implicated in neuropsychiatric disorders, suggesting a pathway for MIA to increase psychiatric risk. Our findings provide further evidence that MIA is associated with functional networks implicated in psychiatric disorders and numerous animal models of those illnesses. Future human studies should acquire measures of inflammation and offspring functional connectivity at multiple time points throughout development. They should also manipulate MIA levels experimentally, presumably in prevention trials, to demonstrate more definitively the causal influences of MIA on neurodevelopmental outcomes.

Keywords: Immune Biomarkers, Infant, Cognition, Fetal Neurobehavior

Disclosure: Nothing to Disclose.

W51. Vulnerability of DHCR7+/- Mutation Carriers to Aripiprazole Exposure

Karoly Mirnics*, Zeljka Korade

Munroe-Meyer Institute for Genetics and Rehabilitation, University of Nebraska Medical Center, Omaha, Nebraska, United States

Background: Smith-Lemli-Opitz Syndrome (SLOS) is an inherited neurodevelopmental disorder characterized by multiple congenital malformations and behaviors typical of autistic spectrum disorder. SLOS is caused by mutations in both copies in the gene encoding the last enzyme in the cholesterol biosynthesis pathway – 7-dehydrocholesterol reductase (DHCR7). Defective DHCR7 causes accumulation of 7-DHC, which is a highly oxidizable and toxic compound. It is believed that heterozygous DHCR7 mutation carriers have >1% frequency in the human population, and they are considered healthy. However, we recently found that heterozygous carriers have elevated 7-DHC levels. In addition, we also found that that aripiprazole (ARI), haloperidol (HAL) and trazodone (TRZ) treatment strongly increase 7-DHC levels, even in individuals carrying no mutant copies of the DHCR7 gene. Based on these data, we hypothesized that ARI/HAL/TRZ exposure and single allele DHCR7+/- mutations potentiate each other, elevating 7-DHC levels into a pathological range.

Methods: Time pregnant Dhcr7+/- and WT mice (5/group) were exposed to ARI (5 mg/kg) and vehicle from gestational days GD12 – GD20. Pup brains, lungs and liver were dissected and analyzed at GD20 for sterol levels by our novel GC-MS PTAD method. In a parallel study, we analyzed 1,067 blood samples from the UNMC Biobank for cholesterol, lanosterol, desmosterol, 7-DHC, and 8-DHC.

Results: We found that ARI treatment during embryonic development resulted in 1) fewer number of litters; 2) decrease in number of pups per litter; and 3) strongly and significantly increased 7-DHC in brain, liver, and lung of newborn pups (p<0.001 for all comparisons). This finding was true regardless of Dhcr7 genotype of mother and pup. In context of pup Dhcr7+/- genotype we observed that 7-DHC levels were significantly increased in the ARI-treated brains of pups with the Dhcr7+/- genotype when compared to the ARI-treated controls (p<0.01). In context of combined maternal and pup Dhcr7+/- genotype we found that the highest elevation in 7-DHC was observed when in the group of ARI-exposed Dhcr7+/- pups when the mother also carried the mutant Dhcr7+/- allele (p<0.01 for all group comparisons).

The parallel UNMC Biobank sterol profiling study revealed that individuals treated with ARI and/or TRZ had significantly increased 7-DHC and 8-DHC levels in comparison to untreated individuals (p<0.001 for both).

Conclusions: In the context of our current studies and parallel studies in human dermal fibroblasts, we suggest that treatment with 7-DHC elevating substances (such as ARI, HAL and TRZ) might be unsafe for the population that carries single-allele disruptions of the DHCR7 gene. In addition, it appears that the vulnerability to 7-DHC-elevating compounds is perhaps most pronounced during pregnancy and brain development, especially when both the mother and the fetus carry a single mutant, potentially disruptive DHCR7 allele. This complex drug exposure*maternal genotype*fetus genotype*developmental time point interaction may elevate 7-DHC levels into a toxic range comparable to that seen in SLOS patients, resulting in deleterious developmental outcomes.

Keywords: SLOS, DHCR7, 7-DHC, Cholesterol Biosynthesis, Aripiprazole

Disclosure: Nothing to Disclose.

W52. Altered Brain Structure and Connectivity in Infants Exposed to Serotonin Reuptake Inhibitors During Pregnancy

Claudia Lugo-Candelas*, Jiook Cha, Susie Hong, Vanessa Bastidas, Myrna Weissman, William Fifer, Michael Myers, Ardesheer Talati, Ravi Bansal, Bradley Peterson, Catherine Monk, Jay Gingrich, Jonathan Posner

Columbia University Medical Center/New York State Psychiatric Institute, New York, New York, United States

Background: Selective serotonin reuptake inhibitor (SSRI) use among pregnant women has increased in recent decades (Cooper et al. 2007; Andrade et al. 2016), yet their effect on fetal neurodevelopment continues to be a topic of considerable debate. In the fetal brain, serotonin (5-HT) signaling affects cell proliferation, differentiation, neuronal migration, network formation, and synaptogenesis (Chooi et al. 2007; Velasquez, Goeden, and Bonnin 2013; Kiryanova, McAllister, and Dyck 2013). Atypical serotonergic signaling resulting from prenatal SSRI exposure may thus alter fetal brain development (Hermansen and Melinder 2015). Animal studies document both celuclar changes (e.g. disruption in thalamocortical organization, reduced dorsal raphe neuronal firing, reduced arborization of 5-HT neurons, and altered limbic and cortical circuit functioning) and behavioral consequences (e.g. including increases in anxiety and depression like behaviors in adulthood) in rodents post perinatal SSRI exposure (Olivier et al. 2011; Ansorge et al. 2004; Simpson et al. 2011; Xu, Sari, and Zhou 2004). Human studies have yielded mixed findings; some suggest increased internalizing and externalizing behaviors during early childhood (Oberlander et al. 2006; Casper et al. 2003), whereas others fail to find such associations (Misri 2006; Suri et al. 2011). We examined associations between SSRI exposure and brain development.

Methods: Associations between prenatal SSRI exposure and brain development were examined in 98 infants (16 with parental SSRI exposure, 21 with untreated prenatal maternal depression (PMD) exposure, and 61 healthy controls (HC). Pregnant women were assessed for mood symptoms between 19 and 39 weeks gestation (mean gestation= 33.18±4.57 days) and sleeping, non-sedated infants underwent a MRI session a few weeks after birth (average 3.43±1.50 weeks). Grey matter (GM) morphology and white matter (WM) connectivity (based on numbers of fibers estimated) were assessed using deformation-based GM morphometry and diffusion probabilistic WM tractography, respectively.

Results: Compared to non-SSRI exposed infants (i.e. PMD+HC), SSRI exposed infants showed a significant GM volume expansion in the right amygdala and insula (right amygdala: Cohen's d=0.65; right insula: Cohen's d=0.86) as well as in the right superior frontal gyrus, and the left occipital gyrus at a whole brain corrected p<0.05 (adjusted for infant sex, age at scan, birth weight, and mother's postnatal depression score). Further, probabilistic diffusion tractography showed a highly significant increase in structural connectivity between the right amygdala and right insula in the SSRI exposed group, relative to all non-SSRI exposed infants (t=4.82, p=0.000007, Cohen's d=0.98; adjusted for the aforementioned covariates).

Conclusions: In line with animal studies, the present study suggests a potential influence of prenatal SSRI exposure, likely via aberrant serotonin signaling, on the development of the amygdala-insula circuit in the fetal brain. Amygdala-insula circuitry plays a key role in adaptive emotion processing, and abnormalities in this circuit have been linked to anxiety and depression (Paulus and Stein 2010). Thus, GM volume expansion in the amygdala and insula and WM hyperconnectivity in prenatally SSRI exposed infants could signal an increased vulnerability to anxiety and/or other mood disorders. Further study is required to better elucidate the effects of gestational SSRI exposure on fetal brain development and later life susceptibility to depressive, cognitive, and motor abnormalities.

Keywords: Prenatal SSRI Exposure, Amygdala, Insula, Fetal Brain Development, Maternal Depression

Disclosure: Nothing to Disclose.

W53. Modulation of the Neural Circuitry Underlying Trait Hypnotizability With Spaced Continuous Theta-Burst Stimulation

Nolan Williams*, Keith Sudheimer, Katy Stimpson, Dalton Duvio, Christina Chung, Danielle DeSouza, Booil Jo, Leanne Williams, David Yeomans, David Spiegel

Stanford University School of Medicine, Palo Alto, California, United States

Background: Human neural traits are thought to be stable across the span of adulthood. The neural trait of hypnotizability has been demonstrated to be a reflection of the functional connectivity (fc) between the left dorsolateral prefrontal cortex (L DLPFC—specifically BA9) and the dorsal anterior cingulate cortex (dACC) where higher fc between L DLPFC and dACC has been demonstrated to be correlated with higher hypnotizability scores. Repetitive transcranial magnetic stimulation (rTMS) has been demonstrated to be a functional connectivity modulator that is frequency-dependent. Continuous theta-burst stimulation (cTBS) is an inhibitory form of rTMS and when applied to the L DLPFC, increases fc between L DLPFC and dACC. Spaced, patterned TBS produces long-lasting changes to targeted neural networks.

Methods: Normal healthy control participants were recruited for this pilot study. Each participant received pre-rTMS resting state and structural MRI scans. A hierarchical clustering targeting approach was applied to the pre-TMS resting state scan to identify the area within the L DLPFC (BA 9) with the highest fc to the dACC. Spaced continuous theta-burst stimulation was applied to the identified downstream network target within L DLPFC to dACC. Immediately post-rTMS, the subject received follow-up resting state and structural MRI scans. Pre-rTMS/post-rTMS Hypnotic Induction Profile (HIP) scores were assessed in all subjects.

Results: 7 of 8 subjects experienced an increase in their HIP scores after the application of rTMS with an average increase of 2 points (+/- 1.5) on the 16-point HIP scale (p=0.003). Preliminary functional connectivity analysis demonstrated an increase in the functional connectivity between L DLPFC and dACC.

Conclusions: A human neural trait can be modulated with rTMS. Spaced theta-burst stimulation can produce stable changes over the course of an hour. This is the first evidence that the neural network underpinnings of trait hypnotizability can be modulated thereby potentially expanding hypnotherapy to a larger group of patients.

Keywords: Theta Burst Transcranial Magnetic Stimulation, Neuromodulation, Hypnosis

Disclosure: Nothing to Disclose.

W54. Self-Biases and Neural Activations in Anorexia Nervosa: Linking DSM and RDoC

Carrie McAdams*, Jessica Harper, Erin Van Enkevort

University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: The National Institute of Mental Health created the Research Domain Organizational Criteria (RDoC) as an initiative to advance understanding of the biological units contributing to psychiatric illnesses. Translation of research data using RDoC into clinical intervention and treatment requires relating these constructs to psychiatric illnesses as defined by the Diagnostic and Statistical Manual (DSM). In the DSM, anorexia nervosa is defined in part by cognitive distortions about one's own body; RDoC includes a construct, the Perception and Understanding of Self, in the domain of Social Processes. Here, in concert with clinical assessments of depression, anxiety, and eating disorder symptoms, we evaluated relationships between brain function and cognitive measures related to self-perception in anorexia nervosa.

Methods: A total of 59 women, including 19 healthy comparison women (HC), 18 women with anorexia nervosa but in sustained weight-recovery (AN-WR), and 22 women recently with anorexia nervosa (AN-C), completed office-based assessments and a functional magnetic resonance imaging (fMRI) task. The task required the completion of verbal appraisals in different contexts, including a direct self-appraisal (Self, “I believe I am responsible”) and a reflected-self appraisal (Reflected, “My friend believes I am cynical”); the contrast of these conditions relates to the process of perspective-taking or mentalization. Using whole-brain regressions (PFWE<0.05), we explored whether regional brain differences during mentalization related to either DSM clinical symptoms (eating, depression, or anxiety symptoms) or to the RDoC measures of self-perception (externalizing bias of the Internal Personal Situational Attribution Questionnaire, alexithymia, or self-esteem). Finally, longitudinal clinical and cognitive measures were obtained two to six years later from women with anorexia nervosa.

Results: The externalizing bias, a measure of one's positive relative to negative tendencies during self-attributions, was related to utilization of the left inferior frontal gyrus and insula during mentalization for all participants (4620 mm3, MNI -40, 0, 8, Z=3.70, pFWE<0.001). Pearson's bivariate correlations confirmed this relationship for all subjects (r=0.59, p<0.001), as well as for each group independently (AN-C, r=0.37, p=0.09; AN-WR, r=0.55, p=0.02; HC, r=0.56, p=0.01). Clinical relevance of the externalizing bias is supported by cross-sectional comparisons demonstrating that the AN-C group had more negative externalizing biases than both the HC and AN-WR groups. Finally, the longitudinal dataset showed that the AN-C group that achieved a year of weight-recovery at long-term outcome showed trending changes in their externalizing biases.

Conclusions: The relationship between the externalizing bias and neural function in the insula during mentalization in all three groups provide evidence for biological validity of a self-perception construct in RDoC. Our findings that engagement of the insula during mentalization depend on the valence of one's self-biases, are consistent with literature proposing that the insula is involved in the evaluation of social norms. Eating disorders are characterized by high sensitivity to social expectations; this sensitivity may manifest by altered utilization of the insula during self-evaluations. Second, the relationships between the externalizing bias and clinical state (AN-C vs AN-WR) provide evidence that disturbances in self-perception may contribute to symptoms in anorexia nervosa. Finally, our longitudinal data suggest changes in the externalizing bias are related to clinical outcome in anorexia nervosa, supporting evaluation of the externalizing bias as a therapeutic target in the treatment of anorexia nervosa.

Keywords: Social Cognition, Self-Other Processing, Functional MRI (fMRI), Eating Disorders, Anorexia Nervosa

Disclosure: Nothing to Disclose.

W55. New Evidence that Core Symptoms of Attention Deficit Hyperactivity Disorder (ADHD) are Associated With Both Binge/Disinhibited Eating and Restrictive Eating Behaviour: Implications for Therapy

Colin Dourish*, Panagiota Kaisari, Pia Rotshtein, Suzanne Higgs

P1vital, Wallingford, United Kingdom

Background: Individuals with Attention Deficit Hyperactivity Disorder (ADHD) may be at risk of developing disordered eating. However, it is unclear whether core symptoms of ADHD relate to specific types of disordered eating behaviour and little is known about the mediating mechanisms. In a recent online study, we showed for the first time that the inattentive symptoms of ADHD relate to binge/disinhibited and restrictive eating, both directly (for binge/disinhibited eating) and indirectly via negative mood and awareness of internal hunger and satiety signals. These results have potentially important clinical implications for therapy, and therefore, in the present study, we investigated whether our findings are robust and could be replicated in a well-defined clinical laboratory sample. We also investigated the specific contributions of hyperactivity and impulsivity to disordered eating using both self-report and performance-based measures. In addition, we examined whether interoceptive sensitivity assessed objectively using a heartbeat perception task provides another pathway of association between inattentive symptoms of ADHD and disordered eating as identified in our previous study where we used a self-report measure of awareness of internal hunger and satiety signals.

Methods: We conducted a laboratory based study of 142 student volunteers aged 18-32 from the University of Birmingham, UK (80.3% female). Body Mass Index (BMI) range was 14.7-34.4 kg/m2, mean 21.4 kg/m2 (SE=0.3). The majority of participants (66%) had a BMI within the normal range; 18% were classified as underweight; 14.5% as overweight and 1.5% as obese. Symptoms of ADHD were assessed using the Conner's Adult ADHD Rating Scale Self Report-Screening Version (CAARS: S-SV). The Binge Eating Scale (BES), the Bulimic Investigatory Test, Edinburgh (BITE), the Loss of Control over Eating Scale (Brief version), the Dutch Eating Behaviour Questionnaire and the Eating Attitudes Test (EAT-26) were used to assess disordered eating. A composite measure based on the total score of the Hospital Anxiety and Depression (HADS) scale and the Perceived Stress Scale was created to evaluate the potential mediating role of negative mood. The Reliance on Hunger/Satiety cues sub-scale of the Intuitive Eating Scale was used to assess interoception. A food-based Go/No Go task and the Barratt Impulsivity scale (BIS) were used to assess impulsivity (as distinct from hyperactivity) and a heartbeat perception task was used to objectively assess interoception.

Results: The mean, SE, minimum and maximum scores on the 4 subscales derived from the CAARS: S-SV were Inattentive Symptoms (0-27), 9.23, 0.43, 0-24; Hyperactive-Impulsive Symptoms (0-27), 7.25, 0.40, 0-22; Total ADHD Symptoms (0-54), 16.49, 0.75, 0-45; ADHD Index (0-36), 11.71, 0.4, 2-31. Principal component analysis reduced the disordered eating measures to two components: “binge/disinhibited eating’’ and “restrictive eating’’. Controlling for age, gender, Body Mass Index (BMI) and alcohol use, both inattentive and hyperactive/impulsive symptoms of ADHD related to both binge/disinhibited and restrictive eating. In addition, negative mood meditated the relationship between core symptoms of ADHD and both types of disordered eating. Reliance on internal hunger/satiety signals (but not interoceptive sensitivity assessed by accuracy in the heartbeat perception task) mediated the relationship between inattentive symptoms of ADHD and binge eating and restrictive eating. We also found a direct relationship between inattentive symptoms of ADHD and binge/disinhibited eating behaviour and hyperactive/impulsive symptoms of ADHD were directly related to both binge/disinhibited and restrictive eating. Impulsivity as assessed using the BIS, directly predicted binge/disinhibited eating but there was no evidence to suggest a direct association with restrictive eating, providing indirect evidence for a specific contribution of impulsivity symptoms of ADHD to binge/disinhibited eating and hyperactivity symptoms to restrictive eating. There was no moderating effect of age, gender, ADHD medication and/or BMI in any of the relationships between core symptoms of ADHD and disordered eating.

Conclusions: We provide the first evidence from a laboratory study of an important role for inattentive symptoms in mediating disordered eating in ADHD. These results confirm and extend our previous findings from an online study of ADHD and eating behaviour that used the same self-report measures. ADHD symptoms were positively related to disordered eating, including both binge/disinhibited and restrictive eating and negative mood, reflecting aspects of anxiety, depression and stress mediated the relationships. Deficits in awareness and reliance on internal hunger/satiety signals provided another mechanistic pathway of association between inattentive symptoms of ADHD and disordered eating, especially binge/disinhibited eating. Interestingly, inattentive symptoms of ADHD were directly related to disordered eating, specifically binge/disinhibited eating. These findings could have important implications for prevention and early intervention programmes, which should focus on mood regulation in individuals with ADHD symptoms at risk for developing disordered eating. Further investigation of the role of inattentive, hyperactivity and impulsivity symptoms in disordered eating may be helpful in developing novel treatments for both ADHD and disordered eating.

Keywords: Eating disorders, Binge Eating Disorder, ADHD, Inattentive Symptoms, Impulsivity

Disclosure: Part 1: P1vital, Employee, Board Member, Stock / Equity, Part 2: P1vital, Stock / Equity, Part 3: P1vital, Employee, Stock / Equity, Part 5: P1vital, Employee.

W56. Exaggerated Neural Anticipation of Aversive Interoceptive State Change is Associated With Past Symptom Severity in Adult Women Remitted From Bulimia Nervosa

Laura Berner*, Alan Simmons, Christina Wierenga, Amanda Bischoff-Grethe, Martin Paulus, Ursula Bailer, Miki Ogasawara, Laura Greathouse, Walter Kaye

University of California, San Diego, La Jolla, California, United States

Background: The pathophysiology of bulimia nervosa (BN) is currently unknown, but hallmark binge eating and purging symptoms and high levels of emotional instability suggest that individuals with BN have difficulty regulating internal homeostasis. Integral to both physiological state regulation and emotion regulation is interoception, or the detection and integration of body signals (e.g., hunger, heartbeat, respiration). Women with BN show altered anticipatory neural signaling for pleasant interoceptive events (i.e., palatable tastes), implicating reward-based learning abnormalities in the disorder. However, little is known about how the brains of individuals with BN process non-food-specific, aversive interoceptive stimuli. Since BN symptoms both create sensations that would be unpleasant to individuals without an eating disorder and involve extreme intolerance of unpleasant sensations (e.g., purging to eliminate fullness), abnormal anticipation and processing of aversive body states also may promote faulty learning and recurrent maladaptive eating behaviors in BN. In the present study, we aimed to 1) determine whether women remitted from BN (RBN) and control women (CW) differ in their neural response to the anticipation, experience, and termination of a non-painful aversive breathing load and 2) examine the relationship between past BN symptom severity and altered interoceptive response. Because we hypothesized that interoceptive disturbance plays a joint role in physiological state and emotion dysregulation in BN, our exploratory analyses focused on the amygdala. We studied only adults who were remitted from BN to avoid the confounding effects of active symptomatology.

Methods: Functional magnetic resonance imaging (fMRI) measured neural responses during an event-related, intermittent breathing load restriction paradigm in 24 RBN and 25 CW group-matched for age, body mass index, and years of education. AFNI's 3dttest ++ compared the groups’ neural activation during three conditions of interest: anticipation of breathing load, 40 cm H20/L/sec inspiratory breathing load, and termination of breathing load. As in our prior investigation of remitted anorexia nervosa (Berner et al., 2017), we restricted our fMRI analyses a priori to bilateral regions of interest involved in non-painful aversive interoception: the insula, anterior/mid-cingulate (ACC), posterior cingulate (PCC), lateral PFC, dorsal and ventral striatum, and the amygdalae. Intrinsic smoothness was estimated using the spatial autocorrelation function (acf) option in AFNI's 3dFWHMx. Each ROI was treated as a search region, and small-volume multiple comparison correction was applied using Monte Carlo-based simulations in AFNI's 3dClustSim (voxel-wise p<0.001, cluster-wise alpha=0.05). Poisson generalized linear models examined whether altered anticipatory amygdala activation in RBN was associated with past binge eating and purging frequency.

Results: During breathing load anticipation, the RBN group showed increased activation compared with CW in left (t=3.88) and right (t=3.77) mid-insula, left posterior insula (t=3.67), left superior frontal gyrus (t=3.75), dorsal ACC (t=3.71), PCC (t=3.76), bilateral putamen (ts 3.61-3.71), and right amygdala (t=4.25). There were no group differences during or after breathing load. Exploratory analyses within the RBN group revealed that right amygdala activation during breathing load anticipation was positively associated with worst past binge eating frequency (p=0.004) and self-induced vomiting frequency (p<0.0001).

Conclusions: Results add to evidence suggesting that BN is associated with altered processing of salient stimuli. Although women with BN show reduced anticipatory responses for pleasant food stimuli, we observed an increased preparatory signal during anticipation of non-painful aversive body state changes. Individuals with depression show similar hyperactivation during the anticipation of painful and visually aversive stimuli, but in this non-depressed RBN sample, anticipatory amygdala hyperactivation was uniquely associated with markers of past binge eating and purging severity. In contrast to our prior findings in remitted anorexia nervosa, we did not find evidence of increased activation during or after the experience of a non-painful aversive state. Abnormal anticipatory brain activation may contribute to an impaired ability to modulate response to interoceptive and affective states, ultimately interfering with effective learning from body-related experience during binge/purge episodes and maintaining these maladaptive eating behaviors. Whether these alterations are a pre-existing trait or a brain-based abnormality resulting from BN symptoms is unclear. However, if replicated in ill BN, these preliminary results suggest that altered interoceptive anticipation may be an effective target for novel interventions.

Keywords: Bulimia Nervosa, Interoception, Functional MRI (fMRI)

Disclosure: Nothing to Disclose.

W57. Repurposing Medications to Target Core Behavior and Comorbid Symptoms in Eating Disorders

Terry Schwartz*, Walter Kaye, Mary Ellen Trunko, Emily Gray, Anne Cusack, Laura Berner, Tiffany Brown, Tiffany Nakamura

University of California, San Diego, California, United States

Background: Anorexia nervosa (AN), bulimia nervosa (BN), and Avoidant and restrictive food intake disorder (ARFID) tend to be chronic disorders with considerable morbidity and mortality. There are no FDA approved medications for AN or ARFID. While more is known about medication for BN, there are no proven treatments for those with severe emotional and impulse dysregulation. Our center has been involved in a series of investigations testing existing medications to determine if they reduce eating disorder (ED) core symptoms, as well as reduce common comorbid emotional, cognitive and behavioral dysfunctional patterns.

Methods: We have used open trial or chart review designs to test the efficacy of several medications. We also sought to find standardized assessment instruments that are sensitive to pharmacological induced changes in behavior in ED. We report on open trials of 1) Lamotrigine, as a mood stabilizer in patients with eating disorders and affective dysregulation who do not have DSM V bipolar diagnosis’ 2) ketamine in ED patients with severe depression; 3) Aripiprazole and SSRI's in AN; and 4) mirtazapine, an atypical antidepressant with gastrointestinal symptom relief properties as well as appetite stimulant properties in ARFID.

Results: Nine ED patient who did not respond adequately to antidepressant medications took lamotrigine for 147+79 days (mean final dose=161±49 mg/day) while in day and evening treatment. Time on lamotrigine (ps<0.0001) and higher lamotrigine doses (ps<0.005) were associated with reduced self-reported affective and behavioral dysregulation. Eating disorder and mood symptoms decreased moderately. 2) Five ED patients were treated with multiple doses of IV or IM ketamine over 2 to 12 months. Three patients with severe treatment resistant depression had a substantial reduction in their depression, but there did not appear to be a clear or consistent reduction in ED core symptoms in any patient. 3) Preliminary impressions suggest that aripiprazole may be more effective than other atypicals in reducing anxiety and improving eating and weight gain in AN, and that concomitant administration of SSRI's may potentiate Aripiprazole response. 4) In 14 ARFID patients on weight restoration, the average change in BMI per week prior to initiating mirtazapine was.10 BMI points per week (SD =.08) and the average change in BMI per week post mirtazapine was.23 BMI points per week (SD =.14) (p <.05). These results suggest that mirtazapine may facilitate weight gain.

Conclusions: Recent imaging and genetic studies provide new insights into the powerful neurobiological underpinnings that contribute to ED and can be used to nominate existing medication to target new specific symptoms. 1) Lamotrigine showed initial promise in reducing emotional instability and behavioral impulsivity in severely dysregulated ED patients. 2) The use of ketamine for treatment resistant severe depression may be extended to treat depression in ED. 3) Targeting dopamine D2 receptors and the serotonin system via the unique pharmacological profile of Aripiprazole may reduce anxiety, improve rigid thought processes and facilitate weight gain in AN; and 4) Mirtazapine was safe and well tolerated in ARFID patients and was associated with a greater rate of weight gain than compared to our treatment as usual weight restoration program. Although our findings are limited by the confounds inherent in an open series, these preliminary results support further investigation of these medications for a variety of ED patients using rigorous controlled trials.

Keywords: Eating Disorders, Pharmacology, Neuroimaging, Neurocircuitry

Disclosure: Nothing to Disclose.

W58. Sex-Specific Alterations in Dopamine Transporter Function From Food Restriction and/or Exercise are Amplified During Adolescence

T. Lee Gilman*, W. Anthony Owens, Lauren Metzel, Christina George, Lynette Daws

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

Background: Afflicting at least 3 percent of teenagers, eating disorders such as anorexia nervosa, bulimia nervosa, and binge eating disorder entail severe health consequences in addition to their psychological toll. Yet no effective treatments for eating disorders exist. An established regulator of both eating behaviors and physical activity, the dopaminergic system undergoes a sensitive maturation period during adolescence. Eating disorders, which are 2.5-fold more prevalent in females than males, typically emerge during adolescence. However, studies into the role of the dopaminergic system in ontogeny of eating disorders are lacking, as are investigations into dopaminergic system maturation in adolescent females.

Methods: Here we measured function of the dopamine transporter (DAT), a critical regulator of dopaminergic signaling, using both in vivo high-speed chronoamperometry and locomotor assays of cocaine response. Using an activity-based anorexia paradigm we explored how food restriction, exercise on a running wheel, or the combination thereof impacted DAT function in adult (postnatal day 90) and adolescent (postnatal day 30) Sprague-Dawley rats of both sexes.

Results: Broadly, rats exposed to both food restriction and free exercise showed leftward shifts in the dose-response to the locomotor-promoting effects of cocaine, regardless of sex or age. We further observed enhanced locomotion in response to cocaine and heightened striatal function of DAT in females compared to males, and in adolescents compared to adults. Similarly, free running wheel access and restricted food access produced dramatic reductions in adolescent, but not adult, DAT-mediated dopamine uptake. Adolescent females in particular exhibit the most pronounced exercise- and diet-induced decreases in striatal DAT function.

Conclusions: Together, these findings suggest that adolescent plasticity of the dopaminergic system confers – particularly in females – vulnerability to eating disorders and persistence of associated unhealthy behaviors (e.g., compulsive exercise). Therefore, drugs that enhance dopamine uptake or otherwise reduce dopamine signaling duration may prove efficacious in the treatment of adolescent eating disorders. Ongoing experiments are evaluating striatal DAT expression, as well as circulating insulin and leptin levels, as both of these metabolic hormones influence DAT function. Future experiments will examine how adolescent changes in DAT function and sensitivity, as a result of exercise and/or food restriction, affect voluntary access to drugs of abuse or highly palatable foods.

Keywords: Dopamine Transporter, Anorexia Nervosa, Adolescence, Sex Difference, Eating Disorders

Disclosure: Nothing to Disclose.

W59. Taste Reward Learning Response, its Relationship to Cortisol, and Effective Connectivity in Adolescent Anorexia Nervosa

Guido Frank*, Marisa DeGuzman, Megan Shott

University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States

Background: Anorexia nervosa is a severe psychiatric disorder that predominantly occurs in females, and that has its onset during adolescence. No medication has been approved for the disorder. Previous research in adults with anorexia nervosa indicated elevated brain activation to a taste temporal difference reward-learning paradigm. In that paradigm, unexpected delivery or omission of stimuli elicit a neuronal prediction response error in dopaminergic but also other neuronal populations. A recent study from our group in a smaller group of adolescents with anorexia nervosa and using monetary stimuli further supported heightened prediction error response. In the present study, we wanted to test in a large group of adolescents whether we would also find elevated prediction error response in adolescent anorexia nervosa in response to taste stimuli. We further wanted to test whether such an elevated response would be observed before and after treatment as an indicator of lasting neurobiological abnormalities beyond weight restoration. Lastly, anorexia nervosa is associated with elevated cortisol levels, and we hypothesized that the stress system could functionally be involved in elevated prediction error response in anorexia nervosa based on basic research.

Methods: We recruited 108 adolescents, 56 adolescents with anorexia nervosa (age adjusted body mass index below the 10th percentile) and 52 age matched healthy controls. All study participants were carefully assessed for clinical measures and underwent blood oxygen level-dependent functional magnetic resonance brain imaging (fMRI) at the begin of treatment in a highly structured specialized eating disorder treatment program. During fMRI, the study participants learned the association between conditioned visual stimuli and unconditioned taste stimuli (sucrose solution, tasteless water), as well as the unexpected violation of those learned associations. A subset of participants, 28 adolescents with anorexia nervosa and 32 healthy controls completed a second scanning session; the adolescents with anorexia nervosa were scanned at the end of treatment, and the healthy control group 6 weeks after scan one, which approximates the time in treatment for the anorexia nervosa group. Study participants also provided samples for salivary cortisol and amylase. Individual trial-by-trial computational prediction error regression with fMRI brain response was conducted, and both a whole brain and region of interest analysis was used to identify within and between group activation. Salivary cortisol was tested whether it would be related to prediction error response. In addition, an independent multisample greedy equivalence search algorithm tested effective connectivity between brain taste reward regions during sucrose tasting.

Results: An unmasked whole brain analysis indicated higher prediction error response (p<0.05, FWE corrected) in adolescents with anorexia nervosa in bilateral caudate head (L: x=-8, y=12, y=0; R: x=8, y=8, y=-2) and bilateral pulvinar (L: x=-16, y=-28, y=10; R: x=16, y=-28, y=10). An anatomical region of interest (ROI) based analysis (Bonferroni corrected) also found elevated bilateral caudate prediction response (L p<0.005; R: p<0.001) in anorexia nervosa, as well as in the left antero-ventral insula (p<0.011) and bilateral nucleus accumbens (L: p<0.005; R: p<0.029). The repeated measures ANCOVA in participants with two scans showed a significant (Bonferroni corrected) effect for group in the left caudate head (p<0.012) with higher activation in the anorexia nervosa group. Salivary cortisol regression with prediction error response indicated a significant cluster (whole brain p<0.05 FWE corrected, x=18, y=58, y=6) in the right medial prefrontal cortex. The effective connectivity analysis showed in controls that the hypothalamus drove ventral striatal activity, however, in adolescents with anorexia effective connectivity was directed from anterior cingulate via ventral striatum to the hypothalamus.

Conclusions: This study further implicates brain prediction error response in anorexia nervosa in a relatively large sample of adolescents. Cortisol has repeatedly been found elevated in anorexia nervosa, and this study implicates cortisol in prefrontal cortex activity in anorexia nervosa, which could be related to dopamine function in the disorder. Lastly, the effective connectivity results support the hypothesis that anorexia nervosa is associated with a strong top down control mechanism over hypothalamic signals, a brain region that integrates signals of body homeostasis. Anterior cingulate cognitive-emotional top down control could affect food reward and eating drive, override hypothalamic inputs to the ventral striatum and enable prolonged food restriction.

Keywords: Anorexia Nervosa, Reward Prediction Error, Effective Connectivity, Chronic Food Restriction Stress, Cortisol

Disclosure: Nothing to Disclose.

W60. Sex-Based Differences in Neuronal Response to Hedonic Foods

Kristina Legget*, Marc-Andre Cornier, Daniel Bessesen, Brianne Mohl, Jason Tregellas

University of Colorado School of Medicine, Aurora, Colorado, United States

Background: Sex-based differences are observed in obesity rates, eating behaviors, and eating disorders, likely reflecting differences in neurophysiology underlying food-related behaviors in men and women. Hedonic eating, or eating beyond homeostatic needs, may be particularly associated with obesity. As such, the current study investigated sex-based differences in neuronal responses to images of food with high hedonic value, compared to foods of neutral hedonic value. This was assessed in both fasted and fed states, using functional magnetic resonance imaging (fMRI). The study also sought to determine if sex-based differences in neuronal response to food cues varied based on propensity to obesity, with the inclusion of both obesity-prone and obesity-resistant study groups. Identifying how sex influences neurobiological responses to food cues, particularly those related to hedonic eating, and how this relates to obesity, may help to inform the development of more effective, sex-specific weight loss and maintenance strategies.

Methods: Fifty-six adults completed the study, half of whom had a propensity to be resistant to weight-gain and obesity (obesity-resistant [OR]) and half of whom were prone to weight gain and obesity (obesity-prone [OP]). The final analysis included 25 OR participants (11 women, 14 men) and 28 OP participants (14 women, 14 men). After completing a four-day, macronutrient-controlled, eucaloric run-in diet, participants underwent functional magnetic resonance imaging (fMRI) while viewing visual stimuli in both fasted and fed states. Visual stimuli consisted of foods with high hedonic value, foods with neutral hedonic value, or non-food objects. The primary contrast of interest compared responses to high hedonic foods vs. neutral hedonic foods.

Results: In the fasted state, a significant effect of sex on the response to hedonic compared to neutral foods was observed, such that women showed greater response than men in the nucleus accumbens (p=0.0002) and insula (p=0.010). Sex-based differences in response to hedonic compared to neutral foods were not observed in the fed state. No significant effects of group (OP vs. OR) or significant interactions between sex and group when comparing hedonic to neutral foods were observed in the fasted or fed states.

Conclusions: The current study found a greater response to hedonic food cues in the fasted state in reward-related brain regions in women compared to men, suggesting that women may be more sensitive to reward value of hedonic foods than men when fasted. As hedonic eating is a key contributor to obesity, this may relate to higher rates of binge eating and obesity observed in women compared to men in the general population.

Keywords: Sex Differences, Obesity, Brain Imaging, fMRI

Disclosure: Nothing to Disclose.

W61. Bariatric Surgery in Obese Patients Reduced Resting Connectivity of Brain Regions Involved With Self-Referential Processing

Yi Zhang*, Guanya Li, Qianqian Meng, Jizheng Zhao, Karen von Deneen, Ehsan Shokri-Kojori, Dardo Tomasi, Nora Volkow, Gang Ji, Gene-Jack Wang

Xidian University, Xi'an, China

Background: Obese patients exhibit increased brain activation in motivation-reward neurocircuits encompassing limbic and striatal regions and in interoceptive processing brain regions while showing reduced activation in executive frontal regions when exposed to food stimuli. Bariatric surgery is the most effective intervention to sustain weight reduction in morbid obesity. However, how it influences neurocircuitry following weight loss is largely unknown. Functional connectivity density (FCD) mapping has been used to assess the spatial distribution of brain connectivity hubs, whereas seed-voxel correlation analyses have been used to identify connectivity patterns relative to specific brain regions during resting-state. Here we employed FCD mapping to examine the effects of bariatric surgery on brain neurocircuitry and their relation to relevant clinical variables including weight loss.

Methods: Twenty-two obese patients who underwent bariatric surgery (BS) and 19 obese controls (Ctr) underwent 3.0T MRI. BS group was tested before and one month after bariatric surgery, and Ctr group was also tested twice without surgey: at baseline and one month later. For each image scan, we conducted a 6-min resting-state fMRI scan (T2*-weighted gradient echo planar imaging sequence in a 3T GE MRI scanner, TR/TE=2000/30 ms, 64 x 64 matrix size, 256 x 256 mm2 FOV, 90° flip angle, 4 mm2 in-plane resolution, 4 mm slice thickness and 32 axial slices) under a 12-hour fasting condition. The images were realigned, normalized to MNI space, 0.01–0.08 Hz band-pass filtered, and time points with excessive motion (framewise displacement>0.5 mm) were removed. Nuisance signal fluctuations in white matter, cerebrospinal fluid and the whole brain, as well as 6 motion parameters (obtained from realignment) and their temporal derivatives were removed from data. FCD mapping was employed to calculate local (lFCD) and global (gFCD) voxelwise connectivity metrics using the standard correlation threshold (r=0.6). Two factor (BS, Ctr) repeated (Baseline, 1 Month Later) ANOVA measures were used to assess the main effects and interactions in lFCD and gFCD; regions of interest were identified for subsequent seed-voxel correlation analyses to assess resting-state functional connectivity (RSFC) and to examine the association with weight loss. Statistical significance was based on family-wise error (FWE) corrections for multiple comparisons at the cluster level (PFWE<0.05) using a minimum cluster size of k=100 and a cluster forming a threshold of P<0.001.

Results: Baseline FCD metrics did not differ between BS and Ctr groups. In both groups, higher lFCD in the ventral medial prefrontal cortex (VMPFC) was associated with higher body mass index (BMI). Bariatric surgery significantly decreased lFCD in the VMPFC, posterior cingulate cortex (PCC)/precuneus, and dorsal anterior cingulate cortex (dACC)/dorsomedial prefrontal cortex (DMPFC) and decreased gFCD in the VMPFC, right dorsolateral prefrontal cortex (DLPFC) and right insula in the BS group (PFWE<0.05). Decreased lFCD in the VMPFC and PCC/precuneus was correlated with a reduction in BMI after surgery. Seed voxel connectivity analyzed in these regions showed that after surgery the VMPFC had stronger connectivity with the left DLPFC and weaker connectivity with the hippocampus/parahippocampus, and that the PCC/precuneus had stronger connectivity with the right caudate and left DLPFC. There were no significant lFCD or gFCD changes between the repeated measures in the Ctr group.

Conclusions: Bariatric surgery significantly decreased FCD in regions involved in self-referential processing (VMPFC, DMPFC, dACC, precuneus), and interoception (insula), and the changes in the VMPFC and precuneus were associated with a reduction in BMI implicating them in the improvement in control of eating behaviors following surgery.

Keywords: Obesity, Bariatric Surgery, Self-Referential, Sensory-Emotional-Memory, fMRI

Disclosure: Nothing to Disclose.

W62. Chemogenetic Inactivation of Corticostriatal Projections Differentially Disrupts Impulsive Choice in Rats Selected for High or Low Trait Impulsivity

Natalie Zlebnik*, Jennifer Wenzel, Mary Patton, Brian Mathur, Joseph Cheer

University of Maryland School of Medicine, Baltimore, Maryland, United States

Background: Impulsive choice is characterized by preference for smaller immediate rewards over larger delayed rewards, and trait impulsive choice behavior is a determining factor in the development and persistence of drug abuse. However, the neuroanatomical substrates of impulsive choice and how they may differ in impulsive vs. controlled phenotypes are still under investigation. Emerging evidence from human and animal studies suggests frontal cortical regions exert influence over striatal reward processing areas during decision-making in impulsive choice or delay-discounting tasks.

Methods: To examine how these circuits are involved in decision-making in animals differing in trait impulsive choice, we used chemogenetic tools to selectively and reversibly target corticostriatal projections during the performance of a delay-discounting task in rats previously screened for high (impulsive) vs. low (controlled) impulsivity. The prefrontal cortex (PFC) was injected with a viral vector expressing inhibitory designer receptors exclusively activated by designer drugs (Gi-DREADD), and then PFC projections to the nucleus accumbens (NAc) were selectively suppressed by intra-NAc administration of the Gi-DREADD activator clozapine-n-oxide.

Results: Inactivation of the PFC-NAc projection elicited a robust increase in impulsive choice in controlled rats without affecting decision-making in impulsive rats. This demonstrates a critical role for PFC afferents to the NAc during controlled choice behavior and suggests that maladaptive hypofrontality underlies decreased executive control in animals with high trait impulsivity.

Conclusions: Results such as these may have important implications for the pathophysiology and treatment of drug addiction and impulse control disorders such as attention deficit/hyperactivity disorder.

Keywords: Impulsivity, Delay Discounting, Medial Prefrontal Cortex, Nucleus Accumbens, Chemogenetics

Disclosure: Nothing to Disclose.

W63. Altered Corticostriatal Activations and Connectivity During Reinforcement Learning in Unmedicated Adults With Obsessive-Compulsive Disorder

Marilyn Cyr*, Sophie Schiff, Martine Fontaine, Emily Steinberg, Helen Simpson, Rachel Marsh

New York State Psychiatric Institute, Columbia University, New York, New York, United States

Background: Obsessive-compulsive disorder (OCD) is characterized by intrusive thoughts (i.e., obsessions) and repetitive behaviors (i.e., compulsions) to avoid or reduce distress associated with these thoughts. OCD has a lifetime prevalence of 2%-3% [Robins et al., 1984; Kessler et al., 2005] and is classified by the World Health Organization as one of the top ten most debilitating illnesses of any kind in terms of lost earnings and diminished quality of life [Bobes et al., 2001]. Animal and human research suggest that dysfunction in cortico-striato-thalamo-cortical (CSTC) circuits may underlie OCD symptoms [Graybiel & Rauch, 2000; Maia et a., 2008; Aouizerate et al., 2004; Menzies et al., 2008]. Specifically, the compulsive, ritualistic behaviors that characterize OCD are thought to reflect deficits in and an imbalance between two decision systems: habit and goal-directed [Voon et al., 2014; Gillan et al., 2015]. Habits are automatic behaviors acquired through stimulus-response (S-R) learning and involve the sensorimotor CSTC circuit, whereas goal-directed behaviors are acquired through purposeful, reward-based learning and involve the limbic/associative CSTC circuit [Balleine & O’Doherty, 2010]. Although abnormal functioning of mesolimbic and ventral striatal circuitry during reward-based spatial learning have been documented in OCD [Marsh et al., 2015], it is unclear whether the functioning of CSTC circuits that during S-R learning is altered or preserved.

Methods: In the present study, we used computational fMRI with a reinforcement learning (RL) model to assess striatal activation and CSTC connectivity in 28 unmedicated adults with OCD and 28 healthy controls (HC) during their performance on a virtual reality-based task analogous to the “win-stay” version of a radial maze task used in rodents. A Q-learning model [Watkins and Dayan, 1992], which involves learning the value Q(s, a) of an action (or response) a in a state s, was fitted to each participant's choice behavior of entering a lit (baited with a reward) or unlit (unbaited) arm. Based on a priori hypotheses that posterior putamen codes for S-R learning, we examined group differences Q-learning related signals in the striatum. We also examined changes in corticostriatal connectivity with Q-learning by computing a psychophysiological interaction (PPI) [Friston et al., 1997] using a functionally defined seed point within the posterior putamen.

Using the same task paradigm and RL model with healthy adults, we previously showed that Q-learning was associated with (1) increases in posterior putamen activity, (2) increases in connectivity between the posterior putamen and cortical sensorimotor regions, and (3) decreases in connectivity between the posterior putamen and several nodes of the limbic CSTC circuit [Horga et al., 2014]. In the present study, we hypothesized that, relative to HC, patients with OCD would show 1) greater engagement of and connectivity within sensorimotor CSTC, and 2) less connectivity within limbic CSTC with S-R learning.

Results: Both groups showed significant engagement of right posterior putamen and sensorimotor regions with Q-learning, consistent with previous findings from healthy adults [Horga et al., 2014]. Unlike HC, patients with OCD significantly deactivated bilateral ventral striatum with Q-learning. These findings are compatible with abnormal recruitment of mesolimbic and ventral striatal circuitry during reward-based learning [Marsh et al., 2015]. Connectivity analyses revealed greater learning-related increases in connectivity between right posterior putamen and sensorimotor regions in patients with OCD than in HC. These results support our hypothesis of increased S-R-related CSTC functioning in OCD.

Conclusions: These findings suggest that, in addition to the well documented deficits in reward-based (goal-directed) learning in OCD [Gillan & Robbins, 2014], S-R learning processes may also be abnormal (i.e., enhanced), possibly contributing to the habitual nature of their compulsive, ritualistic behaviors that characterize the disorder. Future studies should assess whether S-R learning processes are also altered in early-onset (i.e., pediatric) OCD and whether such alterations precede the disorder in the form of a risk marker.

Keywords: Reinforcement Learning, Obsessive Compulsive Disorder, Functional MRI (fMRI), Computational Modeling, Cortico-Striatal Circuit

Disclosure: Nothing to Disclose.

W64. Exploring Top-Down Control of the VTA and Lateral Habenula by the mPFC in Reward Prediction and Impulsive Decision Making

Adina Buxbaum*, Sage Aronson, Joshua Chandra, Roberto Malinow

University of California, San Diego, La Jolla, California, United States

Background: The medial prefrontal cortex (mPFC) plays an integrative role in decision making and goal directed behaviors, furthermore, lesions in the mPFC increase impulsive behaviors. Altered impulsivity may result in a range of psychiatric disorders such as attention deficit/hyperactivity disorder (ADHD) and substance abuse, among others. Consistent with the important role of the mPFC in top down cognitive control, downstream projections from the mPFC coordinate activity in midbrain and epithalamic regions. In the case when alternate decisions yield advantages in different forms, mPFC circuits may ‘battle out’ for downstream activation or inhibition of regions involved in disappointment or reward, specifically the lateral habenula (LHb) and Ventral tegmental area (VTA). We propose that local collaterals in the cortex may laterally inhibit excitatory projections to different brain regions, resulting in altered behavior, depending on which population of mPFC neurons ‘wins’.

Methods: 1. Electrophysiology: We address these questions with acute slice physiology and live activity imaging of specifically projecting cortical neuron populations. The recent development of retrograde virus strategies allows us to target cortical neuron populations with projections specifically to the LHb or VTA. By combining patch clamp physiology and optogenetic stimulation of light activated receptors expressed in cortical neurons we are investigating the inhibitory connectivity of local collaterals in LHb or VTA projecting mPFC neurons. This allows us to optogenetically activate LHb and VTA targeting populations, while recording responses from other neuron populations. 2. Live neuron activity imaging in behaving rats: ‘Temporal Discounting of Reward’ behavioral paradigms, where rats are trained to choose between a small, immediate reward or a larger, delayed reward is being used in conjunction with live activity imaging of LHb or VTA projecting cortical neuron populations in free moving rats using Fiber Photometry (FP). FP utilizes brain-implanted fiber optics to detect fluorescent fluctuations in calcium or voltage sensors in labeled neuron populations. We are imaging calcium indicator activity in LHb or VTA projecting afferents in real time while the animal performs an impulsivity-testing task. This will elucidate afferent specific mPFC neuronal activity prior to, during and after decision-making, revealing insights into the function and output of the prefrontal cortex during impulsivity.

Results: Patch clamp electrophysiological recordings of mPFC neurons in acute slices, combined with optical stimulation of LHb projecting mPFC neurons have suggested a mechanism of lateral inhibition in cortical networks, confirming our primary hypothesis. We are currently investigating a relationship in the activity of LHb and VTA projecting cortical fibers using simultaneous imaging in awake, behaving rats during the temporal discounting of reward behavioral paradigm.

Conclusions: Taken together, these multi-disciplinary results should confer a description of the function and mechanism by which cortical neurons take in environmental cues and the mechanisms by which they activate or suppress other cortical neurons to determine the activation of downstream midbrain regions, shaping behavior and decision making.

Keywords: mPFC, Ventral Tegmental Area (VTA), Lateral Habenula, Reward-based Decision-Making, in vivo Imaging

Disclosure: Nothing to Disclose.

W65. Serotonin Modulation of Impulsive Behavior

Kate Nautiyal*, Rene Hen

Columbia University, New York, New York, United States

Background: Impulsivity is a core feature of many psychiatric disorders including substance use disorder, pathological gambling, and attention deficit hyperactivity disorder. Impulsive behavior is modulated by serotonin but the circuit level mechanisms through which these effects occur are largely unknown. Our previous work shows that an absence of 5-HT1B receptors in mice results in increased impulsive action that is due to a lack of heteroreceptor (but not autoreceptor) expression during adulthood (rather than development).

Methods: We explored the mechanisms through which 5-HT1B receptor signaling influences the neural and behavioral mechanisms of impulsivity using our transgenic mouse line (floxed tetO-htr1b) which allows for inducible and tissue-specific knockout of 5-HT1B receptors. Using various Cre mouse lines, we identified the cell-type expression of 5-HT1B receptors relevant for impulsive behavior. Impulsivity was measured with operant paradigms including differential reinforcement of low-rate responding and Go/No-Go. Using additional operant paradigms, we also assessed the involvement of phenotypes related to impulsivity including motivation and habit formation. Finally, using implanted endoscopes to image calcium activity of single cells in vivo during operant behavior, we assessed the impact of the absence of 5-HT1B receptors on circuits that are known to influence impulsivity.

Results: Tissue-specific knockouts revealed that an absence of 5-HT1B receptors on GABAergic cells throughout the brain resulted in increased impulsive behavior. There was no significant effect of the absence of 5-HT1B autoreceptors (expressed on serotonin neurons) on impulsivity. This suggests that serotonin can influence impulsivity through modulation of inhibitory tone, in addition to via alterations in serotonin signaling directly (as previously proposed). Behavioral screens revealed that an absence of 5-HT1B receptor expression results in increased motivation for reward but not increased habit formation or apparent compulsive behavior. Current in vivo calcium imaging studies are aimed at investigating the mechanisms underlying 5-HT1B receptor-dependent inhibitory control focusing on the modulation of dopamine circuits known to be involved in impulsivity. Specifically, we’re identifying the effect of an absence of global 5-HT1B receptor expression on the cellular activity of dopamine cells and also dopamine-responsive cells which express 5-HT1B.

Conclusions: Overall, our results point to a role for serotonin modulation of impulsivity via influence on GABAergic signaling. Our current work is aimed at identifying the neural circuit mechanisms through which 5-HT1B receptors mediate this behavioral regulation. Furthermore, these effects are dependent on adult expression suggesting a potential avenue for pharmacological treatment for disorders with pathological impulsivity.

Keywords: Impulsivity, 5-ht1b, Calcium Imaging, Inhibitory Control

Disclosure: Nothing to Disclose.

W66. Developmental Malleability of Dopamine Projections to the Lateral Septum Impacts Adult Aggression

Darshini Mahadevia, Deepika Suri, Nao Chuma, Annette Ziolkowski, Mark Ansorge*

Columbia University, New York, New York, United States

Background: Pathological aggression has been associated with hyperactivity of the dopamine (DA) neurotransmitter system. To that end, our lab has identified an adolescent (postnatal days 32 to 41) sensitive time window during which increased DA signaling enhances adult DA function as well as aggressive behavior in mice. Here we elucidate underlying neurocircuit-based mechanism.

Methods: We used electrophysiology in acute brain slices as well as in vivo, and optogenetics as well as local pharmacology in awake behaving animals.

Results: We demonstrate that adult aggression elicited by P32-41 DAT blockade is associated with hyperactivity of ventral tegmental area (VTA) DAergic neurons, both in vivo and in acute brain slices. Furthermore, by using in vivo optogenetics, we dissected the DA-associated circuitry in adult aggression and find that ChR2-based activation of VTA, and not substantia nigra pars compacta DAergic neurons, increases aggressive behavior. In investigating VTA-target regions, we find that direct activation of DAergic terminals in the lateral septum (LS) mimics the enhanced aggression observed with VTA DAergic stimulation. Next, we investigated the necessity of the LS pathway for VTA DA induced aggression, by locally infusing D2 antagonist (Sulpiride) into the LS. We find that this infusion blocked VTA DA elicited aggression, demonstrating that VTA DA drives aggression primarily by signaling to LS. Connecting the VTA-LS pathway to the classic aggression circuitry, we find that train photo stimulation of DA terminals in the LS causes hyperpolarization of LS neurons via activation of D2 receptors, thereby inhibiting LS projections to the ventromedial hypothalamus.

Conclusions: Together, our data provide insight into a DA period that may determine the developmental trajectory of key DA aggression neurocircuitry. Our findings might ultimately aid prevention and treatment approaches for neuropsychiatric disorders characterized by unprovoked violence, an issue that carries serious consequences in society today.

Keywords: Dopamine, Brain Development, Aggression, Neurocircuits, Optogenetics

Disclosure: Nothing to Disclose.

W67. Serotonin (5-HT) 5-HT2C Receptor (5-HT2CR) in the Nucleus Accumbens Controls Motor Impulsivity and the Incentive-Salience of Cocaine Reward and Cue Reactivity

Kathryn Cunningham*, Noelle Anastasio, F. Gerard Moeller

University of Texas Medical Branch, Galveston, Texas, United States

Background: The 5-HT2R receptors (5-HT2AR, 5-HT2BR, 5-HT2CR) are colocalized to nodes of corticoaccumbens circuitry involved in impulsivity and cocaine use disorder (CUD). We have recently demonstrated that 5-HT2CR knockdown in rat medial prefrontal cortex (mPFC) evokes impulsivity and cue reactivity, an upregulation of mPFC 5-HT2AR protein, and a leftward shift in the potency of a selective 5-HT2AR antagonist to suppress motor impulsivity. These data suggest that 5-HT2CR hypofunction disrupts a local 5-HT2AR:5-HT2CR balance. The 5-HT2AR and 5-HT2CR are also colocalized to the nucleus accumbens (NAc) and the purpose of the present study was to test the hypothesis that an engineered 5-HT2CR loss in the NAc shell (NAcSh) will impact the expression of motor impulsivity and the incentive-salience of cocaine reward and cocaine cue reactivity.

Methods: Recombinant AAV vectors were constructed with a separate expression cassette for eGFP and shRNA directed at the 3’ untranslated region of the rat 5-HT2CR to decrease expression of all endogenous 5-HT2CR isoforms. The 5-HT2CR shRNA-AAV-eGFP viral vector was bilaterally infused into the NAcSh of male rats, while control animals received bilateral intra-NAcSh infusions of non-silencing control (NSC)-AAV-eGFP. Rats were trained in the 1-choice serial reaction time (1-CSRT) task then trained to self-administer a low dose of cocaine (0.25 mg/kg/inf; 180 min/day; FR1-5). Rats were subsequently challenged with three doses of cocaine (0.05, 0.125 or 0.75 mg/kg/inf) to establish the dose-effect relationship for cocaine intake. Rats were subjected to 14 days of forced abstinence, followed by a 1-hr cue reactivity test session in which lever presses on an FR1 resulted in delivery of the discrete cues previously paired with cocaine).

Results: Premature responses in the 1-CSRT task were elevated in rats with a demonstrated knockdown of 5-HT2CR in the NAcSh. The loss of 5-HT2CR in the NAcSh triggered a functional disruption of the local 5-HT2AR:5-HT2CR balance as evidenced by a leftward shift in the potency of M100907 to suppress impulsive behavior. Acquisition of cocaine self-administration in knockdown and control rats did not differ. However, loss of 5-HT2CR in the NAcSh resulted in a downward shift in the cocaine dose-response curve and higher cue reactivity relative to the NSC rats.

Conclusions: These data suggest that 5-HT2CR in the NAcSh, like the mPFC, plays a critical role in motor impulsivity and the regulation of the incentive-salience value of rewarding stimuli. However, there is a differential role for the NAc 5-HT2CR in the control of the reinforcing effects of cocaine as loss of the 5-HT2CR in the mPFC did not alter acquisition of cocaine self-administration. The altered response to the 5-HT2AR antagonist and the shift in potency of the 5-HT2AR antagonist suggests that there is a compensatory shift in the NAc mechanisms that control behavior under conditions of 5-HT2CR hypofunction.

Keywords: Serotonin 5-HT2C Receptor, Impulsivity, Cue Reactivity, Nucleus Accumbens

Disclosure: Nothing to Disclose.

W68. Association for Disrupted VGLUT3 Homeostasis Upon High Inherent Impulsivity

Noelle Anastasio*, Veronica Campbell, Brionna Davis-Reyes

University of Texas Medical Branch at Galveston, Galveston, Texas, United States

Background: Impulsivity, broadly defined as behavior without sufficient forethought, has been noted in cocaine-dependent human subjects, and contributes to relapse in cocaine use disorder. There is evidence that both serotonin (5-HT) and glutamate (Glu) neurotransmission within the raphe-corticoaccumbens circuit are critical drivers of the cognitive and/or behavioral dimensions underlying impulse-control disorders. Vesicular glutamate transporters (VGLUTs) sequester cytosolic Glu into vesicles for release from the presynaptic terminal. Of note, VGLUT3, one of the three members of this family of transporters, is localized on heterogeneous neurons that have been demonstrated to co-release Glu and 5-HT within the raphe-corticoaccumbens circuit. Further, VGLUT3 mediates 5-HTergic terminal density in the forebrain, suggesting that VGLUT3 plays a role in 5-HT signaling strength, likely due to vesicular synergy. In rodent models, VGLUT3 has been causally linked to anxiety-like behaviors and cocaine-mediated behaviors in part through strengthened postsynaptic receptor plasticity (e.g., AMPA receptor/NMDA receptor ratio). We hypothesized that an imbalance in VGLUT3, Glu-receptive AMPA receptor and 5-HT-receptive 5-HT2CR homeostasis in the NAc associates with individual differences in impulsivity.

Methods: Outbred, male Sprague Dawley rats were identified as high (HI) or low (LI) impulsive using the one-choice serial reaction time (1-CSRT) task in which nose-pokes after presentation of a visual stimulus resulted in food pellet delivery. Nose-pokes before presentation of the visual stimulus (i.e., premature responses) indexed motor impulsivity. The upper and lower quartile of animals were identified as HI or LI rats, respectively. Following phenotypic identification, NAc synaptosomal protein was extracted and VGLUT3, AMPAR, and 5-HT2CR protein levels determined via immunoblot.

Results: The HI/LI phenotype was stable in that HI rats made significantly more premature responses than LI rats across 70 days of training (p<0.001). HI rats expressed higher NAc synaptosomal VGLUT3 vs. LI rats (p<0.05); there was a positive correlation between NAc VGLUT3 expression (r=0.643, p=0.02) and impulsivity. Conversely, HI rats exhibited lower NAc synaptosomal expression of the GluA1 AMPA receptor subunit and the 5-HT2CR vs. LI rats (p<0.05). An inverse correlation with NAc GluA1 (r=-0.549, p=0.05) or 5-HT2CR (r=-0.443; p<0.05) and impulsive responding was observed.

Conclusions: These data putatively suggest in HI rats that higher VGLUT3 expression may augment Glu/5-HT tone resulting in a compensatory downregulation of NAc Glu- and 5-HT-responsive receptors, as a component of homeostatic synaptic plasticity. Thus, alterations in the morphological and/or functional properties of NAc VGLUT3 may result in shifts in 5-HT:Glu neurotransmission within the raphe-accumbens circuit that ultimately resculpts the pre- and/or postsynaptic milieu differentially in HI vs. LI rats and may contribute to the cognitive impairments seen in cocaine use disorder.

Keywords: Impulsivity, Vesicular Glutamate Transporter (VGLUT), Nucleus Accumbens

Disclosure: Nothing to Disclose.

W69. The Effects of Chronic Administration of the D2/3 Receptor Agonist Ropinirole on Gambling-Like Behaviour in Rats: Role of the Nigrostriatal Dopaminergic Pathway

Melanie Tremblay, Michael Barrus, Paul Cocker, Mason Silveira, Brittney Russell, Brett Hathaway, Wendy Adams, Sukhbir Kaur, Catharine Winstanley*

University of British Columbia, Vancouver, Canada

Background: Chronic administration of full agonists at D2/3 receptors, such as pramipexole and ropinirole, have been associated with the onset of problem gambling and impulse control disorders in a significant minority of patients with movement disorders who are treated with these compounds. A clearer understanding of the neurocognitive mechanisms through which such drugs induce these psychiatric complications could contribute to adjunctive treatments designed to prevent the development of these unwanted side-effects. Such information may also shed light on the neurobiological basis of gambling and impulsive behaviour.

Methods: In a series of studies, we assessed the effects of chronic administration of ropinirole (5 mg/kg/day for 28 days via osmotic minipump) in rodent models of different types of gambling-related cognitions. Specifically, we used the rat gambling task (rGT), an animal model of the Iowa Gambling Task used clinically, in which rats sample between options associated with differing amounts of reward and punishing time-outs in order to maximise sugar pellets earned. In a cued variant of this task (crGT), reward delivery was accompanied by salient audiovisual stimuli that scaled in complexity with the size of the win, similar to commercial gambling products. We also used the rat betting task (rBT), a paradigm specifically designed to capture the escalation of commitment bias, in which animals choose between options matched for expected utility, one of which delivers a guaranteed reward, the other double that reward or nothing with 50:50 odds. Any significant effects on decision making were replicated in a model of early stage Parkinson's disease, namely in rats with 6-hydroxydopamine lesions of the dorsolateral striatum. ex vivo, tissue samples from the ventral and dorsal striatum were subjected to Western blot analyses to assay levels of dopamine receptors and related intracellular signaling pathways. The effects of chronically up or downregulating activity in the nigrostriatal pathway on performance of the rBT was then determined using DREADDs.

Results: As expected, risky choice was higher in rats performing the cued rGT rather than the uncued task. However, this form of decision making under uncertainty was not altered by chronic ropinirole, even though premature responding (an index of motor impulsivity comparable to that obtained from the five-choice serial reaction time task) did transiently increase during drug treatment. Interestingly, this ropinirole-induced increase in impulsive responding was significantly more pronounced in rats performing the cued rGT. Chronic ropinirole also robustly increased preference for uncertain options on the rat betting task, both in intact rats and those with dorsal striatal dopamine depletion. This increase in preference for uncertainty was associated with markers of elevated GSK3beta signalling in the dorsal striatum. Furthermore, chronic up regulation of activity within the nigrostriatal dopamine pathway could partially replicate the effects of chronic ropinirole, whereas chronic down-regulation was without effect.

Conclusions: Chronic ropinirole increased motor impulsivity and preference for uncertain outcomes on the rBT, yet did not alter cost/benefit decision making in the rGT. The ability of ropinirole to modulate choice on the rBT appears to involve activation of the nigrostriatal dopamine system, suggesting an important role for D2 receptors in this region in mediating biases in decision making under uncertainty, potentially through GSK3beta.

Keywords: Dopamine (D2, D3) Receptors, Gambling, Impulsivity, DREADDs

Disclosure: Part 1: Hogan Lovells, LLP, Consultant, Part 2: Hogan Lovells, LLP, Consultant, Part 3: Hogan Lovells, LLP, Consultant.

W70. A Genetic Study of Psychosis in Huntington Disease: Evidence for the Involvement of Glutamate Signaling Pathways

Debby Tsuang*, Tiffany Greenwood, Suman Jayadev, Marie Davis, Andrew David-Shutes, Thomas Bird

VA Puget Sound HCS, GRECC (182B), Seattle, Washington, United States

Background: Psychotic symptoms of delusions and hallucinations occur in about 5% of persons with Huntington Disease (HD). The mechanisms underlying this occurrence are unknown, but given that the same symptoms also occur in schizophrenia, genetic risk factors for schizophrenia may be relevant to the development of psychosis in HD. This study is to investigate the possible role of genes associated with schizophrenia to the occurrence of psychotic symptoms in HD.

Methods: Subjects with manifesting HD were divided into those with psychotic symptoms (HP+P; 50 subjects) and those without (HP – P; 180 subjects). DNA was genotyped using the Infinium Psych Array – 24 v1.1 Bead Chip. We compared the allele frequencies of SNPs previously associated with schizophrenia and related psychiatric disorders in subjects with HD+P versus those with HD – P.

Results: Of the 22 candidate genes tested, 10 revealed an association with psychosis in HD. The majority of these genes are involved in glutamate signaling, including CTNNA2, DRD2, ERBB4, GRID2, GRIK4, GRM1, NRG1, PCNT, RELN, and SLC1A2.

Conclusions: We report the association of several previously identified candidate genes for schizophrenia with the occurrence of psychotic symptoms in HD. These data support the potential role of glutamatergic and related genes in HD psychosis.

Keywords: Huntington's Disease, Schizophrenia-Like Behavior, Schizophrenia, Synaptic Aberrations

Disclosure: Nothing to Disclose.

W71. Role of Gut-Brain Axis in Chemotherapy-Induced Behavioral Comorbidities

Savannah Bever, Jacob Allen, Jasskiran Kaur, Michael Bailey, Leah Pyter*

Ohio State University, Columbus, Ohio, United States

Background: Recent evidence indicates that the gut microbiome can influence brain and behavior, including cognition. Approximately one-third of cancer survivors report cognitive impairments after receiving chemotherapy. In addition, many cancer survivors who receive chemotherapy also experience adverse effects on their gut, including diarrhea and vomiting, along with a shift in diversity of their natural gut microbiota. However, very little is known about the potential role of the gut microbiome in the enduring and prevalent consequences of chemotherapy on the brain and behavior (e.g., cognitive impairments). The hypothesis for this study is that chemotherapy shifts diversity in the gut microbiota, leading to neuroinflammation and cognitive impairments.

Methods: To test this hypothesis, murine mammary cancer cells were injected into the fourth mammary fat pad in immunocompetent Balb/c female mice. After tumors developed, they were surgically resected. One week after tumor resection, half of the mice received 6 doses of paclitaxel chemotherapy (30 mg/kg; i.p.; every other day); the others received vehicle. After chemotherapy, cognition was tested in all mice using spontaneous alternation, contextual and cued fear conditioning, and novel object recognition tests. Circulating baseline concentrations of corticosterone and cytokines, neuroinflammation, and 16S fecal microbiotal profiles were determined.

Results: Preliminary data suggest that chemotherapy treatment impaired cued conditioning and increased Gram-negative gut bacteria which were positively related to brain inflammation.

Conclusions: The gut microbiota may exert indirect effects of chemotherapy on the brain, thereby potentiating behavioral comorbidities. By establishing the role of the gut microbiota in the behavioral consequences of cancer potential novel and non-invasive treatment targets may be determined.

Keywords: Gut Microbiota, Gut-Brain Axis, Breast Cancer, Inflammation

Disclosure: Nothing to Disclose.

W72. Estradiol Add-Back in BDNF Val66Met Mice Mimics Some Behavioral and Transcriptional Traits of Premenstrual Dysphoric Disorder

Jordan Marrocco*, Gordon Petty, Nathan Einhorn, Neelima Dubey, Jessica Hoffman, Karen Berman, David Goldman, Peter Schmidt, Bruce McEwen

Rockefeller University, New York, New York, United States

Background: Premenstrual dysphoric disorder (PMDD), a severe form of premenstrual syndrome, affects over 5% of women, with symptoms similar to generalized anxiety disorder and major depression. Clinical data indicates that women suffering from PMDD are differentially sensitive to circulating ovarian hormones compared to healthy women. Altered brain-derived neurotrophic factor (BDNF) levels in the limbic system, such as the hippocampus, are correlated with depression and anxiety in both humans and rodent models. The BDNF Val66Met single-nucleotide polymorphism is a common human variant of the gene, with Met carriers displaying altered sensitivity to stress and ovarian hormones. We hypothesized that 17β estradiol (E) add-back treatment could modulate behavior and RNA transcription in Met carriers in a way that recapitulates menstrual cycle-related behavioral sensitivity reminiscent of PMDD.

Methods: We ovariectomized a knock-in mouse heterozygous for the Met allele of the BDNF gene (BDNFMet/+) and their matched wild type (BDNF+/+). After a 10 day-recovery period, mice were treated with either estradiol (E2, 200nM E/0.1% ethanol) or vehicle (0.1% ethanol), administered via drinking water for six weeks. A battery of behavioral tests was performed during treatment. RNA- sequencing (Illumina NextSeq 500 Sequencing) was performed on the ventral hippocampus dissected from ovariectomized BDNFMet/+ and BDNF+/+ mice treated or not with E2.

Results: In the open field test, E treatment induced an anxiogenic effect on BDNFMet/+ mice, and in the splash test E treatment induced depressive-like behavior in BDNFMet/+ mice, but not in BDNF+/+. This indicates that Met carriers display a heightened emotional sensitivity to E treatment. Thus, the behavioral phenotype of Met carriers resembled that of women suffering from PMDD, who experience increased depressive symptoms after E add-back treatment. The behavioral similarities between BDNFMet/+ mice and women with PMDD prompted us to perform a comparative whole-genome RNA-sequencing analysis between mouse ventral hippocampus and lymphoblastoid cell cultures from healthy women and women suffering from PMDD. RNA sequencing revealed a greater overlap in E-induced gene expression between BDNFMet/+ mice and women with PMDD than between BDNF+/+ mice and healthy women. Several genes belonging to the ESC/E(Z) complex, such as SIRT1,EZH1, and MTF2 were differentially altered by E treatment in wild-type and BDNFMet/+ mice. Interestingly, these same genes have been shown to also be differentially expressed in response to E treatment in healthy women and women with PMDD.

Conclusions: Our sequencing data supports the understanding that the ESC/E(Z) complex plays a critical role in modulating estradiol sensitivity. In combination with our behavioral findings, our data indicate that BDNFMet/+ mice recapitulate PMDD both behaviorally and transcriptionally, suggesting the Met allele as a risk factor for PMDD.

Keywords: BDNF Val66Met, Estradiol, Premenstrual Dysphoric Disorder, RNA Sequencing

Disclosure: Nothing to Disclose.

W73. Alterations in the Mitochondrial Biogenesis Pathway in Postmortem Brains From Patients With HIV-Associated Neurocognitive Disorders

Brandi Quintanilla*, Emma Martine, Sarah Gough, Cristian Achim, Adam Fields

University of California, San Diego, San Diego, California

Background: Over 1.3 million people in the USA are infected with Human Immunodeficiency Virus (HIV), and about half of these individuals will develop HIV associated neurocognitive disorders (HANDs). HAND remains prevalent, despite reduced or non-detectable viral load achieved with combined anti-retroviral therapy (cART) regimens. HAND patients experience a diminished quality of life because of neurocognitive deficits, including impaired memory, executive function, language, and attention. While a multitude of pathogenic mechanisms likely contribute to neurodegeneration during HAND, mitochondrial dysfunction has long been implicated in this process. Healthy mitochondria are crucial to overall brain function, and are especially important for the integrity of neurons, which are highly energy dependent. The maintenance of a healthy pool of mitochondria depends upon functional mitochondrial biogenesis and mitochondrial dynamics (fission/fusion and transport of mitochondria). Our group has recently reported that alterations in mitochondrial fission/fusion may contribute to neurodegeneration during HAND; however, the role of mitochondrial biogenesis in HAND has yet to be explored. Here, we hypothesized that altered mitochondrial biogenesis may contribute to neurodegeneration associated with HAND.

Methods: Seventy brains from carefully characterized HIV+ decedents that were all on cART regimens were acquired from the National NeuroAIDS Tissue Consortium (NNTC) and analyzed for expression levels and patterns of mitochondrial biogenesis-related genes. These patients had extensive neuromedical and neurocognitive analyses performed within one-year of death. Frontal cortices were processed for immunoblot and immunohistochemical analyses for expression of transcription factors which regulate mitochondrial biogenesis: peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1-α) and mitochondrial transcription factor A (TFAM). PGC1-α and TFAM protein expression levels were determined by densitometry analyses of immunoblot band intensity. Cellular localization and overall expression of PGC1-α and TFAM were determined using immunohistochemical analyses of brain sections. RNA and DNA were collected from all brains to determine PGC1-α and TFAM mRNA levels and mitochondrial (mt) DNA quantity per cell, respectively.

Results: Total protein levels of PGC1-α and TFAM were significantly increased in brains from HAND patients with more severe neurocognitive impairment, compared to brains from HIV+ patients with no impairment (p<0.05, respectively). However, neuropathological analyses revealed that TFAM was decreased in cortical gray matter of HAND decedents, and increased in subcortical white matter (p<0.05). Interestingly, immunohistochemical analyses of PGC1-α revealed an opposite pattern compared to TFAM, with PGC1-α increased in gray matter and decreased in white matter, however, these differences did not reach statistical significance. PGC1-α and TFAM mRNA levels trended toward an increase in HAND verses control. In these same tissues, mtDNA quantity was lowest in brains from patients with the most severe forms of HAND, compared to patients with mild HANDS, or those lacking a HAND diagnosis.

Conclusions: These findings suggest that alterations in mitochondrial biogenesis in HIV+ patients with HAND persist in the cART era. Alterations found in the mitochondrial biogenesis pathway may be the result of a compensatory response to mitochondrial damage, decreased mtDNA levels, or HIV proteins that directly alter PGC1-α activity. Targeting altered mitochondrial biogenesis-related gene expression in HAND patients may provide a novel therapeutic solution in halting the progression of HAND and other neurodegenerative disorders.

Keywords: HIV Associated Neurocognitive Disorder, Mitochondrial Dysfunction, Transcription Factors

Disclosure: Nothing to Disclose.

W74. Amygdala Habituation: A Key Biomarker for Post-Traumatic Stress Disorder

Jennifer Stevens*, Ye Ji Kim, Timothy Ely, Kerry Ressler, Tanja Jovanovic

Emory University School of Medicine, Atlanta, Georgia, United States

Background: Amygdala function is central to neurobiological models of post-traumatic stress disorder (PTSD). Amygdala hyper-reactivity has been shown to be a pre-disposing risk factor for PTSD. However, little research has addressed dynamic amygdala responses to repeated threat stimuli (habituation/sensitization) in PTSD. Interestingly, amygdala habituation has recently been shown to be a more stable biomarker of amygdala function than reactivity to emotional stimuli, showing better test-retest reliability. Although many cortical brain areas habituate to repeated presentations of similar stimuli, the amygdala has been shown to sensitize in response to repeated threat stimuli in healthy samples. Here we investigated potential alterations in amygdala habituation / sensitization to threat in trauma-exposed participants.

Methods: In Study 1 (Chronic sample), N=100 trauma-exposed African American women were recruited from the Grady Trauma Project, an epidemiological sample of participants recruited through the primary care waiting rooms of a large publicly funded urban hospital. Women participated in MRI and reported lifetime trauma histories (CTQ, TEI) and current PTSD symptoms (PSS). In the fMRI task, women passively viewed fearful and neutral face stimuli, and amygdala habituation was quantified by comparing responses to the first 5 blocks>the last 5 blocks for each emotion type. In Study 2 (Prospective sample), N=80 participants were recruited in the Emergency Department within 24 hours of trauma and participated in the same fMRI task 6 weeks later. PTSD symptom trajectories were measured for the following 12 months.

Results: In the Chronic sample, PTSD symptom severity was associated with increased amygdala habituation (reduced sensitization) to fearful face stimuli (R2ΔPSS=0.13, β=0.38, p<.001), with the strongest association for numbing symptoms (r=.47, pFWE=.008). This association held after controlling for childhood and adult trauma load (β=0.30, p=.004), and depression symptoms (β=0.26, p=0.002). In the Prospective sample, we investigated whether this was a predisposing risk factor for PTSD, by testing whether amygdala habituation in the peri-traumatic period predicted heightened symptom levels over the following year. Preliminary analyses showed no association between amygdala habituation and later PTSD symptoms (p>.05).

Conclusions: Findings in the chronic PTSD sample in Study 1 were consistent with two previous small studies (N=11, Protopopescu et al., 2005; N=19, van den Bulk, 2016), showing that PTSD was associated with greater habituation to fearful faces. We additionally showed that this was not related to trauma load or depression symptoms, and may be related to emotional disengagement as the strongest correlation was for numbing symptoms. Initial analyses in the prospective sample in Study 2 suggested that amygdala habituation is not a pre-disposing risk factor for PTSD, but may instead index chronic PTSD symptoms. The findings support the utility of a relatively novel biomarker for chronic PTSD, and are in concordance with psychophysiological findings showing increased sympathetic autonomic responses to threat in single-trauma PTSD, but not in a group with PTSD after multiple lifetime traumas. Amygdala habituation may be an interesting new target for interventions designed to enhance cognitive behavioral therapy strategies such as prolonged exposure (PE), in which habituation plays a central role across repeated imaginal exposures to the index trauma.

Keywords: Post Traumatic Stress Disorder, Functional MRI (fMRI), Cognitive Phenotypes, Habituation, Amygdala

Disclosure: Nothing to Disclose.

W75. Unique Biphasic NMDA Receptor Modulation by Rapastinel: The Role of Positive Modulation

John Donello*, Yong-Xin Li, Roger Kroes, Joseph Moskal, Pradeep Banerjee

Allergan, Inc., Irvine, California, United States

Background: Rapastinel (GLYX-13, AGN-241659) is an N-methyl-D-aspartate receptor (NMDAR) modulator currently in late-stage development as a treatment for major depressive disorder. Previous studies have shown that rapastinel modulates NMDA-induced intracellular calcium responses in primary rat cortical neurons in a concentration-dependent manner, and this effect of rapastinel is independent of the glycine co-agonist site. The voltage-dependent activity of NMDARs is jointly determined by the presence of endogenous ligands for the receptor (i.e., glutamate and the ligand for the co-agonist site [glycine/D-serine]), and it is unknown if rapastinel's effect on NMDAR activity is influenced by NMDAR ligand concentrations. We, therefore, evaluated the effects of rapastinel on calcium responses in the presence of varying concentrations of an agonist (NMDA) and a co-agonist (D-serine) in rat primary cortical neurons, and confirmed rapastinel's effect on NMDAR function using whole-cell patch-clamp recordings in cultured rat cortical neurons.

Methods: Rat brain cortical neurons (embryonic day 18 and 19) were grown on poly-D-lysine–coated glass coverslips. For calcium imaging, neurons were loaded with the calcium indicator Fluo-4 AM at 14-28 days post-seeding and mounted on the stage of an Olympus BX61WI upright microscope equipped with a confocal laser scanning system. Varying concentrations of D-serine and NMDA were applied to neurons in the presence and absence of rapastinel, and responses were measured. All test compounds were dissolved in magnesium-free extracellular medium (containing TTX and NBQX) and applied to cells using a customized, rapid drug application system. Change in fluorescence was normalized to the maximum response, which was defined as response to 10 μM NMDA+3 μM D-serine. For current measurement, whole-cell patch-clamp recordings were conducted on neurons 3 weeks post-seeding. The amplitude of sustained currents was measured as the mean current between 12 and 14 seconds after agonist application over the 2-second interval.

Results: In the presence of 10 μM NMDA and varying D-serine concentrations (30-1000 nM), rapastinel (100 nM) moderately enhanced the NMDA-induced calcium response. In the presence of a maximal concentration of NMDA and D-serine (3000 nM), rapastinel no longer potentiated the response. Similarly, in the presence of a maximal D-serine concentration (3000 nM) and varying concentrations of NMDA (3-5 μM), 100 nM rapastinel enhanced the NMDA-induced calcium response. Higher concentrations of rapastinel (≥1 μM) partially inhibited the response at all D-serine and NMDA concentrations. In cultured rat cortical neurons, application of 10 μM NMDA, in the absence of magnesium, induced a small inward current when the cell membrane was voltage-clamped at -60 mV and a small outward current at +40 mV. At both membrane potentials, 100 nM rapastinel significantly increased NMDA current amplitudes. However, in the presence of magnesium, 10 μM NMDA activated an outward current at +40 mV, but did not induce significant current at -60 mV due to the magnesium block. Application of 100 nM rapastinel significantly increased outward NMDA current amplitude at +40 mV.

Conclusions: These results demonstrate that rapastinel modulates NMDAR activity in primary rat cortical neurons, and that it acts independently of the glycine co-agonist site and co-agonist concentrations. Low concentrations of rapastinel moderately potentiate NMDAR-mediated ionotropic activity, whereas high concentrations partially inhibit this activity. These effects are non-competitive with endogenous NMDAR ligands and persist under physiological conditions in the presence of magnesium. Moderate NMDAR modulation is a unique feature of this compound and may explain its favorable tolerability observed in nonclinical and completed clinical studies compared to NMDAR antagonists like ketamine. In summary, rapastinel exhibits a unique, concentration-specific biphasic modulation of NMDAR function and is the first molecule that can moderately enhance and inhibit NMDAR activity independent of the co-agonist concentration. Antidepressant-like behavior is associated with brain concentrations of rapastinel that positively modulate NMDARs, demonstrating a new pharmacological mechanism to rapidly relieve depression.

Keywords: GLYX-13, Depression, Calcium Imaging, Neuronal Culture

Disclosure: Part 5: Allergan, Employee.

W76. Improvement in Disease Severity in Patients With Treatment Resistant Depression Following Treatment With Intranasal Esketamine

Abigail Nash*, May Shawi, Jaskaran B. Singh, Ella Daly, Kimberly Cooper, Pilar Lim, Rosanne Lane, Jagadish Gogate, Allitia DiBernardo, David Hough, Larry Alphs

Janssen Global Services, Titusville, New Jersey, United States

Background: Recognizing the importance not only of the clinician's opinion but also of the patient's experience and perspective, Sequenced Treatment Alternatives to Relieve Depression (STAR*D) utilized both clinician-reported and patient-reported outcomes in a large-scale multi-step study on antidepressant effectiveness in real-world settings. Both approaches indicate that <17% of Major Depressive Disorder (MDD) patients respond to novel oral treatments after two prior antidepressant failures. To address this low response rate and continue to investigate the use of patient-rated outcomes in clinical trials, an antidepressant with a new mechanism of action is being investigated for efficacy and safety utilizing both clinician-rated and patient-reported scales.

Methods: This is a post-hoc analysis of a Janssen R&D Phase 2a clinical trial (ESKETINTRD2003). Subjects aged 20-64 with MDD without psychotic features (DSM IV) and a history of inadequate response to ≥2 antidepressants were randomized [3:1:1:1] to 1 week of twice-weekly treatment with intranasal placebo (n=33), esketamine 28 mg (n=11), 56 mg (n=11), or 84 mg (n=12). Participants taking oral antidepressants at study entry continued treatment during the study. Changes in depression severity were measured using the Clinical Global Impression Severity (CGI-S) and the Patient Global Impression Severity (PGI-S) scales.

Results: At all esketamine doses (28 mg, 56 mg, 84 mg), subjects reported a one-point mean change in PGI-S from baseline to week one compared to no change on placebo (p-values 0.005, 0.001, 0.032 respectively). Similarly, mean CGI-S scores improved for subjects receiving esketamine at all doses (p-values 0.028, 0.004, 0.049 respectively) compared to no change in placebo subjects. These data are consistent with previously reported data based on the Montgomery Åsberg Depression Rating Scale (MADRS) and support positive correlation between patient-reported and clinician-reported outcomes.

Conclusions: Initial results from this Phase 2a study suggest clinically relevant improvement in depression symptoms in as early as one week when treated with twice-weekly intranasal esketamine as reported by both clinicians and patients. This work will help guide future investigations of esketamine in larger populations to provide better therapeutic options for treatment resistant MDD patients.

Keywords: Antidepressant Trials, Esketamine, Treatment Resistant Depression, Patient Reported Outcomes, Clinician Reported Outcomes

Disclosure: Part 1: Janssen Global Services, Employee, Part 2: Janssen Global Services, Employee, Part 3: Janssen Global Services, Employee, Part 4: Janssen Global Affairs, Employee, Part 5: Janssen Global Affairs, Employee.

W77. Poster Withdrawn W78. The Hyperactive (HYPER) Rat: An Animal Model for Bipolar Disorder

Jay Weiss*, Katherine Williams

Emory University, Atlanta, Georgia, United States

Background: Development of new treatments for bipolar disorder (BD) has been hindered by the unavailability of animal models for their discovery and testing. We describe here a rat line that shows a variety of behavioral, physiological, genetic, and pharmacological attributes that are similar to what humans with BD show. This is the first animal that shows such an extensive list of similarities to human BD, and these data suggest that this animal, which is a natural and spontaneously-occurring genetic mutation, is an animal analogue of human BD.

Methods: The rat line described here was generated from a single litter of rats all of which were discovered to be hyperactive in the home cage (i.e., showed higher-than-normal nocturnal ambulatory activity). After brother-sister breeding of these initial progenitor rats, their offspring, now for more than 50 generations, have remained distinctly hyperactive. This rat line was named the Hyperactive (HYPER) rat. These rats, when exposed to a stressor, show a hyperactive outburst lasting several days, thereby mimicking a manic–like episode. Much of the basic data derives from recording of ambulatory activity in the home cage around the clock (i.e., 24-hour recording). In such experiments, animals were singly housed in standard colony cages directly on bedding with access to ad libitum to food and water, and activity was recorded by interruption of photocell beams mounted externally along the long axis of the colony cage. Activity was typically recorded for 10-14 days, but in some experiments for up to 3.5 months. To induce stress, early experiments used a session of electric shock delivered via electrodes affixed to the rat's tail. In later experiments, a briefer session of less intense grid shock was used, and eventually no shock was needed as it was sufficient simply to put animals into individual housing (social isolation) to elicit a hyperactive outburst. Increased sensitivity of the rat line developed because selective breeding was done across generations – that is, ambulatory activity was measured in each generation and the males and females that showed the most pronounced post-stress hyperactive outbursts were bred to produce the next generation. Selective breeding increased sensitivity of HYPER rats to show a hyperactive outburst and also the percentage of HYPER rats in each generation that showed this outburst. As studies continued, other behavioral and physiological measurements relevant to BD were made. Anxiety in HYPER rats was assessed using an elevated plus maze and/or open field. Propensity to consume alcohol was examined by presenting HYPER rats with various concentrations of ETOH for voluntary consumption. Anhedonia in HYPER rats was assessed by measuring sucrose consumption in both stress and non-stress conditions. Cognitive processes were studied in the Morris Water Maze. Activity of neurons in many brain regions, including forebrain dopaminergic regions, was assessed by measuring c-fos expression, as well as by measuring concentration of monoamines and their metabolites post mortem in numerous brain regions, including dopaminergic cell-body regions and dopamine-rich forebrain regions.

Results: Salient observations are as follows: (1) as stated above HYPER rats respond to stress with an outburst of extreme hyperactivity mimicking a manic-like response that begins 2-3 days after the stressful event and lasts 4-7 days. This is not seen in normal, randomly-bred rats. Crossbreeding studies have now demonstrated that susceptibility of HYPER rats to showing this hyperactive outburst is a genetically-based trait. (2) One-year old male HYPER rats also show, following tail shock and subsequent recording for 2-3 months, long-lasting depression of motor activity (i.e., depressed activity lasting 21-30 days). (3) Following tail shock and subsequent recording for 2-3 months, one-year old HYPER female rats and some older male HYPER rats cycle between episodes, often lasting multiple days, of extreme hyperactivity and depressed activity; this is the first animal model of BD that shows “cycling” between hyperactivity and hypoactivity as in human BD. (4) Lithium and valproate treatment of HYPER rats blocks manic-like hyperactive outbursts [described above under (1)] and reduces cycling [described above under (3)]. Other characteristics of HYPER rats found in human BD are (5) elevated anxiety, (6) increased alcohol consumption, (7) anhedonia (reduced preference for sucrose), (8) cognitive abnormalities, and (9) reduced dopaminergic function in forebrain regions. Also, an additional observation is potentially of considerable significance – beginning in Generation 28 of the HYPER line, a small number of HYPER rats in subsequent generations show (10) keratotic skin lesions that are similar to the dermatological disorder Darier's disease in humans which has been found to co-segregate with human BD, pointing to similar genetic abnormalities being present in the HYPER rat and in humans with BD.

Conclusions: The HYPER rat shows similarity to human BD with respect to a large number of behavioral symptoms as well as underlying pathophysiology and genetic determinants present in humans with BD. BD-like behavior of HYPER rats is also ameliorated by treatment with lithium or valproate. The HYPER rat shows the most similarities to human BD of any potential animal model of BD reported to date. Study of this unique natural, genetically-determined animal model may help development of new treatments for BD.

Keywords: Bipolar Disorder, Mania, Depression, Animal Model

Disclosure: Nothing to Disclose.

W79. OnabotulinumtoxinA for the Treatment of Major Depressive Disorder in Adult Females: A Phase 2, Randomized, Double-Blind, Placebo-Controlled Trial

Suresh Durgam*, Mitchell Brin, Arlene Lum, Lynn James, Jeen Liu, Michael Thase, Armin Szegedi

Allergan, Jersey City, New Jersey, United States

Background: Prior published single-center clinical studies suggest that a single therapeutic intervention with onabotulinumtoxinA (onabotA; BOTOX), injected into facial muscles, may represent a novel, safe, and well-tolerated therapeutic option that could provide symptomatic relief for several weeks in patients with major depressive disorder (MDD). The working hypothesis is that patients with depression may display relative overactivity of the corrugator muscles and that onabotA could reduce depressive symptoms by altering dynamic facial expressions and proprioceptive sensory input. This study was designed to evaluate the safety and efficacy of onabotA vs placebo as treatment for MDD in adult females (NCT02116361).

Methods: This was a large, phase 2, multicenter, randomized, double-blind, placebo-controlled, parallel-group, 24-week study. Female patients (18-65 years) had moderate to severe MDD (DSM-IV-TR criteria), current depressive episode ≥4 weeks, Hamilton Rating Scale for Depression 17-item Version (HAM-D17) score ≥18 at baseline, and Clinical Global Impression-Severity (CGI-S) score ≥4. Patients were randomized 1:1:2 to onabotA 30 U, onabotA 50 U, or placebo. Two treatment regimens were evaluated: (1) 6 intramuscular (IM) injections of onabotA 30 U (n=65) vs matching placebo (n=58) and (2) 8 total injections (6 IM and 2 subcutaneous) of onabotA 50 U (n=65) vs matching placebo (n=67). All injections were administered into the glabellar region during a single treatment session on Day 1. The primary endpoint was change from baseline to Week 6 in the Montgomery-Åsberg Depression Rating Scale (MADRS) total score, and was analyzed using a mixed-model repeated-measures (MMRM) analysis with missing scores imputed using a modified last observation carried forward (mLOCF) method. The model used within each dose cohort included treatment, visit, treatment by visit interaction and pooled investigator center as factors, and baseline clinic MADRS total score, duration of illness, and number of previous depression episodes as covariates. Secondary efficacy measures included change in CGI-S score and HAM-D17 total score and were analyzed using observed data with no imputation of missing values. Safety was assessed through AEs, physical examination, clinical laboratory tests, vital sign measurements, and Columbia-Suicide Severity Rating Scale (C-SSRS).

Results: A total of 258 patients were randomized; 255 received treatment. Premature discontinuations before Week-6 and Week-12 visits were similar between groups (Weeks 6/12 [%]: placebo=10.2/19.5, onabotA=7.7/16.2). The onabotA 30 U group demonstrated numerically greater change from baseline to Week 6 in MADRS total score compared with placebo (least squares [LS] mean change: placebo=–7.9, 30 U=–11.6). Although differences from placebo did not reach statistical significance (P=.053), the magnitude of treatment difference (LS mean difference=–3.7) was consistent with effect sizes generally observed with other antidepressants. Evaluation of MADRS results by study visit showed consistent numerical improvement compared with placebo in the 30 U group from Week 3 to Week 15, with LS mean differences between groups ranging from -3.3 to -5.1 at all visits. Results were similar for secondary efficacy measures, with consistently numerically greater decreases on the CGI-S (-0.5) and HAM-D17 (≥-1.6) observed in the onabotA 30 U group compared with placebo at every study visit through Week 24 (except Week 18 in HAM-D17); results would have been considered statistically significant (P<.05) had the primary endpoint been met. The onabotA 50 U group did not show numerical advantage over placebo on primary or secondary efficacy parameters. Approximately 52% (132/255) of patients experienced ≥1 treatment-emergent AE (TEAE; onabotA, 54.6%; placebo, 48.8%). Most TEAEs were local to the site of injection and mild or moderate in intensity. The only TEAE that occurred in ≥10% of patients in any treatment group was headache, which occurred with similar frequency between combined groups. Discontinuation due to AEs occurred in 4 onabotA- and 1 placebo-treated patient; none were considered treatment related. Eyelid ptosis, a local AE known to be associated with onabotA glabellar injections, occurred in 4.6% of onabotA- and 0 placebo-treated patients. Serious TEAEs occurred in 6 placebo- and 3 onabotA-treated patients, including 1 fatal outcome in the onabotA 30 U group following elective surgery 108 days after treatment; none were considered treatment related. No treatment-related changes were observed in clinical laboratory values, vital signs, physical examination results, or C-SSRS results.

Conclusions: In female patients with MDD, onabotA treatment was well tolerated at single doses of 30 U and 50 U. The safety and tolerability profile was consistent with events in previous studies across all doses. While statistical significance was not achieved in the primary efficacy endpoint, the overall efficacy results suggest benefit for onabotA 30 U in the treatment of MDD in adult female patients. Overall, onabotA 30 U, administered in a standardized injection paradigm, may represent a novel treatment of depressive symptoms with a potentially superior safety and tolerability profile compared with approved antidepressant drugs currently used for treatment of patients with MDD.

Keywords: Botox, Major Depressive Disorder (MDD), Antidepressant

Disclosure: Part 3: Allergan, Stock / Equity, Part 5: Allergan, Employee.

W80. ALKS 5461 (a Buprenorphine-Samidorphan Combination) for Adjunctive Treatment of Major Depressive Disorder: Results From Analyses of US Patients From Three Independent Placebo-Controlled Trials

Maurizio Fava*, Michael Thase, Madhukar Trivedi, Sanjeev Pathak, Arielle Stanford, Asli Memisoglu, William Martin, Amy Claxton, Elliot Ehrich

Massachusetts General Hospital, Boston, Massachusetts, United States

Background: Major depressive disorder (MDD) is a complex syndrome including emotional symptoms, social deficits and impaired ability to cope with stress (Han and Nestler, 2017; Jones et al, 2015; Scheurich et al, 2008). A substantial proportion of patients suffers from inadequate response to available antidepressants and continues to experience clinically significant symptoms (Rush et al 2009). All approved antidepressants, including atypical antipsychotics, work through modulation of monoamine (serotonin, norepinephrine, and dopamine) pharmacology. New medicines with distinct mechanisms of action that target brain systems with known relevance to depression, such as the opioid pathway, are needed to improve clinical outcomes (Hsu et al, 2015; Kennedy et al, 2006). However, the development of novel opioid-based therapies has been challenging due to the risk of abuse and dependence.

We studied ALKS 5461, an opioid modulator, in the adjunctive treatment of MDD. It is a fixed dose combination of buprenorphine (BUP), a μ-opioid receptor partial-agonist and κ-opioid receptor antagonist, and samidorphan (SAM), a potent μ-opioid receptor antagonist. The purpose of SAM in this combination is to counteract BUP's μ-opioid receptor agonist activity associated with its subjective and rewarding effects while preserving its potential antidepressant properties. Four multicenter, placebo-controlled, double-blind studies of adjunctive ALKS 5461 have been conducted. Two studies met their pre-specified efficacy endpoints, namely ALK5461-202 (NCT01500200) and FORWARD-5 (NCT02218008). One study, FORWARD-4, (NCT02158533) narrowly missed the pre-specified endpoint, but demonstrated efficacy for ALKS 5461 at multiple time points and thus provides support for efficacy. One study, FORWARD-3 (NCT02158546), did not produce an efficacy signal in the presence of higher than expected placebo response.

The objective of this post-hoc analysis is to evaluate consistency of efficacy of the ALKS 5461 2/2 mg dose (the intended therapeutic dose) among similar populations, namely US region, utilizing similar endpoints across the three studies which demonstrated efficacy and had similar study designs.

Methods: ALK5461-202, a US-only study, and the US population from FORWARD-4 and FORWARD-5 were analyzed. All 3 studies evaluated 2 doses of ALKS 5461 and utilized the Sequential Parallel Comparison Design (SPCD, Fava et al. 2015). ALKS 5461 2/2 mg was common to all 3 studies and its efficacy was evaluated for this report. Efficacy evaluations included MADRS-6AVG (change from baseline to average of Week 3 through end of treatment), MADRS-10AVG (change from baseline to average of Week 3 through end of treatment), and MADRS-10 at the end of treatment (MADRS-10EOT). Treatment-emergent adverse events (TEAEs) were monitored throughout the study. Furthermore, Columbia-Suicide Severity Rating Scale (C-SSRS) and Clinical Opiate Withdrawal Scale (COWS) data were collected to monitor subjects for emergence of suicidal ideation during treatment and opioid withdrawal following discontinuation of study drug.

Results: Demographics and baseline characteristics were generally similar across studies. While ALK5461-202 was conducted only in the US (N=142), FORWARD-4 (N=385) and FORWARD-5 (N=407) included a US patient population comprising n=323 (84.1%) and n=327 (80.5%), respectively. Statistically significant improvement with ALKS 5461 versus placebo was observed for all endpoints for each study: in Study ALK5461-202, FORWARD-4, and FORWARD-5, Least Square Mean Differences in MADRS-6AVG score (SE) versus placebo were -2.3 ([1.10]; P =0.035), -2.2 ([0.82]; P =0.007), and -1.9 ([0.69]; P =0.006) respectively. Likewise, statistically significant reductions in MADRS-10AVG score were observed: -3.6 ([1.52]; P=0.019), -2.7 ([1.19]; P=0.025), and -2.7 ([0.95]; P=0.005) versus placebo. In addition, there were statistically significant reductions in MADRS-10EOT: -5.2 ([1.66]; P=0.002), -3.5 ([1.40]; P=0.012), and -2.4 ([1.07]; P=0.028), versus placebo.

The most common TEAEs in the entire population of each study were gastrointestinal (most frequently nausea and/or constipation), and they occurred around initiation of treatment. There was no pattern of TEAEs indicative of abuse potential or opioid dependence and withdrawal. Furthermore, there was no evidence of withdrawal as assessed by COWS and no pattern of treatment emergent suicidal ideation or behavior by C-SSRS or TEAE.

Conclusions: Adjunctive ALKS 5461 2/2 mg demonstrated statistically significant, clinically meaningful and consistent efficacy versus placebo across three independent SPCD studies in US patients. Overall, ALKS 5461 was generally well tolerated with no evidence of abuse liability in these studies.

Keywords: Major Depressive Disorder, Treatment Resistance, Opioid Receptors

Disclosure: Part 1: Abbott Laboratories; Acadia Pharmaceuticals; Alkermes, Inc.; American Cyanamid; Aspect Medical Systems; AstraZeneca; Avanir Pharmaceuticals; AXSOME Therapeutics; BioResearch; BrainCells, Inc.; Bristol-Myers Squibb; CeNeRx BioPharma; Cephalon; Cerecor; Clintara, LLC; Covance; Covidien; Eli Lilly and Company; EnVivo Pharmaceuticals, Inc.; Euthymics Bioscience, Inc.; Forest Pharmaceuticals, Inc.; FORUM Pharmaceuticals; Ganeden Biotech, Inc.; GlaxoSmithKline; Harvard Clinical Research Institute; Hoffman-LaRoche; Icon Clinical Research; i3 Innovus/Ingenix; Janssen R&D, LLC; Jed Foundation; Johnson & Johnson Pharmaceutical Research & Development; Lichtwer Pharma GmbH; Lorex Pharmaceuticals; Lundbeck, Inc.; Marinus Pharmaceuticals; MedAvante; Methylation Sciences, Inc; National Alliance for Research on Schizophrenia & Depression (NARSAD); National Center for Complementary and Alternative Medicine (NCCAM);National Coordinating Center for Integrated Medicine (NiiCM); National Institute of Drug Abuse (NIDA); National Institute of Mental Health (NIMH);, Grant, Self, Neuralstem, Inc.; NeuroRx; Novartis AG; Organon Pharmaceuticals; PamLab, LLC.; Pfizer, Inc.; Pharmacia-Upjohn; Pharmaceutical Research Associates., Inc.; Pharmavite® LLC; PharmoRx Therapeutics; Photothera; Reckitt Benckiser; Roche Pharmaceuticals; RCT Logic, LLC (formerly Clinical Trials Solutions, LLC); Sanofi-Aventis US, LLC; Shire; Solvay Pharmaceuticals, Inc.; Stanley Medical Research Institute (SMRI); Synthelabo; Taisho Pharmaceuticals; Takeda Pharmaceuticals; Tal Medical; VistaGen); Wyeth-Ayerst Laboratories, Grant, Part 2: Belvoir Media Group, Honoraria, Part 3: Abbott Laboratories; Acadia; Affectis Pharmaceuticals AG; Alkermes, Inc.; Amarin Pharma Inc.; Aspect Medical Systems; AstraZeneca; Auspex Pharmaceuticals; Avanir Pharmaceuticals; AXSOME Therapeutics; Bayer AG; Best Practice Project Management, Inc.; Biogen; BioMarin Pharmaceuticals, Inc.; Biovail Corporation; BrainCells, Inc; Bristol-Myers Squibb; CeNeRx BioPharma; Cephalon, Inc.; Cerecor; CNS Response, Inc.; Compellis Pharmaceuticals; Cypress Pharmaceutical, Inc.; DiagnoSearch Life Sciences (P), Ltd.; Dinippon Sumitomo Pharma Co., Inc.; Dov Pharmaceuticals, Inc.; Edgemont Pharmaceuticals, Inc.; Eisai, Inc.; Eli Lilly and Company; EnVivo Pharmaceuticals, Inc.; ePharmaSolutions; EPIX Pharmaceuticals, Inc.; Euthymics Bioscience, Inc.; Fabre-Kramer Pharmaceuticals, Inc.; Forest Pharmaceuticals, Inc.; Forum Pharmaceuticals; GenOmind, LLC; GlaxoSmithKline; Grunenthal GmbH; Indivior; i3 Innovus/Ingenis; Intracellular; Janssen Pharmaceutical; Jazz Pharmaceuticals, Inc.; Johnson & Johnson Pharmaceutical Research & Development, LLC; Knoll Pharmaceuticals Corp.; Labopharm, Inc.; Lorex Pharmaceuticals; Lundbeck, Inc.; Marinus Pharmaceuticals; MedAvante, Inc.; Merck & Co., Inc.; MSI Methylation Sciences, Inc.; Naurex, Inc.; Nestle Health Sciences; Neuralstem, Inc.; Neuronetics, Inc.; NextWave Pharmaceuticals; Novartis AG;Nutrition 21; Orexigen Therapeutics, Inc.; Organon Pharmaceuticals; Osmotica; Otsuka Pharmaceuticals; Pamlab, LLC.; Pfizer, Inc.; PharmaStar; Pharmavite®, LLC.; PharmoRx Therapeutics; Precision Human Biolaboratory; Prexa Pharmaceuticals, Inc.; PPD; Purdue Pharma;, Consultant, Self, Puretech Ventures; PsychoGenics; Psylin Neurosciences, Inc.; RCT Logic, LLC (formerly Clinical Trials Solutions, LLC); Rexahn Pharmaceuticals, Inc.; Ridge Diagnostics, Inc.; Roche; Sanofi-Aventis US, LLC.; Sepracor Inc.; Servier Laboratories; Schering-Plough Corporation; Shenox Pharmaceuticals; Solvay Pharmaceuticals, Inc.; Somaxon Pharmaceuticals, Inc.; Somerset Pharmaceuticals, Inc.; Sunovion Pharmaceuticals; Supernus Pharmaceuticals, Inc.; Synthelabo; Taisho Pharmaceuticals; Takeda Pharmaceutical Company Limited; Tal Medical, Inc.; Tetragenex; Teva Pharmaceuticals; TransForm Pharmaceuticals, Inc.; Transcept Pharmaceuticals, Inc.; Vanda Pharmaceuticals, Inc.; VistaGen, Consultant, Part 4: Patents for Sequential Parallel Comparison Design (SPCD), licensed by MGH to Pharmaceutical Product Development, LLC (PPD) (US_7840419, US_7647235, US_7983936, US_8145504, US_8145505); and patent application for a combination of Ketamine plus Scopolamine in Major Depressive Disorder (MDD), licensed by MGH to Biohaven. Patents for pharmacogenomics of Depression Treatment with Folate (US_9546401, US_9540691)., Patent, Copyright for the MGH Cognitive & Physical Functioning Questionnaire (CPFQ), Sexual Functioning Inventory (SFI), Antidepressant Treatment Response Questionnaire (ATRQ), Discontinuation-Emergent Signs & Symptoms (DESS), Symptoms of Depression Questionnaire (SDQ), and SAFER; Lippincott, Williams & Wilkins; Wolkers Kluwer; World Scientific Publishing Co. Pte.Ltd., Royalties.

W81. Efficacy of Cariprazine in Patients With Bipolar Depression and Mixed Features: Post Hoc Analysis of a Randomized, Double-Blind, Placebo-Controlled Phase II Trial

Stephen Stahl*, Willie Earley, Yan Zhong, Mehul Patel, Istvan Laszlovszky, Gyorgy Nemeth, Roger McIntyre

Arbor Scientia Group, Carlsbad, California, United States

Background: Bipolar depression with mixed features is associated with more severe symptoms, mood episode recurrence, higher rates of comorbidity, worse clinical outcomes, and suicidality. Antidepressants are often ineffective and come with risk of manic episodes. Cariprazine is a dopamine D3/D2 receptor partial agonist that is approved for the treatment of adults with schizophrenia and manic or mixed episodes associated with bipolar I disorder; it is currently in development as monotherapy in bipolar depression and as adjunctive treatment in major depressive disorder. In a recent 8-week Phase II, randomized, double-blind, placebo-controlled trial (NCT01396447), cariprazine 1.5 mg/d demonstrated effectiveness versus placebo in the treatment of bipolar depression. This post hoc analysis of that study was conducted to evaluate the efficacy of cariprazine in a subset of patients presenting with mixed features.

Methods: Adult patients with a DSM-IV-TR diagnosis of bipolar I disorder and currently experiencing a major depressive episode were randomized to placebo or cariprazine 0.75 mg/d, 1.5 mg/d, or 3.0 mg/d for 8 weeks of double-blind treatment. Patients were required to have a total score ≥20 on the 17-item Hamilton Depression Rating Scale (HAMD) and a total score ≤10 on the Young Mania Rating Scale (YMRS) at screening. Patients with a baseline YMRS score ≥4 were considered to have mixed features and were included in the post hoc analysis. Efficacy was assessed for each cariprazine dose group compared with placebo; pooled cariprazine doses of 1.5 mg/d and 3.0 mg/d were also evaluated versus placebo. Efficacy analyses included change from baseline to week 6 (a priori timepoint for efficacy assessment in original study) in HAMD total score, analyzed using a mixed-model for repeated measures. HAMD remission (HAMD total score ≤7) at week 6 was also analyzed using an observed cases approach. Numbers needed to treat (NNTs) were calculated.

Results: A total of 353/571 (61.8%) randomized patients with bipolar I depression met the criteria for mixed features (placebo=86; 0.75 mg/d=87; 1.5 mg/d=90; 3.0 mg/d=90; pooled 1.5-3 mg/d=180). At week 6, the least squares (LS) mean change from baseline in HAMD total score was statistically significantly greater at α=0.05 for the cariprazine 3.0 mg/d (-11.5; P=.0436) and pooled 1.5-3 mg/d (-11.0; P=.0439) groups compared with placebo (-9.1). LS mean change from baseline to week 6 in the cariprazine 0.75 mg/d (9.3) and 1.5 mg/d (10.9) treatment groups did not differ significantly from placebo. The percentage of patients meeting HAMD remission criteria at week 6 was numerically (not significantly) greater in the cariprazine 1.5 mg/d (30.7%), 3.0 mg/d (22.9%), and 1.5-3.0 mg/d (26.9%) treatment groups versus placebo (17.9%), but not in the cariprazine 0.75 mg/d group (17.6%). Cariprazine 1.5 mg/d had the largest treatment effect, with an NNT of 8 (odds ratio [95% CI]: 2.0 [0.9, 4.5]; P=.081).

Conclusions: Treatment with cariprazine was effective versus placebo in reducing depression symptoms in patients with bipolar I depression and mixed features. In this analysis, higher doses of cariprazine appeared to be more effective in this subpopulation. Cariprazine may be an appropriate treatment option for patients with bipolar I depression and mixed features.

Keywords: Cariprazine, Mixed Features, Bipolar Depression

Disclosure: Part 1: Acadia, Grant, Alkermes, Grant, Allergan, Grant, Arbor Pharmaceuticals, Grant, AstraZeneca, Consultant, Axovant, Consultant, Biogen, Grant, Biopharma, Grant, Celgene, Grant, Forum, Grant, Genomind, Stock / Equity, Innovative Science Solutions, Grant, Intra-Cellular Therapies, Consultant, Jazz, Grant, Lundbeck, Advisory Board, Merck, Grant, Otsuka, Consultant, Servier, Consultant, Shire, Advisory Board, Sunovion, Grant, Takeda, Grant, Teva, Grant, Part 2: All the above, Grant, Part 3: None of the above, Grant, Part 4: See part 2, Grant.

W82. A Complex Epigenetic Switching is Critical for TNF-A Upregulation in Prefrontal Cortex of Suicide Subjects

Yogesh Dwivedi*, Qingzhong Wang, Bhaskar Roy, Richard Shelton, Gustavo Turecki

University of Alabama at Birmingham, Birmingham, Alabama, United States

Background: Several lines of evidence suggest that abnormalities in pro-inflammatory cytokines may be important contributory risk factor in suicidality. Tumor necrosis factor (TNF)-α, a member of TNF superfamily, is one of the most important pro-inflammatory cytokines which plays a critical role in innate and adaptive immune responses. Studies have shown that not only TNF-α is increased in serum of depressed suicide attempters, but cerebrospinal fluid (CSF) TNF-α levels can serve as a predictor of suicidal ideation. This is supported by another study showing that a composite score of the inflammatory markers including TNF-α is associated with high suicidal ideation among patients with major depressive disorders (MDD). Early-life adversity, a significant risk factor in suicidal behavior, is also associated with persistent high levels of TNF-α. We recently reported that expression of TNF-α is significantly increased in prefrontal cortex of adolescent suicide subjects. How the expression of this gene is regulated in brain of suicide subjects has not been studied. The present study was undertaken to examine the precise molecular mechanisms associated with TNF-α gene dysregulation in suicide brain.

Methods: TNF-α expression was examined in dorsolateral prefrontal cortex (dlPFC) of suicide subjects with and without major depressive disorder (MDD) and non-suicide MDD subjects using two different cohorts of brain samples. Roles of putative miRNAs targeting TNF-α and RNA-binding protein HuR were tested using in-vitro and in-vivo approaches and by examining expression of TAR-RNA binding protein (TRBP). Genetic influence on TNF-α expression was determined by eQTL analysis and by genotyping three SNPs in promoter region of TNF-α gene. Promoter methylation of TNF-α was determined using MeDIP assay. Expression of miR-19a-3p and TNF-α was also determined in peripheral blood mono nuclear cells (PBMC) of suicidal patients.

Results: TNF-α expression was significantly higher in dlPFC of suicide subjects regardless of psychiatric diagnosis. Its expression level was also increased in MDD subjects who died by causes other than suicide. On the other hand, expression of miR-19a-3p was upregulated specifically in suicide subjects. In a preliminary observation, similar upregulation of TNF-α and miR-19a-3p was noted in PBMC of suicidal patients. Despite its ability to directly target TNF-α in-vitro, miR-19a-3p showed no interaction with TNF-α in dlPFC. HuR stabilized TNF-α transcript presumably by sequestering its 3’UTR from miR-19a-3p-mediated inhibition. Furthermore, decreased TRBP expression supported abnormality in interaction between miR-19a-3p and TNF-α. Additionally, TNF-α transcriptional upregulation was associated with promoter hypomethylation, whereas no genetic influence was noted on altered TNF-α or miR-19a-3p expression in suicide subjects.

Conclusions: Our study shows that not only TNF-α is significantly upregulated in dlPFC of suicide subjects but there is a complex epigenetic regulatory mechanism that govern this increase. It appears that several molecules may significantly contribute to this upregulation such as miR-19a-3p, TRBP, and HuR. HuR may affect TNF-α expression independently to miR-19a-3p. Epigenetic modification mediated by hypomethylation of the promoter region of TNF-α also appears to participate in this upregulation. Genetic factors presumably do not play a role regulation of TNF-α. Interestingly, as with dlPFC, we observed that expression of both TNF-α and miR-19a-3p were significantly increased in the PBMCs of depressed individuals who had severe suicidal ideation. It indicates that both TNF-α and miR-19a-3p can be examined in blood tissues, which can presumably be associated with suicidality. Overall, our study provides a novel concept in understanding the molecular basis of altered expression of TNF-α in suicidal behavior, which could be critical in its etiopathogenesis.

Keywords: TNF-Alpha, Suicide, Depression, Gene Regulation

Disclosure: Nothing to Disclose.

W83. Poster Withdrawn

W84. Social Stress Induces Neurovascular Pathology Promoting Depression

Caroline Menard*, Madeline Pfau, Georgia Hodes, Veronika Kana, Victoria Wang, Sylvain Bouchard, Aki Takahashi, Meghan Flanigan, Hossein Aleyasin, Katherine LeClair, William Janssen, Benoit Labonté, Eric Parise, Zachary Lorsch, Sam Golden, Mitra Heshmati, Carol Tamminga, Gustavo Turecki, Matthew Campbell, Zahi Fayad, Cheuk Ying Tang, Miriam Merad, Scott Russo

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

Background: Clinical studies suggest that heightened peripheral inflammation contributes to the pathogenesis of major depressive disorder (MDD). The prevalence of MDD is two to three-fold higher in patients suffering from cardiovascular disease and conversely, MDD is associated with 80% increased risk of cardiovascular morbidity and mortality. Chronic inflammation and sustained increases in circulating pro-inflammatory cytokines have been associated with atherosclerotic plaque formation, progression, and rupture, likely contributing to the pathogenesis of cardiovascular disease and heart failure. Concomitantly, clinical studies report higher levels of circulating pro-inflammatory cytokines in MDD patients, a pattern that has been replicated in preclinical animal models of depression. Individual differences in the peripheral immune system and modulation of cytokine release, notably interleukin-6 (IL-6), are associated with susceptibility versus resilience to chronic social stress.

Methods: We investigated the effect of chronic social defeat stress, a mouse model of depression, on blood-brain barrier (BBB) permeability and infiltration of peripheral immune signals. Our study combines imaging studies (magnetic resonance imaging scans, electron and confocal microscopy) to region-specific viral-mediated manipulations and behavioral experiments.

Results: Chronic social defeat stress induced changes in the expression of genes involved in endothelial cell biology, angiogenesis, vascular permeability, and BBB formation and function. Interestingly, we found reduced expression of endothelial cell-specific tight junction protein claudin-5 (cldn5) and abnormal blood vessel morphology in the nucleus accumbens (NAc) of stress-susceptible but not resilient mice. The NAc plays a crucial role in stress responses and mood disorders including MDD. CLDN5 expression was also decreased in postmortem NAc of depressed patients. Cldn5 down-regulation using an adeno-associated virus (AAV) short hairpin RNA (shRNA) combined with a subthreshold social stress was sufficient to induce depression-like behaviors while chronic antidepressant treatment rescued cldn5 loss and promoted resilience. Reduced BBB integrity in NAc of stress-susceptible or AAV-shRNA-cldn5-injected mice resulted in direct passage of peripheral cytokine IL-6 across the BBB and subsequent expression of depression-like behaviors.

Conclusions: These findings suggest that chronic social stress alters BBB integrity through loss of tight junction protein cldn5, promoting peripheral IL-6 passage across the BBB and depression. Our study is the first to characterize and functionally interrogate the neurovascular pathology associated with social stress vulnerability. To our knowledge, it is also the first direct evidence that chronic stress is sufficient to allow infiltration of circulating cytokines in the absence of local mechanical trauma or a brain tumor. Greater understanding of the mechanisms by which chronic stress activates the immune system and undermines BBB integrity may promote the design of more effective antidepressant strategies, either by augmenting current treatment protocols or by informing the discovery of new therapeutics that enhance neurovascular health within stress-related brain regions. A major impediment to this goal will be the development of novel therapeutic agents that can enhance Cldn5 expression in order to repair endothelial damage as, to date, no such compound exists.

Keywords: Chronic Stress, Blood-Brain Barrier, Cytokines, Major Depressive Disorder (MDD)

Disclosure: Nothing to Disclose.

W85. Remission and Recovery in Patients With Bipolar I Disorder (BP-I) Treated With Aripiprazole Once-Monthly (AOM 400)

Jessica Madera*, Na Jin, Joan Amatniek, Brian Johnson, Ross A. Baker, Pedro Such

Otsuka Pharmaceutical Development & Commercialization, Inc, Princeton, New Jersey, United States

Background: After symptomatic treatment of acute episodes of bipolar I disorder (BP-I), the treatment goal is remission and ultimately, recovery. Reported here are the results for secondary endpoints assessing remission and recovery rates in a population of patients being treated with aripiprazole once-monthly 400 mg (AOM 400) from one double-blind, placebo-controlled 52-week maintenance study, and one 52-week open-label study.

Methods: Assess the remission and recovery rates from two clinical trials with AOM 400. Remission was defined as subjects with YMRS and MADRS total scores≤12. The definition for recovery was meeting criteria for remission for 8 consecutive weeks (i.e. sustained remission). The first study (NCT01567527) was a double-blind, placebo-controlled, randomized withdrawal study which consisted of a screening phase followed by 4 treatment phases. Subjects underwent screening for eligibility for 6 weeks, followed by a conversion to oral aripiprazole monotherapy for 4 to 6 weeks, if needed (Conversion Phase), an Oral Stabilization Phase for 2-8 weeks, an AOM 400 Stabilization Phase for 12 to 28 weeks, and a Double-blind, Placebo-controlled Phase where patients were randomized to continue treatment with either AOM 400 or placebo for 52 weeks. The second study (NCT01710709) was an open-label, uncontrolled, multicenter study that enrolled de novo patients with a diagnosis of bipolar I disorder (BP-I) and rollover patients who participated in the double-blind placebo controlled study. The open-label trial was composed of a screening phase followed by 3 treatment phases: a 4 to 6-week conversion phase (if applicable), a 4 to 12-week oral aripiprazole stabilization phase, and a 52-week open-label aripiprazole AOM 400 maintenance phase. De novo subjects entered the trial at screening, and proceeded to the conversion phase if they were currently being treated with an antipsychotic (except oral aripiprazole), mood stabilizer, or antidepressant (ADT) and thus required a switch to oral aripiprazole. Subjects who were not currently being treated with an antipsychotic, mood stabilizer, ADT or oral aripiprazole were able to proceed to the oral stabilization phase. Subjects from the double-blind, placebo-controlled trial who were eligible for entry into this open-label trial, and who chose to continue treatment upon completion of trial, entered directly into the AOM 400 Maintenance Phase. Statistical comparisons of drug vs. placebo were made using the Cochran-Mantel-Haenszel general association test, stratified by region; descriptive statistics are used for other data.

Results: In the double-blind study, compared with placebo, a significantly higher proportion (p=0.0169) of AOM 400-treated patients achieved protocol-defined remission at last visit (74%, n=97/131) vs. 60.2 % for placebo (n=80/133) and a significantly higher proportion (p=0.0122) were in recovery by last visit (64.9%, n=85/131) vs. 49.6% for placebo (n=66/133). Roll-over patients from the primary study entered the open-label study and the observed remission rate remained stable (98.8%, n=84/85) by last visit and 80.0% (n=68/85) completed the open-label study.

In the open-label study, de novo patients also had a high proportion of patients meeting the criteria for remission (87.2%, n=328/376) and recovery (77%, n=292/379) by last visit, and 58.8% (n=223/379) completed the study. The overall safety profile showed that AOM 400 was well tolerated.

Conclusions: Current data showed high rates of remission and recovery in patients with BP-I taking AOM 400 both in placebo-controlled randomized withdrawal study and in a more naturalistic open-label study. The results support AOM 400 as a viable once-monthly option for maintenance treatment of BP-I. The injectable once-monthly formulation may contribute to the high rates of remission and recovery observed here, perhaps due to adherence benefits of once-monthly formulations, giving clinicians another option in maintenance treatment of adult patients with bipolar disorder.

Keywords: Bipolar Disorder, Long-Acting Injectable Antipsychotics, Remission, Recovery

Disclosure: Part 1: Otsuka Pharmaceutical Development & Commercialization, Inc., Employee, Part 2: Otsuka Pharmaceutical Development & Commercialization, Inc., Employee, Part 3: Otsuka Pharmaceutical Development & Commercialization, Inc., Employee, Part 5: Otsuka Pharmaceutical Development & Commercialization, Inc., Employee.

W86. Sex Differences in Ventral Hippocampal Outputs Driving Responses to Stress

Andrew Eagle, Claire Manning, Elizabeth Williams, Rachel Neve, Alfred Robison*

Michigan State University, East Lansing, Michigan, United States

Background: Depressive disorders disproportionately affect women, and the symptoms of these disorders are sexually dimorphic, with female patients more likely to experience anhedonia and social withdrawal. The ventral hippocampus (vHipp) regulates emotional and motivated behaviors through projections to nucleus accumbens (NAc) and amygdala (BLA), and the vHipp-NAc circuit is critical for resilience to social withdrawal after chronic social defeat stress in male mice (CSDS; Bagot et al., 2015). However, sex differences in this circuit and the potential role of this circuit in female susceptibility to the social withdrawal aspects of depression are unknown. Moreover, it is unknown how CSDS alters vHipp gene expression and function, and whether such alterations may underlie resilience. Using a novel dual-virus CRISPR system, we investigate stress-induced alterations in gene expression specifically in vHipp-NAc cells underlying aspects of the susceptible phenotype in male and female mice and provide evidence that neuronal excitability in this circuit drives sex differences in behavior.

Methods: Male and female 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. Chronic Social Defeat Stress (CSDS), subchronic SDS, and witness SDS were performed essentially as previously described (Vialou et al., 2010; Iñiguez et al., 2017). Briefly, male mice were exposed to aggressors for 5 min sessions once per day for 10 days while being co-housed but physically separated from the aggressor for the remaining time each day (CSDS), or were exposed to aggressors for three 5 min sessions in a single day (subchronic SDS). Male or female mice witnessed these encounters through a perforated plexiglass partition (witness CSDS). Social interaction tests were performed one day after the final defeat session. The open field apparatus consisted of a custom-made, square white polyvinylchloride foam box (38 cm x 38 cm x 35 cm) and elevated plus maze was performed using a custom-built apparatus based on plans from ANY-maze (www.anymaze.com). Viral vectors were injected by stereotaxic surgery into vHipp (3° angle; -3.2 mm anteroposterior (AP), ±3.4 mm mediolateral (ML) and -4.8 mm dorsoventral (DV)), NAc (10°; +1.6 AP; ±1.5 ML; -4.4 DV), or (0°; -1.3 AP; ±3.4 ML; -4.5 DV). Slices of vHipp were stained with a polyclonal FosB antibody (5G4, Cell Signaling), and positive cells were counted by a double-blind experimenter. Western blots were performed on tissue punches from vHipp using the same antibody and quantified using ImageJ (www.NIH.gov). For electrophysiology, coronal slices containing vHipp (250 μm thick) were cut in ice-cold sucrose artificial cerebrospinal fluid (ACSF), and recordings were performed at 30–32 °C in ACSF. vHipp was identified under visual guidance using infrared differential interference contrast video microscopy with a 40 × water-immersion objective (Olympus BX51-WI). Whole-cell voltage-clamp recordings were performed with a computer-controlled amplifier (MultiClamp 700B), digitized (Digidata 1440), and acquired with Axoscope 10.1 (Molecular Devices) at a sampling rate of 10 kHz. Circuit-specific translating ribosomal affinity purification (TRAP) was performed using Rosa26eGFP-L10a mice and antibodies 19C8 and 19F7 (Memorial Sloan-Kettering Monoclonal Antibody Facility) as described (Heiman et al., 2014).

Results: We demonstrate that female mice have increased excitability of vHipp-NAc neurons at baseline, and that loss of FosB gene expression in male mice conveys a similar difference in function of this circuit. Moreover, we find that specific silencing of FosB gene expression in vHipp cells projecting to NAc promotes susceptibility to the social withdrawal phenotype, while silencing in vHipp-BLA cells does not. Conversely, silencing FosB in vHipp-BLA cells promotes anxiety behaviors, while vHipp-NAc silencing does not. Finally, we use circuit-specific TRAP to uncover a variety of potential molecular mediators of these cellular and behavioral phenomena.

Conclusions: We demonstrate higher baseline excitability of vHipp-NAc projection neurons in females compared to males and present evidence that this may underlie increased vulnerability to the social withdrawal aspects of depression documented in female patients. Moreover, we use novel CRISPR-Cas9 technology to uncover a transcriptional mechanism for stress-dependent alterations in excitability of this circuit that may underlie susceptibility and resilience to multiple aspects of chronic social stress in both sexes. Finally, the target genes revealed using our circuit-specific TRAP represent potential therapeutic inroads for the sex-specific, personalized treatment of individual aspects of depression-related disorders.

Keywords: Depression Risk, Electrophysiology, CRISPR/Cas9, Neurocircuits

Disclosure: Nothing to Disclose.

W87. Investigation of the Possible Reinforcing Effects of Samidorphan and Naltrexone by Fixed and Progressive Ratio Intravenous Self Administration Testing in Rats

David Heal*, Sharon Smith, Reggie Dean, Mark Todtenkopf

RenaSci Ltd., Nottingham, United Kingdom

Background: Samidorphan (SAM) is a μ-opioid receptor antagonist. ALKS 5461, is a proprietary, investigational, sublingual medication that is being developed as an adjunctive treatment for major depressive disorder. ALKS 5461 contains SAM and buprenorphine, a partial μ-opioid receptor agonist and κ-opioid receptor antagonist.

We investigated whether SAM or naltrexone (NTX; another opioid receptor antagonist) served as positive reinforcers in rats trained to self-administer a low dose of the highly abused Schedule 1 (C-I) opiate, heroin. We also evaluated their relative reinforcing effects using a progressive ratio (PR) schedule to determine the break-points for drug reinforcement and compared them to heroin.

The use of PR/break point determinations to assess the relative reinforcing effects of novel CNS-active drugs is recommended in the EMA guidelines on assessing the abuse potential of novel drugs for human use (CHMP/EMA, 2006).

Methods: Male Sprague Dawley rats (200-225g; Charles River, UK) were mildly food restricted and trained to lever press for food rewards. Once responding was stable, rats were implanted with a jugular vein catheter. Rats were subjected to heroin (15μg/kg/injection[inj]) acquisition followed by saline extinction before being divided into 3 groups for testing additional doses of heroin (7.5 and 25μg/kg/inj), SAM (13.6, 40.8 and 68μg/kg/injection) or NTX (13.6, 40.8 and 68μg/kg/inj). Training and testing were on a fixed ratio-5 (FR5) schedule in 2 hr sessions. When stable

self-administration was observed over 3 consecutive sessions (inj/session did not vary by >25%, or mean ≤6 inj/session for non-reinforcement), the break point for operant responding was determined in a 4 hr PR session. Results were analysed parametrically after angular transformation with each test compound compared to saline and heroin by separate Dunnett's tests. Results are reported as mean±SEM for ≥7 rats/group.

Results: Heroin maintained self-administration (inj/session) on FR5 at levels significantly (p<0.001) greater than saline (4.3±0.2; n=31) at all doses (7.5μg/kg/inj=17.5±1.6 [n=8]; 15μg/kg/inj=18.8±0.4 [n=31]; 25μg/kg/inj=15.4±2.4 [n=7]). On the PR schedule, the break-points (lever-presses/inj) for responding to heroin, 41.5±7.6 for 75μg/kg/inj (n=8), 60.2±8.1 for 15μg/kg/inj (n=31), and 50.6±20.9 for 25μg/kg/inj (n=7) were all significantly (p<0.001) greater than saline (10.4±0.8; n=31).

SAM did not serve as a positive reinforcer on FR5 at 13.6 or 40.8μg/kg/inj. The number of inj/session of SAM (68μg/kg/inj), ie 9.2±2.1, was significantly (p<0.01) greater than saline, ie 4.3±0.2. The break points for all doses of SAM (13.8±2.5, 13.3±1.2 and 17.9±3.3 lever-presses/inj [n=7-8] for 13.6, 40.8 and 68μg/kg/inj, respectively) were all significantly (p<0.01) lower than those for all doses of heroin. Break-points for the 2 lower doses of SAM were not significantly different from saline. The break-point for SAM (68μg/kg/inj) was significantly (p<0.05) greater than saline.

NTX did not serve as a positive reinforcer on an FR5 at any of the doses. However, NTX (13.6μg/kg/inj) maintained 8.1±2.3inj/session which almost reached statistical significance versus saline (p=0.053). Break points for NTX (17.2±3.8, 15.0±3.6 and 16.0±4.1 (n=7-8) for 13.6, 40.8 and 68μg/kg/inj, respectively), were not significantly different from the vehicle break point. NTX's break-points were all significantly (p<0.01) lower than those of all doses of heroin.

When the FR5 and PR/break-point results for SAM were compared with those of NTX, there were no differences between the 2 drugs.

Conclusions: Heroin served as a powerful reinforcer in rats which is consistent with its profile as a recreational drug of abuse. This was demonstrated by (i) number of doses that served as a reinforcer, (ii) number of injections rats took in test sessions, and (iii) high break-points for reinforcement. SAM and NTX elicited weak signals as positive reinforcers at a single dose in heroin-maintained rats. With the exception of a single dose of SAM, their relative reinforcing effects were not significantly different from saline. All of the break-points for SAM and NTX were significantly lower than heroin. Overall, SAM and NTX have similar reinforcing profiles in heroin-maintained rats. Based on head to head comparison results, it can be concluded that the rewarding properties of SAM and NTX are comparable. If these results translate to man, they predict SAM poses a minimal risk for recreational abuse.

CHMP/EMA (2006). Guideline on the Non-Clinical Investigation of the Dependence Potential of Medicinal Products.


Keywords: Samidorphan, Naltrexone, Heroin Self-Administration, Rats

Disclosure: Part 1: RenaSci, Ltd., Stock / Equity, RenaSci, Ltd., Employee.

W88. The Mediating Role of Cholesterol Efflux on the Relationship Between Serum Cholesterol and Suicide Risk

Emma Knowles*, Joanne Curran, Peter Meikle, Kevin Huynh, Harald Göring, Ravi Duggirala, Laura Almasy, John Blangero, David Glahn

Yale University, Hartford, Connecticut, United States

Background: The link between serum cholesterol and suicide attempts is well established; several studies suggest that lower levels of cholesterol are associated with increased risk of mortality by suicide. However, the etiology of this association remains unknown. It might be caused by the influence of genetic factors and/or environmental exposures. We present the first study to directly address this question.

Methods: All genetic analyses were conducted in SOLAR in a sample of 552 Mexican American individuals from extended pedigrees. The standardized genetic covariances between 23 lipid classes, acquired from 10 μl of plasma, and attempted suicide were calculated. The suicide phenotype was taken from the Suicidality section of the Mini International Neuropsychiatric Interview (MINI). In addition, plasma-based cholesterol efflux capacity (CEC; total, non-ABCA1, and ABCA1-specific) and Lecithin:cholesterol acyltransferase (LCAT) levels were available for all participants. Multilevel mediation analyses were conducted in R using the lme4 and lmerTest packages, this allows the clustering due to family structure to be taken into account.

Results: Both free cholesterol (β=-0.70, p=2.9x10-04) and lyso-phosphatidylcholine (β=-0.65, p=2.0-03) exhibited significant genetic overlap with attempted suicide after the application of a multiple comparison correction. Given that the movement of free cholesterol and lyso-phosphatidylcholine have established interactions in the glycerophospholipid metabolism pathway via efflux and LCAT respectively, we investigated these potential relationships in the data. Neither free cholesterol nor lyso-phosphatidylcholine showed an association with LCAT, and lyso-phosphatidylcholine did not show an association with CEC. However, there was a large and significant genetic correlation between free cholesterol and ABCA1-specific CEC (rhog=0.64, p=1.38x10-06). Mediation analysis indicated that the relationship between free cholesterol and attempted suicide was significantly mediated by ABCA1-specific CEC (β=0.07, p=0.035).

Conclusions: While alterations in cholesterol levels have been previously associated with attempted suicide the present study is the first to demonstrate a shared etiology between these two phenotypes. Moreover, the results of this study imply that cholesterol efflux, an initial step in the process of reverse cholesterol transport, may be key to the association between free cholesterol and attempted suicide.

Keywords: Suicide, Cholesterol Biosynthesis, Lipids, Human Genetics

Disclosure: Nothing to Disclose.

W89. Statistical Detection of Subpopulations of Treatment-Resistant Depression: An Application of Latent Class Mixture Models in a Series of Ketamine Clinical Trials

Cristan Farmer*, Elizabeth Ballard, Lawrence Park, Carlos Zarate

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

Background: Under DSM-5 criteria, over 200 symptom combinations can lead to a diagnosis of a depressive episode. Thus, the symptom profiles of individuals who meet criteria for depression vary widely. This heterogeneity may reflect differences in etiology, and subsequently, predict differences in response to intervention. The promise of precision medicine is that the identification of more homogeneous sets of patients who may be more responsive to targeted intervention will yield greater overall success rates in the treatment of depression. In this analysis, we use a person-centered latent variable method to parse the heterogeneity of a pooled sample of depressed individuals, and explore whether these subgroups moderate response to a rapid-acting antidepressant.

Methods: Data were obtained from four clinical trials of ketamine involving 158 adults with major depressive disorder (n=114) or bipolar depression (n=44). All studies were conducted on the same research unit and participants were assessed and treated by the same clinical and research staff. The primary analytic method was latent profile analysis, which is a type of finite mixture model that expresses the overall distribution of several variables as a mixture of two or more distributions. In other words, it is an empirical method for identifying classes of patients. In this study, the classes were formed based on scores on unidimensional constructs, derived from baseline ratings on the Beck Depression Inventory, the Hamilton Depression Rating Scale, the Montgomery-Asberg Depression Rating Scale, and the Snaith-Hamilton Pleasure Rating Scale. The eight unidimensional constructs were Depressed Mood, Tension, Negative Cognitions, Impaired Sleep, Suicidal Thoughts, Reduced Appetite, Anhedonia, and Amotivation. Using data from the three studies that were placebo-controlled crossover evaluations of ketamine (n=101), we assessed the resulting class grouping as a moderator of the effect of ketamine.

Results: The best-fitting model comprised two classes, distinguished by both means and variances of the indicator variables. The profile of the first class (35%, called the Anxious class) was characterized by more severe scores on Tension relative to the other subscales (and compared to the other class). The second class (65%, called the Severe class) was characterized by more severe scores on Depressed Mood, Anhedonia, and Amotivation. The classes were further characterized using other baseline characteristics; the Anxious class was approximately 5 years older (p=.02) and was less likely to have been previously hospitalized for depression (55% versus 72% in the Severe class, p=.04). Although class membership was not related to likelihood of having bipolar or major depression, the Anxious class was significantly more likely to have the atypical subtype of major depression (33% versus 19% in the Severe class) whereas the Severe class was more likely to have the melancholic subtype (33% versus 13% in the Anxious class) (p =.02). While the difference did not reach statistical significance, the Anxious class was more likely to have a family history of alcohol abuse (65% versus 49% in the Severe class, p=.06). No differences between classes were observed for sex, diagnosis (major depression versus bipolar disorder), body mass index, length of current episode, personal alcohol abuse, history of suicide attempt, or family history of mood disorder. Most-likely class membership was entered into a mixed model as moderator of treatment response, using both the Hamilton Depression Rating Scale and the Montgomery-Asberg Depression Rating Scale as outcomes. A significant study-by-class-by-drug interaction was observed for both outcomes (p<.0001). This three-way interaction was explained by a significant drug-by-class interaction in the study of bipolar disorder, but not in the two studies of major depression. In the bipolar study, the effect of ketamine was significantly larger in Severe than in the Anxious class. However, in the two major depression studies, the effect of ketamine did not differ between classes.

Conclusions: Even in the context of clinical trials, where inclusion criteria have created a relatively homogeneous sample, the data do suggest the presence of at least two classes of individuals distinguished primarily by the relative severity of anxious and depressive symptoms. Among individuals with bipolar disorder, the individuals most likely to belong to the Anxious class experienced less improvement with ketamine. However, class membership was not predictive of improvement for the individuals diagnosed with major depression. Thus, these results suggest that clinical symptom-derived subgroups may moderate the effect of ketamine, though this effect may be limited to individuals with bipolar disorder.

Keywords: Statistical Methods, Heterogeneity, Depression Subtypes, Ketamine

Disclosure: Nothing to Disclose.

W90. Inflammatory Resolution Regulator for Cytochrome P450 Eicosanoids may be a Key Prognostic Biomarker for Major Depressive Disorder in Adolescent Females

Pei-an Shih*, Jun Yang, Christophe Morisseau, Chadi Calarge

University of California, San Diego, La Jolla, California, United States

Background: Major depressive disorder (MDD) in adolescents is common; it is characterized by frequent recurrence and chronicity that can lead to impairment, disability, and premature mortality. Depressed individuals exhibit higher levels of proinflammatory cytokines and C-reactive protein compared to nondepressed individuals. While systemic inflammation may play a role in mediating and eliciting depressive symptoms, little is known about the role of inflammation resolution in MDD. Eicosanoids are bioactive lipid mediators produced by the oxidative metabolism of polyunsaturated fatty acids such as n-6 arachidonic acid (ARA) and n-3 docosahexaenoic acid (DHA). Soluble epoxide hydrolase (sEH) is a key enzyme that modulates inflammatory cascade of cytochrome P450-derived epoxy-eicosanoids to their diols. Together, eicosanoids and their regulators affect numerous key physiological processes including endocrine signaling, immune regulation, and inflammation. To gain insights into the role of inflammatory resolution in MDD, eicosanoids markers were studied in adolescent females having completed a longitudinal MDD study.

Methods: Subjects were selected from a completed naturalistic longitudinal study examining the skeletal effects of MDD. In the original study, all subjects were female and were un-medicated or within a month of starting SSRI therapy. A latent class analysis generated MDD trajectory classifier for all subjects. For the retrospective comparative analysis, subjects were selected from two “trajectory groups”: a “high chronicity trajectory” group (n=6; age [SD]= 21.5±1.3) and a “low chronicity trajectory” group (n=6; age [SD]= 21.3±1.7). The severity of MDD at each study visit was measured using Inventory of Depressive Symptomatology [IDS] and Beck Depression Inventory [BDI]. MDD chronicity was assessed as the proportion of weeks with a depression severity rating meeting DSM criteria for episode. Forty eicosanoid markers derived from omega 6 linoleic acid (LA) and ARA, and omega 3 α-Linolenic acid (ALA), eicosapentaenoic acid (EPA), and DHA were measured using the GC/MS and LC/MS/MS systems. Diol-eicosanoid:epoxy-eicosanoid ratios (e.g., DiHETrE – to - EpETrE ratios) were calculated as proxy markers of in vivo sEH activity. The retrospective analysis was conducted using clinical and eicosanoid data from the baseline and the 6th study visits. Nonparametric test statistics (Wilcox test and Spearman's rank correlation coefficient) were used to assess the strength of association in R. Principal component analysis was applied to convert all eicosanoid markers into principal components and used in partial least squares discriminant analysis (PLS-DA). The study received IRB approval & informed consents/assents were obtained.

Results: Although there was no significant baseline difference in any patient factors that are known to influence MDD trajectory such as symptom severity and antidepressant use, at the final visit (6th), the subjects in the “high chronicity” group had significantly worse depression scores (IDS: 29 vs 6.5, p=0.0005; BDI: 24.5 vs 1.33, p=0.002) and higher MDD chronicity (the proportion of weeks with significant depression) compared to the patients in the “low chronicity” group (88.7% versus 17.7%, p<0.0001). Non-parametric inferential statistics revealed eicosanoids 13-HODE and 9.10-DiHOME of LA, and 15-oxo-ETE and 14.15-EpETrE of ARA to be significantly associated with the chronicity status (p-values <0.05). PLS-DA confirmed these associations and demonstrated 9-HODE of LA and 15.16-DiHODE of ALA to also have high power to discriminate high chronicity subjects from low chronicity subjects. Inflammatory resolution process was implicated in MDD chronicity as all 7 baseline diol-eicosanoid:epoxy-eicosanoid ratio markers (proxy of in vivo sEH activity) were uniformly elevated in high chronicity subjects compared to low chronicity subjects (p-values: 0.008–0.39). Given the striking consistency of directional correlation in 7 sEH proxy markers derived from eicosanoids of 4 polyunsaturated fatty acids, Spearman's rank correlation was used to determine the strength and direction of monotonic association of baseline sEH proxy markers with MDD severity at the final visit, and MDD chronicity in all subjects combined. Rho statistics for final visit BDI score range from +0.2 to +0.72 with 7 proxy markers, of which 4 were statistically significant. Rho statistics for chronicity range from +0.21 to +0.76, of which 3 were statistically significant. To confirm the significance of our findings, eicosanoids obtained from the final study visits were analyzed by PCA and PLS-DA. The top 5 most significant markers discriminating high chronicity subjects from low chronicity subjects were all diol-eicosanoids, the by-products of sEH catalyzed reactions.

Conclusions: We identified a number of eicosanoids that are differentially expressed in adolescent females with high chronicity MDD. The most interesting finding was that elevation of sEH proxy marker was consistently correlated with higher MDD chronicity and severity, despite a small sample size. sEH is known as a regulator of inflammatory resolution by its potent and complex mechanisms in the formation/catabolism of epoxy- and diol eicosanoids under varying physiological conditions. We have previously shown that patients with remitted anorexia nervosa have higher levels of sEH compared to healthy controls, implicating sEH in neuropsychiatric disorders. This study is the first report linking high sEH to adverse longitudinal MDD outcomes. Molecular prognostic biomarker for MDD is much needed especially in adolescents. Our data show that a number of eicosanoids are associated with MDM chronicity in adolescent females. Longitudinal data further support that sEH, a pivotal regulator of inflammatory resolution, plays a role in the longitudinal outcome of MDD, and should be researched further as a prognostic biomarker.

Keywords: Adolescent Depression, Inflammation, Eicosanoids, Fatty Acids, Soluble Epoxide Hydrolase

Disclosure: Nothing to Disclose.

W91. [18F]JNJ-64511070, a Novel and Specific PET Radioligand for AMPAR/gamma8 TARP

Hartmuth Kolb*, A. Katrin Szardenings, Greg Chen, Terry Lebold, Brian Lord, Michael Maher, Nancy Wu, Chunfang Xia, Wei Zhang, Gilles Tamagnan, Oliver Barret, Christian Constantinescu

Janssen Pharmaceutical, San Diego, California, United States

Background: The transmembrane receptor regulatory protein TARP gamma8, which is preferentially expressed in the hippocampus, controls hippocampal AMPA receptor expression, distribution and synaptic plasticity. TARP gamme8 antagonists are being developed for the treatment of epilepsy, major depressive disorders (MDD), anxiety disorders, bipolar disorders and schizophrenia. To support our internal TARP gamme8 discovery program, [18F]JNJ-64511070 was identified as a PET ligand to assess target engagement, determine receptor occupancy, and aid dose selection in clinical trials.

Methods: JNJ-64511070 was profiled in a series of standard in vitro, ex vivo, and in vivo assays. Specific binding of [18F]JNJ-64511070 was evaluated in in vitro autoradiography (ARG) experiments on rat brain tissue sections. PET imaging of the tracer in rats was performed. The tracer was further studied in rhesus macaques with JNJ-284, a selective AMPAR TARP gamme8 antagonist, as the pretreatment. PET images were analyzed with arterial plasma input function based Logan graphical analysis (LGA) to estimate the volume of distribution.

Results: JNJ-64511070 is a highly potent (IC50 for human AMPAR gamme8 TARP, FLIPR™ is 0.05 nM) and selective PET ligand. ARG experiments in rat brain sections including blocking studies with the antagonist JNJ-284 confirmed target specificity. Rat micro-PET imaging demonstrated the expected uptake pattern in the hippocampus and frontal cortex, which was completely blocked by JNJ-284. Non-human primate PET imaging studies also indicated highest uptake of [18F]JNJ-64511070 in the hippocampus and frontal cortex regions with minimal background binding. Pretreatment with JNJ-284 demonstrated a dose-dependent blocking effect of the tracer's uptake in the brain.

Conclusions: The preclinical evaluation of [18F]JNJ-64511070 suggests that this tracer is a suitable candidate PET ligand for AMPAR gamme8 TARP and ought to be explored as a biomarker of synaptic density and plasticity.

Keywords: PET Imaging, AMPA Receptors, AMPA TARP gamma8, Synaptic Protein Imaging

Disclosure: Part 5: Janssen R&D, Employee.

W92. Prophylactic Ketamine Alters Nucleotide and Neurotransmitter Metabolism in Brain and Plasma

Josephine McGowan*, Collin Hill, Alessia Mastrodonato, Christina LaGamma, Rebecca Brachman, Niven Narain, Michael Kiebish, Christine Denny

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

Background: Recently, we have shown that ketamine given prior to stress exposure protects against the development of depressive-like behavior in mice. These data suggest that it may be possible to prevent the induction of affective disorders before they develop by administering prophylactic pharmaceuticals, a relatively nascent and unexplored strategy for psychiatry.

Methods: Here, we performed metabolomics analysis of brain and plasma following prophylactic ketamine treatment in order to identify markers of stress resilience enhancement. We administered prophylactic ketamine in mice to buffer against fear expression. Following behavioral analyses, untargeted metabolomic profiling was performed on both hemispheres of the prefrontal cortex (PFC) and the hippocampus (HPC), and plasma following prophylactic treatment.

Results: Prophylactic ketamine reduced fear expression, as we have previously found. Eight metabolites were changed in both hemispheres of the PFC and HPC upon ketamine treatment. Purine and pyrimidine metabolism were most significantly changed in the HPC and PFC. Interestingly, they were also significantly changed in plasma two weeks after prophylactic administration. Moreover, most precursors to inhibitory neurotransmitters were increased whereas precursors to excitatory neurotransmitters were decreased.

Conclusions: Our results suggest that prophylactic treatment differentially affects purine and pyrimidine metabolism and neurotransmission in brain and plasma, which may underlie the long-lasting resilience to stress induced by a single injection of ketamine. These data may provide novel targets for prophylactic development. To our knowledge, this is the first study that identifies metabolomic alterations and biomarker candidates for prophylactic ketamine efficacy in mice.

Keywords: Metabolomics, Ketamine, Prophylactic, Stress, Psychiatric Disorders

Disclosure: Nothing to Disclose.

W93. Relationship of the Kynurenine Pathway to the Improvement of Depression in Esketamine-Treated Subjects With Major Depressive Disorder at Imminent Risk for Suicide

Lynn Yieh*, Derrek Hibar, Carla Canuso, Hartmuth Kolb, Maura Furey, Ziad Saad

Janssen Research & Development, LLC, San Diego, California, United States

Background: Dysregulation of the kynurenine pathway has been observed in patients with Major Depressive Disorder (MDD) and to a greater extent in MDD patients with suicidal ideation. Induction of the kynurenine pathway can be triggered by proinflammatory factors associated with MDD; catabolism of Tryptophan (TRP) to kynurenine (KYN) is mediated by indoleamine 2,3-dioxygenase (IDO), which is inducible by inflammatory factors, such as IFNγ. Induction of TRP degradation can have deleterious effects by shunting TRP away from the serotonin pathway and by affecting processes thought to contribute to the pathophysiology of MDD through its metabolites. For example, kynurenine pathway metabolites modulate the N-methyl-D-aspartate receptor (NMDAR), and hyperactivity of the NMDAR is believed to contribute to development of depressive symptoms.

While correlation of kynurenine pathway metabolites in plasma and CSF of patients with depressive and suicidal symptoms have been reported, there is little information about kynurenine pathway response following treatment. Here we describe the relationship between baseline, pretreatment concentrations of kynurenine pathway elements to the subsequent improvement of depressive symptoms following treatment with esketamine in subjects at imminent risk for suicide.

Methods: PeRSEVERe, a randomized, 25 day double-blind, placebo-controlled study of investigational esketamine in subjects with MDD at imminent risk for suicide was designed to evaluate the efficacy of esketamine (84mg intranasal). Kynurenine pathway components (TRP, KYN, KYNA, KYN3OH, QUINA) were measured by LC-MS/MS from the plasma of 48 subjects (Esketamine N= 26, Placebo N=22) at baseline (pre-treatment) as well as Day 25 post-treatment. SNPs in kynurenine pathway enzyme genes (IDO, KAT, KMO) were also analyzed.

Baseline kynurenine biomarkers were modeled against change in the Montgomery-Asberg Depression Rating Scale (MADRS) at 4h, Day3, and Day 25 (pre-and 4h post-treatment). Baseline BMI, sex, and age were included as covariates. Significance was defined as p<0.05. Significant fits with baseline biomarkers were further analyzed to determine if post- treatment changes in biomarker levels correlated with symptom improvement during the double-blind phase.

Results: Linear models of change in MADRS showed significant contribution by TRP and KYN3OH that differed between placebo and esketamine treated subjects. Lower levels of TRP at baseline correlated with improved response to esketamine at an early timepoint (4h post-treatment, p=0.02), showed a trend towards significance at later timepoints (Day 25 pre- and 4h post dose, p=0.05 and 0.09, respectively), and did not correlate with placebo response (p> 0.10). Lower levels of KYN3OH at baseline, however, correlated with a better response to esketamine treatment at later time points (Day 25 pre- and 4h post dose, p=0.01 and 0.003 respectively), and did not correlate with placebo response (p> 0.10). Subjects with the largest increases of TRP at the end of the study were more likely to benefit from treatment; correlation of changes in TRP from baseline to Day 25 with change in MADRS at 4h post treatment trended towards significance (p=0.09). In addition to metabolites, the IDO SNP with most significant correlation (p=0.002) with MADRS at 4h post-treatment indicated that each additional copy of the minor allele was associated with increased response to treatment compared to the wild type. The KMO SNP with most significant correlation (p=0.003) with change in MADRS at Day 25 post treatment suggested that subjects homozygous for the major allele had increased response to treatment compared to heterozygotes.

Conclusions: Our current findings suggest that some metabolites of the kynurenine pathway may identify subjects who are most likely to respond to esketamine treatment. Subjects with lower levels of baseline TRP were more likely to show an improvement in MADRS at 4h post treatment, suggesting that MDD patients with overactive TRP catabolism could benefit from treatment with esketamine. This hypothesis is further supported by evidence that subjects with the largest increases of TRP levels post treatment (mechanistically possible by shunting TRP away from the kynurenine pathway) were more likely to respond to esketamine treatment. Variants in genes involved in kynurenine metabolism including IDO and KMO are associated with differential response to treatment with esketamine. Further research is needed to understand the functional impact of these variants on kynurenine metabolites in response to treatment with esketamine. Correlation of changes in both TRP and KYN3OH metabolites with improvement in depressive symptoms were significant in esketamine-treated subjects but not placebo, suggesting that esketamine treatment may act to regulate the kynurenine pathway.

Keywords: Kynurenine Pathway, Suicidality, Esketamine

Disclosure: Part 5: Janssen Research & Development, LLC, Employee.

W94. Effects of Buprenorphine on Responses to Emotional Stimuli in Individuals With a Range of Mood Symptomatology

Anya Bershad*, Nicholas Ruiz, Harriet de Wit

University of Chicago, Chicago, Illinois, United States

Background: The opioid drug buprenorphine has been shown to modify responses to emotional stimuli and may have antidepressant properties. In preclinical studies, it shows antidepressant-like and anxiolytic-like effects, and a handful of clinical studies suggest it may reduce symptoms of depression in patients. We have shown that low doses of buprenorphine reduce responses to negative emotional stimuli in healthy adults. Here we extended these findings to individuals with symptoms of depression and anxiety.

Methods: We examined the effects of buprenorphine on i) attention to emotional facial expressions and ii) ratings of and psychophysiological responses to emotional images in adults with a range of mood symptomatology. Volunteers (N=38) were recruited with low, mild, moderate, and severe scores on the Depression-Anxiety-Stress Scale (DASS). They attended two laboratory sessions during which they received either placebo or 0.2mg sublingual buprenorphine in randomized order under double-blind conditions. During peak drug effect, participants completed a visual attention task assessing responses to emotional faces and a picture-rating task assessing responses to emotional images with and without social content.

Results: Buprenorphine reduced attention to fearful facial expressions as it did in our previous study, and the emotion-specific effect was especially pronounced in individuals with high DASS scores. The drug also increased ratings of positivity of images with social content, but this effect was less strong in individuals with higher DASS scores.

Conclusions: These results suggest low doses of buprenorphine may reduce some dimensions of responses to negative emotional stimuli in individuals high on depression or anxiety, while leaving other dimensions unaffected.

Keywords: Buprenorphine, Psychophysiology, Opioids, Depression, Anxiety

Disclosure: Nothing to Disclose.

W95. Rapastinel, a Rapid-Acting Antidepressant, Does Not Produce Adverse Effects on Sleep: An Electroencephalographic Sleep-Wake Cycle Study in Rodents

Pradeep Banerjee*, Elise Esneault, John Donello

Allergan, Inc., Jersey City, New Jersey, United States

Background: Major depressive disorder (MDD) is known to be associated with abnormal sleep quality, such as decreased slow-wave sleep (SWS) and dysregulated rapid eye movement (REM) sleep. Most antidepressants, including selective serotonin reuptake inhibitors (SSRIs), disrupt REM sleep and SWS. Suppression of REM sleep by SSRIs is believed to result from increased synaptic levels of serotonin. SWS is a well-characterized physiological indictor of sleep quality. Increased time spent in SWS reflects enhanced synaptic plasticity and is considered an electrophysiological correlate of improved mood. Rapastinel (GLYX-13, AGN-241659), an N-methyl-D-aspartate receptor (NMDAR) modulator, produces rapid and sustained antidepressant effect in rodents and patients with MDD. Brain microdialysis studies in rodents indicate that rapastinel does not elevate extracellular levels of serotonin. Electrophysiological studies in rodents further suggest that rapastinel enhances cortical and hippocampal long-term potentiation, a cellular correlate of synaptic plasticity, which may underlie its antidepressant effects. In the present study, we evaluated the effects of a single antidepressant dose of rapastinel on sleep-wake cycle in male rats. The effects of rapastinel were compared with that of ketamine, an NMDAR antagonist with known rapid acting antidepressant effects in rodents as well as in humans.

Methods: Rats were surgically implanted with two surface electrodes in frontoparietal cortex and two depth electrodes in the CA1 region of hippocampus (AP +5.0 mm, L ±2.5 mm, V +7.0 mm) under anesthesia. Following recovery from surgery, telemetric transmitters were connected to the implanted electrodes for electroencephalographic (EEG) recordings. On the day of experiment, rats were injected at the beginning of the light phase with rapastinel (3, 10, or 30 mg/kg IV), ketamine (30 mg/kg, IP), or the positive control, zolpidem (12 mg/kg, PO), using a treatment schedule based on a cross-over design (n=9-10/dose/group). Sleep-wake activity was measured continuously following rapastinel treatment for the next 23 hours. Telemetric outputs were digitized online and analyzed using EMKA Technologies software. Physical activity of the rat was evaluated by integration of head movements via telemetric transmitter. Sleep-wake cycles were analyzed based on a spectral power analysis. Latencies to first occurrence of SWS and REM sleep were calculated for the total recording sessions (23 hours), the half cycles (0-12 and 12-23 hours), and 5 time segments (0-4, 4-8, 8-12, 12-18, and 18-23 hours), as were percent time spent in SWS, REM sleep, and wakefulness. Depth of SWS (defined as the ratio between alpha and delta spectral power) was also calculated. Differences between vehicle and test drugs were calculated for all variables, for each animal, and for the means of the animals per treatment group, and analyzed using a paired t-test.

Results: Assessment of sleep latency paradigms indicated that rapastinel (3 mg/kg) slightly but significantly decreased latency to SWS onset, with no effect on latency to REM sleep onset or number of REM periods. Higher doses of rapastinel did not exhibit these effects. By contrast, ketamine (30 mg/kg) significantly increased latencies to SWS and REM sleep onset without affecting the number of REM periods. Zolpidem (12 mg/kg), a sedative-hypnotic, significantly decreased latency to SWS onset and increased latency to REM sleep onset without affecting the number of REM periods. Assessment of sleep architecture revealed that rapastinel (3, 10, or 30 mg/kg) had no effect on the duration of waking, REM sleep, or SWS; the percent of active waking; or the alpha/delta ratio between 0 and 23 hours post-dose, with the exception of a small but significant decrease in percent of active waking between 8 and 12 and between 18 and 23 hours post-dose at 30 mg/kg. However, ketamine increased the duration of waking and decreased duration of REM sleep and SWS with a concomitant decrease in the alpha/delta ratio 0-4 hours post-dose. This was followed by a compensatory increase in sleep 12-23 hours post-dose. Zolpidem significantly decreased active waking and the duration of REM sleep during the first 12 hours and increased alpha/delta ratio 12-23 hours post-dose.

Conclusions: These results suggest that rapastinel does not adversely affect sleep latencies or sleep architecture in the rat over the dose range of 3 to 30 mg/kg. In contrast, ketamine displays several adverse effects on sleep within the first 4 hours of administration, including marked reduction in latencies to SWS and REM sleep onset and a pronounced wakening effect within the first few hours after administration with a concomitant decrease in alpha/delta ratio, suggesting a deeper sleep albeit with a shorter duration. Overall, these results suggest that rapastinel represents a novel rapidly acting antidepressant that may lack the sleep-disrupting properties of ketamine.

Keywords: Major Depressive Disorder (MDD), Slow-Wave Sleep, REM Sleep

Disclosure: Part 5: Allergan, Employee.

W96. Circadian Rhythm Abnormalities in the Amygdala of Subjects With Bipolar Disorder: Altered Period 2 Expression, and Relationship to Lithium Treatment

Harry Pantazopoulos*, Sabina Berretta

Harvard Medical School, Belmont, Massachusetts, United States

Background: Sleep and circadian rhythm dysfunction is commonly associated with mood disorders. Several genetic studies, including GWAS, have shown gene variants for several clock molecules are associated with bipolar disorder (BD) and lithium responsiveness. Notably, lithium and valproic acid directly modulate expression of core clock molecules and circadian period in cultured cells. Consistent with these findings, sleep dysfunction and weaker physical activity rhythms, characterized by reduced amplitude, have been reported in BD, suggesting that genetic vulnerabilities affecting clock molecules may in fact have an impact on circadian rhythms in BD. Remarkably, anxiety and depression in subjects with BD commonly peak in the morning, consistent with a circadian dysregulation of factors impacting these symptoms. Recent findings from our group support this possibility, showing that somatostatin (SST) expression in the healthy human amygdala varies according to circadian rhythms and is disrupted in BD. Decreased SST expression in the amygdala was selectively detected in the morning. Given the anxiolytic effects of SST in the amygdala, we speculate that this decrease may coincide with increased morning severity of anxiety and depression. What may be the molecular mechanisms regulating circadian SST expression? Circadian rhythms in individual cells are regulated by a set of core clock molecules, including Period 2 (Per2). Several studies in rodents have reported rhythms of Per2 protein expression in the amygdala. Furthermore, lithium treatment has been shown to increase expression of Per2 in rodent brain samples. Together, these considerations raise the possibility that altered rhythmic expression of Per2 may contribute to a disruption of SST rhythms in the amygdala of subjects with BD. With these studies we tested the hypothesis that the rhythmic expression of Per2 is altered in the amygdala of subjects with BD, follows a pattern similar to that observed for SST, and is affected by lithium.

Methods: We used Western blot analysis on a separate cohort of 14 subjects with BD and 14 control subjects to test the hypothesis that Per2 expression is altered in the amygdala of subjects with BD. Time of death was used as a proxy for circadian time, as previously published by our group and others to test the hypothesis that the circadian rhythm of Per2 expression is altered in subjects with BD.

Results: Per2 expression was significantly decreased in the amygdala of subjects with BD (F=-2.57, p <0.02), with a significant effect of lifetime exposure to lithium (F=2.17, p =0.04). Lithium exposure was positively correlated with Per2 expression (R square=0.37, p =0.04). Quartic regression analysis of Per2 expression plotted by time of death demonstrated a circadian rhythm of Per2 expression in the amygdala of control subjects, with a peak of expression at 10 am, and a low point of expression at 8 PM. In comparison, subjects with BD had a peak level of expression at 8 PM, and a low point of expression at 10 AM. This altered expression pattern is identical to our previously reported rhythm of somatostatin in the amygdala of subjects with BD.

Conclusions: Our results show that the rhythmic expression of Per2, a key clock molecule regulating circadian rhythms, is altered in the amygdala of subjects with BD. These findings support the hypothesis that molecular clock rhythms are disrupted in this brain region in BD, and indicate that lithium exposure may ameliorate decreases of Per2 expression, as also suggested by reports that lithium increases Per2 expression in rodent tissues and cultured human cells. Parallel Per2 and SST rhythmic patterns in subjects with BD suggests that altered clock rhythms may drive SST expression rhythms. Ongoing studies are testing whether glucocorticoid signaling rhythms may play a role in these mechanisms.

Keywords: Circadian Rhythm, Bipolar Disorder, Lithium, Amygdala, Postmortem Brain Tissue

Disclosure: Nothing to Disclose.

W97. Disrupted Resting-State Functional Connectivity in Medication-Free Women With Postpartum Depression and Anxiety

Kristina Deligiannidis*, Christina Fales, Vanessa Villamarin, Nina Jaitly, Janet Hall, Peter Schmidt, Blaise Frederick, Anthony Rothschild, Constance Moore

Zucker Hillside Hospital, Feinstein Institute for Medical Research, Glen Oaks, New York, United States

Background: Peripartum depression (PPD) commonly affects approximately 1 in 8 women, with undetected or undertreated illness with significant negative effects on the woman, her child and family. Antepartum depressive or anxiety symptoms are a risk factor for PPD and may represent an early manifestation of the disorder which can have a heterogeneous clinical presentation. The pathophysiology of PPD is unknown however limited data, mainly from task-based fMRI, indicate that PPD is associated with changes in functional connectivity (FC) involving the default mode, salience and central executive networks. This study examined resting state (rs) FC within the default mode network (DMN) in postpartum women with PPD and anxiety as compared to healthy postpartum women.

Methods: A prospective observational cohort study evaluated 88 women of 18-40 yrs. of age twice between 26-36 weeks gestation and thrice up to 8 weeks postpartum. Serial depression and anxiety assessments were completed at each of 5 study visits. Women were excluded for: multiple gestation pregnancy, lifetime history of manic episode or any psychotic disorder, elevated suicidal risk, alcohol, tobacco or substance abuse/dependence in the 6 months prior to study entry or use during the study, significant current medical illness, concomitant use of any pharmacotherapy with psychotropic activity, contraindication to MRI and positive urine pregnancy test at time of MRI. A subgroup of 49 postpartum women [healthy comparison (HCS) (n=27, mean age: 28.7±4.8) and medication-free subjects who developed unipolar PPD (PPD) (n=22, mean age: 29.0±4.6)] completed a single postpartum fMRI scan. Scans were completed at: 38.8±18.1 and 29.9± 13.4 postpartum days for HCS and PPD respectively. Women were eligible for a postpartum fMRI scan if they either: (1) remained euthymic throughout the peripartum period (i.e. HCS) or (2) developed minor or major depression with peripartum onset (i.e. PPD), based on study entry and exit diagnostic SCID interview. Thus, 39/88 women did not meet criteria for postpartum MRI due to symptoms less severe than a SCID diagnosis of minor/major depressive episode, study withdrawal due to severe depressive symptoms requiring urgent pharmacotherapy or subject preference not to undergo MRI.

Data were acquired on a 3.0 T Philips Achieva whole-body MR system (Philips Healthcare, Best, the Netherlands) with an 8 element phased-array head coil. T1-weighted anatomical MRI (MPRAGE sequence, 256 × 252 voxels, TR: 6.76 ms, TE: 3.1 ms, FOV: 244 mm × 256 mm × 204 mm, 170 slices, 560 s) were collected. All of the subjects underwent the rs MRI scan with open eyes and were instructed to attend to a static frame projected onto the screen which was visible through a mirror mounted on the head coil. The static frame contained a white plus-sign superimposed on the middle of a black background. Resting-state scan images were obtained using an EPI sequence (84 × 81 voxels, FOV: 256 mm × 256 mm, TR: 2500 ms, TE: 30 ms, flip angle: 75°, slice thickness: 3 mm, 50 slices). To examine group differences in rsFC, we conducted a seed-based analysis using the FMRIB Software Library (FSL) (www.fmrib.ox.ac.uk/fsl). As the seed region, we used a map of the DMN generated by ICA from an unrelated pool of healthy control subjects. During preprocessing, resting-state scans with excessive movement were scrubbed from the data and global signal regression was conducted. To analyze results, a cluster-forming threshold of z=2.81 was used, with random-field theory (RFT) thresholding of clusters at p<0.05. We performed univariate ANOVA to compare rsFC between HCS (n=27) and women who developed PPD (n=22) as well as demographic and clinical measures between groups.

Results: rsFC analysis showed that women with PPD had more positive connectivity between a region of dorsal anterior cingulate cortex (D-ACC, with peaks at MNI coordinates (-4, 32, 46) and (-4, 14, 52)) and the DMN than healthy postpartum women. There were no significant differences between cohorts in age or days since delivery to the day of MRI. At the time of MRI, 86.4% (19/22) of women with PPD and 85.2% (23/27) of HCS were fully or partially breastfeeding. 86.4% (19/22) of women with PPD had a history of a depressive disorder diagnosed by SCID compared to none in the HCS group. 81.8% of women with PPD met SCID criteria for an anxiety disorder at time of MRI. Women with PPD had higher mean HAM-D17 and STAI-S total scores [(PPD: 16.18±6.48, HCS: 3.26±3.13 (p<0.0001); and PPD: 47.73±11.38, HCS: 25.56±7.94 (p<0.0001)], respectively) at time of MRI.

Conclusions: The present report is the largest rsFC investigation in healthy postpartum women and women with PPD. Anxiety disorders were highly co-morbid in this sample of women with PPD. In the early postpartum period, rsFC patterns are disrupted in women with PPD. Given the known role of D-ACC in vigilance and conflict-monitoring, this result suggests women with PPD and co-morbid anxiety may have an increased awareness to threats in their environment which may contribute to the negativity bias seen in patients with depression. Additionally, as the D-ACC has extensive connections with the lateral prefrontal cortex, future studies should examine cognitive and executive functioning in women with PPD and anxiety.

Keywords: Postpartum Depression, Resting State Functional Connectivity, HAM-D, Postpartum Anxiety, STAI

Disclosure: Nothing to Disclose.

W98. Gonadal Steroids Trigger Neural Reward Circuit Dysregulation and Depression in Hormone Sensitive Women

Crystal Schiller*, Aysenil Belger, Sarah Slightom, Erin Richardson, Joshua Bizzell, Peter Schmidt, Gabriel Dichter, David Rubinow

University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States

Background: Neuroendocrine factors are purported to play a role in the etiology of postpartum depression (PPD), but direct evidence for this role is lacking. We investigated the effects of changes in gonadal steroid levels on brain function and mood by simulating two hormonal states related to pregnancy and parturition, respectively, in non-pregnant, euthymic women with a history of PPD and in those without such a history. By using this experimental pharmaco-fMRI paradigm, we are able to experimentally select for a relatively “pure” hormone sensitive phenotype, which we operationally define as at least a 30% increase in mood symptoms during the hormone challenge in order to examine the underlying neural mechanisms of behavioral sensitivity to hormones.

Methods: Participants included women with a history of PPD (hxPPD; n=15) and those without such a history (controls; n=15). The supraphysiologic gonadal steroid levels of pregnancy and withdrawal from these high levels to a hypogonadal state were simulated by inducing hypogonadism with the gonadotropin-releasing hormone agonist leuprolide acetate, adding back supraphysiologic doses of estradiol and progesterone for 8 weeks (“addback”), and then withdrawing both steroids under double-blind conditions (“withdrawal”). FMRI sessions were conducted at the baseline session, which took place during the early- to mid-follicular phase, and during withdrawal (two weeks after the last hormone dose). Participants performed a monetary incentive delay task (MID), to probe the neural circuit implicated in reward processing. ROI analyses were conducted using fsl to examine group differences in percent signal change in the BOLD response during the anticipation of reward. Group-level activation maps were thresholded using a z score of 2.3 (p<.01) to define contiguous clusters of activation. The Inventory of Depression and Anxiety Symptoms (IDAS) was used to assess self-reported mood during each hormone state (baseline, addback, and withdrawal). Repeated measures ANOVA was used to examine effects of group and hormone state on mood.

Results: There were significant group x hormone state interaction effects on the IDAS dysphoria, ill temper, panic, and wellbeing scales (p's<.05). hxPPD reported increased mood symptoms during hormone addback and withdrawal (compared with baseline), whereas controls showed no change in mood symptoms over time. Ten of the 15 hxPPD showed a 30% increase in mood symptoms during hormone addback or withdrawal and were therefore classified as hormone sensitive. One of the 15 controls was hormone sensitive, and her data were excluded from fMRI analyses. In both hormone sensitive women (n=10) and controls (n=14), the MID task activated key nodes of the reward circuit, including the thalamus, caudate, putamen, and nucleus accumbens. At baseline, hormone sensitive women showed less activation of the posterior cingulate and hippocampus than controls during reward anticipation (p's<.01). There also were significant group x hormone state interaction effects in the thalamus, putamen, hippocampus, and insula (p's<.01), such that hormone sensitive women showed decreased activation during hormone withdrawal relative to baseline (p's<.05), whereas controls showed no change with hormone withdrawal.

Conclusions: Our data provide direct, experimental evidence for the role (and potential mediating mechanisms) of the gonadal steroids estradiol and progesterone in the development of PPD in a subgroup of hormone sensitive women. Reproductive hormone changes were associated with dysregulation of the neural circuit underlying reward processing and consequent depressive symptoms in hormone sensitive women but not controls. Thus, the reward circuit may underlie both the susceptibility to and mediation of hormone-related affective dysfunction. Understanding these neurobiological mechanisms will subsequently improve our ability to identify those at risk for PPD.

Keywords: Estradiol, Progesterone, Depression, Functional Neuroimaging, Reward

Disclosure: Nothing to Disclose.

W99. NYX-783 is a Novel NMDA Receptor Modulator With Therapeutic Potential for the Treatment of Post-Traumatic Stress Disorder (PTSD)

Jeffrey Burgdorf*, Elizabeth Colechio, Tushar Bhattacharya, Jacqueline Dunning, Amanda Gross, Jessica Priebe, M Amin Khan, Patric Stanton, Xiao-lei Zhang, Cassia Cearley, Roger Kroes, Joseph Moskal

Northwestern University, Evanston, Illinois, United States

Background: Aptinyx has developed a novel class of small molecule NMDA receptor (NMDAR) modulators with broad applicability across neurologic and psychiatric diseases including post-traumatic stress disorder (PTSD). NYX-783 is a spiro-β-lactam NMDAR modulator and is distinct from known NMDAR agonists or antagonists such as d-cycloserine, ketamine, MK-801, kynurenic acid, or ifenprodil. In recombinant human NMDAR2A-2D-expressing HEK cells, partial agonist activity of NYX-783 was demonstrated at all four receptor subtypes. NYX-783 exhibits high oral bioavailability (94%F), high brain penetration (0.11 CSF / plasma ratio), and has shown efficacy after oral administration in several animal models of psychiatric disease without demonstrating side effects. Here, the in vitro and in vivo pharmacological properties of NYX-783 are further characterized.

Methods: NMDAR current: The pharmacologically-isolated NMDAR component of Schaffer collateral-evoked EPSCs was measured in CA1 pyramidal neurons in slices. Metaplasticity: Hippocampal slices were prepared from rats 24 hours after PO dosing and long-term potentiation (LTP) was measured at Schaffer collateral-CA1 synapses after three sub-maximal high-frequency stimulus trains (2 × 100 Hz/800 ms). Forced swim test: Rats were trained during a 15-min swimming session. Twenty-four hours later, the rats were returned to the swim cylinders for a video-recorded 5-min test. Male rats were tested 1 hour, 24 hours, 1 week, or 2 weeks post-dosing. Female rats were tested 1 hour or 1-week post-dosing. Floating time was manually scored. Positive emotional learning (PEL): Rats underwent a single 3-min trial (1 or 24 hours post-dose) that consisted of six 15-sec periods of tickling interleaved with six 15-sec periods during which the rat was left undisturbed. The number of frequency-modulated 50-kHz ultrasonic vocalizations emitted during the fifth and sixth undisturbed periods was used as an index of PEL. Fear conditioning: On the training day, rats were placed in a shock chamber for 7.5 minutes. Shocks (0.5 mA) were delivered for 1 sec at 90, 210, and 330 sec after the animal was placed into the chamber. For extinction trials, rats were placed in the chamber for 5 min. Extinction trial 1 began 1-hour post-dosing with NYX-783, d-cycloserine, or vehicle. Extinction trials occurred on six consecutive days and then again 14 days post-training. Time spent freezing was quantified for the last 3 min of each extinction trial. Rota-Rod: Animals were tested 15, 30, 60, and 120 min post-dosing in the fixed speed version of the Rota-Rod test (16 RPM) for 300 sec and the latency to fall off the Rota-Rod was recorded. Rats received three habituation sessions before testing.

Results: NYX-783 (100 nM) enhanced NMDAR current in rat hippocampal slices. Single-dose ex vivo studies showed increased metaplasticity in a hippocampal LTP paradigm and structural plasticity 24 hours after administration (0.1 mg/kg PO). NYX-783 reduced floating time in the forced swim test across a wide dose range (10 ng/kg – 100 mg/kg PO) and with a long duration of effect, with similar results in males and females. NYX-783 also facilitated PEL at the 3 highest dose levels (1, 10, and 100 μg/kg PO). In a contextual fear conditioning assay, NYX-783 (1 mg/kg PO) both facilitated extinction and inhibited spontaneous fear recovery whereas d-cycloserine only facilitated extinction. NYX-783 did not show sedative/ataxic side effects (1-100 mg/kg PO) in the Rota-Rod test.

Conclusions: NYX-783 activates an NMDAR-mediated synaptic plasticity process and may have therapeutic potential for a variety of NMDAR-mediated CNS disorders including PTSD.

Keywords: NMDA Receptor, Synaptic Plasticity, Post-Traumatic Stress Disorder

Disclosure: Part 1: Aptinyx, Inc., Employee, Part 2: Aptinyx, Inc., Employee, Part 3: Aptinyx, Inc., Employee, Part 5: Aptinyx, Inc., Employee.

W100. Cortisol Trajectory, Melancholia, and Response to Electroconvulsive Therapy

Brian Mickey*, Yarden Ginsburg, Adam Sitzmann, Clara Grayhack, Srijan Sen, Clemens Kirschbaum, Daniel Maixner, James Abelson

University of Utah, Salt Lake City, Utah, United States

Background: To optimally match patients with the best treatment, clinicians need reliable predictors of response to antidepressant interventions. Electroconvulsive therapy (ECT) is a unique intervention often considered for patients with severe depression, but roughly one-third of individuals do not respond. Excessive hypothalamic-pituitary-adrenal (HPA) axis activity has previously been linked with depression, but reliable and predictive HPA biomarkers have not yet been identified. We tested the hypothesis that cortisol trajectory during the weeks leading up to ECT would predict subsequent clinical response.

Methods: Hair cortisol provides a unique window into the recent history of an individual's HPA function. Cortisol concentrations in each 1-cm segment of hair reflect the average cortisol level over a specified 4-week period. An individual's cortisol trajectory over recent weeks can therefore be reconstructed by measuring cortisol concentrations in sequential hair segments. We measured average cortisol concentrations over the previous 12 weeks using scalp hair sampled from 39 participants with treatment-resistant depression just before ECT treatment.

Results: In contrast to ECT responders, non-responders showed a downward cortisol trajectory during the 8 weeks preceding ECT (p=0.004, group-by-time interaction). Analyses of clinical features revealed associations between cortisol trajectory, melancholia, and global severity, but cortisol trajectory provided information about subsequent ECT response that was independent of clinical features. A classification algorithm showed that cortisol trajectory predicted ECT response with 82% accuracy (sensitivity 88%, specificity 73%) suggesting that this biomarker might be developed into a clinically useful test for ECT-responsive depression.

Conclusions: We conclude that cortisol trajectory varies among patients with depression, maps onto symptoms of melancholia, and predicts response to ECT independently of clinical features. Our findings raise the possibility that ECT-nonresponsive depression represents a distinct form of illness involving declining HPA function, and they highlight the potential utility of cortisol trajectory as a novel biomarker for psychiatric disorders.

Keywords: Translational Biomarker Development, Treatment Resistant Depression, Hypothalamic-Pituitary-Adrenal Axis, Cortisol, Electroconvulsive Therapy

Disclosure: Nothing to Disclose.

W101. Antidepressant Effectiveness Study in Major Depressive Disorder in STAR*D Patients

Wei Guo, Liping Hou, Augustus John Rush, Francis McMahon, Yin Yao*

National Institute of Mental Health, Bethesda, Maryland, United States

Background: Major depressive disorder (MDD) is one of the most common mental disorders in the United States, affecting about 6.7% of the population in a given year. The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study was designed to assess effectiveness of treatments in patients with nonpsychotic major depressive disorder. In the STAR*D data, four levels of the study were designed to assess the effectiveness of depression medication in patients with major depressive disorder. Antidepressant response has been shown as a polygenic trait and involves a large number of SNPs with small effect sizes, and approximately 42% of phenotypic variance in antidepressant response was explained by common SNPs across the human genome. To detect single-nucleotide polymorphisms (SNPs) associated with antidepressant response, we re-analyzed the antidepressant efficacy in MDD patients of STAR*D data using well-imputed dosage data.

Methods: The STAR*D study aimed at assessing the effectiveness of depression treatments in 4,041 outpatients with major depression, recruited from 18 primary care and 23 psychiatric clinical sites across the United States. Of the initial 4,041 participants, 1,165 were excluded because they either did not qualify having “at least moderate” depression or they chose not to participate for unknown reasons. In this study, we included individuals who completed the evaluation of their antidepressant efficacy at level one. To detail, a total 2,876 participants were treated with the antidepressant citalopram, a selective serotonin reuptake inhibitor (SSRI) from 2 to 14 weeks. And QIDS-SR16 was used as the primary measure to evaluate citalopram's efficacy. The following analytical approaches were conducted. First, the most stringent quality control (QC) was carried out for the STAR*D data at both SNP-level and sample-level before imputation. The STAR*D samples were previously genotyped using the Affy 5.0 platform. The imputation of genotypes was conducted using the IMPUTE2 software program using the 1000 Genomes Phase 3 dataset as the reference panel. Furthermore, we re-genotyped a subset of STAR*D-Omni1S samples by Illumina Omni1S BeadChip to inspect the imputation concordance within this replication samples to evaluate the quality of imputation. In addition, the heritability was estimated by the GCTA software. Finally, the whole genome association analysis (GWAS) was conducted in patients of European ancestry.

Results: When the imputed dosages of STAR*D data were compared with the STARD-Omni1S data, we observed that about 80% of the SNPs showed Pearson correlation >0.8 between the imputed dosage of STAR*D (Affy 5.0) data and the STARD-Omni1S genotype data.

Four individuals (0.6%) among 665 overlapped CEU samples showed Pearson correlation ≤0.8 between two data sets, and these individuals were thereby removed in the GWAS analysis. In addition, the mean correlation was decreased from the common to rare SNPs. The heritability was estimated as 52% of QIDS-SR16 on exit week using the genome data. The SNP rs17071498 (P =1.49 × 10-8) was found to be statistically significant in the GWAS study, and the SNP was associated with a eQTL that regulating the gene RP11-521J24.1 in GTEx Nerve-Tibial tissues. The gene RP11-521J24.1 was linked to both tibial nerve and cerebellar hemisphere and cerebellum brain regions.

Conclusions: In an effort to identify single-nucleotide polymorphisms (SNPs) associated with antidepressant response, we re-analyzed the antidepressant efficacy in MDD patients of STAR*D data using well-imputed dosage data. The most stringent quality control (QC) was carried out for the STAR*D data and then the imputation was conducted by utilizing 1000 Genomes data as a reference panel. Furthermore, we re-genotyped a subset of STAR*D samples by Illumina Omni1S BeadChip to inspect the imputation concordance within this replication samples. To conclude, our study revealed a SNP rs17071498 associated with an antidepressant effect at exit and showed some level significant enrichment for brain eQTLs in tibial nerve and cerebellum.

Keywords: STAR*D, GWAS, Heritability

Disclosure: Part 1: No, Employee, Part 2: No, Employee.

W102. Longitudinal Neurocognitive Effects of Combined Electroconvulsive Therapy and Pharmacotherapy in Geriatric Major Depressive Disorder

Sarah Lisanby, Shawn McClintock*, George Alexopoulos, Samuel Bailine, Elisabeth Bernhardt, Mimi Briggs, C. Munro Cullum, Zhi-De Deng, Emma Geduldig, Robert Greenberg, Mustafa Husain, Styliani Kaliora, Rebecca Knapp, Lauren Liebman, Vassilios Latoussakis, W. Vaughn McCall, Martina Mueller, Georgios Petrides, Joan Prudic, Peter Rosenquist, Matthew Rudorfer, Shirlene Sampson, Abeba Teklehaimanot, Kristen Tobias, Richard Weiner, Robert Young, Charles Kellner

University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: Major depressive disorder is a chronic neuropsychiatric disease that results in significant morbidity and mortality. The rate of MDD in elderly adults has been steadily increasing and there are limited antidepressant options to achieve and sustain efficacy. Electroconvulsive therapy (ECT) is an effective antidepressant treatment for MDD, particularly in elderly adults. ECT has been found to result in transient neurocognitive adverse effects, but relatively newer ECT parameters such as right unilateral (RUL) electrode configuration and ultrabrief pulse width (UB) help to minimize the impact on cognitive function. To date, there is limited longitudinal information regarding the neurocognitive outcomes of RUL-UB ECT in combination with pharmacotherapy particularly in geriatric depression. Thus, the purpose of this study was to determine the neurocognitive effects of a continuation course of right RUL-UB ECT, with administration based on presenting depressive symptoms, and combined with medications (venlafaxine and lithium) in elderly adults with MDD.

Methods: The Prolonging Remission in Depressed Elderly (PRIDE) study was a multicenter, randomized study of an individualized continuation ECT schedule combined with pharmacotherapy to enhance long-term outcomes in elderly adults with MDD. The study had two phases. In Phase 1, patients received acute ECT 3x weekly combined with venlafaxine. In Phase 2, those who remitted in Phase 1 were randomized to receive pharmacotherapy (venlafaxine and lithium) alone or the combined modalities (pharmacotherapy and continuation ECT). ECT parameters were standardized as: RUL electrode placement using a Somatics Thymatron System IV with an ultrabrief pulse width of 0.25ms and current of 0.89Amps or a MECTA SPECTRUM device with an ultrabrief pulse width of 0.3ms and current of 0.8Amps. Elderly adults (age>60) with MDD, based on semi-structured diagnostic interviews, were enrolled in the study. All participants provided written informed consent for this IRB approved investigation before completing study procedures. Specific neurocognitive instruments included the Autobiographical Memory Interview-Short Form (AMI-SF), California Verbal Learning and Memory Test-II (CVLT-II), Delis-Kaplan Executive Function System Verbal Fluency Test, Dementia Rating Scale-2nd Edition Initiation Perseveration Index, Stroop Color and Word Test, and the Trail Making Test. In Phase 2, the all neurocognitive measures were administered at baseline and at the 3 and 6-month follow up time points, while the CVLT-II and AMI-SF were administered every 4 weeks. The intention-to-treat (ITT) sample was used for primary neurocognitive outcome analyses. A mixed effect model (MEM) approach compared the longitudinal neurocognitive scores over the 6-month treatment period. The neurocognitive variables were used separately as the dependent variable with treatment condition, time, and timeXtreatment interaction as primary fixed independent variables. Covariates were included as needed. Random subject effects were incorporated using random intercepts and slopes. A difference in trajectories (rate of change) of neurocognitive variables over the full-time period for STABLE vs PHARM was indicated by a significant timeXtreatment interaction term in the MEM. The final time point (6-months) represented the cumulative effect of treatment on neurocognitive function and was considered the primary analysis time point. Effect sizes for neurocognitive scores were determined from the MEM as differences in adjusted least squares means, with corresponding 95% confidence intervals, at study end (6-months). All statistical tests were two-tailed with alpha=0.05.

Results: There were no significant differences between the two treatment groups with regard to outcome at the 6-month time point on neuropsychological measures of psychomotor (p=0.90) or verbal (p=0.31) processing speed, autobiographical memory consistency (p=0.85), short-term (p=0.83) and long-term (p=0.51) verbal recall of learned words, phonemic fluency (p=0.54), inhibition (p=0.84), and complex visual scanning and cognitive flexibility (p=0.28). Within each treatment group, with the exception of processing speed, there was statistically significant and qualitative improvement across most neuropsychological measures (p-values range from 0.02 to <0.0001).

Conclusions: To our knowledge, this is the first study to characterize the neurocognitive outcomes over a 6-month time period of depressive-symptom dose titrated RUL-UB ECT combined with pharmacotherapy, relative to pharmacotherapy alone in elderly adults with MDD. The study found no difference in neurocognitive outcome between ECT plus pharmacotherapy versus pharmacotherapy alone over the 6-month continuation phase. These findings are consistent with our prior research and provide new evidence for the safety of RUL-UB ECT parameters in combination with psychotropic medications. Interestingly, within each treatment condition, patients on average showed significant improvement in most neurocognitive functions, except processing speed. This substantiates prior research that suggested decreased processing speed is a trait marker of MDD. Future research is warranted to confirm the neurocognitive effects of combined ECT and pharmacotherapy over longer periods of time, their association with other outcomes, and elucidate underlying neural mechanisms.

Keywords: Electroconvulsive Therapy, Major Depressive Disorder (MDD), Neuropsychology, Geriatric Depression

Disclosure: Part 1: TMS Health Education, Honoraria, Pearson, Consultant, Part 4: National Institutes of Health, Grant.

W103. Pretreatment Functional Neuroimaging Correlates of Antidepressant Response to Ketamine

Jessica Reed*, Allison Nugent, Maura Furey, Joanna Szczepanik, Jennifer Evans, Carlos Zarate

National Institute of Mental Health, Bethesda, Maryland, United States

Background: In major depressive disorder (MDD) there are multiple treatment options, each of which may be effective for some patients but not for others. Predicting whether a specific treatment will be beneficial in reducing depressive symptoms for a particular patient is difficult, and identifying measurable biological markers, or biomarkers, that could be assessed in individuals to help identify an effective treatment would be valuable. Previous research has used neuroimaging measures at baseline to test for potential correlations with treatment response in depression. Neural biomarkers of response may be challenging to identify for most pharmacological treatments for MDD, due to the length of time it takes for many antidepressants to show effects (potentially several weeks). The rapid antidepressant effects of ketamine, a glutamatergic modulator, make this compound well-suited for investigating biomarkers that correlate with treatment response. We examined blood-oxygen-level dependent (BOLD) signal using functional MRI (fMRI) during an emotional processing task in patients with MDD to determine if pretreatment activation correlated with subsequent antidepressant response after ketamine.

Methods: Participants included 31 patients with MDD, ages 18 to 65, who took part in this study as part of a large double-blind placebo-controlled protocol on the mechanism of action of ketamine. Depressive symptoms were rated using the Montgomery-Asberg Depression Rating Scale (MADRS) before and after participants received an infusion of ketamine (0.5 mg/kg). At baseline, BOLD signal was measured using a 3T GE fMRI scanner during an emotional processing task. In this mixed block/event-related task, participants viewed faces with emotional expressions (angry, sad, neutral, and happy), randomly presented either right-side up or upside-down. Participants were instructed to press a button to judge the emotion in one block and to identify the gender in the other block. The fMRI data were processed and analyzed using AFNI. Preprocessing included despiking, aligning images, registering echo-planar to structural images, blurring to 6 mm, normalizing, and motion correction. First-level individual analyses were conducted using regressors for each stimulus and block type. Group analysis used a multivariate model with the task factors and with a quantitative variable for the percent change in MADRS score from baseline to one day after the ketamine infusion. Results were thresholded at p<0.05, family wise error corrected (with voxel-wise p<0.01).

Results: We found a significant correlation between BOLD signal during the emotional processing task at baseline and percent change in MADRS score after ketamine. This was found in a large cluster in the left insula, extending into left inferior and middle frontal gyri, as well as in a cluster in the right precuneus. For both regions, greater activation at baseline was associated with better antidepressant response to ketamine (decrease in MADRS score of depressive symptoms). This was found across the entire task, not related to any specific task factors. We did not find any brain areas that showed the opposite pattern of greater activation correlating with poorer response.

Conclusions: Our findings showed that pretreatment brain activity was correlated with response to ketamine, with the most significant result found in the left insula. The insula plays a role in emotional processing and has been previously associated with neuroimaging findings related to depression. While prior research has shown some associations between insula activation and response to other medications, there have been mixed findings. This could be related to different tasks being used or differential response depending on the specific treatment. Using functional neuroimaging as a biomarker to predict the likelihood of antidepressant response to ketamine would be greatly beneficial for patients with MDD. This would also be a valuable way for clinicians to work towards individualized treatment for psychiatric illness.

Keywords: Ketamine, Functional MRI (fMRI), Major Depressive Disorder (MDD), Emotional Processing, Antidepressant Response

Disclosure: Nothing to Disclose.

W104. Can the MADRS Measure Rapid Changes in Depressive Symptoms in Response to Esketamine Treatment?

C. Yavorsky, Jaskaran Singh, Nina Engelhardt*

Cronos CCS, Inc., Lambertville, New Jersey, United States

Background: Treatment-resistant depression (TRD), defined as a failure to adequately respond to two or more courses of antidepressant treatment at adequate dosage and duration, represents a serious and ongoing health problem globally. Studies evaluating compounds such as esketamine, an NMDA-receptor antagonist, have shown that esketamine can lift depression in hours, or even minutes in subjects with TRD – much faster than the most commonly used antidepressant medications now available, which often require weeks to take effect.¹ These studies typically use depression severity rating scales, such as the Montgomery-Asberg Depression Rating Scale,2 that were developed for assessment over weeks, and may have limited capabilities for capturing rapid antidepressant effects, especially for anti-suicidal effects.

Methods: We analyzed the behavior of the MADRS scale on total score and at individual item level at critical time points in a trial of esketamine (ESKETIVTRD2001). We also evaluated the rebound characteristics of esketamine as reflected in MADRS total and item scores. ESKETIVTRD2001 was a double-blind, double-randomization, placebo-controlled study of the efficacy of intravenous esketamine in adults with treatment-resistant depression. Treatment groups were placebo, 0.2 mg/kg esketamine, 0.4 mg/kg esketamine. Critical time points were defined as pre-dose and 2- and 4-hr post dose (visit 3), and visits 4, 5 and 6, which spanned three days.

Results: All three treatment groups showed an immediate reduction in mean score in the 2 and 4 hours following infusion, followed by some degree of rebound over the following 3 days. The high dose esketamine group had the largest change in mean MADRS total score. The placebo group showed a small initial reduction of ~15% after the 4-hour post infusion at visit 3, and this remained stable through visit 6. The largest overall improvements in scores from Visit 3 pre-dose to Visit 6 post dose were in Reduced Appetite (item 5) and Pessimistic Thoughts (item 9), with both the 0.2mg/kg group and 0.4mg/kg group showing >50% reductions. The items that had the least treatment response from Visit 3 pre-dose to Visit 6 post dose were Reduced Sleep (item 4) and Suicidal Thoughts (item 10).

Conclusions: Rapid onset of effect as well as rebound characteristics of esketamine are significantly different than what may be expected in clinical trials with more traditional antidepressants (e.g., SSRIs, SNRIs, MAOIs or atypicals). The ability of the MADRS to capture rapid change in mood symptoms appears to be at least as sensitive to overall depression severity as when administered at the more common interval of one week. Mood items appear to be more sensitive to change than both the Reduced Sleep and Suicide items. Further research should be conducted to determine the impact of this compound on suicidality, as well as the differential effect on MADRS items.

Keywords: Esketamine, MADRS, Depression

Disclosure: Nothing to Disclose.

W105. Microglial Activation in Affective Disorders

Jeffrey Meyer*, Elaine Setiawan, Sophia Attwells, Pablo M. Rusjan, Sylvain Houle, Alan Wilson, Romina Mizrahi, Peggy Richter, Steve Kish, Cynthia Xu, Laura Miler

Centre for Addiction and Mental Health, Toronto, Canada

Background: Major depressive disorder (MDD) frequently exhibits increasing persistence of major depressive episodes (MDE) that may become more treatment resistant. Microglial activation, an important component of neuroinflammation, is implicated in neuroprogression. Translator protein total distribution volume, TSPO VT, a marker of microglial activation, may be quantified with the new generation of TSPO PET radioligands such as [18F] FEPPA PET. With this approach, we previously showed that TSPO VT is elevated throughout the grey matter in unmedicated MDE. Here we present new directions of this technology into medication treated, treatment resistant MDD, and unmedicated OCD, a common comorbid illness in treatment resistant MDE.

Methods: 20 medication free MDE, 22 medication treated MDE, 20 medication free OCD and similarly numbered matching healthy controls underwent [18F]FEPPA PET scanning. Prioritized regions were the anterior cingulate cortex, insular cortex and prefrontal cortex in MDD and the dorsal caudate, orbitofrontal cortex, thalamus, ventral striatum, dorsal putamen and anterior cingulate cortex in OCD.

Results: In medication free MDE, TSPO VT was significantly elevated in all brain regions examined (analysis of variance, effect of diagnosis, p=0.001 to p=0.01). The magnitude of TSPO VT elevation was 26%, 32% and 33% in the prefrontal cortex, anterior cingulate cortex and insula respectively. In OCD TSPO VT was significantly elevated in the prioritized brain regions (mean 32%; range 31% to 36% except anterior cingulate cortex 24%; analysis of variance, effect of diagnosis, P<0.001 to 0.004). Slightly lower elevations in TSPO VT (22-29%) were present in other grey matter regions. Posthoc comparisons between these two groups suggest a differential ratio of ACC TSPO VT to caudate TSPO VT. Recruitment of medication treated, treatment resistant MDE subjects is ongoing and interim results show regional elevations of 20 to 25% (analysis of variance, effect of diagnosis, p<0.01 to 0.05).

Conclusions: Microglial activation is present throughout grey matter in a substantial proportion of MDD and OCD subjects, but not in completely identical distributions. Microglial activation may be interpreted as a marker of ongoing neuroinflammation and neuroprogression. Anti-inflammatory and immunomodulatory treatments should be considered for development for cases with prominent microglial activation. Further study is needed to improve identification of cases with more prominent microglial activation in clinical settings.

Keywords: Neuroinflammation, Translocator Protein (TSPO), Positron Emission Tomography Imaging, Microglial Activation

Disclosure: Part 1: Lundbeck-Takeda, Advisory Board, Part 4: Janssen, Grant.

W106. Lack of Antidepressant Effects of (2R,6R)-Hydroxynorketamine in a rat Learned Helplessness Model: Comparison With (R)-Ketamine and (R)-Norketamine

Yukihiko Shirayama*, Kenji Hashimoto

Teikyo University Chiba Medical Center, Ichihara, Japan

Background: The N-methyl-D-aspartate receptor (NMDAR) antagonist ketamine is one of the most attractive antidepressants since this drug can produce rapid-onset and sustained antidepressant effects in treatment-resistant patients with major depression and bipolar disorder. Ketamine (Ki=500 nM for NMDAR) is a racemic mixture containing equal parts of (S)-ketamine (esketamine: Ki=300 nM) and (R)-ketamine (Ki=1,400 nM). (R)-ketamine showed greater potency and longer lasting antidepressant effects than (S)-ketamine in the animal models of depression (Zhang et al, 2014; Yang et al, 2015; Zanos et al, 2016; Fukumoto et al, 2017). Unlike (S)-ketamine, (R)-ketamine does not induce psychotomimetic-like behavioral side effects in rodents (Yang et al, 2015; 2016). Taken together, it is likely that (R)-ketamine could be a safer antidepressant without psychotomimetic side effects in humans than (S)-ketamine (Hashimoto, 2016).

Zanos et al. (2016) reported that metabolism of ketamine to hydroxynorketamine (HNK)(Ki >10,000 nM for NMDAR) was essential for ketamine mediated antidepressant activity, suggesting that NMDAR inhibition may be irrelevant to ketamine's actions. (R)-ketamine is known to be metabolized to (R)-norketamine by P450 enzyme in the liver, subsequently to (2R,6R)-HNK. Recently, we reported that (R)-ketamine showed greater potency and longer lasting antidepressant effects than (2R,6R)-HNK in a social defeat stress model (Yang et al, 2017). The purpose of the present study is to compare the antidepressant effects of (R)-ketamine, (R)-norketamine and (2R,6R)-HNK in a rat learned helplessness (LH) model.

Methods: Male Sprague-Dawley rats (200-230 g, 7 weeks old; Charles-River Japan, Co., Tokyo, Japan) were used. To create an LH paradigm, the animals are initially exposed to uncontrollable stress. When the animal is later placed in a situation where the shock is controllable (escapable), the animal not only fails to acquire the escape response, but also often makes no efforts to escape the shock at all. The LH behavioral tests were performed using the Gemini Avoidance System (San Diego Instruments, San Diego, CA). This apparatus was divided into two compartments by a retractable door. On day 1 and 2, rats were subjected to 30 inescapable electric foot shock (0.65 mA, 30 sec duration, at random intervals (mean 30 sec, average 18-42 sec)). On day 3, a two-way conditioned avoidance test was performed as a post-shock test to determine if the rats would show the predicted escape deficits. This screening session consisted of 30 trials in which electric foot shocks (0.65 mA, 6 sec duration, at random intervals (mean 30 sec, average 18-42 sec)) were preceded by a 3 sec conditioned stimulus tone that remained on until the shock was terminated. Subsequently, saline (2 ml/kg), (R)-ketamine (20 mg/kg), (R)-norketamine (20 mg/kg), or (2R,6R)-HNK (20 and 40 mg/kg) was administered intraperitoneally into LH rats. On day 8, a two-way conditioned avoidance test was performed. This test session consisted of 30 trials in which electric foot shocks (0.65 mA, 30 sec duration, at random intervals (mean 30 sec, average 18-42 sec)) were preceded by a 3 sec conditioned stimulus tone that remained on until the shock was terminated. The numbers of escape failures and the latency to escape in each 30 trial were recorded by the Gemini Avoidance System.

Results: In the rat LH model of depression, (R)-ketamine (20 mg/kg) significantly attenuated the failure of LH and escape latency of LH, consistent with the previous report (Yang et al, 2015). In contrast, neither (R)-norketamine (20 mg/kg) nor (2R,6R)-HNK (20 and 40 mg/kg) did alter the failure of LH and escape latency of LH.

Conclusions: (R)-ketamine showed antidepressant effects in the LH rats, whereas its metabolites (R)-norketamine and (2R,6R)-HNK did not show antidepressant effects in the LH rats. Interestingly, we reported that bilateral infusion of (R)-ketamine into the infralimbic region of medial prefrontal cortex and hippocampus of LH rats showed antidepressant effects (Shirayama and Hashimoto, 2017), suggesting that (R)-ketamine itself can exert antidepressant effects. It is, therefore, unlikely that the metabolism of (R)-ketamine to (2R,6R)-HNK is necessary for antidepressant actions of (R)-ketamine.

Keywords: Ketamine, HNK, LH

Disclosure: Nothing to Disclose.

W107. Possible Role of the Gut-Microbiota-Brain Axis in the Antidepressant Effects of (R)-Ketamine in a Social Defeat Stress Model

Chun Yang, Youge Qu, Qian Ren, Min Ma, Dong Chao, Kenji Hashimoto*

Chiba University Center for Forensic Mental Health, Chiba, Japan

Background: The N-methyl-D-aspartate receptor (NMDAR) antagonist ketamine is one of the most attractive antidepressants since this drug can produce rapid-onset and sustained antidepressant effects in treatment-resistant patients with major depression and bipolar disorder. Ketamine is a racemic mixture containing equal parts of (S)-ketamine and (R)-ketamine. (R)-ketamine showed greater potency and longer lasting antidepressant effects than (S)-ketamine in the animal models of depression (Zhang et al, 2014; Yang et al, 2015; Zanos et al, 2016; Fukumoto et al, 2017). Unlike (S)-ketamine, (R)-ketamine does not induce psychotomimetic-like behavioral side effects in rodents (Yang et al, 2015; 2016). Taken together, it is likely that (R)-ketamine could be a safer antidepressant without psychotomimetic side effects in humans than (S)-ketamine (Hashimoto, 2016).

The gut–microbiota–brain axis is a complex multi-organ bidirectional signaling system between the microbiota and brain that plays a fundamental role in host physiology, homeostasis, development, and metabolism. Several studies suggest that the gut–microbiota contributes to the pathogenesis of depression and the antidepressant actions of certain compounds (Dinan and Cryan, 2016; Burokas et al, 2017; Yang et al, 2017; Zhang et al, 2017). Therefore, the present study examined whether the gut-–microbiota plays a role in the mechanisms underlying the antidepressant actions of (R)-ketamine and (S)-ketamine in a chronic social defeat stress (CSDS) model of depression.

Methods: Male adult C57BL/6 mice, aged 8 weeks (body weight 20-25g, Japan SLC, Inc., Hamamatsu, Japan), male CD1 mice, aged 14 weeks (body weight 40-45g, Japan SLC, Inc., Hamamatsu, Japan) were used in the experiments. Saline (10 ml/kg), (R)-ketamine (10 mg/kg) or (S)-ketamine (10 mg/kg) was administered intraperitoneally (i.p.) into the susceptible mice after CSDS. Saline (10 ml/kg) was also administered i.p. into control mice. Behavioral tests, including locomotion test (LMT), tail suspension test (TST), forced swimming test (FST) and 1% sucrose preference test (SPT), were performed as reported previously. The role of the gut–microbiota in the antidepressant effects of ketamine enantiomers in a CSDS model of depression was examined using 16S ribosomal RNA gene sequencing of fecal samples.

Results: At the phylum level, CSDS susceptible mice showed alterations in the levels of Tenericutes and Actinobacteria; however, neither ketamine enantiomers influenced these alterations. At the class level, both ketamine enantiomers significantly attenuated the increase in the levels of Deltaproteobacteria in the susceptible mice after CSDS. Furthermore, (R)-ketamine, but not (S)-ketamine, significantly attenuated the reduction in the levels of Mollicutes in the susceptible mice. At the genus level, both ketamine enantiomers significantly attenuated the decrease in the levels of Butyricimonas in the susceptible mice. Notably, (R)-ketamine was more potent than (S)-ketamine at reducing the levels of Butyricimonas in the susceptible mice.

Conclusions: This study suggests that the antidepressant effects of two enantiomers of ketamine in CSDS model may be partly mediated by the restoration of the gut– microbiota. Furthermore, the specific effect of (R)-ketamine on the levels of Mollicutes and Butyricimonas may explain its robust antidepressant action.

Keywords: Ketamine, Esketamine, Microbiota-Gut-Brain Axis, Antidepressant, Social Defeat Stress

Disclosure: Part 4: Otsuka, Grant, Taisho, Grant.

W108. Cortical Gaba Reduction Mediates Effort-Based Dopamine Release Deficits in Anterior Cingulate Cortex

Stefan Kolata, Kazuhito Nakao, Vivek Jeevakumar, Yuko Fujita, Kenji Hashimoto, Kazutoshi Nakazawa*

University of Alabama at Birmingham, Birmingham, Alabama, United States

Background: Corticolimbic GABAergic dysfunction is implicated in the etiology of schizophrenia, depression, and related neuropsychiatric diseases (Benes and Berretta, 2001; Lewis et al, 2005; Nakazawa et al, 2012). For example, the level of Gad1 mRNA encoding GAD67 (glutamic acid decarboxylase-67) falls below detectable levels in approximately 50% of parvalbumin (PV)-positive prefrontal cortex (PFC) interneurons in schizophrenia (Hashimoto et al, 2003). However, the impact of cortical GABA level reduction on the neuropsychiatric phenotypes has remained elusive.

The anterior cingulate cortex (ACC) is known to be critically involved in effort-based decision making (Bailey et al., 2016; Kurniawan et al, 2011; Miller et al, 2013; Salamone et al, 2016). For example, lesions or inactivation of the ACC results in animals avoiding previously preferred high-effort options, leading to a higher likelihood of choosing low effort options. In particular, ACC dopamine level appears to modulate the willingness of animals to exert physical effort, because dopamine D1 receptor blockade in the ACC resulted in effort avoidance during cost-benefit decision making (Aly-Mahmoud et al, 2017; Schweimer and Hauber, 2006). It has been unclear how ACC dopamine levels are regulated to execute the effort-based behavior. GABA level reduction in ACC or mPFC has been linked to schizophrenia (Egerton et al, 2017) and depression (Schur et al, 2016). We hypothesized that cortical GABA reduction including ACC elicits cortical dopamine abnormalities, leading to a subset of negative symptom-like phenotypes.

Methods: To explore the impact of cortical GABA reduction following Gad1 reduction on physiology and behavior, we genetically eliminated the Gad1 gene from approximately 50% of cortical and hippocampal interneurons; a majority (70%) of which are PV-positive. This was accomplished by crossing a loxP-flanked Gad1 mouse line (Chattopadhyaya et al, 2007) to a previously characterized Ppp1r2-Cre driver line (Belforte et al, 2010), in which Cre recombinase expression is largely confined to GABA neurons of the cortex and hippocampus. We first measured the tissue dopamine and its metabolites by HPLC, and then subjected the animals to a behavioral batter test. Since the immobility was more prominent in the Gad1 homozygous KO mutant mice compared to the floxed-littermate controls, we simultaneously measure the extracellular dopamine by in vivo microdialysis from awake behaving mice. Dialysis samples were analyzed for dopamine by HPLC. Data are given as mean±SEM. Statistical comparisons have been performed with the unpaired Student's t-test for tissue HPLC and repeated ANOVA for behavior and microdialysis.

Results: We confirmed by tissue HPLC that tissue GABA levels are reduced in medial prefrontal cortex (mPFC) including ACC (t-test, p=0.0013), frontal cortex (p=0.0004), and hippocampus (p=0.046), but not in striatum (p=0.91) or cerebellum (p=0.31) of the mutant mice at 11-13 weeks old. In the behavioral battery test, mutant mice differed significantly in the home-cage wheel-running behavior (genotype x time interaction: F(1,120)=2.68, p<0.001) and their peak wheel running activity was attenuated (post-hoc planned comparisons, F(1,15)=1.23, p<0.02). Male mutants also showed increased immobility duration in tail suspension [t(7)= -2.42, p=0.046)], and in forced swim chamber [t(14)= -2.77, p=0.015)]. However, we found no genotypic differences in their preference for saccharine sweetened water at 0.01% (t(22)=0.48, p=0.64) and 0.03% (t(16)=0.02, p=0.98) solution, nor in breakpoint achieved in the progressive ratio session (t(10)= -0.33, p=0.75) following the acquisition training in fixed ratio 1 (FR1) and FR5 schedule of reinforcement to obtain chocolate pellets. These results suggested that the male mutant mice are normal in hedonic and reinforcement behavior, but exhibit a strong effort avoidance behavior in risk-taking situations. Next, we measured extracellular levels of dopamine in the ACC before and after the tail suspension test by in vivo awake brain microdialysis. in vivo microdialysis revealed that control mice (15-20 weeks of age) augmented the dopamine level in ACC by 1.7-fold from the baseline by tail suspension for 6 min (paired t-test, t(10)= -7.99, p=1.18E-05), regardless of gender (male mice, t(4)= -9.02, p=0.0008; female mice, t(5)= -5.5, p=0.002). While the mutants also showed small dopamine release (paired t-test, t(5)= -3.49, p=0.017), robust difference in tail hanging-triggered dopamine release was observed between controls and mutants (repeated ANOVA, F(1,45)=2.44, p=0.07, post-hoc planned comparisons, at 0-20 min; F(1,15)=22.1, p=0.00028). Furthermore, we found a robust inverse correlation was observed between ACC dopamine increase and tail suspension-elicited immobility across genotypes (Pearson's correlation coefficient, r=-0.75, P=0.00053).

Conclusions: in vivo microdialysis revealed that tail suspension triggered the extracellular release of dopamine in ACC of the control mice. However, the mutant mice with increased immobility showed a robust deficit in effort-triggered dopamine release in AAC. These results suggest a novel link between cortical GABA reduction and impaired effortful behavior through ACC dopamine deficiency. These results also shed on the light in understanding a role of cortical GABA reduction in a subset of negative symptom-like phenotypes of major psychiatric disorders.

Keywords: Schizophrenia, Depression, Transgenic Mice, Effort-Cost Benefit Task, Microdialysis

Disclosure: Nothing to Disclose.

W109. Dopamine Release in Antidepressant-Naïve Major Depressive Disorder: A [11C]-(+)-PHNO Positron Emission Tomography Study

Franklin Schneier*, Mark Slifstein, Alexis Whitton, Diego Pizzagalli, Patrick McGrath, Dan Iosifescu, Anissa Abi-Dargham

The New York State Psychiatric Institute, New York, New York, United States

Background: Major depressive disorder (MDD) and anhedonia have been hypothesized to be associated with dysfunction of the mesolimbic dopamine (DA) system based on studies in animal models of depression and human studies of DA metabolites, DA depletion, activation of mesolimbic circuitry, and clinical response to treatment with DA agonists. However, prior molecular neuroimaging studies relying on nonselective D2/D3 receptor antagonist radioligands and samples heterogenous for antidepressant use have yielded mixed results for striatal receptor availability and DA release. [11C]-(+)-PHNO is an agonist radioligand with preferential selectivity for D3 over D2 receptors, and greater sensitivity to the DA-releasing effects of amphetamine than previously studied antagonist tracers, and it has not been previously utilized to study MDD. We hypothesized that DA release in the ventral striatum would be decreased in MDD, and investigated associations of receptor availability and DA release with anhedonia and clinical response to treatment with the D3-preferring DA agonist pramipexole.

Methods: Twenty adults aged 18-50, with MDD and no history of medication treatment or substance use disorder, and 20 age- and sex-matched healthy comparison subjects participated. All completed positron emission tomography (PET) imaging with [11C]-(+)-PHNO before and after oral administration of dextroamphetamine 0.5 mg/kg. MDD participants were subsequently treated for 6 weeks with open label pramipexole, 0.5-2.5 mg/day. Assessments of BOLD signal with fMRI and behavioral responses to probabilistic reward tasks were conducted in the same sample and are being reported separately. For PET imaging, reference tissue-based kinetic modeling (SRTM) was performed with cerebellum as the reference tissue. Regions of interest (ROIs) including midbrain, thalamus, globus pallidus, pre-commissural dorsal caudate, post-commissural caudate, pre-commissural dorsal putamen, post-commissural putamen, and ventral striatum were drawn on each subject's MRI and transferred to the coregistered PET data. Outcome measures of binding potential relative to non-displaceable compartment (BPND) and percent reduction from baseline (ΔBPND) at the post-amphetamine scan were compared between groups and correlated with clinical features within the MDD group.

Results: MDD and control groups did not differ in mean injected [11C]-(+)-PHNO radioactivity, injected PHNO mass per kg of body weight, regional volumes or plasma amphetamine levels. Baseline BPND and ΔBPND did not differ significantly between groups across all ROIs in a mixed model analysis covarying for age (BPND: F(1, 36.172)=0.501, p =.484, ΔBPND: F(1, 35.702)=1.783, p =.190) or for any single ROI. There were trends for greater DA release in the MDD group in the ventral striatum (-34% vs. -30%, p =.072, d =.58) and the globus pallidus (-27% vs. -22%, p =.096, d =.58). Baseline clinical features in the MDD group, including severity of depression and anhedonia, were not significantly associated with BPND and ΔBPND. The MDD group evidenced significant improvement with pramipexole treatment on the Hamilton Depression Scale (week 0 mean 20.23±2.47 vs. week 6 mean 8.36±5.40, p <.001) and on the Anhedonic Depression subscale of the Mood and Anxiety Symptom Questionnaire (week 0 mean 82.48±12.08 vs. week 6 mean 59.57±18.52, p <.001), but BPND and ΔBPND did not significantly predict clinical response. Exploratory analyses of clinical associations with the ratio of BPND in D2- vs. D3-rich ROIs will also be reported.

Conclusions: This study did not find evidence associating MDD with significant differences in D2/D3 receptor binding or DA release, as measured by [11C]-(+)-PHNO PET with amphetamine. Nor did PET outcomes show significant association with clinical features of the MDD group, including anhedonia and clinical response to pramipexole treatment. These results contrast with evidence (presented separately) from the same samples showing group differences in striatal BOLD response to prediction error and behavioral differences in reward learning, both of which have been previously associated with dopaminergic function. Limitations of this study include the small sample size, multiple factors potentially contributing to BPND (including receptor number, occupancy by endogenous DA, and affinity state), and the limitations of amphetamine induced DA release as a model for physiological DA release. Additionally, specific effects of D2 and D3 receptors could be masked by their combined contributions to PHNO binding in key regions such as ventral striatum.

Keywords: PET Imaging, Dopamine 3 Receptor Imaging, Major Depression Disorder, Dopamine Release, Pramipexole

Disclosure: Part 1: Forest Laboratories, Grant, Part 4: Forest Laboratories, Grant.

W110. Accelerated Epigenetic Aging and Mitochondrial DNA Copy Number in Bipolar Disorder

Gabriel Fries*, Isabelle Bauer, Giselli Scaini, Mon-Ju Wu, Iram Kazimi, Giovana Zunta-Soares, Consuelo Walss-Bass, Jair Soares, Joao Quevedo

University of Texas Health Science Center at Houston, Houston, Texas, United States

Background: Accelerated biological aging is commonly observed in chronic illnesses, increasing the rate of aging-related medical conditions and shortening lifespan in patients. In particular, bipolar disorder (BD) has been previously associated with accelerated aging, yet the mechanisms underlying this association are largely unknown. As a recently developed investigational tool, the DNA methylation-based epigenetic age has been increasingly recognized as a valuable aging marker, although its association with other biological clocks in BD patients and high-risk subjects, such as telomere length and mitochondrial DNA (mtDNA) copy number, has never been investigated. We analyzed these three aging markers in blood from BD patients, siblings of BD patients, and healthy controls, with the ultimate goal of identifying the mechanisms involved in the accelerated aging observed in BD.

Methods: We included 22 patients with BD type I (7 males, 33.9±9.3 years old), 16 siblings of BD patients (6 males, 39±10.6 years old), and 20 healthy controls (8 males, 34.75±10 years old) in this analysis, all matched by age, sex, ethnicity, race, and blood cell count. DNA samples were isolated from peripheral blood and interrogated for genome-wide DNA methylation patterns with the Infinium HumanMethylation450 BeadChip Kit (Illumina). DNA methylation age (DNAm age) was calculated using the Horvath age estimation algorithm, which predicts DNAm age based on the methylation levels of 353 CpGs from the microarray. Accelerated epigenetic aging was calculated by regressing DNAm age on chronological age, forming residuals that were compared between groups. Based on reported cumulative effects of lifetime stress on epigenetic aging, we also investigated whether a difference would be detected when limiting our analysis specifically to older subjects. mtDNA copy number and telomere length were assessed by real-time quantitative PCR.

Results: There was a strong positive correlation between individuals’ DNAm age and chronological age (r=.944, p<.001). When analyzing the entire sample, accelerated epigenetic aging residuals did not differ between groups (p>.05). Older BD patients and siblings (>33 years old, median age in our sample) presented significantly different accelerated epigenetic aging compared to controls (F(2,25)=7.24, p=.003), whereas no difference was detected among the younger subjects (p>.05). BD patients also showed higher levels of mtDNA copy number compared to controls (U=105, p=.01), while no difference was found between controls and siblings (p=.205). mtDNA significantly correlated with epigenetic age acceleration among older subjects (r=.697, p<.001), as well and with global functioning (r=-.364, p=.013) in our sample. Telomere length did not show significant differences between groups (p=.891), nor did it correlate with epigenetic aging or mtDNA copy number, even after splitting the sample into younger and older subjects.

Conclusions: These results suggest that BD and its risk involve an accelerated epigenetic aging, which might represent a novel target for treating BD and subjects at risk. In particular, epigenetic aging in BD seems to be independent of telomere shortening and related to an increase in the mtDNA copy number, suggesting a complex interplay between biological clocks to determine the accelerated aging and its consequences in BD.

Keywords: Bipolar Disorder, Aging, DNA Methylation, Mitochondrial DNA, Leukocyte Telomere Length

Disclosure: Nothing to Disclose.

W111. fMRI Correlates of Lithium Response: A Pilot Study in Bipolar Disorder

Kelly Rootes-Murdy, Kara Glazer, Fernando Goes, Francis Mondimore, Peter Zandi, J. Raymond DePaulo, Jr., Arnold Bakker, Pamela Mahon*

Brigham & Women's Hospital, Harvard Medical School, Boston, Massachusetts, United States

Background: Bipolar disorder is largely treatable and lithium is commonly used as a first-line treatment. However, the process of finding the right medication for a given patient can take months to years as there is currently no identified biomarker that can reliably predict patient response to lithium. The goal of this project was to identify brain features associated with lithium response in patients with bipolar disorder using task-based and resting state functional magnetic resonance imaging (fMRI).

Methods: Participants with bipolar I disorder (N=12) were stabilized on lithium monotherapy and followed for up to two years to make a clinical determination of response to treatment, after which they underwent fMRI. Healthy comparison participants (N=21) were group matched to cases on sex, age and education. All participants were scanned at 7.0 Tesla using EPI BOLD sequences at rest and during an emotional faces task. Activation between groups was compared in at the whole-brain level. Resting state connectivity was analyzed at the whole-brain level using four regions-of-interest as seeds.

Results: Four bipolar participants were classified as non-responders and 8 responders. Results of the whole-brain analysis suggest that lithium non-responders exhibited greater activation to fearful faces in the parahippocampal gyrus (peak T=12.68), whereas lithium responders showed greater activity in prefrontal cortex (peak T=7.48). Lithium non-responders also displayed altered resting state connectivity between hippocampus, ACC and superior temporal lobe compared to responders.

Conclusions: We identified differences in activation and connectivity related to lithium response in regions previously implicated in bipolar disorder pathophysiology. These preliminary findings have several potential implications including contributing to our understanding of bipolar disorder pathophysiology, shedding light on the potential mechanism of action of lithium, and identification of candidate brain features with potential to predict response to lithium treatment.

Keywords: Lithium Response, Bipolar Disorder, Functional MRI (fMRI)

Disclosure: Nothing to Disclose.

W112. Alterations of the Gut Microbiome in Severe Mental Illness: An Analysis of Data From the American Gut Project

Tanya Nguyen*, Max Abramson, Tomasz Kosciolek, Rebecca Daly, Daniel McDonald, Lisa Eyler, Rob Knight, Dilip Jeste

University of California, San Diego, California, United States

Background: Severe mental illnesses, including depression, schizophrenia, and bipolar disorder, are an important public health concern. Individuals with these illnesses suffer from higher rates of chronic medical conditions, such as diabetes and cardiovascular diseases, and much shorter lifespans than the general population. Evidence suggests that these psychiatric disorders are associated with accelerated biological aging, driven by altered inflammatory processes, cellular aging, oxidative stress, and metabolic/vascular dysfunction. The gut microbiome has been implicated in these processes and previously associated with depression and other psychiatric disorders; however, different studies have reported conflicting findings, which may be related to choice of cohorts, confounding effects known to modulate the microbiome, and sample size.

Methods: We examined data from the American Gut Project, the biggest crowd-funded, citizen science project, coordinated at the University of California San Diego. The project set out to understand changes in the human microbiome. Participants provided a physical sample and completed a survey about their health, lifestyle, and diet. Out of over 15,000 individuals who have participated in the project to date, we identified 1,140 adults who provided a fecal sample and endorsed having been diagnosed with depression, schizophrenia, bipolar disorder, and/or posttraumatic stress disorder (PTSD). Subjects were excluded if they endorsed history of inflammatory bowel disease, seizure disorder, kidney disease, or antibiotic use in the past year or if they did not provide responses on matching variables of interest. From the larger sample, we created a 1:1 matched sample (n=111 matched pairs; N=222) of patients and non-psychiatric comparison (NC) subjects based on age (±5 years), body mass index (BMI), history of cardiovascular disease and diabetes, smoking frequency, level of education, and country/region of residence. 16S rRNA sequencing data were analyzed. Alpha and beta diversity measures were calculated using QIIME2 to compare identified cases to NCs. Differential abundance tests were performed using permutation based discrete FDR.

Results: The sample in the present report was comprised of individuals from the United States (n=98), United Kingdom (n=120), and Switzerland (n=4). The mean age of the sample was 46 years (SD=13.3), and the female-to-male ratio was 1.64 (138F/84M). Of the clinical population, 95% endorsed depression, 10% PTSD, 5% bipolar disorder, and 2% schizophrenia (some individuals endorsed more than one disorder). There was significant community level separation between patients and NCs using weighted UniFrac (PERMANOVA p<0.05, pseudo-F=3.04). This difference is notable as groups were 1:1 matched on various demographic factors and clinical conditions known to be present in higher rates among psychiatric patients, which can impact microbial composition. Using a mean difference test with discrete FDR correction, no significant differentially abundant taxa were found at 0.1 control level, but a more permissive threshold of 0.3 showed 5 differentially abundant taxa which overlap with previous findings in human cohorts (order Clostridiales and family Erysipelotrichaceae). No significant differences were observed for measures of within-community (alpha) diversity in terms of phylogenetic diversity (Faith's Diversity Index) or evenness (Shannon Index). Faith's Phylogenetic Diversity and Shannon indexes were positively correlated with age in middle-aged (>46 years) NCs but not in corresponding, age-matched patients or younger NC or patient subjects.

Conclusions: These findings are among the first few multi-national studies to provide evidence for altered gut microbiota in severe mental illness. Global community differences were observed between patients who endorsed a mental illness and NCs. Notably, these differences were observed after controlling for comorbid medical conditions known common in psychiatric disorders, which are also risk factors for microbial abnormalities. Although the analyzed sample size was small and the study was not sufficiently powered to detect significant differences in specific taxa, it is among one of the largest investigations of the gut microbiome to date, in these psychiatric populations. Furthermore, these results indicate the potential to draw clinically-relevant conclusions from self-reported, crowd-sourced data. Increased noise in samples (e.g., from shipping samples) and errors/variability in respondent metadata were minimized by using careful case-control matching accounting for confounding factors. This is one of the first clinical investigations of the gut-brain axis in which samples were not restricted by geographical region or research institution and were collected across the US, UK, and other countries. There is a need for longitudinal studies of microbiome using larger samples of people with serious mental illnesses and rich phenotypic and biomarker data. Microbiome research holds promise for precision medicine through data-driven diagnostics and medical recommendations and cost-effective microbial therapeutics.

Keywords: Gut Microbiome, Schizophrenia, Depression, Bipolar Disorder, PTSD

Disclosure: Nothing to Disclose.

W113. Pharmacokinetic Studies of Scopolamine in Human and Rodent for the Establishment of PK-PD Models

James Shoblock*, Diana Adams, Natalie Welty, Maura Furey, Guang Chen, Pascal Bonaventure, Wayne Drevets, Anindya Bhattacharya

Janssen Research & Development, LLC, San Diego, California, United States

Background: Scopolamine has shown rapid anti-depressant efficacy in several clinical studies at the dose of 4 μg/kg i.v. The robust clinical efficacy begs the question of a muscarinic-subtype selective approach that retains the efficacy profile of scopolamine but improves on the therapeutic margins. The first step in that direction is to mimic clinical exposures of scopolamine in rodents to enable pharmacodynamic and efficacy readouts. As such the aim of this study was to determine the pharmacokinetic profile of scopolamine after various doses and routes of administration in rodents and humans.

Methods: Scopolamine was dosed in mice (4 μg/kg i.v., 0.025 mg/kg i.p) and rats (0.16 mg/kg s.c., 1.5 mg/kg i.p) and plasma and whole brain were collected at various time points. Exposure levels were measured via mass spectrometry. Protein binding in plasma across species and in rat brain was also determined. In humans, scopolamine was given at 4 μg/kg i.v. (15 min infusion) and exposure was measured in plasma at various time points starting at 15 min post-infusion. Based on comparisons between mouse i.v. pharmacokinetics and human plasma levels, we estimated human brain exposure and then tested relevant doses in antidepressant models to establish a PK-PD relationship. Scopolamine was tested in the chronic mild stress model and learned helplessness test in rats and tail suspension in mice.

Results: After 4 μg/kg i.v. in mice, scopolamine achieved a Cmax in plasma of 1.8 ng/ml at 2 min and decreased to 0.02 ng/ml at 2 h. While plasma levels showed a decay typical to i.v. administration, brain levels were sustained over the 2 h, reaching a Cmax of 2.7 ng/g at 1 h and remaining at 2.0 ng/g at 2 h. Similar slower brain clearance was also observed in rat and mouse after non-i.v. systemic dosing. At 0.16 mg/kg s.c. in rat, scopolamine achieved a Cmax in plasma of ~30 ng/ml at 15 min, decaying to ~2 ng/ml by 2 h and below limit of detection by 4 h. The same dose achieved a Cmax of ~50 ng/g in brain at 30 min and remained at ~20 ng/g at 4 h, falling below limit of detection between 8 and 24 h. 1.5 mg/kg i.p. in rat achieved a Cmax of 117 ng/ml in plasma at 30 min and 7.9 ng/ml in plasma at 6 h. In brain, the same dose yielded a Cmax of 202 ng/g at 30 min and 9.7 ng/g at 24 h. In mice, 0.025 mg/kg i.p. scopolamine achieved a Cmax in plasma of 5 ng/ml, with levels falling to 0.02 ng/ml at 24 h. The same dose produced a Cmax in brain of 12 ng/g at 30 min with maintained levels of 4 ng/g at 24 h. Repeated dosing for 3 days every 48 h did not alter exposure levels. Plasma protein binding was 10.4, 6.9, and 17.2% in mouse, rat, and human, respectively, and 31.5% in rat brain. In humans, 4 μg/kg i.v. (15 min infusion) resulted in a Cmax of ~4 ng/ml in human plasma at the first measured time point, 15 min. Because plasma Cmax likely occurred immediately after infusion, we extrapolated the data to estimate earlier plasma levels, and used this data to estimate human brain Cmax, based on the brain:plasma levels observed in mouse after i.v. dosing. Similar to rodent, 4 μg/kg i.v. in human likely achieved sustained exposure in brain, with a possible maximum Cmax of 12 ng/g (estimated). If human brain protein binding is similar to rat, then this would be equivalent to 12 nM free in human brain, suggesting high engagement of central muscarinic receptors given the low nanomolar affinity of scopolamine. In an effort to translate clinical efficacy to preclinical models, we first tested scopolamine with a relatively high dose (1.5 mg/kg i.p., in order to maintain exposure throughout the period of testing) in a chronic stress model in rats. Scopolamine produced rapid onset with complete reversal of the behavioral deficit. In the learned helplessness model, various high doses of oral scopolamine were administered and efficacy was determined. It was found that in order to obtain a dose over the ED50, 60 mg/kg p.o. was needed; due to the poor oral bioavailability of scopolamine, terminal exposures after the high dose were relatively low (25 ng/ml in plasma and 67 ng/g in brain). Finally, in tail suspension test, 0.05 mg/kg s.c. scopolamine at 30 min significantly decreased immobility, and yielded terminal exposures of 3 ng/ml total in plasma and 13 ng/g in brain.

Conclusions: The dose of scopolamine used clinically for depression is likely to engage central muscarinic receptors at high levels for a sustained period of time. Non-intravenous systemic dosing that yields somewhat similar plasma and central exposure in rodents is effective in models related to depression. Current efforts are in progress to tease out the muscarinic receptor subtype(s) involved in sustaining the efficacy of scopolamine yet sparing of the adverse effects of muscarinic receptor antagonism.

Keywords: Muscarinic Acetylcholine Receptor, Pharmacokinetics, Depression

Disclosure: Part 5: Janssen Research and Development, Employee.

W114. Adolescent Anhedonia Following Early-Life Adversity Involves Aberrant Interaction of Reward and Anxiety Circuits and is Reversed by Knockdown of Amygdala Corticotropin-Releasing Hormone

Jessica Bolton, Jenny Molet, Limor Regev, Elizabeth Haddad, Andre Obenaus, Hal Stern, Dewleen Baker, Victoria Risbrough, Tallie Z. Baram*

University of California, Irvine, California, United States

Background: Anhedonia, the diminished ability to experience pleasure, is an important dimensional entity linked to depression, schizophrenia and other emotional disorders, but its origins and mechanisms are poorly understood. We have previously identified anhedonia, manifest as decreased sucrose preference and social play, in adolescent male rats that had experienced early-life adversity as a result of fragmented unpredictable sensory signals from maternal care (FRAG). Here we probed the molecular, cellular and circuit processes underlying FRAG-induced anhedonia and tested them mechanistically.

Methods: We examined functional brain circuits and neuronal populations activated by social play in adolescent FRAG and control rats. Structural connectivity between stress- and reward-related networks was probed using high-resolution diffusion tensor imaging (DTI), and cellular/regional activation was probed using cFos. We employed viral-genetic approaches to reduce corticotropin-releasing hormone (CRH) expression in amygdalar central nucleus (ACe) in anhedonic rats, and tested for anhedonia reversal in the same animals.

Results: Sucrose preference was reduced in adolescent FRAG rats (F[1,21]=8.85, p<0.01). Social play, generally considered an independent measure of pleasure, activated brain regions involved in reward circuitry in both control and FRAG groups. Compared with controls, social play in the FRAG group activated three times as many CRH-expressing neurons in ACe, a nucleus typically involved in anxiety/fear, indicating aberrant functional connectivity of pleasure/reward and fear circuits (t(7)=2.45, p<0.05). DTI-tractography revealed increased structural connectivity of amygdala to medial prefrontal cortex in FRAG rats (t(13)=2.29, p<0.05). CRH-shRNA, but not control shRNA, given into ACe reversed FRAG-induced anhedonia (interaction: F[1,17]=10.42, p<0.005) without influencing other emotional measures.

Conclusions: These findings robustly demonstrate aberrant interactions of reward and fear/anxiety networks after chaotic (high-entropy) early-life experience, and suggest mechanistic roles for CRH-expressing amygdala neurons in emotional deficits portending major neuropsychiatric disorders. Supported by P50 MH096889; NS28912.

NOTE: this poster is linked to Risbrough et al., a poster focusing on how anhedonia predicts PTSD. Please view poster M22, Poster Session I.

Keywords: Anhedonia, Reward Circuitry, Amygdala, Corticotropin-Releasing Hormone, shRNA

Disclosure: Nothing to Disclose.

W115. A Novel Method for Chronic Social Defeat Stress in Female Mice

Alexander Harris*, Piray Atsak, Zachary Bretton, Emma Holt, Raisa Alam, Mitchell Morton, Atheir Abbas, E. David Leonardo, Rene Hen, Scott Bolkan, Joshua Gordon

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

Background: Historically, female subjects have been omitted from preclinical stress studies, despite evidence that women are more liable to develop anxiety and depression. In particular, studies of stress susceptibility and resilience have focused on males since the commonly used paradigm — chronic social defeat stress — does not work in females.

Methods: We used oderants to increase resident male aggressive behavior and induce them to attack intruder female mice repeatedly over the course of 10 days. We then tested the impact of this manipulation on social interaction, sucrose preference and open arm exploration of the elevated plus maze.

Results: We find that female mice that undergo repeated social defeat stress develop social avoidance (Wilcoxan rank sum; p<0.001), decreased sucrose preference (p<0.05) and spend less time in the open arms of the elevated plus maze (p<0.05) relative to control mice. Moreover, a subset of the female mice who undergo repeated aggression display resilience, maintaining control levels of social exploration and sucrose preference.

Conclusions: This method closely follows a standard protocol used for chronic social defeat stress in male mice and produces similar results to those obtained in male mice. We anticipate that the development of this easy-to-use method for social defeat stress in female mice will greatly facilitate the inclusion of female subjects in stress research.

Keywords: Depression, Anxiety, Susceptibility, Resilience, Animal Model

Disclosure: Nothing to Disclose.

W116. Phase 2 and 3 Evaluation of Brexanolone, a GABA A Receptor Positive Allosteric Modulator, in Postpartum Depression

Samantha Meltzer-Brody*, Helen Colquhoun, Robert Riesenberg, C. Neill Epperson, Bassem Maximos, Heather Harrison, Singar Jagadeesan, Kristina Deligiannidis, Haihong Li, Christine Clemson, Stephen Kanes

University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States

Background: Postpartum depression (PPD) is the most common medical complication of childbirth. Brexanolone (USAN; formerly SAGE-547 Injection), is a proprietary formulation of allopregnanolone, a positive allosteric modulator of GABAA receptors. Nonclinical and early clinical data suggest brexanolone may have potential in the treatment of PPD. Randomized, double-blind, placebo-controlled Phase 2 and Phase 3 trials were designed to evaluate the efficacy and safety of brexanolone in subjects with moderate and severe PPD (NCT 02614547, NCT 02942017, and NCT02942004).

Methods: In the Phase 2 study, women with severe PPD (Hamilton Rating Scale for Depression [HAM-D] total score ≥26) were randomized to a 60-hour infusion of brexanolone (n=10) or placebo (n=11). The change from baseline in mean HAM-D total score at Hour 60 in the brexanolone group compared with the placebo group was the primary endpoint, with efficacy/safety measurements obtained up to Day 30. Phase 3 studies were similarly designed for moderate (HAM-D 20-25; N~100 patients) and severe (N~120 patients) PPD.

Results: In Phase 2, a significant reduction in the mean HAM-D total score was observed following brexanolone compared with placebo at Hour 60 (-21.0 vs -8.8, p=0.008). Significant differences from placebo were observed from 24 hours after initiation of treatment (-19.4 vs -8.1, p=0.006) through Day 30 (-20.8 vs -8.8, p=0.010). At 60 hours, 7/10 brexanolone subjects showed remission (HAM-D ≤7), versus 1/11 placebo subjects (p=0.008). There were no deaths, serious adverse events, or discontinuations due to adverse events. The Phase 3 trials, expected to be completed in the second half of 2017, seek to validate and expand upon these findings and will be presented.

Conclusions: In the Phase 2 study, brexanolone was associated with rapid reduction (24 hours after treatment initiation) in depressive symptoms relative to placebo that were sustained through Day 30, and was generally well tolerated. Phase 3 data seek to validate these findings.

Keywords: Brexanolone, Neuroactive Steroid, Postpartum Depression, GABA-A Receptors

Disclosure: Nothing to Disclose.

W117. Major Depression is Associated With Rare Variants in the PHF21B Gene, a Modulator of the Stress Response in Rodents

Julio Licinio*, Mauricio Arcos-Burgos, Sha Liu, Jorge Velez, Bernhard Baune, Catherine Jawahar, Volker Arolt, Udo Dannlowski, Aaron Chuah, Gavin Huttley, Rhys Fogarty, Martin Lewis, Stefan Bornstein, Ma-Li Wong

College of Medicine, State of New York University, Upstate Medical University, Syracuse, New York, United States

Background: Major depressive disorder (MDD) is a gene-environment disorder that affects approximately 350 million people worldwide. However, its underlying genetic basis remains largely unknown. In this study, we investigated the role of functional genetic variants in MDD.

Methods: Whole-genome screening of functional variants were investigated in two cohorts: A discovery Los Angeles Mexican-American cohort (196 controls, 203 MDD) and a replication European-ancestry cohort (499 controls, 473 MDD). Controls had either high levels of stress (Mexican-American controls were primarily recent immigrants) or high stress levels were given higher weights in our analysis (European-ancestry controls stress levels were accessed by questionnaires). Genome-wide association study (GWAS) and rare variant (regression- and permutation-based kernel-based adaptive cluster, KBAC) analyses were performed. Whole-genome sequencing and analyses were obtained in small subset of both cohorts. Pathway and network analyses were performed. Brain expression quantitative trait loci (eQTL) were extracted from the UK Brain Expression Consortium (UKBEC), and an animal model of chronic restraint stress was used to obtain behavioral, quantitative RT-PCR and Western blot data.

Results: In the Mexican-American cohort, we identified 44 common and rare functional variants associated to mild to moderate MDD. Pathway analysis disclosed that overrepresented gene ontology processes included innate immune response, glutamate receptor signalling and detection of smell sensory perception. We replicated the PHF21B gene in the ethnically unrelated European-ancestry cohort. PHF21B variants contributed significantly to the levels of expression of this gene in the hippocampus and other brain areas. We will present new animal behavioral, molecular biology and biochemical data showing decreased hippocampal Phf21b gene and PHF21B protein expression in depressive-like behavior induced by chronic restraint stress.

Conclusions: MDD is a syndrome of considerable genetic heterogeneity, as we identified many common and rare functional variants that confer susceptibility to MDD in the Mexican-American cohort. The inclusion of stress level data allowed the identification and validation of a novel rare functional variant associated with MDD.

Keywords: Major Depression Disorder, Psychiatric Genetics, Chronic Stress

Disclosure: Part 2: Nature Publishing Group, Consultant.

W118. Antidepressants Restore the Long-Lasting Effects of Stress on Body Weight, and Those Effects are Associated With Leptin Levels

Ma-Li Wong*, Su Hyun Lee, Andrew Vincent, Martin Lewis, Claudio Mastronardi, Julio Licinio

South Australian Health and Medical Research Institute and Flinders University, Adelaide, Australia

Background: Major depressive disorder (MDD) and obesity are two common disorders of complex etiology. Antidepressant prescriptions have risen 400% between 1988 and 2008, and obesity rates have doubled in adults during this period. Weight gain is a common outcome of antidepressant treatment; however, the relationship between MDD, obesity and antidepressant treatment is complex and poorly understood. It is plausible that almost 25% of obesity occurrences may be attributable to the association with MDD and current or past antidepressant use. Our lab has been using a new animal model to understand the mechanisms of weight recovery after chronic stress. We hypothesized that the effects of chronic stress on body weight could not only be reversed by antidepressants, but also further worsened by environmental factors, such as high-fat diets, leading to increased body weight and obesity.

Methods: Male Sprague-Dawley rats were subjected to short-term exposure to recurrent restraint stress and antidepressants (fluoxetine or imipramine) for 2 weeks, followed by diet-induced obesity accomplished by high-fat diet intake for 296 days. Measurements: Body weight, food intake, lipid profile (triglyceride, cholesterol, free-fatty acid) and leptin levels.

Results: Chronically stressed, obesity-prone (OP) rats, defined as those within the upper 50% of body weight gain, and obesity-resistant rats (OR, lower 50% body weight) treated with antidepressants had better weight recovery in the post-stress recovery period. The OP subgroup treated with fluoxetine had better body weight recovery than the other OP subgroups, and the OR subgroup treated with imipramine had better body recovery than the other OR subgroups. New metabolic data (leptin, triglyceride, cholesterol, fatty acid) and new analyses (liner regression slopes among obesity prone and obesity resistant animals, and correlation between body measurements and leptin levels) will be presented.

Conclusions: We conclude that the interaction of stress, antidepressant treatment, and high-fat diet may have long-lasting effects on growth, body weight and length, which in turn are associated with leptin levels.

Keywords: Antidepressant Agents, Body Weight, High Fat Diet, Diet Induced Obesity

Disclosure: Part 1: Macmillan Publishers, Ltd., Honoraria, Spouse.

W119. Mu Dependent Effects of Buprenorphine in Combination With Samidorphan on Immobility Behavior in Wistar Kyoto Rats

Karen Smith*, Daniel R. Deaver, Reginald L. Dean, III, Jacobi I. Cunningham, Connie Sanchez, David J. Eyerman

Alkermes, Inc., Waltham, Massachusetts, United States

Background: The endogenous opioid system is thought to play a key role in the regulation of mood [1]. ALKS 5461 is a balanced opioid modulator that represents a novel treatment for depression which combines buprenorphine (BUP) with samidorphan (SAM). We have previously described that combining BUP and SAM produced an antidepressant-like behavioral effect in rats in the forced swim test (FST) [2]. However, the relative contribution of mu opioid receptors to this behavioral effect is unclear. Here, we have characterized the role of mu receptors on immobility behavior using OPRM1 knock out rats. Additionally, we demonstrate that selective pharmacologic blockade of mu receptor activity blocks the BUP: SAM antidepressant-like behavioral effect in FST, in Wistar Kyoto (WKY) rats.

Methods: Female OPRM1 (-/-) knock out rats (SAGE® Labs, Boyertown, PA) and their OPRM1 (+/+) controls were placed in the forced swim test (water; 23±1ºC, 32 cm depth) for a single 5 min test session. For pharmacological studies, male WKY rats were used (n=8/group) as they spontaneously exhibit high levels of immobility in the FST [1] and are insensitive to chronic SSRI treatment [2]. To selectively block mu activity, cyprodime (0, 3, 10 or 30 mg/kg s.c.) was administered 15 min before the BUP: SAM (0.1: 0.3 mg/kg s.c.) combination. Based on previously characterized BUP: SAM neurochemical responses [3], depressive-like behaviors were measured 3 and 24 hr post-drug treatment. To verify the mechanism of BUP: SAM response, we titrated mu receptor activity by increasing SAM drug exposure. Specifically, WKY rats were co-administered vehicle or BUP (0.1 mg/kg s.c.) with increasing concentrations of SAM (0, 0.3, 3 or 10 mg/kg) and tested in the FST 3 and 24hr later. All behavior was video-recorded and scored manually. Data are expressed as mean+SEM and were analyzed using repeating measures ANOVA, with follow-up post hoc analysis or t-test where appropriate.

Results: OPRM1 (-/-) knock out rats had significantly higher levels of immobility in the FST compared to OPRM1 (+/+) controls. BUP (0.1 mg): SAM (0.3 mg) significantly reduced immobility compared to vehicle treated controls 3 and 24hr post-drug administration. This anti-immobility effect was blocked by cyprodime (30 mg/kg) pre-treatment at both time-points (3 & 24hr). Lower doses of cyprodime did not significantly alter the BUP: SAM response. Similarly, decreasing mu activity of the combination with a higher concentration of SAM (10 mg/kg) blocked the BUP: SAM anti-immobility response at each time-point tested.

Conclusions: Selective knock out of OPRM1 increases immobility in the FST, indicating that mu receptors may play a role mediating adaptive responses to learned helplessness. In FST studies using WKY rats, modulation of mu opioid activity with a combination of BUP and SAM produced rapid and sustained reversal of immobility. This effect was blocked by selective mu antagonism. Furthermore, increasing concentrations of SAM in the combination (inducing a corresponding decrease in mu receptor activity), completely abolished the BUP: SAM-induced anti-immobility effect. Taken together, these results indicate that balanced modulation of mu opioid activity is required for the efficacy of the BUP: SAM combination in the FST and this model of depression. These studies were by supported by funding from Alkermes, Inc.

Keywords: Mu-Opioid Receptors, Treatment Resistant Depression, ALKS 5461, Forced Swim Test, Wistar Kyoto Rat

Disclosure: Part 1: Alkermes, Inc., Employee, Part 2: Alkermes, Inc., Employee, Part 3: Alkermes, Inc., Employee, Part 5: Alkermes, Inc., Employee.

W120. Plasma Cortisol Stress Response and in vivo PET Imaging of Human Brain Serotonin 1A Receptor and Serotonin Transporter Binding

Louisa Steinberg*, Harry Rubin-Falcone, Joshua Kaufman, Jeffrey M. Miller, Maria Oquendo, M. Elizabeth Sublette, Thomas Cooper, Eli Min, Todd Ogden, J. John Mann

Columbia University, New York, New York, United States

Background: The hypothalamic-pituitary-adrenal (HPA) axis and the serotonin system have a complex bi-directional relationship (1,2). This relationship is of importance to the pathogenesis of mood disorders, because major depressive episodes are associated with both an abnormal serotonin system and abnormal HPA axis responses to stress. Positron emission tomography (PET) imaging studies in humans have demonstrated mixed results in relating the stress response to markers of serotonergic tone (3–6). In this study, we assayed endogenously secreted cortisol in response to an acute, physical stressor. We hypothesized that higher 5-HT1A receptor binding would be positively correlated with increased cortisol levels in response to the stressor, while SERT binding would show an inverse correlation with the cortisol stress response. Additionally, we expected participants with MDD to demonstrate an increased cortisol stress response compared to healthy controls.

Methods: Thirty-four adult subjects had PET scans with both [11C]WAY-100635 and [11C](+)-McN-5652 for quantification of 5-HT1A and SERT binding, respectively, and stress cortisol level assayed. Fourteen unmedicated depressed participants meeting DSM-IV criteria for MDD in a current major depressive episode were included in the analysis, as well as twenty-one healthy volunteers.

Cortisol was measured in blood samples drawn from subjects immediately prior to PET scans, just after the arterial line was inserted. All samples were drawn between 12pm and 2pm.

PET images were acquired with an ECAT EXACT HR+ scanner (Siemens/CTI, Knoxville, Tennessee); details are described previously (7). A metabolite corrected arterial input function was obtained for both tracers (7), and plasma free fraction (fP) of [11C]-WAY-100635 was assayed in triplicate (8). [11C](+)-McN-5652 does not have a measurable free fraction. Brain regions of interest (ROIs) were chosen a priori based on areas of abundant binding for each tracer and included raphe nuclei, cingulate, dorsal prefrontal cortex, hippocampus, insula, medial prefrontal cortex, parietal cortex, parahippocampal gyrus, occipital cortex, orbital cortex, temporal cortex, dorsal putamen, midbrain, thalamus, anterior cingulate, amygdala and hippocampus. Cerebellar gray matter (CGM) was used as a reference region for [11C](+)-McN-5652 and cerebellar white matter (CWM) for [11C]-WAY-100635.

Distribution volumes (VT) of [11C]WAY-100635 were estimated for each ROI using kinetic analysis with an arterial input function and a two tissue compartment constrained (2TCC) model as previously described (7). Time activity curves were fit with a 2TCC model in which the K1/k2 ratio was constrained to that of the reference region (CWM) for each ROI. BPF was calculated as (VT(ROI) – VT(REF))/fP for [11C]WAY-100635. Derivation of [11C](+)-McN-5652 VT was performed using likelihood estimation in graphical approach (LEGA), previously described (9–11).

The associations between cortisol measurements and 5-HT1A and SERT binding measured in our a priori regions of interest were examined using a linear mixed effects model with cortisol and brain region as fixed effects and subject as a random effect.

Results: Cortisol levels in response to stress were not found to be different between control and MDD groups when co-varying for age, sex, and blood sample time of day (F=2.58; df=1,31; p=0.12). We did not find a relationship between cortisol levels in response to stress and the Beck Depression Inventory score (F=0.485; df=1, 8; p=0.51) or the Hamilton 24-item scale score (F=0.455; df=1, 8; p=0.52), covarying for age, sex, and blood sample time of day.

[11C]-WAY-100635 binding across the a priori ROIs was positively correlated with cortisol levels after accounting for other covariates including the time of day the blood sample was drawn (F=7.28; df=1, 29; p=0.012). The relationship between cortisol levels and [11C]WAY100635 binding was homogenous across a priori brain regions, as the interaction term for brain region was not significant (F =.50; df=12, 384; p =0.91). Post-hoc analysis of relationships between cortisol level and [11 C]WAY100635 binding in each individual ROI indicated a significant positive relationship (p< 0.05) between cortisol and [11C]WAY-100635 binding for all regions except amygdala (p= 0.054) and raphe (p= 0.089).

Cortisol and [11C](+)-McN-5652 BPP were not found to be correlated in a model that included all a priori ROIs (F=1.68; df=1, 28; p=0.21), but a cortisol by region interaction indicated a region-dependent effect (F=7.59; df=5, 155; p< 0.0001). Post-hoc testing showed a negative correlation between cortisol and [11C](+)-McN-5652 BPP in midbrain (p=0.015) and in amygdala (p=.044).

Conclusions: The positive relationship between [11C]WAY-100635 binding and plasma cortisol was observed across all brain regions studied, in both grouped and post-hoc individual ROI analyses except for the raphe nuclei and amygdala, indicating that a more pronounced relationship between cortisol response and 5-HT1A receptor binding is present in the terminal cortical fields. Our data are consistent with the hypothesis that more reactive HPA-axis is associated with greater terminal field 5-HT1A receptor binding.

Limitations of this study include that this was a convenience sample for whom we had a blood sample available to do the cortisol assay and that this cross-sectional study cannot demonstrate causal relationships. As such, there was also no baseline cortisol assessment as part of this study.

Keywords: Acute Stress, Cortisol, Serotonin 1a Receptor, Serotonin Transporter, Major Depressive Disorder

Disclosure: Nothing to Disclose.

W121. Lurasidone Treatment Normalizes the Behavioral Alterations in Rats Exposed to Chronic Mild Stress: A Role for Redox Mechanisms in Modulating GABAergic Homeostasis

Andrea Carlo Rossetti, Maria Serena Paladini, Paola Brivio, Mariusz Papp, Francesca Calabrese, Raffaella Molteni, Marco Andrea Riva*

University of Milan, Milan, Italy

Background: Exposure of rodents to chronic stress represents a valuable strategy to produce a dysfunction of different domains that are associated with psychiatric disorders, such as depression and schizophrenia, in order to investigate the molecular underpinning of such defects as well as the potential for drug intervention.

With these premises, the aim of our study was to establish the impact of chronic mild stress (CMS) on anhedonia and cognition and to evaluate the ability of a chronic treatment with the multi-receptor modulator lurasidone to normalize such defects. Moreover, we aimed at investigating the possible role for GABAergic dysfunction and to explore the contribution of redox mechanisms in the alterations brought about by CMS exposure.

Methods: Adult male Wistar rats were exposed to chronic mild stress (CMS) for 2 weeks and sucrose consumption was used to identify rats that were susceptible to the stressful manipulation. Control and CMS-susceptible rats were then randomized to receive chronic vehicle or the multi-receptor modulator lurasidone (3 mg/kg/day) for 5 more weeks, while continuing the stress procedure. Animals were tested for anhedonia, using the sucrose intake test, and for cognitive abnormalities, using the novel object recognition (NOR) test. At the end of the manipulation, rats were sacrificed and the brain regions of interest were rapidly dissected and stored for the molecular analyses. Specifically, real time PCR and/or western blot analyses were used to investigate the expression of a number of markers related to the GABAergic system and to oxidative stress, using ß-actin as internal standard.

Results: Exposure to CMS produced a persistent anhedonic phenotype as well as a significant deficit in the NOR test. Both behavioral abnormalities were normalized by chronic lurasidone treatment. We next investigated the molecular alterations in dorsal and ventral hippocampus, focusing on the GABAergic system, seen its role in the etiology of psychiatric disorders as well as in controlling cognitive function.

Rats exposed to CMS show a selective reduction in the expression of parvalbumin, which labels a subpopulation of GABAergic interneurons, in dorsal (but not ventral) hippocampus, an effect that was normalized by chronic lurasidone administration. CMS rats also show a significant up-regulation of (NADPH) oxidase 2 (NOX2), which is critically involved in the increase of oxidative stress, in dorsal hippocampus, an effect that was not observed in animals that received lurasidone. Furthermore, the antipsychotic drug was able to increase the expression of Nrf-2, a master regulator of antioxidant defense, while reducing the levels of Keap-1, which exert a repressive control over Nrf-2, thus promoting an antioxidant response that may counteract the effects of stress on PVB neurons.

Conclusions: Our results highlight the ability of lurasidone to normalize different behavioral alterations associated with CMS exposure, suggesting its effectiveness on different ‘dimensions’ (RDoC) of depression or stress-induced disorders, such as anhedonia and cognition.

One molecular system that may be relevant for cognitive dysfunction after stress exposure is represented by reduced expression of PVB, which labels a subpopulation of GABAergic neurons that are vital for the generation of gamma oscillation. Indeed, lurasidone was able to normalize the effects produced by CMS exposure on parvalbumin expression, possibly through its ability in promoting anti-oxidative mechanisms within the dorsal hippocampus.

These effects may be particularly relevant in promoting resilience toward the effects produced by adverse environmental conditions, such as stress, which represents a major vulnerability element in the etiology of psychiatric disorders.

Keywords: Acute and Chronic Stress, Cognition, lurasidone, Oxidative Stress, Parvalbumin Neurons

Disclosure: Part 1: Lundbeck, Honoraria, Otsuka, Honoraria, Sumitomo Dainippon Pharma, Honoraria, Part 4: Sumitomo Dainippon Pharma, Grant, Sunovion, Grant, Lundbeck, Grant.

W122. Vicarious Defeat Stress-Induced Social Dysfunction is Ameliorated by Ketamine and Chlordiazepoxide in Female C57BL/6 Mice

Sergio Iñiguez*, Francisco Flores-Ramirez, Israel Garcia, Mary Kay Lobo, Samuel Castillo, Miguel Arenivar, Omar Lira

University of Texas at El Paso, El Paso, Texas, United States

Background: Stress is a prevailing risk factor for mood-related illnesses, wherein women represent the majority of those afflicted with major depression. Despite the growing literature suggesting that affective disorders can arise after a traumatic event is vicariously experienced, this relationship remains understudied in females at the preclinical level. Thus, the objective of the current investigation was to examine whether exposure to emotional/psychological stress (ES) mediates depression-related outcomes in female mice.

Methods: Female C57BL/6 mice (8-week old, nullparity) vicariously experienced the defeat bout of a male conspecific, by a male CD1 aggressor, for 10 consecutive days. Twenty-four h after the last stress exposure, female mice were tested in the social interaction test. Furthermore, we examined whether ketamine and chlordiazepoxide, pharmacological agents used to treat mood-related disorders in the clinical population, would reverse the ES-induced social dysfunction.

Results: When compared to controls, female mice exposed to ES displayed decreased social behavior, 24 hr post stress exposure. Also, they displayed higher levels of blood serum corticosterone, as well as decreased body weight. Lastly, the ES-induced avoidance-like phenotype was rescued by both ketamine and chlordiazepoxide.

Conclusions: Our data indicate that female mice exposed to ES display a behavioral- and physiologic-profile that mimics symptoms of depression in the clinical population. As such, this experimental model may be adopted to examine vicarious stress-induced mood-related disorders, as well as pharmacological therapeutic response, in a sex specific manner.

Keywords: Ketamine, Benzodiazepine, Anxiety, Depression, Female

Disclosure: Nothing to Disclose.

W123. Neural Processing of Loss in Adolescent Girls at Risk for Major Depression

Sohye Kim*, Will Mellick, Lane Strathearn, Carla Sharp

Baylor College of Medicine, Houston, Texas, United States

Background: Aberrant reward- and loss-related brain functions have emerged as promising endophenotypes for adolescent major depression (MDD). Recent research has shown that adolescents with, or vulnerable to, MDD demonstrate reduced neural activity when anticipating and/or receiving rewards. The study of neural anomalies in loss processing has lagged behind that of reward processing. The present study aimed to shed light on the endophenotypic nature of altered loss processing in adolescent depression by examining depressed and non-depressed girls with a familial vulnerability for MDD, compared with healthy controls.

Methods: We used a well-validated monetary reward task to compare loss-related neural activity during a functional magnetic resonance imaging procedure. Three groups of female biological offspring (N=52) of mothers with differential MDD status were examined: (a) currently depressed daughters of mothers with a history of MDD (MDD group; n=14), (b) age- and socioeconomic status (SES)-matched never-depressed daughters of mothers with a history of MDD (high-risk group; n=19), and (c) age- and SES-matched daughters of mothers with no past or current psychopathology in the mother or daughter (healthy controls; n=19).

Results: While anticipating losses, the MDD group showed heightened activation of the right amygdala, demonstrating this group's enhanced neural sensitivity to prospective losses. Adolescents’ depression severity correlated positively with their amygdala activity. High-risk adolescents displayed amygdala activation comparable to that of healthy controls, but showed blunted activation in the left ventral striatum, suggesting that motivational salience of potential impending losses may be diminished in this group. No group differences emerged when expected losses were delivered. However, when expected losses were omitted and adolescents were provided relief, both MDD and high-risk adolescents showed reduced bilateral ventral striatum activation compared to controls, consistent with attenuated neural responding previously reported while receiving reward.

Conclusions: Findings demonstrate altered loss processing in biological offspring of depressed mothers despite their different behavioral phenotypes (i.e., currently depressed vs. non-depressed) and point to aberrant loss-related neural activity as a promising risk biomarker present even prior to the onset of depression. Findings also shed light on unique patterns of neural responding that may serve as protective factors in at-risk, but currently healthy, offspring.

Keywords: Adolescent Depression, Risk and Resilience, Loss, Amygdala, Ventral Striatum

Disclosure: Nothing to Disclose.

W124. Suicidal Ideation is Associated With Reduced Endogenous Opioid Release During Social Rejection and Acceptance

David Hsu*, Tiffany Love, Kathleen Elverman, Sara Walker, Brian Mickey, Jon-Kar Zubieta, Scott Langenecker

Stony Brook University School of Medicine, Stony Brook, New York, United States

Background: A prominent theory of suicide postulates that “psychological pain” and “thwarted belongingness” are causal factors for suicidal ideation and behavior. In animal models, the endogenous opioids acting at the μ-opioid receptors (MORs) have been shown to be critical for recovering from social distress behaviors, and for promoting social reward. We previously showed overall reduced endogenous opioid release in those with major depressive disorder (MDD) compared to healthy controls during social rejection and acceptance, suggesting that a deficient opioid system may prevent recovery from social “pain” as well as reduce pleasure derived from positive social interactions (Hsu et al., 2015, Mol Psychiatry). This study used data from these participants to test the hypothesis that in MDD, those with suicidal ideation will show exaggerated reductions in opioid release during rejection and acceptance compared to those without suicidal ideation.

Methods: Participants were 16 medication-free adults in a current MDD episode (12 women; mean±SD: age, 30.1±10.3 years; HAMD-17, 18.4±5.8) as assessed by the SCID-I for DSM-IV. Participants were grouped into presence (n=10) or absence (n=6) of suicidal ideation (PSI, ASI), as measured by item #3 of the HAMD-17 (PSI: score of 1 - “feels life is empty, not worth living” or 2 - “recurrent thoughts or wishes about death of self”; ASI: score of 0). Participants were not endorsing suicidal intention or a suicidal plan at the time of study. Prior to positron emission tomography (PET) scanning, participants rated online profiles of preferred-sex individuals with whom they would most like to form a close relationship. A few days later they were given simulated feedback that they were not liked (rejection) or liked (acceptance) by their highest-rated profiles during PET with intravenous administration of the selective MOR radiotracer [11C]carfentanil. A priori regions of interest (ROIs) from our previously published studies were used to examine endogenous opioid release, measured as acute reductions in receptor availability during rejection and acceptance compared to baseline blocks, which did not contain feedback. An analysis of covariance (ANCOVA) was used to compare PSI vs. ASI, while adjusting for depression severity (total HAMD-17 score without suicidal ideation item).

Results: The PSI group showed significantly reduced opioid release during rejection compared to the ASI group in the right nucleus accumbens [F(1,13)=4.73, P=0.049]. Regions approaching statistical significance included the subgenual cingulate cortex [F(1,13)=4.64, P=0.051], and midline thalamus [F(1,13)=3.63, P=0.079). Similarly, the PSI group showed significantly reduced opioid release during acceptance compared to the ASI group in the subgenual cingulate cortex [F(1,13)=5.91, P=0.030], and left amygdala [F(1,13)=4.80, P=0.047]. Regions approaching statistical significance included the right nucleus accumbens [F(1,13)=4.17, P=0.062), right pregenual cingulate cortex [F(1,13)=3.66, P=0.078), and the periaqueductal gray [F(1,13)=3.28, P=0.093). PSI and ASI groups did not differ in MDD severity as measured by the HAMD-17 without suicidal ideation item [t(14)=1.07, P=0.303, two-tailed).

Conclusions: We found preliminary evidence for reduced opioid release during social rejection and acceptance in MDD participants with suicidal ideation, even when controlling for depression severity. Given the role of the opioid system in physical pain and social distress, as well as pleasure from positive social interactions, these findings provide biological evidence supporting theories of “psychological pain” and “thwarted belongingness” in suicidal ideation and behavior. Reduced opioid release during the experience of social rejection and acceptance may impede the ability to regulate negative and positive emotional responses, and may manifest as suicidal ideation.

Keywords: Suicide, Rejection, Endogenous Opioids, PET, Major Depression Disorder

Disclosure: Nothing to Disclose.

W125. A Trial of Adjunctive Probiotic Microorganisms to Prevent Relapse in Patients With Acute Mania

Faith Dickerson*, Robert Yolken

Sheppard Pratt Health System, Baltimore, Maryland, United States

Background: Immunological abnormalities have been identified in individuals with mania and may contribute to the pathophysiology of mania and to bipolar disorder. Previous studies have documented that increased levels of inflammation are a risk factor for relapse in individuals hospitalized for mania. The administration of non-pathogenic microorganisms known as probiotics has been shown to provide a safe means for modulating inflammation. We performed a randomized double-blind placebo-controlled trial to examine whether probiotic supplementation can prevent relapse in patients who have been admitted to the hospital for mania.

Methods: Patients hospitalized for mania were randomized to 24 weeks of double-blind adjunctive probiotic (combined Lactobacillus rhamnosus strain GG and Bifidobacterium animalis subsp. lactis strain Bb12) or placebo therapy. Assignment was 1:1 to each treatment condition. Block randomization was performed based on the baseline mania score. Patients were seen monthly for in person visits and contacted weekly by phone. Treatment was provided by participants' usual outpatient clinicians. The primary outcome was relapse defined as a psychiatric re-admission. The effect of treatment was calculated by determining the time of the first re-hospitalization using the Kaplan Meyer method and odds ratios were calculated using the Cox Proportional Hazard function with age, gender, race, and maternal education as covariates. The relationship between treatment condition and number of readmissions was calculated using ordered logistic regression including these same covariates. We also assessed the number and polarity of mood episodes associated with psychiatric readmission as well symptom severity and other clinical measures during the treatment period.

Results: A total of 83 patients signed consent and completed the baseline visit which was begun when they were hospitalized for mania; of these, 17 terminated their participation before randomization. A total of 66 patients were randomized of whom 79% (52/66) completed the full 24 weeks of the trial. The early terminators were divided equally between the active and placebo groups. The active and the placebo groups did not differ significantly on any demographic or clinical variable at the start of the trial except for race.

A total of 25 participants (38%) had at least one psychiatric readmission during the randomized phase. Of these, 25 individuals, 20 were in the placebo group and 5 in the group were in the probiotic group (OR=0.278, p=.022 adjusted for demographic variables). In terms of time to the first readmission, the hazard ratio for the active group vs. the placebo group was 0.39 (95% CI 0.17, 0.90), p=.028 adjusted for demographic variables. This hazard ratio corresponds to an approximately 2.5 fold reduction in risk of psychiatric re-hospitalization in the probiotic group over the study period. In terms of the symptom presentation associated with readmission, there were significantly decreased odds of re-admissions for depression (p<.05) and for depression or a mixed mood state (p<.006) in the probiotic group. We also found that individuals in the probiotic group had significantly better scores on the Sheehan Disability Scale at the end of the trial measuring the days of work or productive activity lost due to the illness (p=.01 adjusted for demographic variables). Symptom scores as measured by the Brief Psychiatric Rating Scale and the Young Mania Rating Scale did not differ between the treatment groups at the end of the study.

The study medication was well tolerated by study participants and there were not any withdrawals from the study due to complaints about the study medication. The study was overseen by a Data Safety Monitoring Board who reviewed the adverse event data at regular intervals. There were not significant differences between groups in the number of adverse events overall or by any body system.

Conclusions: Probiotic supplementation may help prevent relapse in patients who have been hospitalized for mania. Although the mechanism by which probiotic microorganisms achieve their beneficial effects is not known with certainty, it is likely that they alter the balance of the gut microbiota and prevent the aberrant immune response to food-derived and other harmful antigens. Probiotics are inexpensive and widely available and are considered safe under almost all conditions. The use of adjunctive probiotics might represent a major addition to the therapeutic armamentarium for the management of mood disorders.

Keywords: Mania, Probiotics, Relapse Prevention

Disclosure: Nothing to Disclose.

W126. Neurophysiological Correlates of Depressive Symptoms in Young Adults: A Quantitative EEG Study

Deniz Doruk Camsari*, Lee Poh Foong, Donica Kan, Phang Cheng Kar, Paul Croarkin

Mayo Clinic, Rochester, Minnesota, United States

Background: There is an unmet need for practical and reliable biomarkers for mood disorders in young adults. Identifying the brain activity associated with the early signs of depressive disorders could have important diagnostic and therapeutic implications. In this study we sought to investigate the EEG characteristics in young adults with newly identified depressive symptoms.

Methods: A total of 125 participants with no previous psychiatric history were recruited and depressive symptoms were assessed with the PHQ-9 and DASS-21. Based on the initial screening, a total of 100 participants (n=50 euthymic, n=50 depressive) were further selected and underwent 32-channel EEG acquisition. Mann-Whitney analysis was used to compare the baseline EEG power spectrum between the euthymic and depressive groups. Simple logistic regression and C-statistic were used to explore if EEG power could be used to discriminate between the groups. Finally, we identified the strongest EEG predictors of mood using multivariate logistic regression models.

Results: Comparison of the groups with Mann-Whitney test showed that the depressive group had higher delta power but lower theta, alpha and beta power as compared to the euthymic group. Simple logistic regression analysis with subsequent C-statistics revealed that only high-alpha and beta power originating from the left central cortex (C3) have a reliable discriminative value (ROC curve>0.7 (70%)) for differentiating the depressive group from the euthymic group. Final analysis with multivariate regression analysis showed that the single most significant predictor of group (depressive vs. euthymic) is the high-alpha power over C3 (p=0.03).

Conclusions: The present findings suggest that EEG is a useful tool in the identification of neurophysiological correlates of depressive symptoms in young adults with no previous psychiatric history.

Keywords: EEG Biomarkers, Alpha Oscillation, Depression

Disclosure: Nothing to Disclose.

W127. Do Commercial Pharmacogenetic-Based Decision Support Tools Provide Concordant Medication Recommendations?

Chad Bousman*, Boadie Dunlop

University of Calgary, Calgary, Canada

Background: The use of a person's genetic profile to guide individualised psychotropic therapy decisions has emerged as a promising strategy to improve efficacy and reduce adverse events. Several companies have developed pharmacogenetic-based decision support tools (DSTs), marketed to prescribers to inform the process of psychotropic medication selection and/or dosing for individual patients. However, these DSTs evaluate different genes, and even different variants within the same gene, suggesting it is likely that medication recommendations produced by each tool will not be consistent across all DSTs. Yet, no published data currently exists evaluating the degree of agreement across these DSTs. As such, we evaluated the degree of agreement between commercial DSTs in the context of major depressive disorder (MDD) treatment.

Methods: Four pharmacogenetic-based DSTs (CNSDose®, GeneSight®, Genecept®, Neuropharmagen®) were selected for examination based on peer-reviewed evidence suggesting their potential clinical utility. Five self-identified Caucasian outpatients with a primary diagnosis of MDD and a minimum of two previous antidepressant failures provided buccal swabs or saliva samples for DNA collection, as directed by the instructions enclosed in each DST's sample collection kit. Each kit was then returned to its respective company's laboratory for genotyping, phenotyping, and medication recommendation generation. Genotypes and predicted phenotypes that were reported by two or more of the DSTs were assessed for agreement by dividing the number of observed agreements by the maximum possible agreements across participants. Medication recommendations were extracted for all antidepressants, antipsychotics, anxiolytics/hypnotics, and mood stabilizers provided by each DST for each participant. Medication recommendation agreement was determined using Cohen's kappa and corrected kappa, accounting for random (chance) agreement and the maximum value of kappa, respectively.

Results: None of the DSTs evaluated the same combination of genes or variants on their testing panels. However, seven pharmacokinetic (CYP1A2, CYP2B6, CYP2C19, CYP2C9, CYP2D6, CYP3A4, UGT2B15) and seven pharmacodynamic (BDNF, COMT, HLA-A, HTR2A, HTR2C, OPRM1, SLC6A4) genes were included on two or more of the four DST testing panels. Among these overlapping genes, CYP2C9 was the only gene with 100% genotype and phenotype agreement across the DSTs examined. Perfect genotype agreements were also observed for CYP2C19, CYP3A4, and UGT2B15 but corresponding phenotype agreements ranged from 33% - 90%. Conversely, perfect phenotype agreement for CYP2B6 was observed across the applicable DST pairs examined, despite 73% genotype agreement. Genotypes and phenotypes for all pharmacodynamic genes were in 100% agreement across the DSTs, with the exception of SLC6A4, which had a genotype agreement of 47% and phenotype agreement of 20%. Medication recommendation agreement among all DST pairs examined was greatest for mood stabilizers (84%, Cohens’ corrected kappa, κ=0.405), followed by antidepressants (56%, κ=0.195), anxiolytics/hypnotics (56%, κ=0.065) and antipsychotics (55%, κ=0.030). Approximately one-quarter (26%) of all medication recommendations examined were jointly flagged by two or more DSTs as ‘actionable’ based on the patients tested genetic profile. Among these actionable recommendations, 19% provided conflicting advice (e.g. dosing) for the same medication.

Conclusions: We found modest concordance in medication recommendations provided by commercial pharmacogenetic-based DSTs; divergent recommendations derive from differences in the genes/variants tested and presumably the algorithms used to predict drug-gene interactions. The observed level of disagreement in medication recommendations among the DSTs indicates these tests cannot be considered equivalent and suggests they may differ in their clinical utility. However, determining whether any single DST is superior will require head-to-head testing in clinical trials. Further efforts to standardize genetic-based phenotyping and develop clinical guidelines for the use of DSTs are warranted.

Keywords: Pharmacogenetics, Pharmacogenomics, Antidepressants, Antipsychotics, Cytochrome P450s

Disclosure: Nothing to Disclose.

W128. Relationship Between Serum Cytokines and Brain Morphology in Drug-Naïve, First-Episode Major Depressive Disorder Using Surface-Based Morphometry

Reiji Yoshimura*, Shingo Kakeda, Asuka Katsuki, Yukunori Korogi

University of Occupational Environmental Health, Fukuoka, Japan

Background: The pathogenesis of major depressive disorder (MDD) is only partly understood, although many theories have been proposed regarding its cause; genes, the environment, and endocrine dysfunction are all considered to be factors influencing MDD. Recently, evidence has accumulated for a role of inflammation in the pathogenesis of MDD. For example, several studies have detected higher levels of inflammation in MDD than in healthy controls, although the strength of the evidence varies according to which specific inflammatory markers have been examined; i.e. interleukin (IL)-1β, IL-6, interferon γ (IFN-γ), and tumor necrosis factor α (TNFα). No prior study has however examined the relationship between cytokines and brain morphology in patients with major depressive disorder (MDD). We evaluated the relationship between serum cytokine levels and cortical thinning during the first depressive episode in drug-naïve patients with MDD.

Methods: For 40 patients with MDD and 47 healthy participants (controls), we measured serum cytokine levels (IL-1β, IL-6, IFN-γ, and TNFα) and cortical thickness on brain magnetic resonance imaging (MRI) using surface-based morphometry. MRI data were obtained on a 3T MR system (Signa EXCITE 3T; GE Healthcare, Wankesha, WI, USA) with an 8-channel brain phased-array coil. Original T1 images were acquired by three-dimensional fast-spoiled gradient recalled acquisition with steady state. Acquisition parameters were as follows: repetition time in ms/echo time in ms/inversion time in ms=10/4.1/700; flip angle=10; field-of view=24 cm; section thickness=1.2 mm, and resolution=0.9 × 0.9 × 1.2 mm. All images were corrected for image distortion due to gradient non-linearity using the Grad Warp software program17 and for intensity inhomogeneity with the “N3” function. Forty human tear samples were assayed in singlicate (due to limited sample volumes) using the V-PLEX Human Proinflammatory Panel I (4-Plex), a highly sensitive multiplex enzyme-linked immunosorbent assay used to quantitatively measure cytokines, including IL-1β, IL-6, IFN-γ, and TNFα, from a single small sample volume (25 μL) using an electrochemiluminescent detection method (MesoScale Discovery, Gaithersburg, MD, USA). This study was approved by the Institutional Review Board of [University of Occupational and Environmental Health]. Written, informed consent was obtained from all participants after they received a detailed description of the study purpose study design, and disadvantage of subjects.

Results: In group comparisons using whole-brain vertex-by-vertex analysis with adjustments for age and gender, patients manifested significantly lower thickness of the bilateral superior frontal and medial orbitofrontal cortices (p<0.05, Monte Carlo simulation). There were no regions in which cortical thickness was significantly greater in patients than in controls. Whole-brain vertex-by-vertex correlation analysis of patients showed that thickness of the bilateral superior frontal and medial orbitofrontal cortices was significantly negatively correlated with serum IL-6 levels (p<0.05, Monte Carlo simulation). There were no regions manifesting a significant positive correlation in patients. Furthermore, no brain regions showed significant correlations with other cytokine measurements (IL-1β, IFN-γ, and TNFα). Finally, we did not find any regions manifesting significant correlations between cortical thickness of controls and serum cytokine levels.

Conclusions: e found that serum IL-6 levels were significantly higher in patients with MDD than in controls. Importantly, PFC thickness in patients with MDD was significantly reduced, and also showed a significant inverse correlation with serum IL-6 levels. Since the PFC contains a high concentration of IL-6 receptors, IL-6 receptor-mediated neurotoxicity might occur under the high serum IL-6 levels present in the early stage of MDD. Our results suggest that IL-6 may play a key role in brain morphological changes in the early stage of MDD.

Keywords: Major Depressive Disorder, Cytokines, MRI

Disclosure: Nothing to Disclose.

W129. Role of (R)-Ketamine Metabolites in Antidepressant Effects of (R)-Ketamine

Jun-ichi Yamaguchi, Hidetoh Toki, Youge Qu, Chun Yang, Hiroyuki Koike, Kenji Hashimoto, Akiko Mizuno-Yasuhira, Shigeyuki Chaki*

Taisho Pharmaceutical Company, Ltd., Saitama, Japan

Background: Rapid and longer-lasting antidepressant effects of ketamine in patients with treatment-resistant depression have recently gained much attention. Among optimal isomers of racemic ketamine, (R)-ketamine has been reported to exhibit longer-lasting and more potent antidepressant effects in animal models than (S)-ketamine (Zhang et al, 2014; Yang et al, 2015; Zanos et al, 2016; Fukumoto et al, 2017). (R)-Ketamine is rapidly and stereoselectively metabolized into several metabolites, including (R)-norketamine and (2R,6R)-hydroxynorketamine (HNK). Recently, (2R,6R)-HNK has been reported to mediate antidepressant effects of (R)-ketamine in an NMDA receptor independent manner (Zanos et al., 2016), while there are contrary reports that (R)-ketamine showed greater antidepressant potency than (2R,6R)-HNK (Yang et al., 2016) and local injection of (R)-ketamine itself exhibited antidepressant effects (Shirayama et al., 2017). Therefore, role of metabolites of (R)-ketamine in the antidepressant effects of (R)-ketamine needs to be more investigated. In the present study, we employed pharmacokinetic approaches to elucidate involvement of (R)-ketamine metabolites in the antidepressant actions of (R)-ketamine.

Methods: Male C57BL/6J mice were administered lipopolysaccharide (LPS; 0.5 mg/kg, i.p.), and (R)-ketamine (3 or 10 mg/kg), (R)-norketamine (10 mg/kg) or (2R,6R)-HNK (10 mg/kg) was administered i.p. at 23 h after LPS administration. Antidepressant effects were evaluated by the forced swimming test at 3 h after drug administrations. Locomotor activity was measured at 1 h after drug administrations. Plasma, brain and cerebrospinal fluid (CSF) concentrations of (R)-ketamine, (R)-norketamine and (2R,6R)-HNK were measured by chiral liquid chromatography-tandem mass spectrometry. Inhibition of cytochrome P450 (CYP) enzymes were performed by i.p. administration of a cocktail containing ticlopidine (20 mg/kg) and 1-aminobenzotriazole (ABT) (50 mg/kg) at 1 h prior to administration of (R)-ketamine.

Results: Administration of LPS induced an increase in immobility time of mice in the forced swimming test, indicating depressive-like behavior. (R)-Ketamine (10 mg/kg) significantly attenuated LPS-induced depressive-like behavior, while both (R)-norketamine (10 mg/kg) and (2R,6R)-HNK (10 mg/kg) did not affect the LPS-induced depressive-like behavior. Plasma, brain and CSF concentrations of (R)-norketamine and (2R,6R)-HNK after administration of (R)-norketamine and (2R,6R)-HNK, respectively, were higher than those after administration of the same dose of (R)-ketamine, indicating that enough exposure of (R)-norketamine and (2R,6R)-HNK was achieved at the dose used. In addition, plasma concentrations of (R)-norketamine and (2R,6R)-HNK declined at the similar rate to those of (R)-ketamine following administration of (R)-ketamine. Administration of CYP inhibitors, ticlopidine and 1-ABT, prior to administration of (R)-ketamine markedly increased plasma concentrations of (R)-ketamine, while the same manipulation reduced plasma concentrations of (R)-norketamine and (2R,6R)-HNK. (R)-ketamine (3 mg/kg, i.p.) did not affect LPS-induced depressive-like behavior in the absence of CYP inhibitors, while the same dose of (R)-ketamine significantly exerted antidepressant effects in the same model in the presence of CYP inhibitor. Therefore, prevention of metabolism of (R)-ketamine with CYP inhibitors enhanced antidepressant effects of (R)-ketamine. CYP inhibitors themselves did not affect LPS-induced increase in immobility. Moreover, all manipulations above did not affect locomotor activity.

Conclusions: The present results indicate that (R)-ketamine metabolites, (R)-norketamine and (2R,6R)-HNK, are not involved in the antidepressant effects of (R)-ketamine in the LPS-induced depression model, at least at 3 h after drug administration. Given that plasma concentrations of (R)-norketamine and (2R,6R)-HNK declined at the similar rate to those of (R)-ketamine following administration of (R)-ketamine, it is also unlikely that both metabolites are involved in sustained antidepressant effects of (R)-ketamine. Therefore, the present results suggest that (R)-ketamine, but not its metabolites, is responsible for rapid and sustained antidepressant effects of (R)-ketamine, although we need to confirm in other animal models.

Keywords: R-(-)-ketamine, Antidepressant, Hydroxynorketamine, R-(-)-norketamine

Disclosure: Part 5: Taisho Pharmaceutical Company, Ltd., Employee.

W130. Paraventricular Thalamic Regulation of Habituation to Repeated Stress: Molecular and Network Mechanisms

Seema Bhatnagar*, Brian Corbett

University of Pennsylvania, Philadelphia, Pennsylvania, United States

Background: In the field of stress neurobiology, habituation is defined as a decreasing and learned response to a familiar stressor over time. Disrupted habituation is a signature of post-traumatic stress disorder (PTSD), causing devastating effects for those afflicted. Understanding the molecular and neural substrates underlying habituation may allow for improved therapies for PTSD patients. In rats, we model habituation using the repeated restraint stress paradigm. Exposure to this moderately intense stressor increases the expression of immediate early genes in certain brain regions, induces the production of stress-related hormones, and elicits struggle behavior. All of these responses are highest on day 1 of restraint and attenuate by the 5th to 7th exposure. We have previously identified the posterior paraventricular thalamic nucleus (pPVT) as a crucial brain region that promotes habituation. Using excitotoxic lesions of the pPVT, we showed disruption of habituation to repeated restraint. Furthermore, we showed that the pPVT is a site of negative feedback actions of glucocorticoids that contribute to habituation. These results demonstrate that the pPVT plays an important role in the ability to habituate to repeated stress. However, the underlying molecular mechanisms within the pPVT and the specific neural connections of the pPVT that mediate habituation are unknown. In the present studies in adult male rats, we conducted three experiments to further delineate the role of the pPVT in regulating habituation to repeated stress. These experiments focussed on activity of cells within the pPVT, cellular mechanisms within the pPVT and circuit mechanisms examining whether activity of projections of the pPVT to the medial prefrontal cortex (mPFC) is required for habituation.

Methods: We conducted three experiments to further examine the role of the pPVT in mediating habituation to repeated stress. In Experiment 1, we used Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) to inhibit the pPVT prior to daily restraint and examined the neuroendocrine and behavioral responses on the first compared to the fifth day of restraint. In Experiment 2, we investigated whether activity-regulated cytoskeleton-associated protein (Arc) induction within the pPVT was a molecular mechanism of habituation. Arc is an immediate early gene that reduces excitatory synapse number. In Experiment 3, we studied the role of pPVT to medial prefrontal cortical projections in mediating habituation.

Results: Experiment 1. We demonstrated that chemogenetic inhibition of the pPVT disrupted neuroendocrine habituation to 5 days of 30min daily restraint exposure, consistent with our previous data using neurotoxic lesions.

Experiment 2. We found that Arc expression was increased in the pPVT on day 1, but not day 5 of restraint in naïve rats and that Arc knockdown in the pPVT attenuated habituation.

Experiment 3. Preliminary data indicate that chemogenetic inhibition of a subset of pPVT neurons that project to the medial prefrontal cortex (mPFC), which negatively regulates the stress response, was sufficient to attenuate habituation.

Conclusions: The finding that chemogenetic inhibition of pPVT disrupted habituation is consistent with our previous data using neurotoxic lesions. Thus, these results solidify the importance of the pPVT in habituation to repeated stress. At the cellular level, the results indicate that Arc is a critical substrate for habituation as Arc knockdown attenuated habituation. We hypothesize that Arc-mediated restrictions in excitatory synapse number during the first exposures to restraint are a critical mechanism underlying habituation of the stress response. Finally, at the circuit level, the findings indicate that activation of neurons from PVT to mPFC is important for habituation. Our findings offer new insight into the role of the pPVT in mediating the stress response and are among the first to provide potential molecular and network mechanisms of stress habituation.

Keywords: Acute and Chronic Stress, Habituation, Paraventricular Nucleus of the Thalamus, Synaptic Plasticity

Disclosure: Nothing to Disclose.

W131. The Neurocircuitry of Approach-Avoidance Decisions in Depression: Towards a Cross Species Model of Utility and Motivation

Maria Ironside*, Ken-ichi Amemori, Min Su Kang, Jeffrey Curry, Callie McGrath, Ann Graybiel, Amit Etkin, Diego Pizzagalli

Harvard Medical School, Belmont, Massachusetts, United States

Background: Approach-avoidance (Ap-Av) behaviors are abnormal in major depressive disorder (MDD), related to future symptoms and targeted by psychotherapy. Studies indicate a role for avoidance behaviors in MDD (Trew, 2011), but little is known about the neural correlates that underlie such abnormality (Aupperle, Melrose, Francisco, Paulus, & Stein, 2015). Work in non-human primates has identified different neurons which respond to affective judgement, motivation and reward (Amemori, Amemori, & Graybiel, 2015), in addition microstimulation of the primate anterior cingulate cortex and the administration of anxiolytic drugs can change decision preferences (Amemori & Graybiel, 2012). Research and development of novel treatments are hindered by the lack of well-validated functional measures that are similar across human and non-human model species. To fill this gap, an Ap-Av task was adapted from non-human primate studies so that it could be used in conjunction with functional magnetic resonance imaging (fMRI) and computational modelling of value judgement and motivation.

Methods: Forty-one subjects (17 unmedicated individuals with MDD) completed three runs of an Ap-Av task using reward points and aversive images-sound pairs (35 trials in each run). fMRI data were preprocessed using SPM12 using standard analyses. Hemodynamic responses were modeled using a canonical hemodynamic response function that was convolved with the onset times of task regressors to compute a general linear model (GLM) at the single-subject level. The GLM included six task-related regressors: three conditions related to approach decisions (decision, feedback, reward), and three conditions related to avoid decisions (decision, feedback, no reward). Using computational modelling based on non-human primate research, parametric modulators of chosen value, motivation and reward were used to identify brain regions tracking these components of Ap-Av decision making.

Results: Across all subjects (N=41), whole brain corrected statistics (p=.001, k=20) with the parametric modulators of Chosen Value, Motivation, Reward and Reaction Time revealed four distinct activation patterns across avoid and approach decisions. Chosen Value was modelled to represent the anticipatory value associated with the selected option. When included as a parametric modulator of brain activity at the decision time-point (across all decisions), the following regions were found to track this component: Bilateral orbital gyrus leading into rostral anterior cingulate cortex (ACC), dorsal ACC, parietal sites and superior/ middle frontal gyrus. Motivation was based broadly on omission errors as a proxy for motivation to complete the task. When included as a parametric modulator of brain activity at the decision time-point (across all decisions), the following regions were found to track this component: Multiple large clusters in the bilateral precuneus, a cluster leading from right superior orbital gyrus to rostral to dorsal ACC, to left caudate, right mid cingulate cortex, bilateral middle frontal gyrus (including the dorsolateral prefrontal cortex (DLPFC)). Parametric modulation with trial-by-trial reward points revealed a large bilateral network in regions related to reward, such as the striatum, dorsal ACC and the hippocampus. Finally, parametric modulation with trial-by-trial reaction time revealed a similar pattern to that of Motivation with premotor/motor correlates instead of DLPFC.

Conclusions: This study used computational modelling from non-human primate research to identify brain regions associated with different components of Ap-Av decisions. This can be directly compared to non-human primate work, where data from single cell recordings targeted comparable regions (ACC, DLPFC) to those emerging from a whole brain analysis, using the same parametric modulators and a very similar task design. Thus, it is possible to compare microstimulation of the primate pregenual ACC (Amemori & Graybiel, 2012), which increased avoid decisions to the region tracking anticipatory value of each decision in humans in the current study. In addition, neurons in the primate DLPFC, which fired more specifically during low motivational conditions (Amemori et al., 2015), can be compared to the DLPFC activations which track motivational state in humans in the current study. Furthermore, in this study the offered reward was correlated with large orbitofrontal, striatal and rostral ACC activations, whereas motivation and reaction time were correlated with rostral/dorsal ACC and additional pre-motor and motor activations. This has similarities to further non-human primate research, where orbitofrontal neurons responded preferentially to the valence parameter (offered reward size), whereas premotor cortical neurons responded to motivational parameter (Roesch & Olson, 2004). Identifying the networks that represent the underlying drivers of Ap-Av behavior can help characterize anatomical targets for novel treatments such as non-invasive brain stimulation. This research lays the foundation for a bi-directional model between humans and non-human primates that promises to identify future treatment targets with the long-term goal of improving clinical outcomes.

Keywords: Approach/Avoidance, Functional MRI (fMRI), Nonhuman Primate Models, Computational Modeling, Translational Neuroscience

Disclosure: Nothing to Disclose.

W132. Exploring the Synaptic Basis of Chronotherapy for Major Depressive Disorder

Matthew Klein*, Roberto Malinow

University of California, San Diego School of Medicine, San Diego, California, United States

Background: Depression leads to significant impairment in daily function, and increased risk of suicide. A course of pharmacological treatment typically involves multiple, lengthy trials, and thirty percent of patients will continue to have depression that is refractory to treatment. The prolonged time course necessary for pharmacology may extend suffering, and increase risk of suicide. Therefore, development of more acute treatments for depression may be helpful. One night of total sleep deprivation, part of chronotherapy, results in a reduction in depressive symptoms in around 60% of patients. However, sleep deprivation has limited clinical use as it is labor intensive, and the beneficial effects are diminished by recovery sleep. This study explores changes in dopaminergic signalling following sleep deprivation in order to better understand the synaptic basis of sleep deprivation and how to improve chronotherapy for clinical use.

Methods: An inbred line of congenital learned helplessness (cLH) rats, a validated model of Major Depressive Disorder (MDD), was used to examine the antidepressant effects of one-night total sleep deprivation (SD) using an escapable shock paradigm. We used a variety of neuroscience and molecular biology techniques to examine differences in synaptic transmission resulting from one night of total SD. We recorded synaptic activity in the habenula from ex vivo brain slices using optogenetics, as well as used fiber photometry to record neural activity in vivo. We also employed single molecule imaging techniques to examine alterations in circadian gene expression during SD.

Results: We find that 12 hours of total sleep deprivation acutely reduced depressive-like symptoms in the cLH model. Furthermore, the anti-depressant effects were abolished after recovery sleep. We believe that sleep deprivation decreases activity in the lateral habenula, leading to a reduction in depression-like symptoms.

Conclusions: We have shown that sleep deprivation leads to a reduction in depression-like symptoms in the cLH rat model, similar to results seen in patients with depression. Therefore, we have validated the use of this model in exploring the benefits of chronotherapy on the synaptic and neural circuit level. We are continuing our investigations by examining the effect of sleep deprivation on synaptic activity in the lateral habenula.

Keywords: Major Depressive Disorder (MDD), Sleep Deprivation, Lateral Habenula

Disclosure: Nothing to Disclose.

W133. Major Depressive Disorder is Associated With Blunted Learning Signals in Medial Prefrontal Cortex and Putamen When Seeking Monetary Reward

Jenna Reinen*, Alexis Whitton, Diego Pizzagalli, Mark Silfstein, Anissa Abi-Dargham, Patrick McGrath, Dan Iosifescu, Franklin Schneier

Yale University, IBM Research, Rye, New York, United States

Background: Major depressive disorder (MDD) is characterized by debilitating motivational and affective symptoms, such as anhedonia and social withdrawal, that can severely limit daily function. Although the precise mechanism of these symptoms is unknown, the brain's dopaminergic reward system is a likely candidate given its role in linking predictive cues with valuable outcomes. Impairment in the ability to use learning signals to detect and pursue incentivizing cues could lead to symptoms of anhedonia in MDD and other disorders. However, there is debate in the literature over whether a striatal deficit is associated with depression, and during which temporal stage (anticipation, outcome) such reinforcement blunting may occur. To address this, we examined a dopamine-mediated learning signal (prediction error) thought to be critical for anticipating and updating reinforcement value across three stages of learning: choice, feedback, and reinforcement outcome, in unmedicated MDD patients and matched controls. We hypothesized that a priori-defined regions of interest (ROIs) in the cortico-striatal pathway would demonstrate blunted responses to reward in those with MDD, and would be linked to greater severity of anhedonic symptoms.

Methods: We tested 24 medication-naive patients with MDD and 24 matched control volunteers on a probabilistic reinforcement learning task involving two separate, counterbalanced conditions (Gain and Loss) while undergoing functional MRI (fMRI). This task was designed with jittered intervals to specifically separate and allow individual examination of the stages of choice, feedback, and reinforcement outcome. Participants learned in one condition (a) to choose the cue that was most likely (70%) to yield a positive monetary gain, and in the other condition (b) to choose the cue that was most likely (70%) to yield avoidance of monetary loss. Participants were also evaluated for clinical symptoms, including depression severity, anhedonia, and anxiety. A computational Q-learning model was used to generate behavioral learning parameters for each participant (learning rate, beta), as well as trial-specific learning signal regressors (prediction error) for the fMRI GLM-based analyses.

Results: Behavioral results revealed no significant group differences in raw optimal choice, reaction time, or the parameters generated by the learning model. However, optimal choice in the loss condition was correlated with severity of anhedonic depression (r=0.42, p=0.04), as assessed by the anhedonic depression subscore of the Mood and Anxiety Symptom Questionnaire. Brain imaging results yielded few significant group differences at the time of choice, but several other notable findings were found. (a) In the Gain condition during feedback, the control group showed a learning signal response in ventral striatum, but the MDD group did not. Whole-brain corrected results showed significant group differences (controls>MDD) in medial prefrontal cortex (mPFC). When examining the parametric regressor for prediction error of reward outcome presentation, whole brain and ROI analyses also yielded a group difference (controls>MDD) in bilateral putamen. (b) In the loss condition, no group differences in the regions of interest survived correction. However, in the loss condition we identified an inverse relationship between learning signal response in the putamen and severity of anhedonic depression on the Mood and Anxiety Symptom Questionnaire.

Conclusions: Patients with MDD showed blunted learning signals in medial PFC during reward feedback, and in bilateral putamen during reward outcome. Importantly, this finding was only identified when participants were seeking to earn monetary rewards, but few group differences were found when the participants were avoiding monetary loss, suggesting that motivational context influences reinforcement response in MDD. Further, we found that anhedonic symptoms were associated with optimal choice performance and putamen response in the loss condition. This work implicates a functional role of the dorsal striatum in depression at the time of reinforcement outcome, and supports a link between learning, cortico-striatal response, and the motivational symptoms of MDD.

Keywords: Depression, Striatum, Computational Psychiatry, Learning, Anhedonia

Disclosure: Part 5: IBM Research, Employee.

W134. Functional Mapping of Cortical Dopamine D2 Receptor Expressing Neurons

Jivan Khlghatyan*, Clementine Quintana, Martin Parent, Jean-Martin Beaulieu

Laval University, Quebec, Canada

Background: The D2 receptor (D2R) is a direct or indirect target of antipsychotics and mood stabilizers. Additionally, dopamine receptor 2 (Drd2) gene variants have been identified as risk factors for schizophrenia. Expression and function of Drd2 has mostly been studied in the striatum and prefrontal cortex, considering the involvement of these regions in mental disorders. Limitations of mouse models and technical approaches hindered reliable mapping of Drd2 neurons particularly in cortex.

Methods: We used immunohistochemistry, translational profiling, viral and retrobead tracing, and chemogenetics on a RiboTag mouse model, which express HA tagged ribosomal protein specifically in Drd2 expressing cells.

Results: We mapped brain wide expression pattern of Drd2 and revealed previously unidentified cortical clusters of Drd2 expressing cells. Cell type specific characterization and translation profiling of Drd2 cells indicate their heterogeneity and proportion in various cortical regions. Furthermore, we demonstrate the modulation of translation profiles of mPFC Drd2 neurons after chronic antipsychotic treatment, thus highlight the functional activity of D2 receptor. We constructed a comprehensive connectomic map that may point to a possible functional role of Drd2 neurons from various regions. Ultimately, chemogenetic mimicking of Gi signaling in mPFC Drd2 cells revealed their involvement in the regulation of anxiety related behaviors.

Conclusions: This comprehensive map of Drd2 neurons provides indications for its functional implications in healthy and disease conditions and can be used as a resource for future investigations. Multiple Drd2 neuron containing brain regions and cell types have to be taken into consideration during pharmacological intervention and assessment of functional and behavioral data.

Keywords: D2 Receptor, RiboTag, Neuroanatomy, Cortex, Claustrum

Disclosure: Nothing to Disclose.

W135. (R,S)-Ketamine and (2R,6R)-Hydroxynorketamine are Prophylactic Against Stress-Induced Depressive-Like Behavior in Female Mice

Briana Chen*, Christina LaGamma, Rebecca Brachman, Xiaoming Xu, Shi-Xian Deng, Donald Landry, Christine Denny

Columbia University, New York, New York, United States

Background: Stress exposure is a key risk factor for major depressive disorder (MDD). Some individuals can successfully adapt to stress, while others are particularly susceptible to developing MDD. Females are more than twice as likely as their male counterparts to be diagnosed with MDD. We previously reported that a single sub-anesthetic injection of (R,S)-ketamine prior to stress can induce stress resilience in male mice. However, pharmacological enhancement of stress resilience has yet to be demonstrated in females.

Methods: To develop prophylactics against stress-induced depression in females, we injected (R,S)-ketamine and its metabolites (2R,6R)-hydroxynorketamine ((2R,6R)-HNK) and (2S,6S)-hydroxynorketamine ((2S,6S)-HNK) at various doses one week before one of a number of stressors, including contextual fear conditioning (CFC), learned helplessness (LH), and chronic immobilization stress (CIS). Prophylactic efficacy was validated using the forced swim test (FST). In a separate cohort of mice, to test possible sex-dependent mechanisms of stress resilience, mice were administered ovariectomy surgery (OVX) prior to drug administration, stress exposure, and the FST. We then examined the efficacy of prophylactic compounds in OVX mice administered an estrogen (E2) replacement paradigm.

Results: (R,S)-ketamine and (2R,6R)-HNK, but not (2S,6S)-HNK, significantly decreased stress-induced depressive-like behavior in the FST when administered before stress in a variety of stress models. Interestingly, (R,S)-ketamine was effective as a prophylactic at a lower dose (10 mg/kg) in females than previously found to be effective in males (30 mg/kg). Moreover, prophylactic efficacy was dependent on ovarian-derived hormones, and these effects were not rescued by an E2 replacement protocol.

Conclusions: Our results suggest that prophylactics against stress-induced behavior can be developed in a sex-specific manner. (R,S)-ketamine and (2R,6R)-HNK can be utilized to protect against the development of depressive-like behavior, and the mechanisms by which they induce resilience may be mediated by ovarian-derived hormones. These data provide novel compounds for prophylactic development. To our knowledge, this is the first demonstration of pharmacologically enhanced stress resilience in females as well as the first demonstration that (2R,6R)-HNK may possess the same prophylactic properties of its precursor (R,S)-ketamine.

Keywords: Ketamine, Hydroxynorketamine, Prophylactic, Female, Depression

Disclosure: Nothing to Disclose.

W136. Familial Transmission of Fronto-Accumbal Circuitry: A Model for Testing Stability of RDoC Constructs

Ardesheer Talati*, Jiook Cha, Priya Wickramaratne, Ruixin Zhao, Stewart Shankman, Kelly Correa, Myrna Weissman, Jonathan Posner

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

Background: The NIMH's Research Domain Criteria (RDoC) posit that understanding of psychopathology may be better served by studying functional constructs that integrate molecular, neurobiological and behavioral units of information, rather than through clinical diagnoses alone. Demonstrating stability of these constructs is an important step in validating the RDoC approach. We previously addressed temporal stability, showing stability of cortical thickness and subcortical volume differences across 8 years. Here, we leverage a multi-generational family study to test generational stability. Specifically, we explore the hypothesis that if RDoC constructs represent stable markers of risk for psychopathology, they should be transmitted within families, and should do so independently of clinical course. We apply this hypothesis to test elements of one RDoC construct, frustrative non-reward. Part of the RDoC negative valence system, FNR indexes brain activity associated with response to withdrawal and prevention of reward, and is clinically most relevant to disorders of aggression. We test (Aim 1) whether the neurobiological representation of the FNR construct, namely connectivity between the nucleus accumbens (NA) and the orbitofrontal cortex (OFC) is transmitted from parent to child within families. We then explore (Aim 2) whether clinical representation of the construct FNR (aggression) is also transmitted across generation, and if so, whether neurobiological transmission (NAcc-OFC connectivity) could account for the clinical transmission (aggression).

Methods: Sample: Participants were second (Gen 2) and third (Gen 3) generations at high- or low-risk for depression (where depression was defined by the presence or absence of major depressive disorder in the first generation). There were 46 Gen 3 offspring with DTI data. This preliminary analysis is based on 22 Gen 2:Gen 3 parent:child dyads (total N=37), where both parent and child had complete DTI data. (Gen 2 mean age: 47.4±4.8yrs, 44% female; Gen 3, 21.3±4.8yrs, 52% female).

Imaging: DTI was pre-processed (e.g., bias/noise correction, brain extraction, eddy current correction) using established methods available in MRtrix3 and FSL. We adopted sophisticated analytic pipelines that were designed to improve biological validity of white matter tract estimates. White matter tractography was performed across the whole brain using: multi-tissue constrained spherical deconvolution to estimate voxel-wise fiber orientation distributions, and an efficient probabilistic algorithm (Second-order Integration over Fiber Orientation Distributions) with anatomically-constraint tractography (ACT) option. These initial tractograms were filtered to reduce supurious, false-positive results using Spherical-deconvolution Informed Filtering of Tractograms (SIFT). Anatomical constraints for tractography (i.e., seed and target region masks) were derived from each individual's neuroanatomy: based on T1-weighted structural scans, we segmented the whole brain into 84 regions (Desikan atlas) using Freesurfer imaging analysis suite. We finally generated the whole brain structural connectivity matrix (84x84) in each individual from the tractograms: streamline counts were used to weight connectivity (edge) of given two regions (nodes) following prior studies.

Clinical Measures: Aggression was measured using the Buss Perry Self-Report Questionnaire within two weeks of the MRI scan (6). In addition to the total score, there are four aggression subscales: anger, hostility, physical aggression, and verbal aggression.

Statistical Analyses: Statistical analyses were conducted in SAS using a mixed model approach (for Aim 1) and path analysis (Aim 2). As there were multiple G3 offspring sharing the same G2 parent, nuclear family was included as a random effect in the models.

Results: After accounting for age, sex, and familial risk status of parent and child, connectivity strength between the nucleus accumbens (NAcc) and lateral orbitofrontal cortex (OFC) in the G2 parent significantly predicted the same connectivity in the G3 offspring (left NAcc to left OFC: β=0.64 (std. err.: 0.21), p= 0.028; right NAcc to right OFC; β=0.93 (0.19), p=0.005. No significant transmission was found in the medial OFC on either hemisphere (ps>0.1). Associations remained significant when accounting for lifetime substance use, depression, or anxiety disorders in either parent or child. Finally, we used path analyses to explore whether NA-OFC transmission could account for familial transmission of clinical correlates of FNR (Aim 2). The path model confirmed the findings in Aim 1, namely that NAcc-OFC connectivity was transmitted across generation. In Gen 2 only, lower NA-OFC connectivity predicted higher aggression scores (β= -0.44 (0.18), p=0.01. However, there was no evidence to support intergenerational transmission of aggression or impulsivity or, consequently, for transmission of NAcc-OFC structural connectivity to explain transmission of clinical correlates.

Conclusions: Findings are preliminary and will require replication in larger samples with biological information on both parents. Nevertheless, the results suggest that some of the neural circuitry underlying negative valence systems (frustrative non-reward) is stable in families, and may be transmitted independent of the transmission of psychopathology.

Keywords: Research Domain Criteria (RDoC), Frustrative Non-Reward, Diffusion Tensor Imaging (DTI), Nucleus Accumbens (NAA), Orbitofrontal Cortex (OFC)

Disclosure: Nothing to Disclose.

W137. Mood Outcomes Program: A National Network of Depression Centers (NNDC) Computational Learning Health System to Advance Precision Medicine

David Katzelnick*, Peter Zandi, J. Raymond DePaulo, Paresh Patel, Carolyn Turvey, Patricia Rinvelt, John Greden

Mayo Clinic, Rochester, Minnesota, United States

Background: The National Network of Depression Centers (NNDC) is a collaboration among members in 26 University Depression Centers of Excellence (listed in www.NNDC.org) to help transform treatment and prevention of mood disorders. The Network launched the Mood Outcomes Program (MOP) to establish a computational learning health system for patients with depression, bipolar illness, and related mood disorders, and to link learning health approaches with emerging NNDC research data to aid precision medicine advances. A learning health system is a system of care that makes the best possible use of a large and growing body of evidence identified in longitudinal data collected as a part of clinical care. Precision Medicine is an emerging approach for disease diagnosis, treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person.

Methods: Patients receiving evaluation and treatment in the NNDC Centers are being enrolled with the goal of monitoring patients; an initial target is 25,000 patients over 5 years. Aims are to:

1) collect patient-rated outcomes and measurement-based care by tracking the PHQ-9, GAD-7, C-SSRS screener and the Altman Mania Rating Scale at each visit using a standardized electronic format;

2) feed measurement data graphically back to clinicians in real time to enhance participation and promote measurement based care;

3) utilize emerging computational data bases to enable linkage with emerging biomarker and other research measures;

4) continuously integrate learning health system approaches to facilitate quality improvement at the clinic and population levels;

5) link with mobile monitoring and other personalized treatment tools.

Results: A total of 3,681 NNDC pilot site patients with a psychiatric disorder have enrolled in the Mood Outcomes Program to date; 2,851 have mood disorders of which 2,346 have unipolar depression and 840 have bipolar disorder. A selected illustration of emergent subgroup data that could be employed for clinical and research projects includes: % positive for “ever took steps or attempted suicide”: 32.5 for unipolar depression, 33.7 for bipolar, and 21.2 for non-mood disorder. These subgroups can be compared using research variables.

A summary of our initial data is shown below:

Psychiatric Disorder: Unipolar Depression

% Female: 62.9%

Mean Age: 43.6

Mean First PHQ-9: 12.6

Mean First GAD-7: 10.3

Mean First ASRM: 2.7

% Positive Item #9 PHQ-9: 35.8

% Positive ever Took Steps or Attempted Suicide: 32.5

Psychiatric Disorder: Bipolar Disorder

% Female: 56.8%

Mean Age: 43.9

Mean First PHQ-9: 11.1

Mean First GAD-7: 9.3

Mean First ASRM: 3.6

% Positive Item #9 PHQ-9: 30.0

% Positive ever Took Steps or Attempted Suicide: 33.7Psychiatric Disorder: Non-mood Disorder

% Female: 52.3%

Mean Age: 38.7

Mean First PHQ-9: 9.5

Mean First GAD-7: 9.6

Mean First ASRM: 3.3

% Positive Item #9 PHQ-9: 21.5

% Positive ever Took Steps or Attempted Suicide: 21.2

Conclusions: Requirements for developing precision medicine approaches include:

1. truly large samples (tens of thousands, not tens or hundreds);

2. standardized, longitudinal measurement-based care and sustained treatment trials (years, not weeks);

3. concomitant long-term clinical, biomarker and neuroscience assessments which directly inform clinical decision making;

4. computational strategies for analyses. The Mood Outcomes Program is designed to fulfill these requirements by integrating a learning health system with research assessments. Quality improvement, multi-site research studies, and new monitoring and treatment strategies will be described to illustrate how this integration can enhance the pursuit of precision medicine. LITERATURE REFERENCES

Rush AJ, Am J Psychiatry. Isn't It About Time to Employ Measurement-Based Care in Practice? 2015 Oct;172(10):934-6. doi: 10.1176/appi.ajp.2015.15070928. EpubIOM (Institute of Medicine). 2013. Best care at lower cost: The path to continuously learning health care in America. Washington, DC: The National Academies Press.

Charles P. Friedman, Adam K. Wong, David Blumenthal, Achieving a Nationwide Learning Health System. Science Translational Medicine 10 Nov 2010: Vol. 2, Issue

Keywords: Precision Medicine for Mood Disorders, Learning Health System, Computational Psychiatry

Disclosure: Part 1: Healthcare Technology Systems, Inc., Stock / Equity, Part 2: Healthcare Technology Systems, Inc., Stock / Equity.

W138. Neural Correlates of Impaired Removal of Negative Information From Working Memory are Associated With Rumination and Reappraisal in Individuals Diagnosed With Major Depressive Disorder

Elena Davis*, Melina Uncapher, M. Catalina Camacho, Matthew Sacchet, Anthony Wagner, Ian Gotlib

Stanford University, Stanford, California, United States

Background: Individuals diagnosed with major depressive disorder (MDD) have difficulty removing negative information from working memory, which is related to the impaired emotion regulation and negative cognitive biases, including ruminative thinking, documented in people with this disorder. Previous studies have demonstrated that depressed individuals differ from their nondepressed peers in reaction times during working memory tasks involving emotional content. For example, depressed individuals show response latencies to negative words in a modified Sternberg memory task that are consistent with an intrusion effect, in which irrelevant negative information lingers in working memory (Joormann & Gotlib, 2008). Depressed individuals also exhibit slowed reaction times when attempting to disengage from negative stimuli (Levens & Gotlib, 2010). Investigators seeking to identify the neural correlates of these difficulties have found depression-related activation differences in neural regions within attentional networks – specifically, the dorsal anterior cingulate, superior parietal lobule, and insula – when individuals remove negative information from working memory (Foland-Ross et al., 2013). There is also an association between rumination and hippocampal/amygdala activation during suppression of negative information in MDD (Sacchet et al., 2017). In this study we sought to identify individual-level neural correlates of impairment in gating negative information in working memory that is specific to faces expressing negative emotion for individuals with and without MDD, and also examine associations between patterns of neural activation and measures of emotion regulation and rumination.

Methods: Fifty-eight adults, ages 18-35 years (M=26.7 [4.0] years), were recruited and assessed using the Structured Clinical Interview for DSM-IV to establish a diagnosis of MDD (N=29) or no history of any Axis I disorder (N=29 nondepressed controls [CTLs]). These participants completed an information-gating fMRI paradigm in which pictures of sad faces, neutral faces, and scenes were presented, and subsequent cues indicated whether they should RETAIN or EXPEL the faces from memory. In brief, a target screen was presented displaying a face (either sad or neutral) and a scene image simultaneously for 4 seconds. An 8-second delay followed during which participants were told to retain the face in memory in order to respond to a subsequent prompt. For 1/3 of the trials, an additional 8-second delay period occurred during which a “switch cue” was displayed indicating that on that trial, they needed to EXPEL the face and retain the scene instead. Thus, participants were required to switch their memory focus and shift attention away from the memory of the face. Next, participants viewed 4 probe images and indicated whether the target face or scene image appeared, and accuracy was recorded. Neuroimaging data were processed in SPM8 using standard procedures to extract the neural activation during the 8-second epoch of the “switch cue.” We contrasted EXPEL trials for sad versus neutral faces using a whole-brain voxelwise analysis and family-wise error correction. Finally, we contrasted activation patterns in the MDD and CTL groups and computed correlations between this activation difference and scores on the reappraisal subscale of the Emotion Regulation Questionnaire (ERQ; Gross & John, 2003) and the Ruminative Response Scale (RRS; Treynor et al., 2003).

Results: MDD and CTL participants did not differ in accuracy of task performance (all t(56)<1.2, all p>.3). A cluster in right superior parietal lobule (36 voxels, peak: 30, -56, 58) showed significantly greater activation for the MDD versus CTL group during the “switch cue” delay when sad faces were expelled from working memory, compared to neutral faces (p<.001, 10 voxel cluster extent thresholding). This relatively increased activation to sad faces was positively correlated with higher rumination scores (RRS total; r=.42, p=.001) and with lower self-reported use of reappraisal strategies to regulate negative emotion (ERQ subscale; r=-.37, p=.005), and was driven by the MDD group (rs=.34 and -.31, respectively).

Conclusions: Depressed individuals are characterized by difficulty reappraising negative emotions and heightened rumination on negative self-focused content. We show that these difficulties are related to activation in the superior parietal lobule when depressed individuals attempt to remove sad faces from working memory. The superior parietal lobule is part of the dorsal frontoparietal attention network that underlies top-down control of attention (Corbetta et al. 2008) and is involved in manipulating the contents of working memory (Wager & Smith, 2003). Therefore, relatively greater activation in this region may indicate increased demand on cognitive control resources or more cognitive effort needed to remove irrelevant sad information from working memory in depressed than in nondepressed individuals, particularly if they also have difficulty with reappraisal and rumination. Impairments in the dorsal attention network may be related to the cognitive dysfunction that has been frequently documented in MDD.

Keywords: Major Depressive Disorder, Working Memory fMRI, Cognitive Reappraisal, Rumination

Disclosure: Nothing to Disclose.

W139. Are Selective Serotonin Reuptake Inhibitors Ineffective for Not-Very-Severe Depression?

Elias Eriksson*, Fredrik Hieronymus, Alexander Lisinski, Staffan Nilsson

Goteborg University, Goteborg, Sweden

Background: Several previous trial-based meta-analyses have addressed the possibility that the antidepressant response to selective serotonin reuptake inhibitors (SSRIs) be positively related to baseline severity (1,2), some authors even claiming that these drugs display no meaningful effects in subjects displaying mild or moderate depression (1). The literature on this matter is however far from unanimous, some workers finding no influence of severity on response (3,4). With few exceptions (4), previous reports on this matter however have been based on trial-based meta-analyses rather than patient level-based mega-analyses, and focused on the possible association between severity and response rather than on the actual response (or lack of response) in patients with mild or moderate depression. Moreover, most of them have assessed the effect of active treatment by using a conventional but widely questioned (5,6) measure of response, i.e. the sum rating of the 17 different items of the Hamilton Depression Rating Scale (HDRS). Prompted by our previous observation that the first item of the HDRS, depressed mood, is a more sensitive detector of an antidepressant signal than the conventional effect parameter (7), we have now, using depressed mood as effect parameter, conducted a patient-level analysis aimed at shedding light on the alleged ineffectiveness of SSRIs in subjects displaying relatively low symptom rating at baseline.

Methods: The study was based on patient-level data from adult depressed subjects having participated in drug company-sponsored flexible- or fixed-dose placebo-controlled trials regarding sertraline, paroxetine or citalopram and using HDRS for symptom assessment. Changes in mean scores were evaluated using linear mixed models.

Results: In subjects displaying ≤20 in total HDRS score at baseline, representing 27% (n=2102) of the entire population, SSRI outperformed placebo in reducing depressed mood with an effect size of 0.44 (p<0.0001). Including only those displaying a baseline HDRS score of ≤19 (16%) or ≤18 (8%) yielded similar results (ES 0,39 and 0,42, respectively; p<0.0001 for both comparisons), as did including all subjects displaying ≤21 at baseline (36%, ES: 0.47; p<0.0001).

Conclusions: This comprehensive patient-level analysis reveals that SSRIs display a robust superiority over placebo with respect to reduction in depressed mood also in patients with relatively low HDRS rating at baseline. The results hence rebut the previous claim that SSRIs may exert a meaningful antidepressant effect only in patients with a very severe variant of this disorder (1).

REFERENCES: 1. Kirsch et al, PLOS Med 2008: 5: e45. 2. Khan et al, J Psychiatr Res 2005; 39: 145. 3. Melander et al, Eur Neuropsychopharm 2008: 18: 623. 4. Rabinowitz et al 2016; 209: 427. 5. Bagby et al, AM J Psychiatr 2004: 161: 2163. 6. Bech, 2010: Psychol Med 40: 181. 7. Hieronymus et al, Mol Psych 2016: 21: 523.

Keywords: Depression, Antidepressant, Clinical Trial Rating Methods

Disclosure: Nothing to Disclose.

W140. Features of Dissociation Differentially Predict Antidepressant Response to Ketamine in Treatment-Resistant Major Depression

Mark Niciu*, Bridget Shovestul, Brittany Jaso, Cristan Farmer, David Luckenbaugh, Lawrence Park, Elizabeth Ballard, Carlos Zarate

National Institute of Mental Health, Bethesda, Maryland, United States

Background: Ketamine induces rapid and robust antidepressant effects in major depressive disorder and bipolar depression. Due to its glutamatergic properties, many patients report dissociative effects, which our group has previously shown to be associated with its antidepressant response [Luckenbaugh et al. (2014) J Affect Disord 159:56-61]. Here, we conducted a follow-up with more subjects (inclusive of the previous report) to investigate if the antidepressant response of ketamine is uniquely related to dissociative symptom clusters [including those of Bremner et al. (1998) J Trauma Stress 11(1):125-36].

Methods: 126 treatment-resistant depressed patients (major depressive disorder: n=84, bipolar disorder: n=42) from three studies received a single subanesthetic dose (0.5 mg/kg) infusion of ketamine over 40 minutes. Dissociative effects were measured using the Clinician-Administered Dissociative States Scale (CADSS) at baseline and immediately after infusion. The Hamilton Depression Rating Scale (HDRS-17) was measured at baseline and at 230 minutes, 1 day, and 7 days post-infusion. First, a confirmatory factor analysis for ordered categorical indicators was performed to establish the factor validity of the CADSS subscales described by Bremner et al. (1998) (derealization, depersonalization and amnesia). Second, we fitted a general linear model with maximum likelihood estimation and repeated measures with a compound symmetry covariance structure (selected based on relative fit indices) to HDRS percent change from baseline to 230 minutes, 1 day, and 7 days post-infusion. For the purposes of comparison to our previous publication, we also performed Pearson correlations between the CADSS scores and HDRS percent change, by visit.

Results: The three-factor model was a good fit to the data, with a root mean square error of approximation of .06 (95% CI:.042 –.077), a comparative fit index of.97, and a Tucker-Lewis index of.96. The relationship between CADSS Total Score and HDRS change was negative but non-significant at 230 minutes [B=-0.25, SE=0.14, t(204)=-1.70, p=.09] and Day 1 [B=-0.10, SE=0.14, t(204)=-0.72, p=.47]. A significant effect was found between the total CADSS score and HDRS percent change at day 7 within one study only [B=-0.62, SE=0.30, t(204)=-2.09, p=.04]. Across all studies and time points, the depersonalization subscale was also related to HDRS percent change [B=-0.79, SE=0.38, t(118)=-2.07, p=.04]. A significant effect of derealization on HDRS percent change at day 7 was also found in one study only [B=-1.55, SE=0.65, t(204)=-2.37, p=.019]. No significant effects were found between the amnesia subscale and HDRS percent change (p >.05).

Conclusions: The findings of the current study extended our prior work such that overall levels of dissociation mediated the antidepressant effects of ketamine at day 7, but not at 230 minutes post-infusion, as was previously reported. Further, specific subscales of dissociation differentially correlate with antidepressant treatment response to ketamine. Greater depersonalization was predictive of antidepressant response at all time points across all patients, whereas increased derealization predicted antidepressant response only at day 7, and only in a subset of patients.

Keywords: Major Depressive Disorder, Bipolar Depression, Ketamine, Dissociation, Glutamate

Disclosure: Nothing to Disclose.

W141. Modulation of Nicotine-Induced Cognitive Performance Benefits by Galantamine

Britta Hahn*, Cory Olmstead, Marie Yuille, Megan McComas, Ashleigh Wells, Carolyn Reneski

University of Maryland School of Medicine, Baltimore, Maryland, United States

Background: Several disease states marked by cognitive deficits, most prominently schizophrenia, Alzheimer's disease, and ADHD, involve nicotinic acetylcholine receptor (nAChR) hypofunction and may benefit from nAChR agonist treatment. Cognitive benefits of the prototypical nAChR agonist nicotine are well established. Drug development efforts have been invested into subtype-selective nAChR agonists for the above conditions. Effects have generally been in the expected direction but tended to be of small magnitude and uncertain clinical significance. A way of raising the effect size ceiling may be to co-administer a nAChR positive allosteric modulator (PAM). PAMs do not activate nAChRs on their own but bind to a second, modulatory site and facilitate agonist-induced responses. Thus, combining a PAM with low doses of a nAChR agonist may also be more sparing of native circuit dynamics than larger doses of agonist alone.

The acetylcholinesterase (AChE) inhibitor galantamine (GAL) has been shown to be a PAM at several subtypes of the nAChR, including alpha4beta2, alpha3*, alpha6beta4, and alpha7 nAChRs, at concentrations found in the human brain at clinical doses. The concentration range for GAL's PAM action appears to be slightly below that for AChE inhibition; thus, by testing low doses, we aimed at inducing a bias towards its PAM action. Even at low doses, GAL has additional pharmacological actions, but potentiation of nicotine effects can be explained by allosteric potentiation, and not by AChE inhibition. To test the hypothesis that GAL can potentiate the cognitive-enhancing effects of nicotine, we conducted interaction studies in both humans and rats.

Methods: In the human study, 27 healthy adult never-smokers each underwent cognitive testing on 4 separate days. Across the 4 test days, participants were administered, in a 2 x 2 factorial design and in counterbalanced sequence, (1) a placebo patch and a placebo capsule, (2) a nicotine patch (7mg/24 hrs) and a placebo capsule, (3) a placebo patch and a GAL (4 mg) capsule, and (4) a nicotine patch and a GAL capsule.

In the rat study, 2 groups of 24 male Wistar rats were trained in the 5-choice serial reaction time task (5-CSRTT) for approximately 4 months. The animals were food-reinforced for detecting 1-s light stimuli, presented randomly in one of 5 apertures located along the curved rear wall of the operant conditioning chambers. In a within-subject design similar to the human study, rats were tested across 6 repeat test sessions with 2 doses of GAL (s.c.) and vehicle in the presence of nicotine (0.1 mg/kg s.c.) and vehicle (3 x 2 factorial design).

Results: Human study: (A) In a visuospatial attention task, nicotine significantly reduced both omission errors and reaction time. A non-significant trend suggested small additive effects of GAL on reaction time when attention had to be spread broadly. (B) In the Rapid Visual Information Processing Task (RVIPT), designed to induce information processing overload and tax sustained attention, nicotine significantly increased signal detection. GAL alone impaired signal detection and slowed reaction time; both effects were reversed by nicotine. (C) In a change detection task probing short-term memory, there were no effects on response accuracy, and neither nicotine nor GAL alone had significant effects on reaction time. However, the combination of nicotine and GAL acted synergistically and significantly reduced reaction time, a pattern consistent with positive allosteric potentiation.

Rat study: In both groups, nicotine reduced omission errors and response latency, and increased anticipatory responding in the intertrial interval. Unexpectedly, nicotine reduced the accuracy of responding, an effect explainable by its large rate-increasing effects as measured by anticipatory responses. In one group, GAL (0.5 & 1 mg/kg) dose-dependently reversed the effects of nicotine on anticipatory responding and accuracy, with a similar trend on omission errors. In the other group, lower doses of GAL (0.03 & 0.06 mg/kg) still reversed the effects of nicotine on accuracy, but had no effects the other measures.

Conclusions: A lack of improvement, and even impairment, with GAL alone in healthy subjects is not surprising given previous reports. The antagonistic effects of nicotine and GAL in the human RVIPT and in the rat 5-CSRTT were surprising and may involve residual AChE inhibition or other pharmacological actions of GAL. AChE activity assays, to be performed on human blood samples obtained on each test day, may further the interpretation of these interactions. Importantly, the nicotine x GAL interaction seen on reaction time in the change detection task differed qualitatively, reflecting synergistic effects, and can be explained by GAL's PAM action, but not by AChE inhibition. This finding provides proof-of-principle that positive allosteric potentiation of cognitive-enhancing nAChR agonist effects is possible. Other pharmacological actions of GAL may have overshadowed similar interactions on other measures, which may differ with regard to performance-optimal nAChR tone. Overall, GAL's behavioral profile was not favorable in the unimpaired organism, neither human nor rat. However, the present findings encourage the study of positive allosteric potentiation of cognitive nAChR agonist effects with newly developed pure nAChR PAMs devoid of additional pharmacological actions.

This work was supported by NIH grant R01DA035813 (B. Hahn).

Keywords: Nicotine, Galantamine, Cognition, Positive Allosteric Modulators

Disclosure: Nothing to Disclose.

W142. An Open Label, Non-Randomized Update Suggests Focal Electrically Administered Seizure Therapy (FEAST) may Have a Reduced Time to Re-Orientation Compared to Right Unilateral Ultra-Brief Pulse Electroconvulsive Therapy (UBP-RUL ECT)

Gregory Sahlem*, E.Baron Short, W. Vaughn McCall, Peter B. Rosenquist, James B. Fox, Andrew J. Manett, Ziad H. Nahas, Cecile M. Mazingue, Mark S. George, Harold Sackeim

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Focal Electrically Administered Seizure Therapy (FEAST), is a novel form of Electroconvulsive therapy (ECT), designed to restrict the focus of the electrical stimulus to the frontal lobe, thereby reducing electrical stimulus elsewhere in the brain. Recent small open label studies have suggested that FEAST's enhanced focality may confer a reduced cognitive side effect burden (1, 2). To date there has been no work directly comparing FEAST, to other forms of conventional ECT. We subsequently continued the development of this novel form of ECT by comparing FEAST's effects to those of Right Unilateral Ultra-Brief Pulse (UBP-RUL), which has previously been shown to have the most favorable profile of cognitive side effects.

Methods: In a non-randomized, open label fashion, 23 depressed adults were recruited after being referred for ECT at MUSC, or AU. Sixteen received a course of FEAST (11 women; 2 bipolar disorder; age 44.7 13.8), and 7 received a course of UBP-RUL (5 women; 2 bipolar disorder; age 41.4 15.3). Both forms of ECT were delivered using a MECTA spECTrum 5000Q device (modified in the case of FEAST). Clinical and cognitive assessments were obtained at baseline, and end of the treatment course. The time to orientation recovery, a predictor of amnestic effects, was assessed at each treatment and served as our primary marker of cognitive side effect burden. Efficacy was assessed using the 24-item Hamilton Rating Scale for Depression (HRSD24).

Results: Those participants receiving FEAST, had a numeric reduction in median time to reorientation of 5.4 5.0 minutes (from eyes open until 4/5 orientation), as compared to those participants receiving UBP-RUL, which were oriented in 8.5 8 minutes (p=0.3, t-test). There was a numeric advantage for both average reduction in HRSD24 (average reduction: FEAST=58 18.5%; UBP-RUL=69 16.9%; p=0.18, t-test), and rate of response (50% reduction in HRSD24) in favor of UBP-RUL compared to FEAST (FEAST=62%; UBP-RUL=85%).

Conclusions: This data further supports the hypothesis that treatment with FEAST may have a milder cognitive side effect profile as compared to conventional ECT (UBP-RUL). While FEAST displays a marked antidepressant effect, it continues to remain unclear if it is as efficacious as UBP-RUL. Further study, specifically using randomized, blinded methodology, are needed to make definitive conclusions regarding both the clinical efficacy, and the cognitive side effect burden of FEAST as compared to UBP-RUL.

Keywords: Depression, Electroconvulsive Therapy, ECT, FEAST, Focal Electrically Administered Seizure Therapy

Disclosure: Nothing to Disclose.

W143. Immobility in the Presence of a Social Stressor is Unrelated to Corticosterone Response and General Locomotor Activity in the Balb/c Mouse

Jessica Burket, Jerrah Pickle, Allison Edwards, Bronson Haynes, Stephen Deutsch*

Eastern Virginia Medical School, Norfolk, Virginia, United States

Background: The locomotor activity of the Balb/c mouse, a model of autism spectrum disorder (ASD), is suppressed in the presence of an enclosed or freely-behaving stimulus ICR mouse in the standard 3-chamber apparatus. Moreover, relative to the Swiss Webster comparator strain, Balb/c mice display fewer discrete episodes of social approach, anogenital sniffing and mounting toward the ICR social stimulus mouse, among other deficient prosocial behaviors. Whereas the exploratory behavior of the Balb/c mouse toward an inanimate object did not differ from the Swiss Webster strain, its exploration of an enclosed stimulus mouse was significantly less than the comparator strain. Interestingly, Balb/c and Swiss Webster mice did not differ in their number of open-arm entries in the elevated-plus maze (EPM). The current study examined whether Balb/c and Swiss Webster mice differed in terms of their immobility in the presence of enclosed and freely-behaving ICR stimulus mice, and whether immobility was related to locomotor activity in the absence of the stimulus mouse or the corticosterone response after exposure to the stimulus mouse.

Methods: Animals: 4-week old male Swiss Webster and Balb/c test mice (Harlan Laboratories, Frederick, MD, USA) were housed 2 per cage in hanging clear Plexiglas cages with free access to food and water, and maintained on a 12 h light/dark cycle. The stimulus mice were 4-week old male ICR mice (Charles River Laboratories, Wilmington, MA, USA), housed 4 per cage. All animal procedures were approved by the Eastern Virginia Medical School Institutional Animal Care and Use Committee and conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

Apparatus: The 3-chamber sociability apparatus consisted of a black Plexiglas rectangular box (52.07 cm x 25.40 cm x 22.86 cm), without a top or bottom. The center chamber was smaller (12.07 cm x 25.40 cm) than the two equally-sized end chambers (19.05 cm x 25.40 cm). Inverted wire cups to house the stimulus mouse (Galaxy Cup, Kitchen Plus) were placed in each side of the end chambers during Sessions I and II. Experiments were conducted in dim lighting (< 3.5 lux) and videotaped using Sony HD Video Cameras (Sony Corp., Tokyo, Japan) in nightshot mode with infrared lighting.

Sociability Procedure: In the first 10-min Session, a test mouse was placed in the middle chamber, allowed to acclimate, and transitions between chambers, a measure of locomotor activity, was measured. In the second 10-min Session, a stimulus mouse was enclosed in an inverted wire cup in the side designated the social chamber, and an empty inverted wire cup placed in the nonsocial chamber; the location of the enclosed stimulus mouse was randomly assigned in counterbalanced fashion. The amount of time test mice spent in the social and nonsocial chambers, the amount of time test mice spent engaged in exploring (sniffing) behavior within a 2-cm vicinity of the social and nonsocial inverted cups, and their transitions between chambers were measured. In the third 10-min session, the stimulus mouse was released from the inverted wire cup, and the test and stimulus mice were allowed to interact freely. Transitions between chambers and operationally-defined measures of sociability and stereotypic behavior were reliably obtained and analyzed. Immobility was measured (discrete episodes, seconds, and seconds/discrete episode of immobility) in sessions II and III, which was defined as stationary (i.e. not moving) and not engaged in any social or stereotypic behavior. Serum was prepared from trunk blood obtained≤30 min after the end of session III and stored at -80°C. Serum was diluted 1:70 for measurement of corticosterone levels by Enzyme-linked immunosorbent assay (ELISA; Abcam). Dilutions were made in the sample diluent provided with the kit and the assay procedure followed the manufacturer's protocol.

Results: Balb/c mice showed no significant relationship between transitions during session I and immobility in sessions II and III. On all measures of immobility, Balb/c mice were significantly more immobile than the Swiss Webster comparator strain in both session II and III. Serum corticosterone levels drawn≤30 min after the end of session III were significantly higher in the Balb/c, compared to the Swiss Webster, mice. Corticosterone levels did not differ between groups of Balb/c mice (N=6) selected for the highest and lowest immobility scores (seconds) in either session II or III.

Conclusions: The 3-chamber apparatus provides a platform for reliable assessment of mouse social behavior; it is widely used to characterize models of ASD, the impact of genetic variations on social behavior, and screen compounds for prosocial effects. A limitation of the 3-chamber apparatus is a reliance on locomotor behavior to assess sociability. In the current study, no relationship was found between locomotor activity of the Balb/c mouse in the absence of a social stimulus and its immobility in the presence of an enclosed or freely-behaving stimulus mouse. Balb/c mice were, however, significantly more immobile than Swiss Webster mice, irrespective of whether the stimulus mouse was enclosed or moving freely. Importantly, the intensity of immobility of Balb/c mice in the presence of a stimulus mouse did not appear to be a consequence of their heightened corticosterone response to a social stress as Balb/c mice selected for high or low immobility did not differ in terms of serum corticosterone levels. Thus, in Balb/c mice, the stressfulness of a social encounter alone is not the sole determinant of their increased immobility in the presence of a social stimulus; perhaps, these mice also display an element of social “disinterest”?

Keywords: Balb/c Mouse, Autism Spectrum Disorder, Psychosocial Stress, Immobility, Corticosterone

Disclosure: Nothing to Disclose.

W144. Epigenetic Signatures of One-Month Cortisol Exposure and Emotional Stress in Healthy Individuals

Richard Lee*, Peter Zandi, Kelly Benke, Xiaoju Yang, Olivia Cox, Betsy McCaul, Gary Wand

Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Background: Chronic exposure to the stress hormone cortisol is a significant component of allostatic load, and is associated with cardiovascular and metabolic disorders. Moreover, the brain is especially susceptible to cortisol, as prolonged exposure is associated with an increase in neuropsychiatric symptoms, decrease in hippocampal volume, and cognitive decline. Given the detrimental impact of chronic exposure to cortisol, an easily accessible biological measurement capable of capturing long term cortisol exposure would be immensely useful for testing the predictability of cortisol load on susceptibility to cardio-metabolic and psychiatric disorders. Unfortunately, this measure has yet to be discovered.

While levels of cortisol can be measured from various peripheral sources such as urine, blood, saliva, and hair, cortisol measurements from these sources either do not adequately reflect long term cortisol exposure or are cumbersome to perform. Lack of accuracy of these measurements can be attributed to the pulsatile nature of cortisol secretion. Cortisol measured in urine, blood, and saliva tend to only reflect the most recent circadian and stressful events. While the assessment of 24-hour urine free cortisol levels provides a metric for a single day's cortisol exposure, other cortisol measures in urine, saliva or blood provide even less time-based information. Unless multiple samplings are measured, an accurate assessment of chronic cortisol exposure is not possible. Although measurement of cortisol in hair samples is thought to reflect long term exposure to cortisol, sample processing is cumbersome and not always possible especially in individuals with short hair.

Methods: We sought to determine whether molecules found in blood may be able to successfully capture past cortisol exposure. To this end we created a model of one month of cortisol that could be interrogated to find blood-based biomarkers of chronic cortisol exposure. In this study, awakening and bedtime saliva was collected over 30-39 consecutive days and assayed for cortisol in healthy participants. We then sought to identify molecular targets that can capture 30-day cortisol exposure from one blood sampling. Using genomic DNA extracted from blood after the 30+ days, we implemented Agilent's human Methyl-Seq platform on more than 84 individuals with a wide range of cortisol levels and emotional stress. The genome-wide human Methyl-Seq platform targets 3.7 million CpGs residing in CpG islands and their shores, Gencode promoters, Ensembl regulatory elements, and DNaseI hypersensitive sites. For analyzing the human data, we used the bumphunter analytical package that enabled us to perform SVA (surrogate variable analysis) to account for unmeasured sources of heterogeneity, including cellular heterogeneity in blood.

Results: We identified multiple differentially methylated CpG dinucleotides that correlated significantly with individual 30-day mean awakening or bedtime cortisol and emotional stress measurements: IRF1 (R=0.52, P=1.1x10-5, Emotional Stress), OR1F1 (R=0.53, P=5.5x10-5, Awakening Cortisol), NAV1 (R=0.36, P=0.0016, Bedtime Cortisol), LINC01168 (R=0.56, P=1.0x10-4, Emotional Stress), and GLB1L (R=0.21, P=0.8, Awakening Cortisol). We further validated these findings by an independent, candidate-gene approach called bisulfite pyrosequencing, which measures DNA methylation variations at greater than 90% precision.

Conclusions: We are currently in the process of testing our methylation loci in several cohorts of disease, such as blood DNA from patients with Cushing's Disease, cord blood samples of infants exposed to maternal stress, and blood samples from individuals with mental disorders. Identification and characterization such markers of allostatic load have the potential to provide insightful information on susceptibility to cardio-metabolic and psychiatric diseases.

Keywords: Epigenetics, DNA Methylation, Acute and Chronic Stress, Biomarker

Disclosure: Nothing to Disclose.

W145. Large-Scale Cognitive GWAS Suggests Novel Nootropic Targets

Max Lam Zhan Yang, Anil Malhotra, Todd Lencz*

Hofstra Northwell School of Medicine, Glen Oaks, New York, United States

Background: Neurocognitive deficits represent a critical component of many neuropsychiatric disorders and disease states that can affect health outcomes across the lifespan. Recently, a large (N~78,000) genome-wide association study (GWAS) reported 18 genome-wide significant loci for cognitive performance as measured primarily by brief assessment (Sniekers et al. 2017, Nature Genetics, PMID: 28530673). At the same time, the Cognitive Genomics Consortium (COGENT), utilizing a PCA-derived measure of general cognitive ability, has reported two loci from its own collection of >35,000 subjects in 24 cohorts (Trampush et al. 2017, Molecular Psychiatry, PMID: 28093568). In the present report, we meta-analyze results of these two studies to perform the largest cognitive GWAS to date. To further boost power to detect loci associated with cognitive ability, we utilized a novel approach, Multi-Trait Analysis of GWAS (MTAG; Turley et al. 2017, biorxiv) to combine our results with related data from a large GWAS on educational attainment (Okbay et al. 2016, Nature, PMID: 27225129).

Methods: After accounting for overlapping cohorts, the combined dataset for cognitive ability included N=107,207 unique individuals. All subjects were of European ancestry and were drawn from the general population. Genotype data across all cohorts was imputed using the 1000 Genomes v3 reference panel, resulting in ~8M high-quality SNPs. GWAS summary statistics were combined for meta-analysis using the sample-size weighted method in METAL, with gene-wise and pathway results examined using MAGMA. MTAG was applied to summary statistics from this dataset, combined with data from ~300,000 individuals of European ancestry assessed for educational attainment (years of schooling). Genetic overlap with other GWAS phenotypes was examined using the LD score regression (LDSC) method.

Results: Meta-analysis of the cognitive performance samples (N=107K) yielded 28 genome-wide significant independent loci, 10 of which were novel. Moreover, 88 genes were implicated in the MAGMA analysis, which identified “neuron spine” as the most over-represented biological pathway. Inclusion of the educational attainment in MTAG analysis boosted the signal of our cognitive GWAS by ~500% and resulted in 71 genome-wide significant loci, of which 39 are novel to either phenotype. LDSC analyses of the meta-analytic and MTAG results revealed several novel genetic correlations, including a surprising strong correlation with parental age at death (rg~.4-.5). Transcriptomic analysis (TWAS) suggested the potential for differing genetic effects in cortex (neurite outgrowth), hippocampus (post-synaptic density), and cerebellum (inflammation). Drug related gene-set enrichment analysis indicated that specific drugs targeting calcium and potassium channels may be worthy of investigation for cognitive enhancing effects.

Conclusions: Sample sizes for GWAS of cognitive ability have finally attained critical mass to uncover increasing numbers of significant loci associated with brain function. Genetic overlap analysis reveals very strong correlations with psychiatric disorders, as well as with many health-relevant phenotypes. Gene set enrichment analysis points to significant roles for calcium and potassium channels as treatment targets. Based on the GWAS power curves that have been observed for other complex phenotypes, further increases in sample size will demonstrate substantially enhanced yield.

Keywords: GWAS, Cognition, Cognitive Enhancement

Disclosure: Nothing to Disclose.

W146. Sibling Risk Across Psychiatric and Neurodevelopmental Disorders and Quantitative Brain Phenotypes: Psychiatric Cross Disorder Risk

Mark Weiser*, Daphna Fenchel, Dorit Tzur, Shimon Burshtein, Sven Sandin, Michael Davidson, Eyal Fruchter, Abraham Reichenberg

Sheba Medical Center, Ramat Gan, Israel

Background: While phenomenological classifications such as DSM and ICD define psychiatric disorders as separate diagnostic entities, many symptoms, risk factors and treatments are shared across different diagnostic entities. In addition, many of the genes associated with a given psychiatric disorder are associated with other disorders. The objective of this study was to utilize diagnostic data from screening of an entire population to further understand shared familial, perhaps genetic risk, across psychiatric diagnostic groups.

Methods: Subjects were adolescents (ages 16-17) undergoing mandatory screening for eligibility to serve in the Israeli military, between the years 1998-2014. We compared the risk of psychiatric disorders in siblings of probands with psychotic disorder (n=9028), mood disorders (n=12,639), anxiety disorders (n=16,394), personality disorders (n=25,211), intellectual disability (n=13,667), autism spectrum disorders (ASD) (n=3400) and low cognitive ability (defines as IQ<2 standard deviations below population mean, n=23,226) to the risk of psychiatric disorders in siblings of controls matched for age and sex of the probands. Non-psychiatric controls were siblings of probands with type-1 diabetes (n=3638), and hernia (n=36,806). Odds ratios (OR's) were adjusted for sex, socio-economic status and year of birth.

Results: Siblings of probands with all psychiatric disorders were at increased risk for all psychiatric disorders examined and for low cognitive ability (most ORs ranging 2-3). Higher risks were found among siblings of probands with psychotic disorder for psychotic disorder (OR=9.277, 95% CI=8.668-9.929), among siblings of probands with intellectual disability for intellectual disability (OR=9.537, 95% CI=8.817-10.317) and for ASD (OR=7.539, 95% CI=6.335-8.970) and among siblings of probands with ASD for intellectual disability (OR=6.879, 95% CI= 5.972-7.924) and for ASD (OR=11.53, 95% CI=9.239-14.403). In comparison, siblings of probands with non-psychiatric illnesses (type-1 diabetes or inguinal hernia) were at increased risk for concordant disorders, but not for any of the psychiatric diagnoses.

Conclusions: In this large population based study, there appears to be a large shared risk among different psychiatric diagnostic groups, with specifically increased genetic risks in psychotic disorder, intellectual disability and ASD. Molecular studies should continue in their attempts to identify both the shared and the specific genetic variations associated with different psychiatric diagnostic groups.

Keywords: Psychiatric Cross Disorder Risk, Epidemiology, Neurodevelopmental Disorders, Psychotic Disorders

Disclosure: Nothing to Disclose.

W147. Antidepressant Effects on the Pain Network in Patients With Depression

Yun Wang*, Joel Bernanke, Bradley Peterson, David J. Hellerstein, Jonathan Posner

Columbia University/New York State Psychiatric Institute, New York City, New York, United States

Background: Depression is a common disorder affecting 9.5% of the population and conferring significant negative long-term sequela. Antidepressant medications offer an effective treatment, yet nearly 10%-30% of patients either do not respond or have side effects rendering them unable to continue the course of treatment. More specific and targeted treatments are needed to address this significant public health concern. An important pathway to foster new treatment development is to identify potential targets for novel interventions by isolating the mechanisms by which existing treatments exert their salutary effects. Coupling resting state functional MRI (rs-fMRI) with placebo-controlled, randomized clinical trials of antidepressants, we aimed to identify antidepressant treatment mechanisms and assess the reproducibility of the findings based on two, independent datasets.

Methods: We conducted two independent antidepressant treatment studies with two different SNRI medications: duloxetine and desvenlafaxine. Both studies included a 10-week (duloxetine) or 12-week (desvenlafaxine) prospective, double-blinded, placebo-controlled trial with pre- and post-treatment MRI scans. The duloxetine cohort consisted of n=48 adults with depression, n=24 randomized to duloxetine, and n=24 to placebo. The active treatment arm (i.e. duloxetine) had 42 usable pre-treatment rs-fMRI scans, and 36 usable post-treatment scans. The placebo arm had 42 usable pre-treatment scans, and 36 usable post-treatment scans.

The desvenlafaxine cohort consisted of n=44 adults (n=21 desvenlafaxine; n=23 placebo). The active treatment arm (i.e. desvenlafaxine) had 38 usable pre-treatment scans, and 36 usable post-treatment scans. The placebo arm had 44 usable pre-treatment scans, and 38 usable post-treatment scans.

For both studies, rs-fMRI scans were obtained on a 3T GE scanner with two 5-minute resting state runs (TR=2.2s, echo time/flip angle=30ms/90, voxel resolution=3.753.753.5mm). Standard imaging preprocessing methods were performed with SPM12. To address potential spurious correlations caused by head motion, we used the ART toolbox to detect outlier images defined as frame-wise displacement (FD)>0.5mm or signal intensity changes>3 standard deviations.

Treatment effects were tested with the following three steps. First, we used a data-driven approach, multi-voxel pattern analysis (MVPA), to examine functional connectivity across all voxels in the brain. MVPA calculates whole-brain connectivity between each voxel and the rest of the brain and reduces data dimensionality with principal component analysis. Second, we used a meta-analysis method, multilevel kernel density analysis (MKDA), to evaluate the consistency of findings across our two studies. This revealed several brain regions in which there was a significant Group Time interaction across both studies; 14 out of 19 of these regions were within a previously defined neural network – the pain network (e.g., insula, supramarginal gyrus, thalamus, superior temporal gyrus, periaqueductal gray). Third, based on results from a meta-analysis of our two studies, we conducted post-hoc testing using graph theory to examine antidepressant effects on the pain network.

The pain network regions-of-interest (ROI) were constructed with radius=5mm and coordinates described elsewhere. The average time series were extracted from 16 ROIs separately (i.e. all of the nodes that co-activate with pain). The time series were correlated region by region for each participant yielding a 1616-correlation matrix per subject. Network density, defined as the mean of all possible connections, was calculated using UCINET.

Results: Repeated-measure analysis of network density showed a significant Group Time interaction in both studies (duloxetine: F=4.34, p=0.04; desvenlafaxine: F=18.45; p=0.0002) Post-hoc t-tests confirmed (i) non-significant pre-treatment differences between the active vs placebo arms (p's>0.1) (ii) non-significant pre- vs post-treatment changes in network density in the placebo arm (p's>0.6) and (iii) significant pre- vs post-treatment reductions in network density in the active treatment arms (p=0.022 for duloxetine, p=0.00015 for desvenlafaxine).

We also combined the two datasets together into one and performed similar analyses as above. A significant interaction (F=9.02, p=0.0039) and reduction in pre- vs post-treatment network density (p<0.0005) was found. Reduction in pain network density correlated with improvement in clinical symptoms, as measured by the 24-item HAM-D (r=0.425, p=0.017)

Conclusions: Depressed patients frequently describe depression as ‘psychic pain’ and often experience increased physical pain; and after successful antidepressant treatment, both depressive and pain symptoms often improve4. However, mechanisms underlying these effects have been unclear. This is the first study demonstrating a potential causal pathway by which antidepressant treatment reduces depressive symptoms that are reliably detected across two, independent placebo-controlled studies. Subsequent research may examine several persisting questions such as whether other antidepressants have similar effects on the pain network, and whether reductions in pain network density cause improvements in depressive symptoms through direct effects on mood, or if these improvements are indirectly mediated by reductions in pain (or somatic) sensitivity.

Keywords: Adolescent Depression, Pain, Antidepressants, MVPA, Meta-Analysis

Disclosure: Nothing to Disclose.

W148. A Decade of Psychiatric GWAS Research

Chunyu Liu*, Tanya Horwitz, Katie Lam

University of Illinois at Chicago, Chicago, Illinois, United States

Background: After more than ten years of accumulated efforts, genome-wide association studies (GWASs) have accumulated many findings, most of which have been deposited into the GWAS Catalog. The major questions about the GWAS have been: 1) Are those findings reliable? 2) Are those associated SNPs causal? As such GWAS findings are often selectively and mistakenly interpreted by the media, the public, clinicians, patients, and even some investigators in related fields, we feel the obligation to provide a scholastic, holistic, and stringent summary of the GWAS data. In doing this, we aim to identify SNPs that have been repeatedly reported in association with the same disorder and that should thus bear better confidence. We also hope to draw capture hidden connections between disorders or traits by highlighting genomic regions that could be associated with multiple conditions.

Methods: Between GWAS's inception and the time the dataset was downloaded, the catalog had collected a total of 2,430 papers, 1,818 phenotypes, and 28,463 SNPs. We reclassified the psychiatry-related traits into 198 reclassified phenotypes, which accounted for 472 papers and 6,564 SNPs.

We summarized the phenotypes relevant to psychiatric disorders by identifying SNPs replicated across studies for the same phenotype. We also identified hidden connections between disorders or traits by highlighting SNPs that were associated with multiple phenotypes.

Results: In total, 151 SNPs were replicated for the same psychological trait in different studies. The SNPs rs2075650 and rs4420638 were the most replicated SNPs within a single reclassified phenotype; both were associated with Alzheimer's disease. Schizophrenia was associated with 87 SNPs, although many of these replications came from the same two studies. Alzheimer's disease and schizophrenia were also linked to a large volume of physical phenotypes such as blood pressure, cholesterol, and BMI. Alzheimer's disease also shared risk SNPs with age-related phenotypes such as age-related macular degeneration, cognitive decline, and longevity. Alcohol-related phenotypes shared SNPs with many physical phenotypes related to diseases—coronary heart disease, esophageal cancer, oral cavity and pharyngeal cancer, and cardiovascular disease risk. Nicotine-related phenotypes were linked to lung cancer, and a series of respiratory functions. Two separate studies also identified the same risk SNP for bipolar disorder and educational attainment, a link that has been previously discussed in genetic and non-genetic research alike.

Conclusions: A SNP association does not translate directly to a causal relationship, the phenotypes can be the direct consequence of the associated variants, can also be another phenotype that is further downstream in the cascade. Many of the uncovered the cross-phenotype associations do cluster by the apparent pathological relevance. These findings indicate the importance of studying the complex relationship between genotypes and phenotypes, as well as those among different phenotypes. Meanwhile, the replicated SNPs identified in this review can be the best candidates for functional studies to get to the biological mechanisms.

Keywords: GWAS, Neuropsychiatry, Population Genetics

Disclosure: Nothing to Disclose.

W149. Mesocorticolimbic Connectivity in Humans With a DCC Mutation

Daniel Vosberg*, Yu Zhang, Amanda Chalupa, Dominique Allard, France Durand, Alain Dagher, Chawki Benkelfat, Myriam Srour, Ridha Joober, Franco Lepore, Guy Rouleau, Hugo Theoret, Cecilia Flores, Marco Leyton

McGill University, Montreal, Canada

Background: The axon guidance molecule receptor DCC (deleted in colorectal cancer) plays a critical role in directing mesocorticolimbic dopamine pathways. Rodent studies indicate that DCC haploinsufficiency leads to increased dopamine projections to the medial prefrontal cortex (mPFC) (Manitt et al J. Neurosci. 2011) and decreased mesolimbic dopamine innervation (Reynolds et al Biol Psychiatry 2017). To investigate this in humans, we have identified and recruited members of a Quebec family who are haploinsufficient for DCC. It was predicted that DCC mutation carriers would exhibit (i) altered structural and functional connectivity between the midbrain and projection sites in the ventral striatum and mPFC, and (ii) a behavioral phenotype consistent with low striatal dopamine transmission, comprising diminished novelty seeking and reward seeking traits, and increased punishment sensitivity.

Methods: 20 DCC+/- mutation carriers, 14 DCC+/+ relatives, and 20 DCC+/+ unrelated healthy controls participated in this study. Anatomical and functional connectivity of the substantia nigra and ventral tegmental area (SN/VTA) were measured using diffusion weighted and resting-state functional magnetic resonance imaging. Anatomical connectivity was then evaluated using probabilistic tractography and functional connectivity using Pearson's correlations of the mean time-course for each area. Personality traits were measured using the Temperament and Character Inventory (TCI) and the Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ).

Results: Compared to the control groups, DCC+/- carriers exhibited reduced anatomical connectivity from the SN/VTA to ventral striatum and ventral medial prefrontal cortex. Functional connectivity analyses revealed similar alterations between midbrain and medial prefrontal cortex. Individual differences in anatomical and functional connectivity co-varied with the personality traits of novelty seeking, sensitivity to punishment, and sensitivity to reward.

Conclusions: Together, these data provide the first evidence that an axon guidance gene is involved in mesocorticolimbic connectivity in humans. By influencing personality traits, they might also affect susceptibility to putative mesocorticolimbic related psychiatric disorders.

Keywords: Axon Guidance Molecule Genes, Neurocircuitry, Neurodevelopment, Corticostriatal Connectivity, Magnetic Resonance Imaging

Disclosure: Nothing to Disclose.

W150. Disrupted Cerebro-Cerebellar Functional Connectivity in High-Functioning Autism Spectrum Disorder: A Whole-Brain, Data-Driven, High Temporal Resolution fMRI Study

Sheeba Anteraper*, Xavier Guell, Gagan Joshi

Massachusetts Institute of Technology, Cambridge, Massachusetts, United States

Background: A recent review of resting-state functional connectivity (RsFc) MRI studies in Autism Spectrum Disorder (ASD) reveals both over- and under-connectivity in ASD (Hull et al., 2016). While the majority of such investigations were carried out with an apriori hypothesis, posing a limit to the number of brain regions examined, we aimed to follow an unbiased whole-brain data-driven approach to look for potential RsFc differences between healthy controls (HC) and high-functioning ASD patients (HF-ASD).

Methods: Connectivity analyses were carried out using CONN (Whitfield-Gabrieli et al., 2012). Whole-brain networks (Default Mode, Sensorimotor, Visual, Salience/Cingulo-Opercular, Dorsal-Attention, Fronto-Parietal/Central Executive, and Language) derived from Independent Component Analyses on the Human Connectome Project (HCP) dataset (n=497) were used for functional connectivity analyses on the Autism Brain Imaging Data Exchange dataset (n=59; 35 HC and 24 HF-ASD) (http://fcon_1000.projects.nitrc.org/indi/abide/). Resting scan images were obtained on 3T Siemens Skyra (TR/TE/TI/Flip Angle were 475 ms/30 ms/60°, with 3 mm voxels, Multiband factor of 8, 947 time points). These high temporal resolution parameters provide an unprecedented statistical analysis power and de-noising capabilities in RsFc studies of HF-ASD. T1-weighted structural images were acquired using 3D-MPRAGE (TR/TE/TI/Flip Angle were 2200 ms/2.88 ms/794 ms/13°, with 0.8 mm voxels). Prior to connectivity analysis, data were realigned, normalized and spatially smoothed with a 6-mm kernel using SPM12. Band pass filtering was executed at 0.008-0.1 Hz. Physiological sources of noise (signals from white matter and cerebrospinal fluid) were regressed out together with movement related covariates, their first-order derivatives, and motion outliers (scan-to-scan threshold of 0.5 mm). Correlation coefficients were converted to normally distributed scores using Fisher's r-to-z transform to allow for second-level analyses. Cerebellum clusters were visualized using the SUIT toolbox (Diedrichsen et al., 2015). There was no significant difference in total number of valid scans between the HF-ASD and HC groups. However, maximum head movement was significantly higher in the HF-ASD group; therefore, we regressed out the effect of maximum motion in the second-level analyses.

Results: Of the seven networks examined, only the mean of the seed regions of interest (ROIs) within the language network (bilateral inferior frontal gyrus and bilateral posterior superior temporal gyrus (pSTG)) stood out as statistically significant (p<0.001) for the between group seed-to-voxel comparison. Compared to HC, distinctly lower correlations between the language network and left cerebellum (Crus I, II and VIIb) were apparent within the HF-ASD group. Of the four ROIs within the language network, the biggest contribution to the between group difference was from bilateral pSTG, and more specifically left pSTG. In addition, a correlation analysis between clinical scores of autism (total Autism Diagnostic Observation Schedule, ADOS) and RsFc of the whole brain (on both language network and bilateral pSTG) revealed significant inverse correlations with left cerebellar functional connectivity (p<0.001).

Conclusions: Whole-brain coverage is a trade-off when attempting to achieve high temporal resolution in RsFc studies. Using a synergistic combination of 32 Channel head coil and high-temporal resolution simultaneous multi-slice imaging, we provide clear evidence of cerebro-cerebellar differences in HF-ASD within the language network. Multiple studies signal that the cerebellum may pay a crucial role in the pathophysiology of ASD (Rogers et al., 2013; Wang et al., 2014). Previous reports show white matter differences (reductions in franctional anisotropy) in autism within the STG using diffusion tensor imaging (Lee et al., 2007), as well as perfusion (regional cerebral blood flow) abnormalities specifically within the left STG with inverse correlations with autism trait severity (Meresse et al., 2005). Within the ASD population, pSTG hypo-activity pertaining to social deficits (Redcay et al., 2013); pSTG hypo-connectivity as a predictor of deficits in emotion recognition (Alaerts et al., 2014); and pSTG differences in the mirror neuron system (Fishman et al., 2014) have also been indicated. However, the cerebellar contributions to such abnormalities, including alterations in the RsFc between cerebellum and pSTG, are yet to be reported.

Our study reveals disruption of cerebro-cerebellar functional connectivity in adults with HF-ASD, specifically between bilateral pSTG and left cerebellar regions. fMRI task-based activity maps from the face matching task (participants decide which of two angry/fearful faces on the bottom of the screen match the face at the top of the screen, contrasted with the same task performed with shapes instead of faces) from an independent cohort (HCP, n=787) largely overlapped with regions of abnormal cerebellar connectivity in our group of HF-ASD patients. This overlap may inform the functional interpretation of our findings, suggesting a disruption of the cerebro-cerebellar connectivity that underlies language, social and emotion processing in the healthy brain. Notably, decreased connectivity between these areas correlated with autism clinical scores. Since the cerebellum is an under reported territory in RsFc studies compared to the neocortex, establishing the cerebellar basis of neural sub-type of autism could serve as a target for therapeutic interventions for aiding diagnosis, and for determining predictors of response to such interventions for managing the disorder. Over all, the findings we report advance our understanding of the neurobiology of HF-ASD, and support the utility of whole-brain, data-driven, high-temporal resolution data in the development of biomarkers and therapeutic approaches in psychiatry.

Keywords: Autism Spectrum Disorder, Resting State Functional Connectivity, Cerebellum

Disclosure: Nothing to Disclose.

W151. The Effects of Ketamine on Dopaminergic Function: A Meta-Analysis and Review

Michelle Kokkinou*, Abhishekh Ashok, Oliver Howes

Psychiatric Imaging Group, Robert Steiner MR Unit, MRC London Institute of Medical Sciences (LMS), Hammersmith Hospital, Imperial College London, London, United Kingdom

Background: Ketamine, a non-competitive N-methyl-D-aspartate receptor antagonist, has recently been found to have rapid and robust anti-depressant effects (Berman et al., 2000, Biological Psychiatry, 47, 351-354). Additionally, ketamine induces psychotomimetic effects (Stone et al., 2014, Psychopharmacology, 231, 2107-2116) and it is a club drug of abuse (Morgan and Curran. 2012, 107, 27-38). These effects have been linked to the dopaminergic system.

Methods: In order to comprehensively review the evidence for the effects of ketamine on the dopaminergic systems, we systematically searched the PubMed and PsychInfo electronic databases for studies from July 1972 to mid-July 2016. Total of forty original peer-reviewed studies which compared dopamine measures in the rodent, human and primate brain following acute and chronic ketamine administration relative to a drug free baseline or control condition were identified. Importantly there were adequate rodent studies of the effects of acute ketamine administration at sub-anaesthetic doses to permit meta-analysis.

Results: Here we show that compared to control conditions, acute ketamine administration in rodents is associated with significantly increased dopamine levels in the cortex (Hedge's g= 1.33, [95% confidence interval (CI), 0.81-1.85], p <0.01), striatum (Hedge's g= 0.57, [95% confidence interval (CI), 0.05-1.10], p <0.05) and the nucleus accumbens (Hedge's g=1.30, [95% confidence interval (CI), 0.14-2.45], p<0.05) and 62-180% increase in dopamine neuron population activity. Interestingly sub-analysis indicated that increases were more pronounced in in vivo (g=1.93, [95% confidence interval (CI), 1.40-2.45]) than ex vivo (g=0.50, [95% confidence interval (CI), -0.03-1.02]) studies. Additionally, there were not enough studies for meta-analysis in other brain regions studied, or of the effects of chronic ketamine administration. Elevations in dopamine levels in the cortex (from 88% to 180%) were reported in the studies of chronic ketamine administration. Of note, no study reported alterations in dopamine levels at anaesthetic doses (>100mg/kg) of ketamine ex vivo, although this remains to be examined in vivo. Lastly, findings of dopaminergic function in non-human primates and in human studies following acute ketamine administration compared to control using positron emission tomography (PET) were not consistent.

Conclusions: Our meta-analysis findings highlight that acute ketamine administration leads to elevation in dopamine levels in the rodent brain. Importantly future studies are needed to determine the dopaminergic function following chronic administration of ketamine and the effects of higher doses of ketamine in primate brain.

Keywords: Ketamine, Dopamine, Meta-Analysis

Disclosure: Nothing to Disclose.

W152. Valbenazine in Tardive Dyskinesia of Older and Younger Adults: Pooled Results From 3 Randomized, Double-Blind, Placebo-Controlled Trials

Martha Sajatovic*, George Alexopoulos, Khodayar Farahmand, Bijal Sheth, Joshua Burke, Scott Siegert, Grace Liang

Case Western Reserve University, Cleveland, Ohio, United States

Background: Tardive dyskinesia (TD) is a persistent movement disorder associated with chronic exposure to dopamine receptor blocking agents (DRBAs) such as antipsychotic and antiemetic medications. Valbenazine (INGREZZA; VBZ), a novel and highly selective vesicular monoamine transporter 2 (VMAT2) inhibitor, is FDA-approved for treating TD in adults, regardless of age. However, older patients are at a higher risk for developing TD when exposed to DRBAs and may also be at higher risk of drug-induced parkinsonism. The effects of VBZ on TD were studied in both older (≥55 years) and younger (<55 years) participants from three 6-week, randomized, double-blind, placebo (PBO)-controlled trials.

Methods: Data were pooled from KINECT (NCT01688037), KINECT 2 (NCT01733121), and KINECT 3 (NCT02274558). Efficacy data were analyzed in the intent-to-treat (ITT) population by pooled VBZ (40 mg, 80 mg) and PBO groups. Subgroups were defined by age at baseline (older, ≥55 years; younger, <55 years). Efficacy assessments at Week 6 included Abnormal Involuntary Movement Scale (AIMS) total score (sum of items 1-7) mean change from baseline, AIMS response (≥50% total score reduction from baseline), Clinical Global Impression of Change-Tardive Dyskinesia (CGI-TD) mean score, and CGI-TD response (score ≤2 [much improved or very much improved]). For AIMS mean score change and CGI-TD mean score, 95% confidence intervals (CIs) were analyzed for the mean difference (MD) between VBZ and PBO. For the AIMS and CGI-TD response analyses, relative risk (RR) with 95% CI was calculated based on the ratio between VBZ and PBO response. Safety assessments included treatment-emergent adverse events (TEAEs).

Results: The pooled ITT population included 219 participants in the older subgroup (PBO, n=93; 40 mg, n=63; 80 mg, n=63) and 154 participants in the younger subgroup (PBO, n=65; 40 mg, n=51; 80 mg, n=38). In both subgroups, the mean improvement from baseline to Week 6 in AIMS total score was greater with VBZ (both doses) than with PBO: ≥55 years (PBO, -0.2; 40 mg, -2.4 [MD=-2.2, 95% CI=-3.5 to -1.0]; 80 mg, -3.4 [MD=-3.2, 95% CI=-4.3 to -2.1]); <55 years (PBO, -0.8; 40 mg, -1.3 [MD=-0.5, 95% CI=-1.7 to 0.8]; 80 mg, -2.8 [MD=-2.0, 95% CI=-3.4 to -0.5]). Both subgroups also had higher rates of substantial AIMS response (≥50% score improvement) with VBZ (both doses) relative to PBO: ≥55 years (PBO, 9.7%; 40 mg, 28.6% [RR=2.95, 95% CI=1.42 to 6.15]; 80 mg, 39.7% [RR=4.10, 95% CI=2.05 to 8.19]); <55 years (PBO, 10.8%; 40 mg, 20.0% [RR=1.86, 95% CI=0.76 to 4.54]; 80 mg, 39.5% [RR=3.67, 95% CI=1.64 to 8.18]). CGI-TD mean scores at Week 6 indicated global improvements in both older and younger participants, with higher rates of CGI-TD response (score ≤2) at Week 6 with VBZ as compared to PBO: ≥55 years (PBO, 19.4%, 40 mg, 30.2% [RR=1.56, 95% CI=0.89 to 2.73]; 80 mg, 41.3% [RR=2.13, 95% CI=1.28 to 3.55]); <55 years (PBO, 18.5%; 40 mg, 35.3% [RR=1.91, 95% CI=1.02 to 3.60]; 80 mg, 39.5% [RR=2.14, 95% CI=1.12 to 4.08]). The incidence of any TEAE was similar between VBZ and PBO in both subgroups: ≥55 years (PBO, 36.9%; 40 mg, 46.9%; 80 mg, 46.4%); <55 years (PBO, 44.0%; 40 mg, 39.1%; 80 mg, 48.8%). A TEAE of akathisia was found in 3 older participants (40 mg, n=3) and 4 younger participants (PBO, n=1; 80 mg, n=3); the TEAE of parkinsonism was not found in any participant in either subgroup. No serious TEAE occurred in >1 participant in either subgroup.

Conclusions: In 3 double-blind, PBO-controlled studies, once-daily VBZ improved TD and was generally well tolerated in both older and younger adults. Consistent and clinically meaningful TD improvements with VBZ 80 mg were found across age groups. In addition, TEAEs were similar across treatment groups and age groups, with no apparent evidence that the older group had a higher risk for treatment-emergent parkinsonism or akathisia.

Keywords: Valbenazine, VMAT2, Tardive Dyskinesia

Disclosure: Part 1: Neurocrine Biosciences, Inc., Consultant, Otsuka Pharmaceutical Development & Commercialization, Inc., Consultant, H. Lundbeck A/S, Consultant, Part 2: Otsuka Pharmaceutical Development & Commercialization, Inc., Consultant.

W153. The Relationship Between Normative Emotional Arousal to Negative Emotional Stimuli and Subsequent Recall in Male and Female Trauma-Exposed Veterans With and Without PTSD

Anne Richards*, Samantha Hubachek, Thomas Metzler, Steve Woodward, Daniel Mathalon, Thomas Neylan

San Francisco VA Medical Center, San Francisco, California, United States

Background: Posttraumatic stress disorder revolves around a highly distressing life event that intrudes upon individuals’ daily life in the form of distressing memories. Existing research in healthy subjects indicates that emotionally arousing stimuli are better remembered. The current analysis is part of an ongoing study of trauma-exposed veterans examining biological sex and PTSD effects on contributors to memory for negative emotional laboratory stimuli. We hypothesize that females and individuals with PTSD will report higher negative emotional responses to laboratory stimuli during encoding and that negative images will be better recalled than neutral images in the group as a whole.

Methods: Fourteen male and female U.S. military veterans with and without PTSD (43% female), aged 24-47 (M 30.6) participated in a test of emotional memory in a lab setting. Participants were presented with a set of 75 negative images (highly arousing and negatively valenced based on gender-specific normative ratings) and 75 neutral images (low-arousing, neutrally valenced based on normative ratings) from the International Affective Picture System in a pseudorandomized fashion. Each image was presented for 1500ms, after which participants rated their response to each image on a scale of 1 (highly pleasant) to 7 (highly unpleasant) before presentation of the next image. During a recall session 5 hours later, participants viewed all previously viewed images plus 30 negative and 30 neutral new images. Participants indicated that the image was “old” (previously viewed) or “new” using a keyboard stroke.

Results: There was a trend effect of sex on subjective ratings, with females reporting more negative responses on average than males (T(12)=1.73, p=.11). There was no PTSD effect on subjective ratings. In the group as a whole, higher normative arousal, but not subjective rating, predicted more accurate identification of old images as old (t(2093)=2.69, p=.007). However, higher normative arousal was also associated with falsely identifying a new image as having been previously presented (t(827)=-4.99, p <.001). Additionally, male subjects (t(824)=-2.89, p=.004) and PTSD+ subjects (t(826)=1.83, p.068) were more likely to falsely remember arousing images as having been previously seen.

Conclusions: Consistent with existing research, female subjects tended to rate images as more unpleasant than male subjects. While subjective ratings did not predict recall, higher normative arousal score was associated with both increased accuracy in identifying old images as well as increased false memories for new stimuli. Interaction effects indicated that male sex and PTSD status (trend) were associated with higher false memories for arousing images. Ongoing data collection will improve power to examine PTSD group, sex, and group by sex interaction effects on memory, to be included in the presentation.

Keywords: Emotional Memory, Sex Differences, PTSD, Veterans

Disclosure: Nothing to Disclose.

W154. Distinct Contributions by Different Regions of the Medial Prefrontal Cortex, Nucleus Accumbens and Basolateral Amygdala in Mediating Risk/Reward Decision-Making Guided by External Cues

Mieke van Holstein*, Maric Tse, Stan Floresco

University of British Columbia, Vancouver, Canada

Background: We often face decisions requiring a choice between options that vary in terms of reward magnitude and uncertainty. In some instances, information about the likelihood of obtaining different rewards are inferred using internally-generated representations, whereas in other instances, external stimuli provide information about the relative amount of reward uncertainty. For example, an experienced Blackjack player knows the odds of winning a hand are larger when the dealer is showing a “6” card compared to an “ace”. Similarly, studies with human participants often present choices between taking a gamble and choosing a safe alternative, with explicit cues informing participants about reward probabilities prior to the decision. Functional neuroimaging studies using these tasks have revealed a role for the nucleus accumbens (NAc) and distinct parts of the prefrontal cortex (PFC), in particular the ventromedial PFC (vmPFC) and the anterior cingulate cortex (ACC), in risky decision making. Preclinical studies with rodents have provided additional insight in the neural basis of risk/reward decision making, however, many of these studies have used tasks where choice is guided by internally-generated information. An exhaustive series of experiments using probabilistic discounting tasks has revealed critical and dissociable roles for the medial PFC, NAc shell and the basolateral amygdala (BLA) in this form of risk/reward decision making. Yet, how these regions may contribute to decision making guided by external discriminative stimuli has not been explored. To bridge this gap between assays in humans and rats we developed a novel assay we have colloquially termed the “Blackjack” task. Using this task, we assessed the roles of the medial PFC, NAc shell, and BLA in cue-guided risky decision making. Due to their potential homology to the human vmPFC and ACC, we were particularly interested in the roles of the infralimbic (IL) and prelimbic (PL) regions of the rodent medial PFC, respectively.

Methods: Animals were trained on the Blackjack task, wherein they were presented with a choice between a small/certain (1 pellet) and a large/uncertain (4 pellet) option. The odds of obtaining the larger reward varied unpredictably from trial to trial between either 50% or 12.5%. Prior to a choice, one of two tones signalled that the odds of obtaining the larger reward were either good (50%) or poor (12.5%). After training, male Long Evan rats were implanted with bilateral cannulae either in the NAc shell, BLA, IL, or the PL. After re-training, each animal received two test days: one after the infusion of saline, and one after the infusion of the GABAa/b agonists muscimol and baclofen.

Results: Under control conditions, well-trained rats selected the large reward option more often when the odds were good vs. poor (~70% vs. 25%). Inactivation of the NAc shell increased risky choice, particularly on trials where cues signalled that the odds were poor. Subsequent studies assessed the involvement of key cortical and subcortical inputs the NAc shell in mediating this form of decision making. Inactivation of the IL increased risky choice in a manner similar to inactivation the NAc shell. Likewise, inactivation of the BLA tended to increase risky choice when the odds were poor. However, this profile was distinct in that it was accompanied by pronounced impairments in the ability to suppress the tendency to be influenced by the rewarded or non-rewarded outcome on the preceding trial. In stark contrast to the above-mentioned findings, inactivation of the PL reduced risky choice selectively when the odds were good, suggesting that this region may contribute to the appropriate pursuit of larger/risky rewards.

Conclusions: These data reveal that the IL and the NAc shell both suppress inappropriate actions toward larger rewards when cues provide information that these rewards are unlikely. In addition, the IL and PL regions of the medial PFC play opposing roles when using cues to guide decisions associated with uncertainty, similar to the roles the vmPFC and ACC might play in humans. Finally, the BLA might act to suppress inappropriate actions toward larger rewards when cues provide information that these actions are unlikely to yield a reward. Moreover, this nucleus appears to tailor the sensitivity to wins and losses to the odds at hand. Collectively, these data demonstrate how distinct prefrontal-subcortical networks mediate dissociable aspects of risk/reward decision making. This task may thus play a valuable role in further elucidating the neural bases of normal and abnormal decision making associated with numerous psychiatric disorders.

Keywords: Reward-Based Decision-Making, Animal Models, Medial Prefrontal Cortex, Nucleus Accumbens Shell, Basolateral Amygdala

Disclosure: Nothing to Disclose.

W155. Challenges and Advances in Ca2+ Imaging in the Nucleus Accumbens of Monogamous Prairie Voles

Jennifer Scribner, Elliott Saslow, Jessica Jimenez, Eric Klein, Rene Hen, Mazen Kheirbek, Zoe Donaldson*

University of Colorado Boulder, Boulder, Colorado, United States

Background: Attachment-related behaviors are disrupted in a number of psychiatric and behavioral disorders. Monogamous prairie voles are a laboratory-amenable species that forms life-long pair bonds between mated individuals, making them an excellent model for studying the neurobiology of attachment. Attachment formation in voles results from a reward association that links the pleasurable aspects of mating and social interaction with the sensory cues of a particular individual. Multiple lines of converging evidence suggest that signaling within the nucleus accumbens (NAc) is critical for the formation of this reward association. Manipulations of oxytocin and dopamine signaling in the nucleus accumbens inhibit bond formation, and neuroplastic changes within this region are thought to underlie long-term bond expression. While the NAc has clear importance for pair bonding, very little is known about the signatures of neuronal activity within this region that change upon bond formation.

Methods: We performed in vivo Ca2+ imaging in 17 adult male and female prairie voles as they interacted with their partner and a novel opposite-sex individual before and after mating. We used commercially available methods to process Ca2+ imaging videos and calculate Ca2+ spiking events (Inscopix Inc). We used open-source IDtracker software to track the test animal's behavior during a 20-minute partner preference test. Custom scripts in Matlab were used to match imaging and behavior data and for subsequent analysis.

Results: After mating, prairie voles prefer to interact with their pair-bonded partner during imaging sessions, and this preference becomes stronger with time. Rates of Ca2+ spiking activity were not uniformly distributed; instead more activity is observed during early bouts of social interaction and at the beginning of social interaction bouts. Social interactions are highly complex, including aggressive and affiliative encounters, which makes it challenging to identify consistent differences in neural activity patterns that can be linked to interaction partner alone. Instead, we have identified a subset of cells that is selectively active when the test animal is approaching a tethered individual, a behavior that is much more uniform than social interaction itself. We have found that the proportion of partner approach-cells increases following preference formation while the proportion of novel-approach cells remains the same. Activity within these cells is not correlated with velocity.

Conclusions: One interpretation of our results is that a subset of cells that is active during social approach may reflect a neuronal signature of social anticipation or social motivation. The increased proportion of neurons that exhibit activity when approaching a pair bonded partner may reflect an increased motivation to or anticipation of interacting with that individual. We are currently developing methods to assess social motivation in prairie voles in order to further explore hypothesis. This work may ultimately shed light on the neuronal basis of attachment and facilitate studies of the neurobiology underlying attachment deficits.

Keywords: Attachment, Social Behavior, Calcium Imaging

Disclosure: Nothing to Disclose.

W156. BDNF Signaling Impacts Maternal Care and Oxytocin Neuron Gene Expression

Kristen Maynard*, John Hobbs, BaDoi Phan, Courtney Williams, Nina Rajpurohit, Joo Heon Shin, Andrew Jaffe, Keri Martinowich

Lieber Institute for Brain Development, Baltimore, Maryland, United States

Background: Oxytocin (OXT) is a neuropeptide critical for regulation of sex-specific social behaviors. Brain-derived neurotrophic factor (BDNF) and its receptor TrkB are highly expressed in the paraventricular hypothalamus, the site of OXT-secreting neurons. Transcription of BDNF is controlled by several promoters, which drive expression of multiple transcripts encoding an identical protein. We previously showed that BDNF promoter I significantly contributes to BDNF production in the hypothalamus. Furthermore, female mice with selective disruption of promoter I-derived BDNF (Bdnf-e1 -/-) show deficits in sexual receptivity and maternal care. These social behavior impairments are associated with reduced OXT transcripts.

Methods: To delineate the role of BDNF signaling in OXT neuron function, we used TRAP (translating ribosome affinity purification) and RNA sequencing (RNAseq) to examine differences in cell-type-specific gene expression in OXT neurons from control and Bdnf-e1-/- females. First, to validate our technique and identify a molecular profile for OXT neurons, we crossed mice expressing Cre recombinase under control of the endogenous OXT promoter to the RiboTag mouse, which expresses a Cre-dependent hemagglutinin (HA) tag on the ribosomal protein RPL22. Ribosome-associated mRNAs were affinity purified using an HA antibody and characterized with high-throughput RNAseq analysis. To elucidate changes in OXT gene expression following perturbations in BDNF signaling, we used an adeno-associated viral (AAV) RiboTag approach to isolate and sequence ribosome-associated transcripts in OXT neurons from control and Bdnf-e1 -/- females with impaired social behavior.

Results: Here we report genes and gene pathways highly enriched in OXT-expressing cells and provide a molecular profile defining the OXT neuron translatome. We find that OXT-enriched mRNAs encode proteins critical for structural and functional plasticity. We also identify transcripts vulnerable to misregulation in OXT neurons following disruption of hypothalamic BDNF signaling. These encode proteins known to function in subcellular compartments crucial for plasticity, including synapses, axons, and mitochondria. Differentially expressed gene sets between control and Bdnf-e1 -/- OXT neurons were also enriched in development-related gene ontology (GO) terms for biological processes including axon development, neuronal differentiation and migration, and synaptic organization and transmission.

Conclusions: These results provide a molecular characterization of a sexually dimorphic hypothalamic cell population that governs complex social behaviors. Our findings identify BDNF-dependent signaling pathways in OXT neurons that can be targeted to modulate female-typical behaviors, including sexual receptivity and maternal care.

Keywords: BDNF, Oxytocin, Maternal Behavior, RNAseq, RiboTag

Disclosure: Nothing to Disclose.

W157. Social Status and the Brain Kappa Opioid Receptor: Finding From a PET Imaging Study

David Matuskey*, Mark Dias, Mika Naganawa, Shannan Henry, Songye Li, Nabulsi Nabeel, Jim Ropchan, Richard Carson, Yiyun Huang

Yale University, New Haven, Connecticut, United States

Background: Previous positron emission tomography (PET) studies by our group and others in non-human and human primates have demonstrated a relationship between striatal dopamine D2/D3 receptors and social status 1,2,3 Overall, these results suggest that social factors affect the brain reward pathways, a finding that is important for vulnerability to addiction and mental health. However, dysphoria and aversive effects also play an important role in assessing the value of a reward and these aspects has not been previously studied in relation to social status in humans.

A mechanism to investigate this is through PET imaging of the kappa-opioid receptors (KOR). Generally, agonists at KOR are aversive, compared to other opioid receptor agonists such as mu and delta that are rewarding and reinforcing.3 KOR and dynorphin, its endogenous ligand, have thus been characterized as a counterbalance or aversive system to the rewarding dopamine system and have been implicated in drug use and stress induced relapse.4 The objective of this work was to investigate whether the KOR system is related to social status in humans using PET imaging with the novel KOR radiotracer [11C]EKAP. We hypothesized that there would be a negative correlation with higher social status having lower [11C]EKAP volume of distribution (VT), interpreted as lower levels of KOR expression.

Methods: Eighteen medically healthy participants (mean age 41.2 ±9.3) were asked to complete the Barratt Simplified Measure of Social Status (BSMSS). Subjects received a MRI and then underwent a 120-min PET scan with [11C]EKAP using a High Resolution Research Tomograph (HRRT). Arterial blood sampling and metabolite analysis were conducted to obtain the input function. Regions of interest were based upon a MR template and included the reward/aversion areas of the amygdala, caudate, pallidum, putamen, and frontal cortical regions. Secondary areas included non-reward areas of thalamus, and parietal, temporal and occipital cortical regions. The multilinear analysis-1 (MA1) method was applied to the regional time-activity curves (TACs) to calculate [11C]EKAP VT. All Pearson correlation coefficients were adjusted for body mass index (BMI), gender and age due to possible or known effects of these variables on KOR.6 Correlations were adjusted for multiple tests using the Bonferroni correction.

Results: Regional [11C]EKAP VT and BSMSS were found to be negatively correlated in the the amygdala (r=- 0.685, p=0.005) caudate (r=- 0.656, p=0.008) pallidum (r=- 0.607, p=0.016), putamen (r=- 0.630, p=0.012) and frontal cortex (r=- 0.531, p=0.042). All p values shown were unadjusted, but remained significant after Bonferroni correction (p=0.01 threshold) except for the pallidum and frontal cortex. In secondary (non-reward) regions, correlations of [11C]EKAP VT and BSMSS were all non-significant.

Conclusions: There was an inverse correlation between social status and binding of the kappa opioid receptor that was specific to the reward/aversion (e.g., saliency) areas of the brain. This finding suggests the KOR system may act in opposing ways to the dopamine system as a mediator for the negative effects of social behaviors in humans.


1. Matuskey D, Gaiser EC, Gallezot JD, Angarita GA, et al. A preliminary study of dopamine D2/3 receptor availability and social status in healthy and cocaine dependent humans imaged with [(11)C](+)PHNO. Drug Alcohol Depend. 2015 Sep 1;154:167-73. doi: 10.1016/j.drugalcdep.2015.06.039. Epub 2015 Jun 30. PMID: 26164205.

2. Martinez D, Orlowska D, Narendran R, Slifstein M, Liu F, Kumar D, et al. (2010): Dopamine type 2/3 receptor availability in the striatum and social status in human volunteers. Biological Psychiatry. 67:275-278.

3. Morgan D, Grant KA, Gage HD, Mach RH, Kaplan JR, Prioleau O, et al. (2002): Social dominance in monkeys: dopamine D2 receptors and cocaine self-administration. Nat Neurosci. 5:169-174.

4. Akil H, Owens C, Gutstein H, Taylor L, Curran E, Watson S. (1998) Endogenous opioids: overview and current issues. Drug Alcohol Depend. 51(1-2):127-40.

5. Wee S, Koob GF. (2010) The role of the dynorphin-kappa opioid system in the reinforcing effects of drugs of abuse. Psychopharmacology. 210:121-135.

6. Vijay A, Wang S, Worhunsky P, Zheng MQ, Nabulsi N, Ropchan J, Krishnan-Sarin S, Huang Y, Morris ED. (2016) PET imaging reveals sex differences in kappa opioid receptor availability in humans, in vivo. Am J Nucl Med Mol Imaging. 6(4):205-14.

Keywords: PET Imaging, Kappa Opioid, Social Status

Disclosure: Nothing to Disclose.

W158. Investigation of Prenatal Stress in the Cerebral Organoid Model: A Focus on Cell-Type Specific Responses Following Glucocorticoid Exposure

Cristiana Cruceanu*, Simone Roeh, Silvia Martinelli, Maik Koedel, Rossella Di Giaimo, Silvia Cappello, Elisabeth Binder

Max Planck Institute of Psychiatry, Munich, Germany

Background: Mental illnesses share increased risk through long-term regulatory changes to the stress hormone system, precipitated as adversities during early life, including the prenatal period. The developing brain is undergoing some of the most important growth and plasticity periods during the prenatal period, and increased glucocorticoid (GC) exposure has been suggested as one of the main factors mediating the effect of stress during this time, with likely effects on epigenetic and transcriptional trajectories. Some progress has been made in elucidating the neurobiology of stress, but many questions cannot be addressed with current animal models – particularly in regards to human-specific early developmental molecular pathways. We aim to use 3-dimensional cerebral organoid culture systems as a research model for identifying prenatal stress-related developmental risk factors for psychiatric disorders.

Methods: Cerebral oganoids were grown and differentiated using standard protocols (Lancaster et al). RNA expression was quantified in homogenized organoids using total RNA sequencing (RNAseq; NEB ribosomal depletion and Illumina HiSeq4000) and real-time polymerase chain reaction (qRT-PCR). To investigate molecular mechanisms of how GCs may impact nervous system development, we subjected organoids to known GC receptor (GR) agonist dexamethasone. To investigate pooled specific cell types, we used Fluorescence-Activated Cell Sorting (FACs) for TUBB3 (neurons) and PROM1(neural progenitors) followed by bulk total RNAseq and qRT-PCR as well as Immunohistochemistry (IHC) for protein expression. For single cell RNAseq (scRNAseq), organoids were dissociated to single cells and the individual transcriptomes were sequenced (Wafergen iCell8 and Illumina HiSeq4000).

Results: Using RNAseq for transcriptome profiling and IHC for protein expression across nine developmental stages (from day 17 to day 210) we showed that important markers of developing as well as maturing brain cells are expressed in the cerebral organoid model, including SOX2, PAX6, FOXG1, MAP2, with trajectories consistent with increasing neuronal differentiation. Furthermore, GR-regulated genes such as FKBP5 were also expressed, and with increasing levels consistent with advancing developmental stages. The transcriptome profiles were in line with increased proportions of mature cells present, which was also validated via FACs, where the percentage of TUBB3-positive neurons increased nearly two-fold from day 30 to day 60 organoids. In order to determine the appropriate GC-stimulation paradigm in organoids we performed a time-course dose-response experiment, and determined that stimulation with 100nM dexamethasone over 12 hours elicits a sustained effect on FKBP5 expression (fold-change=4.7) in organoids. Finally, to tease apart the effect of the GC-stimulation stress paradigm in the very complex cell-type landscape of the human brain – and human cerebral organoids – we used scRNAseq at three consecutive ages representative of human brain development (30, 60, and 90 days) with and without dexamethasone treatment. We profiled in 5000 individual cells per developmental stage, and found dynamic distribution of genes that are hallmarks of brain development as well as genes previously associated with GC-responsive pathways. We are currently investigating differential response to dexamethasone in specialized brain cell types like neurons compared to less differentiated neural progenitors, as well as modified developmental trajectories as a result of GC-stimulation.

Conclusions: Cerebral organoids follow developmental trajectories of the human brain, which are likely in part mediated through epigenetic regulation of GR-responsive pathways. Furthermore, organoids show responsiveness to glucocorticoids in a way that is consistent with previous findings from animal model and human peripheral studies. Based on dynamic distribution of cell types in the maturing organoids and the parallel changes in of GC-responsive genes, we suspect cell-type specific effects of GC stimulation in cerebral organoids, which motivated the use of single-cell sequencing in this model of the brain. scRNAseq techniques will allow us to deconstruct tissue heterogeneity, elucidating the molecular underpinnings of human vulnerability and response to GC-mediated stress. In follow-up analyses, we plan to compare our findings in the cerebral organoid model to evidence from longitudinal human cohorts of mothers and their babies who underwent stress during pregnancy. In this effort we hope to pave the way for translation into clinical research, and expedite transitions from basic research to identification of at-risk populations and intervention strategies.

Keywords: Prenatal Stress, Glucocorticoids, Single-Cell, Transcription

Disclosure: Nothing to Disclose.

W159. Hippocampal Mitochondrial Gene Expression Changes With Development and Early Life Stress

Audrey Tyrka*, Kathryn Ridout, Mizan Gaillard, Chelsea Lopez, Kevin Bath

Butler Hospital / Brown Medical School, Providence, Rhode Island, United States

Background: Exposure to early life stress (ELS) and trauma is a major public health problem that increases risk for psychiatric and other disorders. In the developing hippocampus and cortex, ELS or stress hormone exposure lead to diminished cell proliferation and increased cell turnover and death. Bath et al have shown these key neurobiological changes in a mouse model of fragmented maternal care that also produces precocious fear learning and anxious and depressive-like behaviors. Mitochondria are central to the stress response and also play an integral role in inflammation, cellular aging, and apoptosis. However, no studies have examined normative or ELS effects on hippocampal mitochondrial gene expression during early development.

Methods: C57BL/6N male mice were reared normally (Unhandled Controls) or with restricted access to bedding (ELS) from p4-p11. Time points prior to ELS (p4), immediately after ELS (p12), and across development (p16, p21, p28, p38, p50) were examined. Hippocampal samples (n=4-6/timepoint) from ≥2 different litters were isolated and cDNA synthesized. Real-time PCR was run in multiplex with 18S as standard to measure expression of the 13 protein-coding mitochondrial genes (11 oxidative phosphorylation, 2 tRNA). Developmental and group differences were examined using two-way ANOVA.

Results: Expression of each of the mitochondrial genes changed significantly with development (p's<.0001). ELS significantly altered expression of several mitochondrial genes across development, and age x group interactions were significant for several mtDNA genes (p's<.05).

Conclusions: This is the first study examining mitochondrial gene expression during murine hippocampal development and the first to examine the effects of ELS on mitochondrial gene expression. These results suggest that ELS may alter ontogeny of mitochondrial gene expression, with implications for understanding ELS effects on hippocampal development.

Keywords: Mitochondrial DNA, Early Life Stress, Hippocampus, Mouse Models

Disclosure: Nothing to Disclose.

W160. Region Specific Dysregulation of Dopaminergic Signaling in SAPAP3 Knock out Mice Displaying Excessive Over-Grooming

Daniel Foster*, M. Suhaib Mahmood, Samantha Yohn, Daniel O'Brien, Weimin Peng, P. Jeffrey Conn

Vanderbilt University, Nashville, Tennessee, United States

Background: Converging clinical and preclinical data suggest that striatal circuitry plays a key role in modulating repetitive behaviors and indicate that aberrant signaling in specific striatal pathways may represent a common mechanistic underpinning of numerous diseases including Obsessive Compulsive Disorder (OCD). Imaging studies have found reduced striatal dopamine (DA) receptor binding in OCD patients compared to healthy controls, an indication that dopaminergic signaling may be hyperactive in these individuals. Consistent with this hypothesis, blockade of dopaminergic signaling by administration of atypical antipsychotics can effectively treat some OCD patients that are refractory to serotonin selective reuptake inhibitors (SSRIs). Furthermore, experimentally mimicking or augmenting DA signaling in rodents can induce repetitive behaviors. However, the precise role of DA is regulating repetitive behaviors is poorly understood. The purpose of this study was to determine if dopamine signaling is altered in a mouse model displaying excessive self-grooming and further elucidate the potential utility of compounds targeting the striatal DA system in modulating repetitive behaviors.

Methods: Here, we report studies using fast-scan cyclic voltammetry (FSCV) in mice lacking the postsynaptic protein SAP90/PSD95-associated protein (SAPAP3 KO mice) as well as control littermates. Rodent self-grooming provides a behavioral output with which one can monitor repetitive, self-directed, patterned behavior that has great translational value to OCD-like disorders. Total time spent grooming was monitored in SAPAP3KO mice and control littermates. To further examine the role of DA in regulating repetitive grooming behaviors the magnitude and kinetics of DA transients were assessed using FSCV in ex vivo slice preparations as well as in anesthetized mice in vivo. DA transients were elicited in the dorsolateral striatum (DLS), dorsomedial striatum (DMS); and nucleus accumbens core (NAcc). In some experiments mice were crossed with DAT-Cre animals and channelrhodopsin 2 (ChR2) was virally expressed in DA neurons to allow optical stimulation of DA transients.

Results: As previously reported, SAPAP3 KO mice showed excessive grooming compared to control littermates at the age assessed (4-5 months). DA transients evoked by a single electrical pulse in slices from SAPAP3 KO mice were not significantly different from those observed in slices from control littermates in any of the regions tested including the DLS, DMS and NAcc. However, when four electrical pulses were applied at a frequency of 10Hz to mimic DA neuron bursting, the magnitude of DA transients observed in the DMS and NAcc of SAPAP3 mice were greater than those evoked in control littermates. Interestingly, phasic stimulation produced similar DA transients in the DLS of both genotypes suggesting that phasic DA signaling was not globally altered. To confirm this finding, we crossed SAPAP3 KO mice with DAT-Cre mice and injected ChR2 containing virus into the midbrain to selectively express ChR2 in DA neurons. Transients were then optically evoked resulting in selective activation of DA neurons. Optical stimulation produced a pronounced enhancement of DA release in SAPAP3 KO mice specifically in the DMS and only following phasic-like stimulation.

Conclusions: These exciting findings suggest that DA signaling in SAPAP3KO mice is dysregulated in a very precise manner that is sub-region specific as well as dependent on the pattern of stimulation. These results suggest that targeted therapies that can modulate these specific modes of dopaminergic signaling in these distinct striatal subregions could provide improved efficacy in OCD patients that are resistant to SSRI treatment.

Keywords: Dopamine, Striatum, Repetitive Behavior

Disclosure: Nothing to Disclose.

W161. Reversing Corticosteroid Effects on the Human Hippocampus With Lamotrigine

E. Sherwood Brown*, Nasreen Sayed, Changho Choi, Nicholas Tustison, Jarred Roberts, Michael Yassa, Erin Van Enkevort, Alyson Nakamura, Elena I. Ivleva, Prabha Sunderajan, David A. Khan, Miguel Vazquez, Bruce McEwen, Traci Holmes

University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: In animals, stress and corticosteroid excess are associated with changes in memory and hippocampal structure that are prevented with agents that decrease glutamate release. Humans also demonstrate changes in memory and hippocampus with corticosteroids. In this report the effects of glutamate-release inhibitor lamotrigine on hippocampal structure and function were examined in people receiving medically needed prescription corticosteroid therapy.

Methods: A total of 54 outpatient adults (n=28 women) receiving chronic (≥ 6 months) oral corticosteroid therapy were randomized to lamotrigine or placebo for 48 weeks administered in a double-blind, placebo-controlled, parallel-group fashion. Declarative memory was assessed using the Rey Auditory Verbal Learning Test (RAVLT, a word list task); 3T structural magnetic resonance imaging (MRI) as well as proton MR spectroscopy (1HMRS) focused on hippocampus were obtained at baseline and week 48 using a Philips 3T scanner. Hippocampal subfield volumes were assessed using a multi-atlas label matching automated approach. Utilizing a mixed-model approach, structural and biochemical data were examined by separate ANOVAs, and memory was assessed with a multi-level longitudinal model.

Results: The lamotrigine and placebo groups were similar in age, sex, race, mean prednisone therapy duration, premorbid intelligence estimates, and medical conditions. The lamotrigine group, on average, was receiving a higher prednisone dose, however, dose was not a significant covariate for the outcomes. At baseline (in the combined groups prior to starting study medication), prednisone dose (mg/day) significantly negatively correlated with the right hippocampal choline/creatine ratio (r=-0.28, p=0.045) and the right CA1 volume positively correlated with right glutamine/creatiner (r=0.31, p=0.026). RAVLT total scores demonstrated a significant interaction between treatment group and time with higher scores with lamotrigine (p =.047). A significant between-treatment group hippocampal subfield volume difference was observed in the left subiculum (p=0.020) and right CA1 (p=0.007) demonstrating larger volumes with lamotrigine. The choline/creatine ratio in the left hippocampus was marginally higher with lamotrigine (p =.054). Lamotrigine was well tolerated. Overall number of reported side effects was not significantly different between groups (p=0.50). A total of 73.1% reported greater than 80% study adherence by pill counts with lamotrigine as compared to 57.1% with placebo (χ2(1)=1.50, p=0.22).

Conclusions: Lamotrigine was associated with larger hippocampal subfield volumes and better declarative memory performance than placebo. Findings suggest that, in humans as well as in animal models, glutamate release inhibitors may attenuate some of the effects on the human hippocampus associated with corticosteroids.

Keywords: Corticosteroids, Lamotrigine, Hippocampus, Hippocampal Subfields, Hippocampal Volume

Disclosure: Nothing to Disclose.

W162. D1-Selective Partial Agonist PF-9571 Displayed Robust and Sustained Efficacy and did not Induce Dyskinesias in MPTP Nonhuman Primate Model of Parkinson's Motor Symptoms After Repeat Dosing

David Gray, Kari Fonseca, Wenlei Liu, Rouba Kozak*

Pfizer Global R&D, Cambridge, Massachusetts, United States

Background: Insufficient activation of striatal dopamine receptors underlies the cardinal motor deficits in Parkinson's disease (PD). Increasing activation of D1R's has been a longstanding goal for control of PD motor symptoms. D1R selective agents such as A-77636 and dihydrexidine have demonstrated robust single dose efficacy in the translatable MPTP PD motor symptom model (Taylor, 1991; Pearce, 1999) however, ligands with suitable pharmacokinetics for robust clinical investigation have proved elusive. Furthermore, a handful of preclinical reports emerged suggesting that D1R agonist ligands may have response tachyphylaxsis (Blanchet, 1996) and potential for increased induction of dyskinesias (Westin, 2007; Darmopil, 2009).

Methods: This work reports preclinical assessment of PF-9751, a selective D1R non-catechol partial agonist from a new chemical class with improved pharmacokinetics, after multiple and chronic dosing. Motor efficacy and dyskinesia were scored in MPTP-lesioned macaques (Macaca fasicularis; N=8) that had been previously exposed to levodopa and had developed responsive dyskinesias. Blinded rating occurred on day 1 and day 3 following once-daily administration of PF-9751. We then examined the impact of PF-9751 when administered to macaques with MPTP-induced motor deficits and no prior exposure to levodopa. Motor disability and dyskinesia was measured after 30 consecutive days of once-daily administration of PF-9751.

Results: Following 3 days of once-daily dosing of PF-9751, motor disability scores were significantly improved during the entire 6-hour scoring session vs. single vehicle dose. Similar to single dose levodopa, disability scores were reduced to near zero; however, motor improvement with PF-9751 was maintained for -the full 6-hour assessment period vs 2–3 hours for levodopa. Dyskinesias were measured and were present, but to a lesser degree with PF-9751 compared to levodopa. Upon longer term dosing with PF-9751 to macaques with no prior exposure to levodopa, motor disability was significantly reduced from vehicle condition. Contrary to what would be expected with levodopa dosing, no dyskinesias were observed after 30 consecutive days of PF-9751 administration. Taken together, these data show that tachyphlaxsis of motor response is not observed with the novel ligand class and highlight the potential for D1R agonist, PF-9751, to have continuous behavioral efficacy.

Conclusions: This result confirms the strong motor efficacy of D1 agonists and suggests that continuous activation of D1R by appropriate agonist does not carry increased risk of dyskinesia induction in non-primed animals, and may represent a mechanism for improving motor function in Parkinson's without initiating or exacerbating dyskinesias in chronic disease conditions.

Keywords: D1 Dopamine Receptors, MPTP, NHP, Parkinson's Disease

Disclosure: Part 1: Pfizer, Inc., Employee, Part 2: Pfizer, Inc., Employee, Part 3: Pfizer,Inc., Employee, Part 4: Pfizer, Inc., Employee, Part 5: Pfizer, Inc., Employee.

W163. Anti-NMDAr

Ronald Gurrera*

Harvard Medical School, Brockton, Massachusetts, United States

Background: The most common form of autoimmune encephalitis is caused by antibodies against the N-methyl-D-aspartate receptor (NMDAr). This autoimmunity may be triggered by pregnancy, neoplastic processes, and viral infections, as well as other unknown factors, and it afflicts women more often than men. Complex behavioral disturbances are a hallmark of anti-NMDAr encephalitis and occur so frequently that most cases are first seen by psychiatrists. However, this disorder is commonly misdiagnosed because it mimics primary psychiatric illnesses and neuroleptic malignant syndrome, and is not detected by routine laboratory testing. Without prompt and effective treatment, which is associated with a good prognosis, substantial residual morbidity or death may ensue, so early recognition is paramount. The aim of this study was to examine published reports of anti-NMDAr encephalitis in which behavioral abnormalities were prominent, in order to identify co-occurring clinical features that might prompt clinicians to consider this diagnosis early in the course of evaluation; and to determine whether men and women with anti-NMDAr encephalitis-related behavioral syndromes present differently.

Methods: A systematic search of PubMed and EMBASE databases was conducted to identify any report indexing a NMDAr autoimmune disorder and any form of mental, behavioral or psychiatric disorder. This search yielded 387 unique citations. Seventy-five reports were accessed containing individual case information on 144 unique patients aged 19 years or older who exhibited behavioral signs or symptoms during the course of anti-NMDAr encephalitis. A structured tool was used to collect data regarding 41 behavioral and clinical features observed during the illness episode. These features were aggregated into domains of prodromal symptoms, fever, seizures, insomnia, altered autonomic function, altered cognition, language and speech dysfunction, motoric and sensory abnormalities, and catatonic features (including stereotypies). Sex-specific distributions of clinical domains were analyzed using standard statistical and graphical techniques. All probabilities were 2-tailed.

Results: The sample consisted of 17 men (11.8%) and 127 women (88.2%) (p<<<.001). Age ranges were similar (19-69 years vs. 19-67 years) but mean (S.D.) age for men was significantly older (37.6 (3.2) vs 28.7 (9.3) years, t= 2.17, df= 142, p=.044). Non-behavioral clinical features were common in both groups, and appeared with similar frequencies: seizures (70.6% vs 73.2%), altered cognition (70.6% vs 70.1%), motor abnormalities (58.8% vs 65.4%), catatonic features (47.1% vs 48.8%), altered autonomic function (35.3% vs 40.2%), non-behavioral prodrome (35.3% vs 29.9%), language and speech disorders (35.3% vs 23.6%), insomnia (35.3% vs 15.0%), fever (29.4% vs 22.8%), and sensory abnormalities (5.9% vs 8.7%) (p>.35 for all comparisons). Behavioral symptoms were noted to be the first, or among the first, features on initial presentation in 33 (35.1%) women and 4 (30.8%) men. At least one of the 3 most frequent non-behavioral features – seizures, altered cognition (e.g., disorientation, diminished arousal, memory dysfunction), and motor abnormalities other than catatonic-like signs – was also present in 120 (94.5%) women and 16 (94.1%) men.

Conclusions: Although anti-NMDAr encephalitis is significantly more common in women, its clinical presentation is remarkably similar in men and women. Seizures, cognitive dysfunction, and/or neurological findings (motoric or sensory) were also present in 94.4% of all cases in which behavioral abnormalities were observed, which may cue clinicians to consider this recently discovered disorder in any patient presenting with a new onset or atypical psychiatric syndrome. Prompt and effective treatment usually leads to full recovery, but clinicians need also be aware that this disorder can relapse years later, requiring another course of treatment.

Keywords: Autoimmune Encephalitis, NMDA Receptor, Neuroleptic Malignant Syndrome, Diagnosis

Disclosure: Nothing to Disclose.

W164. Synthesis and in vitro and in vivo Evaluation of Novel Radioligands for LRRK2

Yu-Shin Ding*, Noeen Malik, Daniel Tuchman, Johan Sandell, Andrew Gifford

New York University School of Medicine, New York, New York, United States

Background: LRRK2 (Leucine-rich repeat kinase 2) has recently been proved to be a promising drug target for Parkinson's disease (PD) due to its apparent enhanced activity caused by PD-associated mutations, for example, G2019S mutation in LRRK2 can cause up to 30% of sporadic PD in some populstions.1 To date, there have been no publications in which a LRRK2 inhibitor has been radiolabeled and used for in in vitro or in vivo studies of LRRK2. In the present study, we radiolabeled the LRRK2 ligand, LRRK-IN-1,2 for the purposes of performing in vitro (IC50, Kd, Bmax, autoradiography) and in vivo (biodistribution and blocking experiments) evaluations in rodents and human striatum tissues.

Methods: [3H]LRRK2-IN-1 was prepared with high radiochemical purity (> 99%) and a specific molar radioactivity of 1.52 GBq/μmol via tritium/hydrogen (T/H) exchange using Crabtree's catalyst. For IC50, Kd and Bmax determination, LRRK2-IN-1 was used as a competing drug for nonspecific binding assessment. The specific binding of the tracer was further evaluated via an in vivo blocking study in mice with a potent LRRK2 inhibitor, Pf-06447475 (0, 1, 3, 10 mg/kg). [3H]GNE-9605, one of the most promising second generation LRRK2 inhibitors, was also synthesized and evaluated. Comparative in vivo bio-distribution and blocking studies and in vitro ARG studies were carried out.

Results: In the radioligand binding assays, specific binding of [3H]LRRK2-IN-1 was observed both in the rat tissues (kidney and brain) and in the post-mortem human tissues (brain). The binding of [3H]LRRK2-IN-1 was displaced by unlabeled LRRK2-IN-1 with an IC50 values of 40±4 in rat kidney, 65±3 in rat brain striatum and 73±6 nM in human brain striatum tissues. in vitro binding studies demonstrated a saturable binding site for [3H]LRRK2-IN-1 in rat kidney, rat brain striatum and human brain striatum with Kd of 26±3 and 43±8, 48±2 nM, respectively. In rat, the density of LRRK2 binding sites (Bmax) was higher in kidney (6.4±0.04 pmol/mg) than in brain (2.5±0.03 pmol/mg), however, in human brain striatum, the Bmax was 0.73±0.01 pmol/mg protein. Autoradiography imaging in striatum of rat and human brain tissues gave results consistent with binding studies. In in vivo biodistribution and blocking studies in mice, co-administration with Pf-06447475 dose-dependently reduced the uptake of [3H]LRRK2-IN-1 (%ID/g) with 50-60% reduction at the highest dose (10 mg/kg) in the kidney or brain.

Comparative studies of [3H]GNE-9605 vs. [3H]LRRK2-IN-1 indicated that although LRRK2-IN-1 has a lower ΔG (lower free binding energy determined via docking studies) than GNE-9605, GNE-9605 is more CNS permeable due to its higher lipophilicity than LRRK2-IN-1.

Conclusions: Our studies suggested that LRRK2 binding sites are present at high levels in both peripheral tissues (kidney) and brain and this enzyme can therefore serve as a suitable target for PET studies. Our studies also indicated that [3H]LRRK2-IN-1 is able to specifically label LRRK2 binding sites in vitro and thus can be used as a research tool for radioligand binding assays. However, the low level of brain uptake limits its use as a suitable PET ligand for brain imaging of LRRK2. [3H]GNE-9605 showed more promising properties as a ligand, as it has a higher uptake and higher specific binding in the brain as compared to [3H]LRRK2-IN-1.

Keywords: Parkinson's Disease, Mutation, Neurodegenerative Disease

Disclosure: Nothing to Disclose.

W165. Expression of Serine Racemase Switches From Neurons to Reactive Astrocytes in Advanced Alzheimer's Disease Braak Stages (human) and in a Rat Model of Alzheimer's Disease

Darrick Balu*, Kevin Muszynski, Jacki Rorabaugh, Harry Pantazopoulos, Sabina Berretta, David Weinshenker, Joseph Coyle

McLean Hospital, Harvard Medical School, Belmont, Massachusetts, United States

Background: N-methyl-D-aspartate receptors (NMDARs) play a central role in synapse formation, synaptic plasticity, and learning and memory. A distinctive feature of NMDAR activation, is that in addition to glutamate, it requires the simultaneous binding of a co-agonist to the GluN1 subunit. It is now recognized that D-serine is at least as, if not more important than glycine at modulating NMDARs in the forebrain. Under physiological conditions, the neuronal enzyme serine racemase (SR) synthesizes D-serine in the brain. However, we recently demonstrated the induction of SR and D-serine in reactive astrocytes following traumatic brain injury in mice, which comports with the original identification of SR and D-serine in cultured astrocytes that take on a reactive profile in vitro. We hypothesized that SR might be expressed in reactive astrocytes in Alzheimer's disease (AD), as it is also known to be associated with significant astrogliosis. To test this hypothesis, we examined post-mortem brain tissue from human subjects and from the transgenic rat (TgF344-AD) that expresses the disease-causing mutant human amyloid precursor protein (APPsw) and presenilin 1 (PS1ΔE9). TgF344-AD rats display amyloid plaques and core components of AD that are often missing in transgenic mice, including age-dependent cognitive impairment, astrogliosis, endogenous tau hyperphosphorylation, and apoptotic neuronal loss in the forebrain.

Methods: For human post-mortem studies, we utilized brain tissue that was lightly fixed in paraformaldehyde and then cryoprotected, and consisted of age-matched (>75y) non-psychiatric control subjects that were classified as having either low (I-II) or advanced (V-VI) AD Braak staging. Brain sections that included the amygdala, entorhinal cortex, and hippocampus were processed for dual antigen immunofluorescence using antibodies for detecting SR (validated in SR knockout mice) and glial fibrillary associated protein (GFAP). For animal studies, we examined WT and TgF344-AD rats at 6 and 16 months of age, as the forebrain pathology in TgF344-AD rats does not appear until 16 months. Immunofluorescence was used to test for the co-localization of SR/D-serine with GFAP. High-pressure liquid chromatography was used to measure L- and D-serine. Western blotting was used to compare the expression levels of various proteins in the hippocampus.

Results: Subjects with advanced AD Braak stages (Braak V-VI) had massive astrogliosis in the entorhinal cortex (particularly layer II) and hippocampus, as visualized by GFAP immunoreactivity. Importantly, these GFAP+ reactive astrocytes highly expressed SR. In contrast, subjects with low Braak stages had minimal astrogliosis, with SR being expressed in neurons across all three examined brain regions, but not in quiescent GFAP+ astrocytes. There were a small number of reactive astrocytes observed in low Braak stage subjects that also expressed SR. We found much fewer SR+ neurons in advanced Braak stage compared to low Braak stage subjects across all three brain regions, which is likely due to neurodegeneration. Aged (16mo) TgF344-AD rats showed the same pattern as advanced Braak stage human tissue, with lower levels of neuronal SR and high expression of SR in reactive GFAP+ astrocytes. Interestingly, the total amount of SR protein did not differ between genotypes. We also found that the reactive astrocytes in 16mo TgF344-AD rats were enriched for D-serine. TgF344-AD rats had increased tissue content of L-serine in the hippocampus compared to WT rats. As astrocytes are the major source for brain L-serine, this is consistent with the increased number of astrocytes in TgF344-AD rats. Finally, the levels of phospho-Akt Ser473 was decreased and phospho-GluN2B Ser1303 increased in TgF344-AD rats. These intracellular signaling changes are consistent with increased extrasynaptic NMDAR activation and decreased cell survival.

Conclusions: Our findings are the first demonstration that reactive astrocytes express SR in human brain, specifically in tissue from subjects with advanced AD Braak stages. Furthermore, we observed a similar phenomenon in a transgenic rat model of AD that exhibits astrogliosis. As the amount of SR protein did not differ between WT and TgF344-AD rats, it reinforces the importance of investigating the cellular distribution of proteins, not just their overall expression levels. We hypothesize that D-serine being released from reactive astrocytes binds to extrasynaptic GluN2B-containing NMDARs, thereby driving excitotoxic signaling and ultimately cell death. This hypothesis is supported by the increased phosphorylation of GluN2B at Ser1303 in TgF344-AD rats. In sum, our findings have important implications not only for AD, but for other diseases associated with SR-expressing reactive astrocytes such as traumatic brain injury, and highlight this pathway as a potential a therapeutic target to block NMDAR excitoxicity.

Keywords: Serine Racemase, D-serine, Alzheimer's Disease, Reactive Astrocyte, NMDA Receptor

Disclosure: Nothing to Disclose.

W166. Simultaneous PET/MR Imaging for the Exploration of Serotonin 5-HT1A Receptor Biased Agonists

Benjamin Vidal, Sylvain Fieux, Jérôme Redouté, Didier Le Bars, Adrian Newman-Tancredi, Nicolas Costes, Luc Zimmer*

CERMEP-Imaging Platform, BRON, France

Background: In neuropharmacology, the recent concept of “biased agonism” implies the capacity of highly specific agonists to target specific intracellular pathways in specific brain areas. In the context of serotonin pharmacotherapy, 5-HT1A receptor biased agonists can be of interest to optimize pharmacological treatments of several neuropsychiatric disorders. The aim of this study was to bring additional support to this concept thanks to simultaneous functional MRI and PET molecular imaging (PET-MRI). We compared a highly selective 5-HT1A receptor agonist, F13640 (also known as befiradol or NLX-112, a drug-candidate for the treatment of L-DOPA-induced dyskinesia) with a 5-HT1A receptor biased agonist, F15599 (also known as NLX-101, a drug-candidate for the treatment of breathing deficits in Rett syndrome), at different doses, in anaesthetized cats. The drug occupancy was measured by PET imaging with [18F]MPPF, a 5-HT1A receptor radiotracer, after administration of both agonists and was correlated with agonist-induced brain activation patterns.

Methods: PET and fMRI data were acquired simultaneously using a Siemens Biograph mMR hybrid camera (CERMEP-Imaging Platform). [18F]MPPF was injected in a 90-min perfusion (bolus followed by a constant infusion) in isoflurane-anesthetized cats (n=4). After a post-injection equilibrium period of 50 min, F13640 or F15599 was injected i.p. The 5-HT1A receptor occupancy was quantified by comparing [18F]MPPF binding potential values, before and after the pharmacological challenge. Thirty minutes after the beginning of the PET acquisition, continuous T2*EPI MRI acquisitions were performed to measure the Blood Oxygen Level Dependence (BOLD) signal. The fMRI session was divided into 20 minutes of baseline and 20 minutes after drug injection. A voxel based analysis of the fMRI data was performed at an individual level, followed by a second-level analysis. This PET-MRI protocol was replicated for each cat with increasing doses of F13640 and F15599 (0.04, 0.08, 0.16 mg/kg ip).

Results: PET and fMRI data, taken together, showed clear differences between the two agonists in terms of binding and subsequent activation patterns. F13640 displayed similar binding in brain regions expressing both presynaptic and postsynaptic 5-HT1A receptors and elicited BOLD modifications in multiple brain regions. In comparison, the biased agonist F15599 showed preferential binding and activation in brain regions expressing postsynaptic 5-HT1A receptors, with little capacity to elicit 5-HT1A activation in the raphe nucleus, a brain region that expresses presynaptic 5-HT1A autoreceptors.

Conclusions: This study is the first simultaneous exploration of receptor occupancy and its consequences in terms of brain activation of 5-HT1A receptors. PET-fMRI represents a powerful tool in neuropharmacology and opens new ways to address the innovative concept of biased agonism.

Keywords: Serotonin 1a Receptor, Simultaneous PET-MR, Functional Connectivity, Cats/Foxes

Disclosure: Nothing to Disclose.

W167. Targeted GABA A Modulation in the Mouse Model of Fragile X Syndrome

Craig Erickson*, Sarah Fitzpatrick, Matthew Davenport, Ernest Pedapati, John Sweeney, Tori Schaefer

Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, United States

Background: Fragile X Syndrome (FXS) is the most common inherited form of developmental disability (DD) and most common single gene cause of autism spectrum disorder (ASD). FXS is caused by a CGG triplet repeat expansion resulting in methylation and transcriptional silencing of the Fragile X Mental Retardation 1 gene (FMR1) and subsequent deficient production of Fragile X Mental Retardation Protein (FMRP). FMRP is a known RNA binding protein responsible for translational control of many mRNAs involved in intracellular and synaptic signaling and plasticity. Brain dysfunction in FXS is in part associated with an imbalance in excitatory/inhibitory signaling. Excessive excitatory metabotropic glutamate receptor 5 (mGluR5) signaling has been demonstrated in FXS and has been the target of extensive treatment research in the field including multiple failed late-phase clinical trials. Inhibitory signaling including phasic and tonic inhibition in the Fmr1 knock out (KO) mouse model of FXS is drastically reduced in brain regions important for emotional, fear, and cognitive processing. Specifically, subunits of the gamma-aminobutyric acid receptor A (GABA(A)) which regulate inhibitory control are regulated by FMRP and are reduced in the amygdala, hippocampus, and cortex of Fmr1 KO mice. This suggests potentiation of GABA(A) activity may be a viable treatment mechanism in FXS. Non-specific positive modulation of GABA(A) using traditional benzodiazepines is not suitable for long-term pharmacotherapy in individuals with FXS due to sedating and cognitive dulling effects via modulation at α1 and α5 receptor subunits, respectively. Given these factors, we proposed to study the selective GABA(A) α2,3 agonist AZD7325 in the Fmr1 KO mouse.

Methods: We evaluated 1 mg/kg AZD7325 (low dose), 3 mg/kg AZD7325 (high dose) and vehicle (VEH) in Fmr1 KO mice compared to wildtype littermates. All testing, except audiogenic seizure assessment, was completed in adult mice (2-4 months of age) following 4 weeks of daily treatment. To assess for tolerability of AZD7325, cerebellar and striatal dependent motor tasks including open-field activity, pole descent, rotorod, and an adhesive removal task were completed. The impact of drug on memory was evaluated using a novel object recognition paradigm. Impact on dysregulated sensory responses was evaluated using an acoustic startle paradigm. A conditioned fear paradigm assessed drug impact on excessive freezing behavior noted in the Fmr1 KO mouse. The impact of AZD7325 treatment on the audiogenic seizure phenotype of juvenile Fmr1 KO mouse was additionally evaluated. Finally, the impact of drug on excessive hippocampal extracellular signal-related kinase (ERK) activation in the Fmr1 KO mouse was assessed. All behavior data were analyzed by two-way ANOVA with repeated measures when warranted (2-sided). Data were corrected for multiple comparisons using the False Discovery Rate method with significant results reported as p<0.05 corrected for multiplicity.

Results: No AZD7325-associated motor task deficits were noted. In the novel object recognition task, the Fmr1 KO mice had a significant baseline deficit remembering a novel object that was rescued with low- but not high-dose AZD7325 treatment (n= 14-18 per group). In an acoustic startle paradigm, the Fmr1 KO mice exhibited exaggerated responses to low intensity white noise (82 dB) stimuli and reduced response to high intensity stimuli (120 dB). Low-dose AZD7325 corrected low intensity noise response to WT levels and both low- and high-dose chronic treatment with AZD7325 normalized high intensity acoustic response (n=14-18 per group). Sixty percent of VEH-treated three-week-old Fmr1 KO mice displayed tonic-clonic seizure following a loud siren stimulus. Seizure activity was significantly reduced in the Fmr1 KO mice following a single dose of both 1 or 3 mg/kg AZD7325 given 30 minutes prior to the siren stimulus in a dose dependent manner (n=15 per group). Excessive hippocampal activated (phosphorylated) ERK was attenuated to WT levels following treatment with low-, but not high-dose AZD7325 (n= 6-8 per group).

Conclusions: AZD7325 treatment corrected several deficits that characterize the Fmr1 KO mouse model of FXS. The novel mechanism of action of AZD7325 has not been explored in humans with FXS. Given the extensive human safety data describing use of AZD7325 combined with these promising pre-clinical FXS-specific results, human FXS study of AZD7325 is warranted. To address the many translational drug development failures in FXS in recent years, human work must appreciate the complexity of translating Fmr1 KO drug treatment findings. Doing this will require use of extensive pharmacodynamic measures and a quantitative outcome measure battery to best understand target engagement and potential response in a first in disorder AZD7325 FXS proof of concept study.

Keywords: Fragile X Syndrome, GABAA, ERK

Disclosure: Part 1: Confluence Pharma, Stock / Equity.

W168. NP10679: A Ph-Sensitive GluN2B Inhibitor is Effective in Models of Ischemia and Nicotine Relapse

Robert Zaczek*, Meyers Scott, Cassandra Gipson-Reichardt, Haichen Wang, Lyuboslavsky Polina, Peter Kalivas, Daniel Laskowitz, Kamalesh Ruppa, George Koszalka

NeurOp CNS Drug Discovery Inc., Higganum, Connecticut, United States

Background: Inhibition of N-methyl-D-aspartate receptors, especially the GluN2B subunit, has been proposed as a therapeutic intervention for a number of disorders including stroke, pain and depression. Previous GluN2B inhibitors have shown promise but are limited by poor tolerability and/or bioavailability. NP10679 is a selective, pH-dependent GluN2B inhibitor. In experiments using human GluN2B containing NMDA receptors expressed in frog oocytes, NP10679 is 6.2-fold more potent (IC50=21 nM) at pH 6.9 than at pH 7.6 (130 nM). The NP10679 pH dependence increases selectivity and potency under conditions of lower pH found in ischemic brain tissue and conditions thought to be driven by excess glutamate release, such as neuropathic pain and relapse to drug abuse. NP10679 also interacts with the histamine H1 receptor (40 nM) and the hERG channel (621 nM). However, extensive safety pharmacology experiments indicate adequate margins between the desired effects of NP10679 and any untoward effects associated H1 or hERG inhibition. Mild sedation occurs at plasma levels several fold greater than those associated with efficacy (see below) and will serve as an indirect biomarker of NP10679 activity in planned clinical studies. NP10679 has a 6-8 hr half-life in mouse, rat and dog, excellent brain penetrance and good oral bioavailability (F=75%). Based on acceptable safety and formulation data, an IND for NP10679 was opened 3Q 2016.

Methods: Stroke model - Vehicle or doses of NP10679 (n= 10-14), were tested for ability to reduce neuronal cell death in a mouse MCAo transient ischemia stroke model. Doses were dosed 15 min prior to 60-min occlusion. Brains were fixed, sectioned, and stained for digital measurement of infarct volumes. Throughout the experiment from dosing through data analysis the experimenter(s) were blinded to dose groups.

Subarachnoid hemorrhage (SAH) model - Mice were randomly assigned to treatment groups (n=13-14 per group) and investigators blinded to groups throughout each study. Mice were subjected to aneurysmal SAH injury and administered NP10679 or vehicle immediately following SAH, and at 12 hour intervals over 3 days. Mice were evaluated for sensory function (Neuroseverity Score) on Day 0 (prior to SAH surgery) and daily for three days post SAH.

Reinstatement of Nicotine Seeking - Rats underwent overnight food training followed by daily 2-hr nicotine self-administration sessions in which one response (fixed-ratio 1) on the active lever yielded one intravenous nicotine infusion (0.02 mg/kg/infusion [pH=7.4], followed by a 20-sec timeout period) paired with cue lights and a discrete tone cue (2900 Hz tone). An inactive lever was available throughout the duration of each session to control for nonspecific responding. Following 10 consecutive sessions (≥ 10 infusions/day), rats were placed into daily extinction training sessions (no nicotine delivery or cues) for 14 sessions, or until extinction criteria were met. Reinstatement to the active lever was elicited by contingent conditioned cues. NP10679 was administered 30 min prior to the reinstatement session.

Rotarod - Groups of eight mice were given single IP doses of vehicle, Ifenprodil at 30 mg/kg or NP10679 at 2, 5, 10 mg/kg. Rotarod performance was evaluated at 20, 45, 70, and 95 minutes post-dose.

Overt Neurobehavioral Signs – Studies were conducted as part of the pre-IND safety work performed at MPI Research (Mattawan, MI). Male rats (n=10 per group) were given single IV doses of vehicle or NP10679 at 3, 8, or 15 mg/kg by 20-minute infusion. Rats were evaluated for clinical signs within 5 minutes after infusion ended and at 24 hours post-infusion.

Results: NP10679 mitigates impairment in a model of SAH and limits neurodegeneration in a MCAo model of stroke. The minimal effective dose in both assays was 2 mg/kg with near maximal effects at 5 mg/kg. NP10679 did not affect mean rotarod performance at≤5 mg/kg, but worsened performance at 10 mg/kg (thought due to mild sedation), indicating a 5-fold efficacy to side effect margin. To determine if NP10679 is effective in a paradigm associated with glutamate hyperfunction rather than ischemic injury, NP10679 was tested for ability to interfere with cue-induced reinstatement of nicotine seeking in rats. NP10679 was highly effective at a dose of 2 mg/kg IP. A clear margin between these effects and overt neurobehavioral signs exists since the only NP10679-related clinical observations were decreased activity in 5 rats and partially/completely closed eyelids in 3 rats at the end of infusion at 15 mg/kg. These were considered manifestations of sedation.

Conclusions: The pH-sensitive GluN2B inhibitor NP10679 is effective in two models of ischemic brain injury and in cue-induced reinstatement of nicotine seeking. The finding that other GluN2B inhibitors block reinstatement of both opiate and nicotine seeking suggests NP10679 may provide therapy for opiate abuse. Pre-IND toxicity studies conducted with NP10679 indicate it can be safely administered to human subjects in a Phase I clinical trial planned to begin in 2018. The selectivity of NP10679 will allow for prophylactic use in patients at risk for an ischemic event and also for other indications like depression, uncontrolled pain and substance abuse. The authors acknowledge the financial support of the National Institute of Neurological Diseases and Stroke (NINDS) and the National Institute on Drug Abuse (DA036569 to C.D.G.) to fund this work.

Keywords: NMDA Antagonists, Drug Relapse, Neurodegeneration, GluN2B

Disclosure: Part 1: NeurOp, Inc., Employee, Part 2: NeurOp, Inc., Employee, Part 5: NeurOp, Inc., Employee.

W169. A Novel Role of the Neuropsychiatric Risk Gene CANA1C in Mitochondrially-Mediated Neuronal Cell Death

Maria Noterman, Anjali Rajadhyaksha, Eric Taylor, Andrew Pieper*

University of Iowa Carver College of Medicine, Iowa City, Iowa, United States

Background: CACNA1C encodes the α1 subunit for Cav1.2, the predominant pore-forming voltage-gated L-type calcium (Ca2+) channel (LTCC) subunit in the brain, and confers much of the channel's functional properties, including drug sensitivity, selectivity, and composition of both the pore-forming unit and voltage-sensing domain. Cav1.2 plays an essential role in coupling plasma membrane depolarization to Ca2+ influx into neurons to initiate signaling cascades and expression of transcription factors, with downstream effects such as enhancing expression of neurotrophic factors. Disrupted LTCC activity has also been implicated in an array of chronic, progressive neuropsychiatric and neurodegenerative conditions. For example, in the most extensive genome-wide association study of patients with mental illness to date, risk loci in CACNA1C were found in common across five major disorders: bipolar disorder, schizophrenia, major depression, attention deficit-hyperactivity disorder, and autism spectrum disorder. We have shown that forebrain-specific Cacna1c-deficient mice display both elevated anxiety-like behavior and elevated cell death of young newborn neurons in the hippocampal dentate gyrus. Given the links between Ca2+ handling and mitochondrially-mediated neuronal cell death, we hypothesized that Cav1.2 may play a role in this process in a manner that protects stressed neurons from cell death.

Methods: We have applied western blotting for Cav1.2 in biochemical preparations of purified neuronal mitochondria from brain-specific Cacna1c mice and their wild type littermates. We have also assessed measures of mitochondrial respiration and calcium loading in both isolated mitochondria and in primary neuronal cultures from brain-specific Cacna1c mice and their wild type littermates. In addition, we have applied indicators of reactive-oxygen species (ROS) damage to neurons in the brains of brain-specific Cacna1c mice and their wild type littermates.

Results: We have observed elevated neuronal cell death in the brains of brain-specific Cav1.2-deficient mice, as well as increased ROS-mediated damage throughout their brains. Elevated ROS damage is compatible with the fact that high levels of mitochondrial calcium can uncouple oxidative phosphorylation, leading to increased generation of ROS. We have also obtained pharmacologic evidence that activating LTCCs increases membrane potential of isolated neuronal mitochondria, while LTCC blockers delay the release of calcium from mitochondria into the cytosol of hippocampal neurons.

Conclusions: Synthesis of our results has allowed us to generate the following model, which is formally presented here for the first time. Under normal polarized neuronal conditions, Ca2+ enters mitochondria via the mitochondrial calcium uniporter (MCU), and is maintained at normal concentration in mitochondria through steady slow exit via Ca2+ exchangers in the mitochondrial membrane. However, when neuronal depolarization leads to rapid Ca2+ influx, such as occurs in excitatory synaptic signaling that can progress to excitotoxicity, there is excessive Ca2+ entry into mitochondria via the MCU. Our data suggests that this stress can be relieved via Cav1.2 activation acting as a mitochondrial release valve to rapidly allow additional flow of Ca2+ out of mitochondria into the cytosol. In circumstances of overwhelming mitochondrial Ca2+ overload, however, there is full oxidative phosphorylation uncoupling and cell death as a consequence of increased ROS formation, as well as formation of the death-inducing mitochondrial permeability transition pore. We thus hypothesize that functional inhibition or genetic elimination of mitochondrial Cav1.2 forces mitochondrial Ca2+ levels to a chronically higher state, rendering neurons poised for death as they are less able to withstand acute increases in Ca2+ influx. If this model is correct, it could help explain why aberrations in CACNA1C render individuals susceptible to such a wide spectrum of neuropsychiatric diseases.

Keywords: Calcium, Neurodegeneration, Mitochondria, CACNA1C

Disclosure: Part 1: Neuroprotective Therapeutics, Consultant, Neuroprotective Therapeutics, Stock / Equity, Neuroprotective Therapeutics, Royalties, Part 2: Neuroprotective Therapeutics, Consultant.

W170. Enhancing Emotion Regulation in Borderline Personality Disorder Patients Through Longitudinal Reappraisal Training: Evidence From Self-Reported Negative Affect and fMRI

Harold Koenigsberg*, Bryan Denny, Jin Fan, M. Mercedes Perez-Rodriguez, Antonia New, Erin Hazlett, Daniel Rosell, Margaret McClure, Jacqueline Trumbull, Maria Lopez

Icahn School of Medicine at Mount Sinai, Bronx, New York, United States

Background: Borderline personality disorder (BPD) is the prototypical disorder of emotion dysregulation. We have shown previously that when BPD patients attempt to employ the highly adaptive emotion regulatory strategy of cognitive reappraisal, they do not engage, as healthy subjects (HCs) do, brain regions known to participate in cognitive reappraisal. In this study, we assessed whether BPD patients could be trained to enhance cognitive reappraisal, thereby improving reappraisal success and normalizing neural activity during reappraisal.

Methods: At each of six sessions over approximately 4 weeks, BPD and healthy control (HC) participants were shown negative social emotional images that were specifically chosen for their relevance to core BPD themes of interpersonal anger, exclusion, sadness, and abandonment. These stimuli were counterbalanced across sessions and conditions, and participants were given instructions to reappraise their responses to half and to look and respond naturally at the other half. Subjects were instructed in the reappraisal by psychological distancing tactic, which involves viewing stimuli as an impartial, objective observer. Emotion self-reports were obtained after each image presentation. Sessions 1-5 were spaced two days apart and afforded training through practice on novel images. Session 6 was a follow-up session using the same training and task approximately 2 weeks following Session 5. fMRI data were acquired at Sessions 1, 5, and 6. The present analyses reflect data from a preliminary sample of 14 BPD's and 16 HC's (Sessions 1-5) and a preliminary subsample of 10 BPD's and 13 HC's who have also completed Session 6.

Results: BPD patients showed significantly reduced negative emotion self-reports during reappraisal-by-distancing over the course of the training period (p<.009) (Session 1 to Session 5) and this was maintained over the follow-up period (p<.004) (Session 1 to Session 6). Further, with active training (i.e. Session 1 to Session 5), BPD patients, who had not decreased amygdala activity on day 1, were able to achieve decreased amygdala activity when reappraising on day 5 (p<.05). Moreover, preliminary data indicates that this attenuation was preserved at the follow-up session two weeks later. With training, BPD patients also increased reappraisal activity in ventrolateral prefrontal cortex, a region shown to be engaged during reappraisal in HC's in this and prior studies.

Conclusions: These data represent the first evidence that longitudinal training can increase reappraisal success and normalize reappraisal neural activity in any patient population and suggest a potential translational role for reappraisal training in BPD treatment.

Keywords: Cognitive Reappraisal, Borderline Personality Disorder, Reappraisal Training, fMRI

Disclosure: Nothing to Disclose.

W171. The Functional Neuroanatomy of Neurofeedback Control

Christian Paret, Jenny Zaehringer, Matthias Ruf, Gabriele Ende, Christian Schmahl*

Central Institute of Mental Health, Mannheim, Germany

Background: Brain-computer interfaces (BCIs) provide real-time feedback from neural activity that is used for therapy purposes, sometimes named ‘neurofeedback’ therapy. The functional neuroanatomy of neurofeedback control is incompletely understood and is examined in this study.

Methods: Healthy female subjects (N=20) participated in a real-time fMRI neurofeedback study (TR=1s) and received feedback from the amygdala's BOLD signal, while they were instructed to up- and down-regulate feedback. We analyzed feedback-related and condition-related activity to dissociate brain responses to feedback from brain responses to up- and down-regulation blocks.

Results: Results revealed a thalamo-striatal-frontal cortex network with response patterns characteristic for different types of neurofeedback information. Ventral striatal (VS) and rostral prefrontal cortex (PFC) activations correlated with amygdala-regulation success. Thalamus showed non-specific responses that may drive attention and perceptual processing. VS responses were guided by the reinforcer value of feedback and provide an essential signal for learning. Rostral PFC responses were modulated by task-differences and may contribute to regulation under varying contextual demands. Insula, dorsal anterior cingulate (ACC) and dorsolateral PFC activation was exclusively related to control, not to feedback processing. Up- and down-regulation activated subgenual ACC in opposite directions, a finding in line with a presumed function in amygdala modulation. Feedback-related and condition-related responding overlapped in thalamus and VS.

Conclusions: This study reveals differential neural networks associated with neurofeedback processing and neurofeedback regulation. The results contribute to the development of more elaborate models on the neuropsychological basis of BCI control. Limitations result from limited coverage of the MRI-volume, lacking parietal and dorsal-to-posterior ACC. Generalization to other neurofeedback signals remains to be shown.

Keywords: Real-Time fMRI Neurofeedback, Amygdala, Prefrontal Cortex

Disclosure: Nothing to Disclose.

W172. Inflammatory Markers and Cognitive Performance in Patients With Schizophrenia Treated With Lurasidone

Brian Miller, Andrei Pikalov, Cynthia Siu, Michael Tocco, Joyce Tsai, Philip Harvey, Antony Loebel*

Sunovion Pharmaceuticals, Inc., Fort Lee, New Jersey, United States

Background: Recent studies have linked inflammation, obesity, and lipid dysregulation with cognitive impairment, a core feature of schizophrenia. Elevated C-reactive protein concentration has been shown to be a reliable biomarker for inflammatory states. We conducted an exploratory analysis to investigate the potential influence of inflammation, obesity and lipid metabolism on changes in symptom severity and cognitive performance in patients with schizophrenia treated with lurasidone.

Methods: Patients with acute exacerbation of schizophrenia were treated with one of two fixed doses of lurasidone (80 or 160 mg/day), placebo, or 600 mg/day quetiapine XR in a 6-week double-blind study. Efficacy was assessed with PANSS, CGI-S and MADRS scores. A wide-range CRP (wr-CRP) assay (equivalent to high sensitivity CRP assay) was used to assess levels of inflammation. CRP was evaluated as a logarithm transformed (log) continuous variable and as a categorical variable divided into low (≤ 2 mg/L), medium (> 2 mg/L and≤5 mg/L) and high (> 5 mg/L) subgroups (Raison et al., 2013). Cognitive function was assessed with the Cogstate computerized cognitive battery at baseline and week 6 endpoint. To address the issue of incomplete data from the non-evaluable sample, an analysis based on the full ITT sample was conducted using rank score, in which non-evaluable scores at baseline and/or week-6, missing cognitive testing scores and/or early dropouts were assigned the lowest possible rank score (=0). Tied ranks were resolved by standard averaging methods. Nonparametric bootstrap resampling method was applied to estimate the main and interactive effects of CRP on ranked cognitive scores based on the 95% percentile bootstrap confidence intervals of 1500 replicates.

Results: Elevated level of wr-CRP (log) was associated with cognitive impairment at study baseline (P<0.05), with significantly lower cognitive performance in the subgroup with high wr-CRP (> 5 mg/L) compared to those with low wr- CRP (< 2 mg/L) at study baseline (P<0.05). Higher level of CRP (log) was also associated with significantly greater symptom severity as assessed by PANSS score, as well as higher BMI/body weight, and lower levels of high-density lipoprotein (HDL) and high hemoglobin A1c (HbA1c) at study baseline (P<0.05). No significant associations were found for wr-CRP (log) with low-density lipoprotein (LDL) and total cholesterol at study baseline. High wr-CRP level (> 5 mg/L) at study baseline predicted less improvement of cognitive composite score at week 6 endpoint for all treatment groups, compared to those with low to medium wr-CRP levels (< 5 mg/L).

The joint effect of wr-CRP (log) and HDL on moderating cognitive efficacy of lurasidone was significant (P<0.05), with greater lurasidone (vs. placebo) effect size in patients with low wr-CRP and high HDL concentration. Likewise, the joint effects of wr-CRP (log) and HOMA IR on moderating cognitive efficacy of lurasidone was significant (P<0.05), with greater lurasidone (vs. placebo) effect size in patients with low wr-CRP and low HOMA IR concentration.

Lurasidone treatment was associated with significant reduction in symptom severity as assessed by PANSS, CGI-S, and MADRS change scores from baseline to week 6, independent of wr-CRP, HDL and HOMA IR levels at study baseline. Lurasidone had no significant effect on change in wr-CRP level from baseline to week 6 endpoint.

Conclusions: Results from cross-sectional analyses suggest that elevated wr-CRP and disturbances in lipid metabolism may affect cognition in patients with schizophrenia. Our findings from this exploratory analysis suggest the joint effects of low wr-CRP level combined with either high HDL or low HOMA IR can predict cognitive improvement in patients treated with lurasidone (vs. placebo). Our overall findings suggest that inflammation and its interactive effects with insulin resistance and lipid parameters in patients with schizophrenia might impact cognition and response to treatment with antipsychotics.

Keywords: Inflammatory Markers, Cognitive Functioning, Lurasidone, Lipids, Schizophrenia; Functional Capacity; Antipsychotic Treatment

Disclosure: Part 1: Sunovion Pharmaceuticals, Inc., Employee, Part 2: Sunovion Pharmaceuticals, Inc., Employee, Part 3: Sunovion Pharmaceuticals, Inc., Employee, Part 4: Sunovion Pharmaceuticals, Inc., Employee, Part 5: Sunovion Pharmaceuticals, Inc., Employee.

W173. Neuroimaging Biomarkers of Response to Cognitive Remediation in Veterans With Schizophrenia

Philip Szeszko*, Margaret McClure, Fionna Graff, Tamar Lichtman, King-Wai Chu, Michelle Feder, Maureen Kilmade, Larry Siever, Erin Hazlett

Icahn School of Medicine at Mount Sinai, Bronx, New York, United States

Background: Schizophrenia spectrum disorders (SSDs) account for the highest proportion of disability adjusted life years caused by all mental disorders with 4% of Veterans having this diagnosis, but occupying more hospital beds than Veterans with any other illness. These individuals demonstrate significant deficits in cognitive and everyday (i.e., adaptive and social) functioning that contribute to lengthy and repeated VA hospital admissions. In addition, magnetic resonance imaging studies are critically needed to elucidate neural changes associated with successful cognitive remediation (CR) to develop new therapeutic targets for the treatment of impairments and for the successful prediction of treatment response using a personalized medicine approach.

Methods: The current study included four cohorts of individuals with SSDs (mean cohort size=4 participants, range: 3-6 participants) with excellent attendance (mean attendance: 79.11%, mean absences: <1.0 per session). The total sample comprised 18 Veterans, mostly men (84%, n=16), with a mean (SD) age=49.3 (8.6). All participants received outpatient psychiatric care at the James J. Peters VA Medical Center (JJPVAMC), and had a diagnosis of schizophrenia or schizoaffective disorder, no active substance dependence and/or symptoms of major depressive disorder. Patients were stabilized on antipsychotic medications (i.e., no changes of medication dose adjustments of >25%) for at least 4-weeks prior to the initiation of CR.

CR consisted of a combination of computer-based cognitive enhancement exercises (Psychological Software Services) and manualized-social skills training modified from the Cognitive Enhance Therapy program (CET; Hogarty and Greenwald, 2006). Specifically, participants met for 60 minutes to complete computer based cognitive enhancement exercises plus 60 minutes of social-skills training twice weekly (4 hours total per week) over 7.5 weeks for a total of 30 sessions (15 computer-based CET+15 social skills training). Ten participants received structural and diffusion tensor imaging exams on a Siemens 3T scanner prior to the initiation of CR. The anisotropy and eigenvector maps were computed off-line. We employed the tract-based spatial statistics (TBSS) pipeline that was adapted by the Enhanced Neuroimaging Genetics by Meta Analysis (ENIGMA) DTI Working Group at the USC to compute white matter regions-of-interest. Gray matter volumes were computed using FreeSurfer.

Results: To assess cognitive functioning, patients were administered selected subtests of the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS)-Consensus Cognitive Battery. There were significant (p<.05) improvements in these scores following CR for visual-motor speed of processing, verbal speed of processing, verbal learning, reasoning and problem solving, social cognition and a measure of visual spatial working memory (DOT Test). Baseline (pre-treatment average diffusion tensor imaging fractional anisotropy) significantly (p <.05) predicted better visual spatial working memory following CR. Moreover, structural imaging analyses revealed that greater baseline cerebellum volume and more middle temporal gray matter volume significantly (p <.05) predicted visual spatial working memory performance following CR.

Conclusions: Taken together, Veterans with SSDs significantly improved in the majority of areas assessing neurocognition and social cognition following 7.5 weeks of CR. Patients were highly engaged in CR with excellent attendance. Deficits in visual spatial working memory have been identified as an important endophenotype for SSDs and may benefit from targeted treatment approaches. Moreover, these findings also highlight potential neuroimaging biomarkers associated with improvements in visual spatial working memory in response to CR among individuals with SSDs that may be useful in identifying therapeutic targets for improving functional outcome.

Keywords: Cognitive Remediation, Psychosis, Neuroimaging

Disclosure: Nothing to Disclose.

W174. The Effects of Oxytocin on EEG and Pupillary Measures of Social Processing: A Dose-Finding Study

Eric Reavis*, Jonathan Wynn, Jennifer Hoy, Jaime Morales, Gerhard Hellemann, Michael Green, Stephen Marder

University of California, Los Angeles, California, United States

Background: Serious mental illnesses such as schizophrenia are associated with impairments in the ability to process social information, referred to as social cognition. These impairments involve skills in detecting basic social cues, such as identifying emotions expressed in faces, as well as detecting human movement in animated figures. These impairments present an attractive target for intervention since they are associated with poor community functioning, including the ability to work, succeed in educational activities, and to sustain social relationships. Oxytocin (OT) is a peptide hormone that plays an important role in regulating mammalian social behaviors and has been shown to support the formation of attachment bonds in animals. These effects are paralleled in healthy humans for whom OT can have prosocial effects that include improving trust and cooperation. There is compelling evidence that intranasal OT has effects on social processes in healthy individuals. OT has been shown to enhance electrophysiological and pupillometry measures sensitive to social processes. Mu-suppression, the decrease in EEG mu activity (8-13 Hz) typically seen when viewing biological motion in comparison to non-biological motion, has been shown to be enhanced with OT. Similarly, pupillary dilation in response to facial expression is increased after OT administration, indicating increased emotional salience attributed to faces in response to OT. Direct effects of OT on brain-based measures of social processing have not been examined in schizophrenia. Hence, there is a need for valid biomarkers to be used in clinical studies to objectively measure the effects of OT. Additionally, important decisions for clinical trials, including fundamental questions about dose, have not been systematically explored. In the current study, we evaluated indicators of OT's effects on the brain by specifically examining the dose-response relationship for target measures of neural engagement on a facial affect processing task and a biological motion task.

Methods: Fifty-one stable outpatients with a DSM-5 diagnosis of schizophrenia participated in a double-blind, within-subjects cross-over study. All participants were taking antipsychotic medication at the time of testing. Eight different doses of OT (8, 12, 24, 36, 48, 60, 72 or 84 IU) were utilized, with 6 participants assigned to each condition. A final sample of 47 participants had valid data for both OT and PL sessions. OT and PL administrations occurred a week apart in randomized order. Doses were administered intranasally thirty minutes prior to testing. Participants first had EEG recorded while viewing a series of movie clips depicting biological or non-biological motion; we examined mu-suppression due to biological motion (in relation to non-biological motion) as a function of dose and drug (OT vs. PL). Next pupillometry was recorded while viewing faces depicting happy, neutral or fearful expressions, in addition to scrambled faces. We examined the ratio of change in pupil size from a 100 ms baseline to the mean of pupil size 1000-3000 ms post-stimulus onset. The EEG and pupillometry biomarkers were evaluated while blind to study condition. To make maximal use of the repeated measures design we used a generalized linear mixed model (GLMM) for all analyses of treatment effects. The GLMM included fixed effects for time, treatment and treatment-by-time interactions, and a random intercept term for each participant to control for individual differences. The primary effect of interest was the drug (OT vs. placebo) X dose interaction, as it models differences in change over time that can be attributed to treatment, with a significant finding at a specific dose(s) reflecting OT target engagement.

Results: For EEG, the task behaved normally, with significant mu-suppression seen for biological motion in relationship to non-biological condition. The results of the GLMM revealed a significant drug X dose interaction for mu-suppression. In particular, there were significant differences between OT and PL for the 36 IU (F1,37.9=5.11, p=0.03) and 48 IU (F1,37.9=7.74, p=0.008) doses. For pupillometry, the task behaved normally, with significantly more dilation to intact vs. scrambled faces, and significantly greater dilation to fearful vs. happy faces. However, there was no significant drug X dose interaction indicating the OT did not have an impact on pupil dilation at any dose.

Conclusions: The results demonstrate target engagement of OT for the biological motion mu suppression task, with greater mu-suppression (i.e., greater social salience) after administration of OT compared to PL. This effect was seen for doses (36 and 48 IU) in the middle range, but not for lower doses or for higher doses. Interestingly, the doses that showed significant effects are similar to those used in prior studies of OT in schizophrenia. However, OT did not affect the pupillary response, which is a more peripheral measure, when viewing faces. The results of this pilot study demonstrate that target engagement of OT can be assessed with brain-based biomarkers and is the first to demonstrate target engagement at specific doses. The results of the current study highlight the ability of OT to directly affect the brain's ability to process social information.

Keywords: Oxytocin, Mu Suppression, Pupillometry, EEG Biomarkers, Social Stimuli

Disclosure: Nothing to Disclose.

W175. Postsynaptic Serine Racemase Regulates Synaptic NMDA Receptors

Jonathan Wong, Eden Barragan, John Gray*

University of California, Davis, California, United States

Background: NMDA receptors (NMDARs) are ionotropic glutamate receptors that play crucial roles in synaptic plasticity and neuronal development and disturbances in NMDAR function have been implicated in a broad range of neuropsychiatric disorders, including schizophrenia. NMDARs are unique among perhaps all receptors in that, in addition to binding glutamate, they have an absolute requirement for a co-agonist, which can be either glycine or D-serine. The identity of the co-agonist is developmentally regulated and spatially restricted in the brain, with many synapses using glycine early on and later switching to D-serine. Despite the importance of these co-agonists in regulating excitatory synaptic transmission, neuronal function, and synaptic plasticity, the source and regulation of D-serine remain controversial. Although D-serine and its biosynthetic enzyme serine racemase (SR) were originally thought to be localized in astrocytes, recent studies using SR constitutive & conditional knock-out mice combined with more selective antibodies instead suggest a predominantly neuronal localization. Furthermore, SR both complexes and colocalizes with PSD-95 suggesting that SR may be localized at the postsynaptic density. Here we examine the functional effects of postsynaptic SR deletion.

Methods: To examine the potential role of postsynaptic SR in synaptic physiology, we eliminated SR in a small subset of hippocampal neurons by transcranial stereotaxic injection of a recombinant adeno-associated virus expressing a Cre:GFP fusion protein (rAAV1-Cre:GFP) into neonatal (P0) transgenic mice with conditional KO alleles of the SRR gene (SRcKO mice, a generous gift of Dr. Joseph Coyle). Acute hippocampal slices were then made from AAV-injected mice at various times post-injection and simultaneous whole-cell patch clamp recordings were obtained from transduced (i.e. GFP-expressing) and neighboring control neurons. This genetic approach utilizes sparse, localized transduction of single cells in order to rigorously analyze the cell-autonomous effects of SR deletion while avoiding presynaptic effects and the circuit-wide effects of global genetic or pharmacologic NMDAR manipulations. Synaptic recordings were obtained at Schaffer collateral-CA1 and perforant path-dentate gyrus (DG) synapses.

Results: Cre:GFP-expressing hippocampal neurons from adult (P45-60) SRcKO mice show an approximately 40% reduction in synaptic NMDAR current amplitude in CA1 and a 70% reduction in DG compared to simultaneously recorded neighboring control neurons. No effects of SR deletion were detectable in earlier development (P14-P21), consistent with the developmental co-agonist switch from glycine to D-serine. Synaptic AMPAR current amplitudes were not significantly different from control at any assessed time point in either CA1 or DG, indicating a selective effect on NMDAR signaling but not overall synaptic maturation or strength. Interestingly, deletion of SR also prolongs the decay time of NMDAR currents suggesting a selective loss of GluN2A-mediated currents.

Conclusions: These results suggest that postsynaptic neuronal SR regulates NMDAR co-agonist site occupancy in a cell-autonomous manner. Combined with recent reports that SR colocalizes with PSD-95, this implies that D-serine is be produced locally at the postsynaptic density, though the mechanism of postsynaptic D-serine release remains unknown. A reduction in postsynaptic D-serine synthesis and release would be predicted to lead to NMDAR hypofunction and the weakening or loss of dendritic spines, hallmarks of the pathophysiology of schizophrenia. Thus, a detailed understanding of the molecular mechanisms regulating postsynaptic SR may lead to the development of novel therapeutic strategies for schizophrenia and other neuropsychiatric diseases.

Keywords: NMDA Receptor, D-serine, Serine Racemase, Synapses, Synaptic Plasticity

Disclosure: Nothing to Disclose.

W176. Anthranilic Acid Elevation in Serum of Schizophrenia and Parkinson's Disorder Patients: Possible Marker of Neurodegeneration

Gregory Oxenkrug*, Paul Summergrad

Tufts University School of Medicine, Boston, Massachusetts, United States

Background: Dysregulations of Kynurenine (Kyn) pathway (KP) of tryptophan (Trp) metabolism were reported in neurodegenerative disorders. Notably, KP dysregulation was not restricted to brain sites but was observed in serum as well. Specifically, activation of the first step of KP, formation of Kyn from Trp, was reported in patients with schizophrenia (SP) and Parkinson's disorder (PD). Less attention was paid to down-stream metabolism of Kyn. i.e., Kyn conversion into kynurenic (KYNA), anthranilic (ANA) acids and 3-hydroxykynurenine (3HK).

Methods: Serum concentrations of Kyn metabolites were evaluated by HPLC-mass spectrometry method in SP (n=16), PD (n=24) and control subjects (n=20). Study was approved by Tufts Medical Center IRB.

Results: Serum concentrations of ANA, KYNA, and Kyn were elevated (several folds) in both PD and SP. Trp was decreased in PD but not in SP.

Conclusions: Dysregulation of KP was observed in SP and, more pronounced, in PD. Both diseases are characterized by abnormalities of dopamine signaling system. Our preliminary data warrant further assessment of serum concentrations of down-stream Kyn metabolites as potential markers of neurodegeneration.

Keywords: Anthranilic Acid, Kynurenic Acid, Parkinson's Disease, Schizophrenia

Disclosure: Nothing to Disclose.

W177. Childhood Adversity and Schizophrenia: The Protective Role of Resilience in Mental and Physical Health, and Metabolic Markers

Ellen Lee*, A'verria Martin, Xin Tu, Barton Palmer, Dilip Jeste

University of California, San Diego, La Jolla, California, United States

Background: Childhood maltreatment has been linked to worse mental and physical health and premature mortality in the general population. Several meta-analyses have shown that schizophrenia, a debilitating mental illness with prominent psychotic and negative symptoms, is strongly associated with childhood adversity, with a two to seven-fold higher prevalence of childhood abuse than in non-psychiatric comparison participants (NCs). Studies in people with schizophrenia and other serious mental illnesses consistently report that a history of childhood abuse is associated with poor outcomes, including: worse psychopathology, worse social functioning, treatment resistance, and lower quality of life.

Schizophrenia is also associated with increased physical morbidity and premature mortality, attributable to increased cardiovascular-related mortality, though the exact biological mechanisms are not clear. Metabolic dysfunction, a major risk factor for cardiovascular-related death, is more common in individuals with schizophrenia than in the general population and dysregulated glucose has been observed even in antipsychotic-naïve, normal-weight, first-episode patients. Moreover, the literature suggests that while obesity and antipsychotic medications may contribute to metabolic dysfunction in persons with schizophrenia, there may be other contributing factors that are intrinsic to having schizophrenia that precede the cumulative metabolic burden of lifestyle behaviors.

Childhood trauma itself may contribute to metabolic dysfunction in schizophrenia. In contrast to the negative effects of trauma, resilience may positively impact mental and physical health in the general population as well as in persons with schizophrenia. In the present study, we focus specifically on the role of resilience in tempering the effects of childhood adversity on outcomes of mental and physical health, specifically metabolic markers, among adults with schizophrenia and non-psychiatric comparison participants (NCs).

Methods: We conducted a cross-sectional study of 114 participants with schizophrenia (DSM-IV-TR criteria) and 101 NCs aged 26-65 years. Sociodemographic, clinical and laboratory measures were examined. Clinical assessments included psychotic symptoms (Scales for the Assessment of Positive and Negative Symptoms) and depression (Calgary Depression Rating Scale for depression). Childhood Trauma Questionnaire (CTQ) was used to assess emotional abuse/neglect, physical abuse/neglect and sexual abuse experienced during childhood. Connor-Davidson Resilience scale was employed to measure resilience. Other variables included mental well-being (Short Form Health Survey – Mental) and physical well-being (Short Form Health Survey – Physical).

Metabolic biomarkers included fasting insulin, glucose, hemoglobin A1c levels. The Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) was calculated to reflect insulin resistance. Independent samples t-tests and Chi-square tests were used to assess differences in baseline variables were compared between the schizophrenia and NC groups. The association of total CTQ scores with baseline variables were compared within each diagnostic group. General linear models were used to analyze the effect of trauma severity and resilience on mental and physical health measures and metabolic biomarkers.

Results: The schizophrenia and NC groups were comparable on age (mean 48-49 years), gender (43-53% female), and race/ethnicity (50-60% Caucasian). Persons with schizophrenia reported more severe childhood trauma, lower resilience, and worse mental and physical health; and had worse metabolic biomarker levels than NCs. Severity of childhood trauma correlated with worse depression in the NCs (r(99)=0.34, p=0.001), but not in the schizophrenia group (r(114)=0.024, p=0.80). In both groups, severity of childhood trauma was associated with worse physical well-being, higher fasting insulin levels and greater insulin resistance.

Notably, resilience appeared to counteract or even outweigh the effects of trauma on mental and physical health. The schizophrenia subgroup with high resilience and severe trauma had comparable or better physical health as NCs with low resilience and severe trauma across a number of measures: mental health well-being scores (t(28)=0.28, p=0.8, d=0.1.), physical well-being (t(28)=0.11, p=0.91, d=0.04), hemoglobin A1c (t(28)=1.07, p=0.29, d=0.42), and HOMA-IR (t(27)=0.77, p=0.45, d=0.32).

Conclusions: To our knowledge, this is the first study to quantitatively assess the effects of both childhood trauma and resilience in schizophrenia on health, notably metabolic function. Consistent with our a priori hypotheses, relative to NCs, people with schizophrenia reported significantly more childhood emotional abuse/neglect, physical abuse/neglect and sexual abuse. In the NC group, such childhood trauma was associated with worse mental and physical health; however, in the schizophrenia group, childhood adversity was associated with worse physical, but not mental, health. Resilience was associated with significantly better mental and physical health outcomes and metabolic biomarkers, regardless of trauma severity.

Interventions to bolster resilience in the general population and in people with schizophrenia may improve outcomes for those with a history of childhood adversity.

Keywords: Schizophrenia, Metabolic Function, Childhood Adversity

Disclosure: Nothing to Disclose.

W178. Disrupting Glutamatergic Transmission During Neurodevelopment Impairs Action-Outcome Associations and Motivation in Adult Rats

Samuel Barnes*, Andre Der-Avakian, Jared Young, Margarita Behrens, Mark Geyer

University of California, San Diego, La Jolla, California, United States

Background: Schizophrenia is a neurodevelopmental disorder characterized by a wide spectrum of symptoms. The negative symptoms of schizophrenia, including deficits in reward-related processing, predict the functional outcome of patients and are currently unresponsive to available medications. The ability to regularly update current representations of reward value and adjust behavior in response to alterations in reward is required to maintain goal-directed behavior. These processes can be quantified across species using the probabilistic reversal learning (PRL) procedure. PRL performance is impaired in schizophrenia. Schizophrenia patients also exhibit impairment in a progressive ratio breakpoint (PRBP) task that assesses motivational performance. Our aim was to determine whether neonatal phencyclidine (PCP; NMDA receptor antagonist) treatment produced deficits in rats performing the PRL or PRBP tasks.

Methods: Male and female Wistar rats received PCP (20 mg/kg, s.c., n=24) or saline (0.9% s.c. n=23) on postnatal day (PND) 7, 9, and 11; behavioral testing began on PND 60. Rats were tested in the PRL procedure using an 80:20 reward ratio once daily for 20 days. PRBP was conducted after PRL testing. All procedures were approved by the University of California San Diego Institutional Animal Care and Use Committee.

Results: Overall PRL performance was impaired in PCP- vs. saline-treated rats; fewer successful reversals [F(1,43)=9.32, p<0.01] were completed. Target win-stay [F(1,43)=13.74, p<0.001] and target lose-shift [F(1,43)=11.12, p<0.01] ratios were reduced in PCP-treated rats over the 20-day test period. Non-target win-stay was reduced [F(19, 817)=1.76, p<0.01] on days 1, 5, 8, 14, 17, and 18 (p<0.05) while non-target lose-shift was reduced [F(19, 817)=1.92, p<0.01] on days 1, 7, 12, 14, 15, 16, 17, 18, 19, and 20 (p<0.05). During the last 10 days of testing, the numbers of trials required to reach criterion for the initial discrimination, first reversal, and second reversal were increased in PCP-treated rats (p<0.01). PCP-treated rats also exhibited a lower breakpoint in the PRBP task [F(1,43)=9.72, p<0.01].

Conclusions: Disrupting NMDA receptor transmission during neurodevelopment led to impaired PRL performance resulting from a reduced ability to correctly utilize response-outcome associations; maladaptive performance resulted from reduced reward sensitivity (decreased target win-stay) and reduced punishment sensitivity (decreased non-target lose-shift). Impairments in reward sensitivity likely contributed the motivational deficits observed in PCP-treated rats. While alterations in punishment sensitivity are not evident in schizophrenia, these findings implicate disrupted NMDA receptor activity during neurodevelopment as a potential mechanism for the impairment in reward sensitivity and motivational abnormalities evident in these patients.

Keywords: Schizophrenia, Action-Outcome Associations, Phencyclidine, Motivation, Reward Sensitivity

Disclosure: Nothing to Disclose.

W179. Auditory and Visual P300 Event-Related Potential Abnormalities in Individuals at Clinical High Risk for Psychosis: Results From the North American Prodrome Longitudinal Study (NAPLS-2)

Holly Hamilton*, Peter Bachman, Aysenil Belger, Ricardo Carrion, Erica Duncan, Jason Johannesen, Gregory Light, Margaret Niznikiewicz, Brian Roach, Jean Addington, Carrie Bearden, Kristin Cadenhead, Barbara Cornblatt, Thomas McGlashan, Diana Perkins, Larry Seidman, Ming Tsuang, Elaine Walker, Scott Woods, Tyrone Cannon, Daniel Mathalon

University of California, San Francisco, California, United States

Background: Abnormalities in the P300 event-related potential (ERP) component reflect early attention-mediated information processing deficits and are a well-established biological marker of schizophrenia. Reductions in P300 amplitude elicited by auditory paradigms, in particular, have been more robustly and consistently observed relative to visual paradigms. We examined whether auditory and visual P300 abnormalities precede illness onset and are associated with future clinical outcomes in individuals at clinical high risk for psychosis (CHR).

Methods: CHR (n=520) and HC (n=228) participants completed baseline EEG recording during auditory and visual oddball tasks as part of the North American Prodrome Longitudinal Study (NAPLS-2). Two subcomponents of P300 were measured: P3b, elicited by infrequent target stimuli and reflecting top-down attention allocation, and P3a, elicited by infrequent non-target novel distractor stimuli and reflecting bottom-up orienting of attention. CHR participants who were followed longitudinally were categorized by clinical status after 24 months of study participation (n=275) and included a group who converted to full psychosis (CHR+C; n=66) and a group who did not convert (CHR-NC; n=209). Of these CHR-NC participants, some continued to meet CHR criteria and experience symptoms (CHR-NC-Symptomatic; n=124), while others obtained full symptom remission (CHR-NC-Remission; n=85). Group effects on baseline P300 amplitudes were examined using ANOVA models with a between-subjects factor of group (CHR, HC), conversion status (CHR+C, CHR-NC) or 24-month clinical status (CHR+C, CHR-NC-Symptomatic, CHR-NC-Remission, HC) and within-subjects factors of modality (auditory, visual) and stimulus (novel, target).

Results: A significant main effect of participant group was qualified by a group x modality interaction [F(1,746)=5.87, p =.016] indicating reduced auditory P300 amplitudes in the CHR group relative to HC [F(1,746)=20.74, p <.001], but no reduction of visual P300 amplitudes in the CHR group (p >.05). In addition, a main effect of conversion status in the absence of any interactions revealed reduced baseline P300 amplitudes in CHR+C relative to CHR-NC across both modalities and stimulus types [F(1,273)=4.33, p =.038]. Finally, examination of the effect of 24-month clinical status on baseline P300 amplitudes yielded a clinical status x modality x stimulus interaction [F(3,449)=2.90, p =.034]. In the auditory modality alone, a main effect of clinical status was qualified by a marginally significant clinical status x stimulus interaction [F(3,499)=2.30, p =.076]. For auditory target stimuli, both CHR+C and CHR-NC-Symptomatic had smaller baseline P3b amplitudes than both HC (ps <.001) and CHR-NC-Remission (p =.003 and p =.004, respectively), while target P3b amplitudes of CHR-NC-Remission did not differ from HC at baseline (p >.05). For novel sounds, CHR+C and CHR-NC-Symptomatic had smaller baseline P3a amplitudes than HC (p =.035 and p =.003, respectively) but did not differ from CHR-NC-Remission (ps >.05), generally reflecting the participant group effect of reduced amplitudes in CHR participants relative to HC. In the visual modality, a group x stimulus interaction was absent, although a main effect of group [F(3,499)=3.14, p =.025] confirmed reduced visual P300 amplitudes in CHR+C relative to all CHR-NC participant groups [CHR-NC-Symptomatic (p =.024), CHR-NC-Remission (p =.011), HC (p =.003)].

Conclusions: Results indicate that auditory and visual P300 amplitudes in CHR individuals are associated with future transition to psychosis. Moreover, results suggest that auditory target P3b, in particular, is sensitive to clinical outcomes including remission from the CHR state. Together, these results implicate P300 as a neurophysiological vulnerability marker for psychosis that may help to enhance the prediction of psychosis risk among CHR individuals.

Keywords: P300, Clinical High Risk, Psychosis

Disclosure: Nothing to Disclose.

W180. PolyIC Response is Dependent on Length: Implications for Preclinical Maternal Immune Activation Models

Melissa Bauman*, Robert Berman, Richard Maddock, Douglas Rowland, Judy Van de Water

University of California, Davis, Sacramento, California, United States

Background: A subset of women who are exposed to infection during pregnancy have an increased risk of giving birth to a child who will later be diagnosed with a neurodevelopmental or neuropsychiatric disorder, such as schizophrenia or autism. A number of factors, including genetic susceptibility, the intensity and timing of the infection, and exposure to additional aversive postnatal events, may influence the extent to which maternal infection alters fetal brain development and which disease phenotype is expressed. Preclinical research using animal models provides a test bed to systematically evaluate the effects of maternal infection on fetal brain development, determine the relevance to human brain disorders, and to evaluate novel preventative and therapeutic strategies. Maternal immune activation (MIA) models in mice, rats, and nonhuman primates suggest that the maternal immune response is the critical link between exposure to infection during pregnancy and subsequent changes in brain and behavioral development of offspring. However, differences in the type, severity, and timing of prenatal immune challenge paired with inconsistencies in behavioral phenotyping approaches have hindered the translation of preclinical results to human studies.

Methods: The Polyinosinic-polycytidylic acid (PolyIC) model, which mimics the acute phase of a viral infection, has become one of the most widely used preclinical MIA models. PolyIC consists of a chain of double stranded inosine (I) and cytidine (C), which can dramatically vary in chain length and therefore molecular weight. Recent studies have shown that in vitro, the ability of PolyIC to stimulate an immune response can vary greatly depending on molecular weight. The present in vivo studies were carried out to determine the relationship between molecular weight, immune system activation and sickness behavior. As an initial step in optimizing the rat MIA model, nulliparous female rats were injected with either saline, 4mg/kg low molecular weight PolyIC (LMW-PolyIC), or 4mg/kg high molecular weight PolyIC (HMW-PolyIC). Cytokine levels were quantified from blood samples collected at 3 and 6 hours post injection. Sickness behavior, including core temperature, activity and basal metabolism, were monitored for 24-hours post injection using the Comprehensive Lab Animal Monitoring System (CLAMS) metabolic cage system. The results of the cytokine responses and sickness behaviors of the nulliparous animals were then used to develop dosing guidelines for pregnant dams. Preliminary behavioral and neuroimaging studies of offspring born to dams that received either LMW- or HMW-PolyIC injections on gestational day 15 will also be presented.

Results: Here we demonstrate an acute increase in serum IL-6 following injection of both LMW- and HMW-PolyIC in nulliparous female rats at both 3hr and 6hr time points. However, the HMW PolyIC was associated with a more pronounced immune response as indexed by increased levels of GM-CSF, IL-4, IL-17, IL-2, IL-5, IL-12, and IL-13 in HMW-injected animals compared to saline-injected controls 3hr post injection and increased levels of IFNg, IL-1α, IL-1β, IL-10, and TNFα at both 3hr and 6hr post injection. LMW-injected animals only differed from controls for IL-1β at 3hr and 6hr, and TNFα at 3 hours. Only animals that received HMW-PolyIC demonstrated weight loss, reduced O2 consumption/CO2 production and reduced feeding and drinking behaviors compared to saline injected animals. Animals that received LMW-PolyIC did not differ from saline injected animals on measures of sickness behaviors.

Conclusions: PolyIC of varying molecular weights can induce significantly different cytokine responses in vivo. Given that methodological variability in the PolyIC model undoubtedly contributes to inconsistent results, this novel finding highlights the need to validate the reagents used, and to interpret results of previous studies while keeping these results in mind.

Keywords: Animal Model, Schizophrenia, Autism

Disclosure: Nothing to Disclose.

W181. Social Cognition Psychometric Evaluation (SCOPE): Results of the Final Validation Study

Philip Harvey*, Amy Pinkham, David Penn

Miller School of Medicine, University of Miami, Miami, Florida, United States

Background: Social cognition is increasingly recognized as an important treatment target in schizophrenia; however, the dearth of well-validated measures that are suitable for use in clinical trials remains a significant limitation. The Social Cognition Psychometric Evaluation (SCOPE) study addresses this need by systematically evaluating the psychometric properties of promising measures. In this final phase of SCOPE, eight new or modified tasks were evaluated.

Methods: Stable outpatients with schizophrenia (n=218) and healthy controls (n=154) completed the battery at baseline and 2-4 weeks later across three sites. Tasks included the Bell Lysaker Emotion Recognition Task (BLERT), Penn Emotion Recognition Task (ER-40), Reading the Mind in the Eyes Task (Eyes), The Awareness of Social Inferences Test (TASIT), Hinting Task, Mini Profile of Nonverbal Sensitivity (MiniPONS), Social Attribution Task – Multiple Choice (SAT-MC), and Intentionality Bias Task (IBT). BLERT and ER-40 modifications included response time and confidence ratings. Eyes was modified to include definitions of terms and TASIT to include response time. Hinting was scored with more stringent criteria. MiniPONS, SAT-MC, and IBT were new to this phase. Tasks were evaluated on: 1) test-retest reliability, 2) utility as a repeated measure, 3) relationship to functional outcome, 4) practicality and tolerability, 5) sensitivity to group differences, and 6) internal consistency.

Results: Hinting, BLERT, and ER-40 showed the strongest psychometric properties and are recommended for use in clinical trials. Eyes, TASIT, and IBT showed somewhat weaker psychometric properties and require further study. MiniPONS and SAT-MC showed poorer psychometric properties that suggest caution for their use in clinical trials. Interestingly, higher levels of confidence in accurate responses predicted worse social outcomes, implicating social meta-cognition as a important topic for later studies.

Conclusions: Most social cognition tests do not meet psychometric standards common to neurocognitive tests. Better tests will be required to improve the assessment of social cognition in treatment studies, including pharmacological treatment. Social-metacognition was the was the best predictor of social outcomes, better than performance on any social cognitive performance-based measure.

Keywords: Social Cognition, Schizophrenia, Meta-Cognition, Insight

Disclosure: Part 1: Boehringer-Ingelheim, Consultant, Allergan, Consultant, Lundbeck, Consultant, Otsuka Digital Health, Consultant, Sanofi, Consultant, Sunovion, Consultant, Takeda, Consultant, Part 4: Takeda, Grant.

W182. Interventions and Cognitive Outcomes in Youth at Risk of Psychosis: A Systematic Review and Meta-Analysis

Jean Addington*, Dan Devoe, Danijela Piskulic, Sylvia Romanowska, Parker Townes

University of Calgary, Calgary, Canada

Background: In youth at clinical high risk (CHR) of psychosis cognitive deficits have been observed across multiple domains of cognition. Currently, there are no reviews evaluating the impact of treatments on cognition in those at CHR. Therefore, we conducted a systematic review and meta-analysis to examine the impact of treatments on cognition in youth at CHR of psychosis.

Methods: The authors conducted database searches of Embase, CINAHL, PsycINFO, Medline, and EBM from 1951 to May 2017. Studies were selected if they included any intervention that reported changes in cognition in youth at CHR. Treatment comparisons were evaluated using both pairwise and paired meta-analyses. Due to the differences in cognitive measures effect sizes were calculated as Hedges g and reported as the standardized mean difference (SMD).

Results: Of 6,612 citations, 16 studies met our inclusion criteria, including a total of 514 CHR participants. Cognitive remediation (CR) was associated with a significant improvement in global cognition (SMD, 0.48; P<0.01), processing speed (SMD, 0.29; P<0.05), reasoning and problem solving (SMD, 0.21; P<0.05), working memory (SMD, 0.28; P<0.01), and visual learning and memory (SMD, 0.29; P<0.05) from baseline to follow-up. Other treatments (i.e. antipsychotics, glycine) had no pooled impact on cognition.

Conclusions: This systematic review and meta-analysis demonstrated small to moderate effect sizes for CR interventions and subsequent improvement of cognition from baseline to follow-up. However, CR studies were not significant when compared to computer games at improving cognitive outcomes.

Keywords: Cognition, Clinical High Risk, Systematic Review, Treatment, Psychosis

Disclosure: Nothing to Disclose.

W183. Correlation Between MMN and Functional Assessment and the Longitudinal Change of MMN in First Episode Psychosis

Xiaojing Li, Sophie Brickman, Polly Huynh, Dost Ongur, Mei-Hua Hall*

McLean Hospital, Belmont, Massachusetts, United States

Background: Mismatch negativity (MMN), an auditory event-related potential (ERP), is robustly impaired in chronically ill schizophrenia (SZ) compared to healthy controls. MMN amplitude reduction typically shows an asymmetry pattern (smaller in the left) in SZ. However, reports of MMN amplitude reduction in first-episode psychosis (FEP) have been inconsistent, and longitudinal studies of MMN in FEP are scarce. This study aims to explore: i) whether MMN amplitude and latency as well as left to right asymmetry to duration deviant tones were impaired in FEP patients at baseline; ii) whether MMN amplitude showed progressive reduction at 6-month follow-up; and iii) whether MMN components were correlated with functional assessment at baseline and/or at the follow-up.

Methods: Baseline MMN were collected from 31 FEP patients and 22 healthy controls during a passive auditory paradigm with duration deviant tones (1600 binaural tones; 1,000 Hz, 80db, 85% standard [25 ms], 15% deviant [50 ms]). MMN components were measured at Fz, Cz, T7 and T8 sites. Currently six-month follow-up data were available on eight FEP patients. All participants had SCID, Functioning Assessment Short Test (FAST), Global Functioning Scale (GFS): Social and Role, UCSD Performance-Based Skills Assessment-Brief Version (UPSA-B), and NEO Five Factor Inventory (NEO-FFI). ANOVA, student T test and partial correlations (controlling for age and sex) were used for statistical analysis.

Results: At baseline FEP patients exhibited shorter MMN latency (t=2.027, p=0.048) but normal amplitude compared to healthy controls. Control subjects showed a significant MMN amplitude left>right asymmetry (T7 vs. T8 site) (t= -2.29, p=0.03) but patients did not. Patients scored significantly more impaired on FAST (F= 37.54, p<0.001), GFS-social (F= 21.69, p<0.001), GFS-role (F= 52.36, p<0.001) than controls but group did not differ on UPSA-B performance. Patients also scored significantly higher on NEO-Neuroticism (F= 7.82, p=0.01) and NEO-Extroversion (F= 6.62, p=0.02) personality dimension than control subjects. At baseline, MMN latency in patients was significantly correlated with FAST total score (partial r= -0.59, p=0.03). At 6-month follow-up, MMN amplitude at Fz was significant decreased in patients (t=-3.320, p=0.013).

Conclusions: The present study corroborates previous results of correlation between impaired MMN and functional aspects of psychotic disorders and of diminished laterality in FEP patients. Our results also indicate a progressive decrease in MMN amplitude over time. We are continuing to follow-up patients longitudinally and to collect a larger sample cohort.

Keywords: First Episode Psychosis, Event-Related Potential, Longitudinal Study

Disclosure: Nothing to Disclose.

W184. Novel MB-Selective COMT Inhibitors Modulate Dopaminergic Activity and Improve Cognitive Function

Gregory Carr*, Spencer Byers, Gongliang Zhang, Cailian Li, Anna Kolobova, Daniel Akuma, Vinh Au, Ingrid Buchler, Michael DePasquale, Pablo de Leon, Glen Ernst, Yifang Huang, Michael Poslusney, Martha Kimos, Huijun Wei, Daniel Weinberger, James Barrow

Lieber Institute for Brain Development, Baltimore, Maryland, United States

Background: Cognitive impairment is a core feature of many psychiatric and neurological disorders. Currently available therapies do not significantly improve cognitive impairment. Dopaminergic neurotransmission in the prefrontal cortex is important for the regulation of cognitive function and therapeutic interventions that modulate dopaminergic function in the cortex could improve cognition. The enzyme catechol-O-methyltransferase (COMT) is a key regulator of dopaminergic tone in the cortex and hippocampus and individual variability in COMT function is associated with differences in cognitive function. Tolcapone, a COMT inhibitor, has been approved as a treatment for the motor symptoms of Parkinson's disease but it has multiple features that prevent its widespread use as a cognitive enhancer including poor pharmacokinetics (PK) that require 3X/day dosing in patients and a black box warning for potentially fatal liver toxicity. We have developed non-nitrocatechol, MB-COMT-selective COMT inhibitors with improved PK and safety profiles compared to tolcapone. Here we describe multiple novel compounds with promising preclinical in vivo profiles that may be suitable for clinical trials testing their cognitive enhancement capabilities.

Methods: To determine the in vivo activity of our novel compounds, we measured the concentrations of dopamine metabolites in cerebrospinal fluid (CSF) and compared the results to the known effects of tolcapone. Male Sprague Dawley rats were administered compounds (IP or PO) and either 1 or 4 hours later, they were anesthetized and placed into a stereotaxic frame. CSF was withdrawn. Following CSF withdrawal, blood was taken via cardiac puncture, rats were transcardially perfused with phosphate buffered saline, and then the brain was removed. Drug levels were measured in CSF, plasma, and brain tissue. The dopamine metabolites homovanillic acid (HVA) and 3,4-Dihydroxyphenylacetic acid (DOPAC), and 3-Methoxytyramine (3-MT) were measured in CSF. We also tested the ability of our COMT inhibitors to improve cognitive flexibility in the rat attentional set-shifting task (ASST). We used a modified version of the digging ASST. Rats were administered COMT inhibitors or vehicle before the first phase of the ASST.

Results: We synthesized multiple COMT inhibitors from two distinct chemical series with significant in vivo activity. Our most promising compounds significantly altered in vivo dopamine metabolism. Our compounds have slightly lower in vitro potency compared to tolcapone, but have much greater brain-penetrance and more favorable PK characteristics. Additionally, our lead compound improved cognitive flexibility by decreasing the number of trials required to complete the ASST extradimensional shift compared to vehicle treatment indicating it may be a viable clinical candidate.

Conclusions: The COMT inhibitor tolcapone has shown promise as a potential cognitive enhancer in short-duration studies across multiple clinical indications, but it has multiple compound-specific negative attributes that would hinder its utilization as a long-term treatment option. Our drug discovery efforts have resulted in the identification of a potent, brain-penetrant COMT inhibitor with similar in vivo effects as tolcapone. We are currently evaluating this compound in additional assays to determine its suitability for clinical studies.

Keywords: Dopamine, Cognitive Enhancement, Prefrontal Cortex, Parkinson's Disease, Mild Cognitive Impairment due to AD

Disclosure: Nothing to Disclose.

W185. Examination of Optically Cleared Postmortem Human Brain Samples for 3D Analysis of Molecular Networks

Eric Chang*, Lok Wan Yeung, Aaron Birnbaum, Anil Malhotra

Zucker Hillside Hospital, Feinstein Institute, Manhasset, New York, United States

Background: Traditional analysis of brain tissue for histopathology and diagnosis is typically restricted to two-dimensions (2D), with a reliance on samples of thin (< 50 microns) tissue sections for molecular characterization. This approach makes it difficult to assess brain components that have complex three-dimensional (3D) architectures, such as blood vasculature or white matter. To gain an improved understanding of the microarchitecture and organization of the human brain, it would be better to examine tissue in its native three-dimensional (3D) state. In recent years, tissue-clearing and immunolabeling techniques such as CLARITY and SWITCH have allowed researchers to examine entirely intact rodent brains for full examination of molecularly defined networks on a brain-wide scale while maintaining subcellular resolution. Here, we applied these clearing techniques to intact blocks of postmortem human brain tissue to enable 3D analysis of specific local circuits.

Methods: Blocks of postmortem human brain samples of varied thickness (5 mm–1.5 cm) were obtained from the North Shore University Hospital pathology department and the NIMH Human Brain Collection Core (HBCC). The blocks of tissue were sub-sectioned for CLARITY or SWITCH processing. Immunolabeling by primary antibodies (GFAP, neurofilament, MBP, iba-1, PV) was followed by secondary antibodies (AF488, AF546, AF647), and in some cases followed by dyes to label vasculature (lectin-488) and cell nuclei (DAPI). Cleared and immunolabeled brain samples were mounted in customized imaging chambers with RIMS media and 3D image stacks were acquired on an Olympus Fluoview FV1000MPE two-photon laser microscope. Analysis was performed using Imaris 8.0 and FIJI software.

Results: We were able to successfully clear large (up to 5000 mm3) blocks of human brain using a passive CLARITY protocol with a modified hydrogel (1.75% acrylamide, 0.01875% bis-acrylamide). Immunolabeled interneurons, astrocytes, microglia, blood vessels, axons, and cell nuclei could all be visualized within up to centimeter-thick blocks of tissue. Processing samples with SWITCH resulted in a mean 24.5% volume reduction (pre vs post: F=8.0) and deformation at the tissue edges, which may be attributed to the absence of a hydrogel mesh for structural support. While passive CLARITY of human blocks required significantly more time (t=11.2, P<0.0001) to achieve optical transparency than equivalent volumes of mouse brain (per ~500 mm3), we found that it was superior for preserving tissue morphology and cellular architecture compared to the SWITCH protocol. Nearest neighbor analyses in 3D using spots and surfaces modules within Imaris indicated that the distance distribution of GFAP-positive astrocytes in relation to lectin-labeled vasculature could be used as a histopathological marker for discriminating patient from control tissue for specific disease conditions (epilepsy, schizophrenia). Astrocyte-to-vessel distances within schlerotic epilepsy tissue averaged 18.1% lower than in control tissue (t=5.9, P<0.0005). Preliminary analyses of PV-positive interneurons and iba-1 spatial relationships suggest that microglia may localize to regions of interneuron loss in schlerotic tissue.

Conclusions: By applying tissue-clearing protocols and advanced tools to examine human brain tissue in 3D, we can begin to visualize molecularly defined circuit elements that may be unique to specific disorders and have eluded traditional 2D pathological diagnosis. These approaches are particularly useful for understanding how complex 3D networks (e.g. vasculature, white matter) relate to the cellular (neurons, astrocytes, microglia) constituents of brain tissue. While current technology limits human brain clearing to volumes of cubic centimeters, these local parcels of intact tissue can be analyzed serially to obtain a better global understanding of the human brain.

Keywords: CLARITY, Immunohistochemistry, Postmortem Brain Tissue, SWITCH, Neurovascular

Disclosure: Nothing to Disclose.

W186. Using Coordinate-Based Meta-Analyses to Explore Structural Imaging Genetics

Thomas Nickl-Jockschat*, Simon Eickhoff, Thomas Mühleisen, Claudia Eickhoff, Hildegard Janouschek

University of Iowa, Iowa City, Iowa, United States

Background: Imaging genetics has become a highly popular approach in the field of schizophrenia research. A frequently reported finding is that effects from common genetic variation are associated with a schizophrenia-related structural endophenotype. Genetic contributions to a structural endophenotype may be easier to delineate, when referring to biological rather than diagnostic criteria.

Methods: We used coordinate-based meta-analyses, namely the anatomical likelihood estimation (ALE) algorithm on 30 schizophrenia-related imaging genetics studies, representing 44 single-nucleotide polymorphisms at 26 gene loci investigated in 4,682 subjects variants. To test whether analyses based on biological information would improve the convergence of results, gene ontology (GO) terms were used to group the findings from the published studies.

Results: We did not find any significant results for the main contrast. However, our analysis enrolling studies on genotype X diagnosis-interaction yielded two clusters in the left temporal lobe and the medial orbitofrontal cortex. All other subanalyses did not yield any significant results. To gain insight into possible biological relationships between the genes implicated by these clusters, we mapped five of them to GO terms of the category "biological process" (AKT1, CNNM2, DISC1, DTNBP1, VAV3), then five to "cellular component" terms (AKT1, CNNM2, DISC1, DTNBP1, VAV3), and three to "molecular function" terms (AKT1, VAV3, ZNF804A). A subsequent cluster analysis identified representative, non-redundant subsets of semantically similar terms that aided a further interpretation.

Conclusions: We regard this approach as a new option to systematically explore the richness of the literature in imaging genetics.

Keywords: Imaging Genetics, Meta-Analysis, Schizophrenia, Schizophrenia Databases

Disclosure: Nothing to Disclose.

W187. White Matter Integrity and Effects of Antipsychotic Medications in Patients With Schizophrenia

Nina Kraguljac*, Thomas Anthony, Frank Skidmore, David White, Adrienne Lahti

University of Alabama at Birmingham, Birmingham, Alabama, United States

Background: A number of studies have reported decreased white matter integrity in patients with schizophrenia, but little is known about the relationship between white matter integrity and antipsychotic medications.

Methods: We enrolled 42 unmedicated patients (thirty were medication-naïve) with schizophrenia in a longitudinal trial with risperidone. Symptom severity was assessed with the Brief Psychiatric Rating Scale (BPRS). We obtained diffusion weighted images before medication was started, and after six weeks of treatment. Healthy controls matched 1:1 on age, gender, and parental socioeconomic status were also scanned twice six weeks apart. 30 diffusion sampling directions spanning the whole sphere were acquired twice and concatenated (in plane resolution 2.2mm, slice thickness 2.2mm, b-value 1000 s/mm2, 5 b0 images). After visual inspection of raw images, we used TORTOISE for correction of bulk motion, eddy currents and susceptibility artifacts using a single interpolation step in DIFF_PREP. For each dataset, the first B0 image was selected as the reference for registration. Prior to registration, diffusion weighted and structural images were upsampled at a factor two and smoothed with a Perona-Malik anisotropic edge favoring gradient based filter to compute the transformations from moving to fixed images. After computation of transformations, original images were used to create the registered images. Bspline correction was done with the T2 weighted image (approximated from a T1 image using AFNI's FATCAT). Diffusion and structural images were resampled to 1.5mm isotropic voxels. Gradient tables were rotated along with motion correction. To obtain a summary measure of motion, the root-mean-square (RMS) was calculated both for absolute (RMSabs) and relative (RMSrel) movement. Datasets with RMSabs of greater than the voxel edge length were excluded from further analysis. Tensors were computed with DIFF_CALC using a linear fitting algorithm. To spatially normalize diffusion images to the Illinois Institute of Technology atlas (IIT2) space, we implemented an optimized non-linear image registration using a modified version of 3dQwarp in AFNI. The warping optimization implements an iterative refinement, where an input image is repeatedly processed through an optimizer in smaller and smaller patches, incorporating convergence criteria at each patch level to better resolve artifacts, with a final patch size of 3x 5x 3 mm. To assess whole brain voxel-wise group differences and changes over time in scalar indices used AFNI's 3dttest++ (age, sex, and RMSrel as covariates) with clustsim, a bootstrapping method used to correct for multiple comparisons.

Results: Mean age of patients was 26.62 years, 62% of subjects were male. Of the 42 patients included here, 33 completed the study. BPRS total scores decreased significantly after six weeks of treatment, average risperidone dose at that time was 3.73+/-1.72mg. Fractional anisotropy (FA) was significantly decreased in the corpus callosum in unmedicated patients (n=40) compared to healthy controls (n=41). Longitudinal analyses showed no changes in FA in healthy controls over time, and no changes in patients after six weeks of treatment with risperidone.

Conclusions: With state of the art data-processing methods we only found a small area of decreased FA in our predominantly medication-naïve patients. This is consistent with prior reports of limited white matter alterations at disease onset that may progress with illness duration (4). Our data suggests that a short-term course of antipsychotic medication may not alter white matter microstructure, but studies with longer follow up durations will be important to determine long term effects of antipsychotic medications.

Keywords: White Matter, Schizophrenia, Antipsychotics

Disclosure: Nothing to Disclose.

W188. Evaluation of the Behavioural and Neurobiological Consequences of Maternal Immune Activation in Male and Female Rat Offspring: A Developmental Model for Schizophrenia

Jo Neill*, Victoria Fasolino, Joanna Oladipo, Ben Grayson, Daniela Cadinu, Maurizio Manca, Giovanni Podda, Michael Harte, Nagi Idris, Katie Murray, Michelle Edye, Eric Prinssen, Irene Knuesel

University of Manchester, Manchester, United Kingdom

Background: Maternal immune activation (mIA) by administration of the viral-mimetic polyriboinosinic-polyribocytidylic acid (poly-I:C) is a key model for neurodevelopmental disorders (NDDs) such as schizophrenia (see Knuesel et al, 2014. Nat Rev Neurol, 10: 643-660 for review). Our overall aim is to establish this model in rats. We have found that the most robust systemic inflammatory response to poly-I:C is produced by an acute dose of 10 mg/kg i.p. in rats of the Wistar strain. Our aim here is to investigate the behavioural and neurobiological consequences of mIA in male and female offspring at specific developmental time points. Our focus is cognitive performance and social behaviour as these aspects of schizophrenia remain an unmet clinical need. Robust animal models such as this are required for evaluation of the developmental progression of these symptoms and associated neuropathological changes.

Methods: Pregnant (10-weeks old) female Wistar rats were injected intraperitoneally (i.p) with poly I:C (Sigma, 10 mg/kg, n=6) or saline (0.9% NaCl, n=5) at gestational day (GD) 15. Changes in core body temperature and body weight were measured prior to treatment, 3h and 6h post-poly I:C. Changes in plasma IL-6 were measured by ELISA 3h after poly I:C. Offspring of both sexes were tested at adolescence (postnatal day-PND-35-41) in the open field (OF) and elevated plus maze tests (EVPM) for anxiety, for social interaction (SI), and novel object recognition (NOR) for short term recognition memory and again at young adulthood (PND 65-68) in SI, NOR and EVPM. Female adults only were tested between PND100-110 in the attentional set-shifting task (ASST) for executive function. Data are expressed as the mean±SEM (n=7-10) and analysed using ANOVA and /or Student's t-test using a nested statistical analysis design incorporating the maternal intervention.

Results: 10 mg/kg poly I:C at GD15 significantly increased IL-6 3h post-injection in the pregnant dams (P<0.05). No significant behavioural effects of poly I:C were observed in either sex at adolescence. However, when the same rats were tested at young adulthood, the vehicle treated female offspring spent significantly (P<0.01) more time exploring the novel compared to familiar object, whereas the poly I:C offspring explored both objects equally in NOR. In SI they showed reduced sniffing (p=0.056) and following (p=0.12) effects however that failed to achieve statistical significance. In the same cohort, but not the same animals, a significant impairment in the extra dimensional shift phase of the ASST was observed in female poly I:C offspring, with a significant increase in trials to criterion (P<0.001). Initial neurobiological analysis in these female animals reveals a reduction in parvalbumin as measured by immunohistochemistry in the prefrontal cortex at PND 130. No significant effects of poly-I:C were seen on EVPM and OF at any time point, in either sex. Furthermore, no significant effects were found in NOR or SI in the adult male poly I:C rats. Males were not tested in ASST or for PV deficits as they had no NOR or SI deficit. In two separate cohorts of pregnant Wistar rat dams (N=8 poly I:C and vehicle treated dams respectively), we have identified significant reductions in placenta weight of both male and female offspring following 10 mg/kg i.p. poly-I:C at GD15 when sacrificed on GD21 (P<0.05). This replicable finding is likely to be very important considering the recent GWAS study identifying an important influence of risk factor genes for schizophrenia on placentation (Ursini et al. 2017 doi: http://dx.doi.org/10.1101/147207).

Conclusions: We have established a robust model of mIA in Wistar rats and used this to identify the longitudinal development of behavioural changes. Poly I:C at GD15 induced an immune response in pregnant Wistar rat dams, at a dose of 10 mg/kg i.p, 3h post-dosing. Behavioural deficits were not detectable during adolescence, but social behaviour deficits seemed to be emerging by adulthood in females and a cognitive deficit phenotype was observed at adulthood in females. This was accompanied by a reduction in prefrontal cortical parvalbumin (PFC PV) suggesting that alterations in GABAergic interneurones contributes to the cognitive deficits observed. No behavioural effects were observed in males. The relationship between the maternal immune response, placenta pathology, anatomical, molecular and cognitive phenotype in the offspring of both sexes requires detailed analysis. Brain and behaviour disturbances induced by mIA are likely to start at the level of the placenta very early in development. This model provides a means to study the longitudinal time course of altered neurodevelopment of relevance of schizophrenia and other NDDs. Work to date has identified placenta and PFC PV as targets for further study.

Keywords: Developmental Neuroscience, Cognition, Schizophrenia

Disclosure: Part 1: GSK, Stock / Equity, Roche, Servier, Autifony, Advisory Board, GW Pharma, Grant, Lundbeck, Honoraria, LB Pharma, Grant, IRLabs, Grant, Boehringer Ingelheim, Grant, Autifony, Grant, Part 2: LB Pharma, IRLabs, BI, Autifony, Grant, GSK, Stock / Equity, Part 4: GW Pharma, Grant, IRLabs, Grant, LB Pharma, Grant, BI, Grant, Autifony, Grant.

W189. in vivo Dopamine D3 and D2 Receptor Occupancy Profile of Cariprazine Versus Aripiprazole: A PET Study

Ragy Girgis*, Anissa Abi-Dargham, Mark Slifstein, Laishun Chen, Antonia Periclou, Nika Adham, Willie Earley

Columbia University, New York, New York, United States

Background: D2 receptor affinity is thought to mediate efficacy in treatment of positive symptoms of schizophrenia; however, negative symptoms and cognitive deficits are generally unresolved. Brain regions believed to be associated with modulation of these domains, such as ventral striatum (VST; including nucleus accumbens and ventral parts of the caudate nucleus and putamen) and thalamus (THA), have relatively high expression of D3 receptors, representing a target for treatment of schizophrenia. Cariprazine is an orally active D3/D2 receptor partial agonist with preferential binding to D3 receptors. In previous studies, cariprazine displayed high in vitro binding affinity for D3 receptors, and high and balanced in vivo occupancy at both D3 and D2 receptors in rats and humans at respective pharmacologically and antipsychotic effective doses. This study characterized D3 and D2 receptor occupancy in various brain regions in humans, after multiple doses of cariprazine compared with aripiprazole, a D2/D3 receptor partial agonist, using positron emission tomographic (PET) imaging.

Methods: Healthy control subjects (N=16) received once daily oral cariprazine (0.5 or 1 mg/d; 4 subjects each) or aripiprazole (2 or 4 mg/d; 4 subjects each) for 14 days. Magnetic resonance imaging (MRI) scans were acquired at screening, and regions of interest (ROIs) were drawn individually for each subject. ROIs included pre- and post-commissural caudate (pre-CA and post-CA, respectively), pre- and post-commissural putamen (pre-PU and post-PU, respectively), VST, THA, globus pallidus (GP), and a midbrain region encompassing substantia nigra and ventral tegmental area (SN/VTA). PET imaging was performed at baseline and following treatment, with [11C]-(+)-PHNO, a D3 preferring D3/D2 receptor radiotracer, to assess D2 and D3 receptor occupancy. PET data were coregistered to each subject's MRI, with ROIs applied to the images, and analyzed using the simplified reference tissue model with cerebellum as reference tissue. Binding potential (BPND) was measured at baseline and post-treatment in each ROI. In each ROI for each subject, the percent change in BPND (ΔBPND) across conditions was computed and entered into a regression model to estimate separate contributions of D2 and D3 receptor occupancy (OccD2 and OccD3, respectively). Two methods of the regression model were implemented: (1) the fraction attributed to D3 receptor binding (fD3) was estimated for each region globally across subjects, with OccD2 and OccD3 estimated for each subject, using a nonlinear least squares estimation; and, (2) values of ΔBPND were fixed to literature values for ROIs and to estimated values from method 1 in the dorsal striatum subregions, with ordinary linear least squares applied to ΔBPND to obtain OccD2 and OccD3. OccD2 values were entered into a one-parameter equation to estimate ED50 for each treatment drug.

Results: Within the pre- and post-CA and pre- and post-PU, ΔBPND ranged between 37% and 66% for both doses of cariprazine and aripiprazole. In the VST, GP, THA, and SN/VST, ΔBPND for both doses of cariprazine ranged between 30% and 72%; however, ΔBPND for both doses of aripiprazole ranged between -11% (indicating increased BPND) and 15%. When fitted using method 1, both cariprazine and aripiprazole showed high D2 receptor occupancy (cariprazine OccD2: 0.5 mg/d=58±7%, 1 mg/d=63±5%; aripiprazole OccD2: 2 mg/d=62±10%, 4 mg/d=67±13%). Cariprazine showed high D3 receptor occupancy (OccD3: 0.5 mg/d=41±12%, 1 mg/d=64±4%), while aripiprazole showed negligible occupancy (OccD3: 2 mg/d=2±3%, 4 mg/d=7±12%). When fitted using method 2, cariprazine and aripiprazole still showed high D2 receptor occupancy (cariprazine OccD2: 0.5 mg/d=55±10%, 1 mg/d=60±4%; aripiprazole OccD2: 2 mg/d=53±4%, 4 mg/d=59±8%). Further, cariprazine showed high D3 receptor occupancy (OccD3: 0.5 mg/d=39±11%, 1 mg/d=69±6%), while aripiprazole showed negligible occupancy (OccD3: 2 mg/d=-14±2%, 4 mg/d=-13±28%). ED50 estimates for D2 receptor occupancy for cariprazine and aripiprazole were 0.44 mg and 1.43 mg, respectively, when fitted with method 1 and 0.52 mg and 2.17 mg, respectively, when fitted with method 2.

Conclusions: Observed PET data show that, in healthy subjects, cariprazine robustly occupies both D3 and D2 receptors and is moderately D3 receptor-preferring. D2 receptor occupancy of cariprazine was similar to aripiprazole across ROIs; the closest standard doses to ED50 were 0.5 mg for cariprazine and 2 mg for aripiprazole. D2 receptor occupancy was only slightly greater at the higher doses of 1 mg cariprazine and 4 mg aripiprazole compared to 0.5 mg cariprazine and 2 mg aripiprazole. Overall, doses of 0.5 mg/d cariprazine and 2 mg/d aripiprazole are predicted to lead to similar occupancy at D2 receptors in the range of 55% to 60% on average. D3 receptor occupancy by cariprazine increased with dose, from 41% at 0.5 mg/d to 64% at 1 mg/d, while D3 receptor occupancy with aripiprazole was negligible at both doses tested. These results are consistent with previous findings, which show that cariprazine occupies both D3 and D2 receptors in vivo and displays preferential D3 receptor binding in vitro. The information obtained here regarding the in vivo occupancy of D3 and D2 receptors by both drugs allows for designing trials to test the proposed added benefits of the D3 receptor-preferring profile of cariprazine in treating negative symptoms and cognitive impairment in patients with schizophrenia.

Keywords: Cariprazine, Occupancy, Positron Emission Tomography Imaging, Dopamine (D2, D3) Receptors, Aripiprazole

Disclosure: Part 4: Genentech, Grant, Otsuka, Grant, BioAdvantex, Grant, Allergan/Forest, Grant.

W190. Aberrations in the Synchronized Oscillatory Activity of Resting State Brain Networks in Schizophrenia and Bipolar Affective Disorder

Scott Sponheim*, Miseon Shim, Seung Suk Kang

Minneapolis VA Medical Center, Minneapolis, Minnesota, United States

Background: Although evidence indicates that the transmission of information through networks of the brain is abnormal in schizophrenia, the mechanisms underlying this dysfunction in neural communication are largely unknown. Recent evidence suggests that oscillatory activity is a means by which neuronal ensembles communicate within and across brain regions. Voltage oscillations of different frequencies appear to carry different functional significance with high frequencies reflecting information passing forward from low to high-level brain regions and lower frequencies reflecting communication from high-level brain regions. The relative timing of oscillatory cycles (i.e., synchronization) is a central determinant of whether neural ensembles effectively communicate information. To better understand aberrant communication within brain networks in schizophrenia we examined the synchrony of oscillations across brain regions in a large sample of individuals affected by severe mental illness, first-degree biological relatives of affected individuals, and healthy control participants.

Methods: Participants were enrolled in a family study of schizophrenia and consisted of 135 patients with schizophrenia, 97 with bipolar affective disorder, 124 biological relatives of individuals with schizophrenia, 75 relatives of individuals with bipolar affective disorder, and 151 control participants. Resting electroencephalography (EEG) data were acquired during eyes open and eyes closed resting states. Dynamics of neural oscillatory activity were operationalized through computing phase locking values for a broad range of EEG frequencies (1 to 55 Hz) for recordings across the scalp. Participants also completed diagnostic interviews, symptom rating assessments, measures of schizotypal personality factors, and cognitive testing.

Results: Principal component analysis (PCA) of phase locking values yielded four fundamental frequencies around which neural synchronization occurred (Delta/Theta, Alpha, Beta, Gamma). Separate spatial PCAs computed on each of the four frequency components revealed five resting state oscillatory networks for each frequency components. The oscillatory networks were highly similar across the frequency components suggesting the existence of neuroanatomical pathways shared by frequencies during resting state cycles. A diagnosis of schizophrenia was associated with increased interhemispheric frontal synchronization in the delta/theta frequency component, as well as increased anterior-posterior synchronization in the delta/theta, alpha, beta, and gamma frequency components. Bipolar affective disorder was associated with decreased synchronization in frontal-temporal networks in the delta/theta, alpha, and beta frequencies. As a group, first-degree biological relatives showed now synchronization abnormalities in resting state oscillations.

Conclusions: Evidence suggests that different oscillatory frequencies share neuroanatomical structures for communication within brain networks during a resting state. Schizophrenia is associated with increased low frequency synchronization between frontal brain regions. Bipolar affective disorder is associated with aberrant desynchronization in frontal-temporal networks. Thus, synchronization of oscillations during resting state may be a means by which to separate forms of severe psychopathology. There were no resting state synchronization abnormalities associated with being a first-degree biological relative of an individual affected by schizophrenia or bipolar affective disorder, and thus aberrant synchronization is related to the clinical expression of a severe mental disorder and not unexpressed genetic liability for the conditions.

Keywords: Resting State Intrinsic Connectivity, Schizophrenia, Bipolar I Disorder, EEG

Disclosure: Nothing to Disclose.

W191. Replication of Significant Improvement in Treatment Resistant Auditory Verbal Hallucinations After 5 Days of Double-Blind, Sham Controlled Inhibitory (Cathodal) tDCS: Evidence for fMRI Guided Treatment

Joshua Kantrowitz*, Guillermo Horga, Odeta Beggel, Michael Avissar, Gail Silipo, Daniel Javitt

Columbia University, New York, New York, United States

Background: Approximately 70% of schizophrenia (Sz) patients experience auditory verbal hallucinations (AVH) as a significant presenting symptom. Recent studies have shown that transcranial direct current stimulation (tDCS) can be highly effective for treating AVH. In tDCS, extremely low (<2 mA) currents are applied to the scalp from devices powered by 9V “transistor” type batteries. Despite the low currents involved, significant modulation of underlying brain function can be observed. Because direct currents are used, the polarity of stimulation significantly modulates the effect. Anodal stimulation applied over a specific brain region, induces an enhancement in underlying activity, while cathodal stimulation induces an inhibition. However, the mechanisms of how tDCS works are not understood. tDCS is known to induce long-lasting effects on glutamate-dependent plasticity that result in neural adaptations at the network level. From a predictive-coding perspective, such effects may enhance the formation of sensory predictions, also thought to rely on glutamate-dependent plasticity. We hypothesize that tDCS stimulation will enhance sensory predictions, thereby normalizing sensory predictive errors (PE) in the auditory cortex that have been shown to be disrupted in patients with AVH.

Methods: Participants were randomized in a 1:1 ratio to 20-minute treatments of inhibitory (cathodal) active vs. sham tDCS per day over 5 consecutive days, following the procedures established by Brunelin 2012. Criteria were age between 18 and 55 with a SCID diagnosis of Sz or schizoaffective disorder, right handed, mean Auditory Hallucination Rating Scale (AHRS) item score >2 (moderate), and on stable dose of antipsychotic medication for at least 1 month.

Results: The primary outcome was the AHRS assessed at baseline, following the final treatment, and at 1- and 3-month follow-up time-points. In a subset (n=15), a validated speech discrimination task based fMRI paradigm (Horga 2014) was administered, to evaluate the degree to which improvement of AVH is associated with normalization of AVH-related physiological abnormalities (PE).

Conclusions: The present report represents the largest study of tDCS for auditory hallucinations. We replicate previous reports and demonstrate significant improvement in AVH in schizophrenia patients’ treatment resistant to antipsychotics. Ongoing MRI analysis will inform future design.

Keywords: Auditory Hallucinations, Functional MRI (fMRI), Transcranial Direct Current Stimulation, Biomarker

Disclosure: Part 1: Alkermes, Grant, Merck, Grant, Lundbeck, Grant, Lilly, Grant, Forum, Grant.

W192. Auditory Steady-State Responses Assessment in New Animal Model of Fragile X Syndrome

Ai Okamura*, Takuma Mihara, Naoki Kozono, Sokichi Honda, Mickey Matsumoto

Astellas Pharma, Inc., Tsukuba-shi, Japan

Background: Fragile X syndrome (FXS) is a developmental disorder stemming from the dysfunction of single gene, fragile X mental retardation 1 (FMR1). It has been reported that the hyper-excitability, intellectual disability, autism-like symptoms, and increased incidence of epileptic seizures present in affected individuals may be caused by underlying alterations in neuronal synchronization. To address this, previous studies have used electroencephalogram (EEG)-based studies, in particular auditory event related potentials (ERPs), to evaluate patient populations and putative preclinical disease models. In addition to providing a counterpoint to electrocorticography (ECoG) studies in human patients, auditory steady state response (ASSR) which is one of ERPs has recently been identified as a possible translational tool in neurophysiological responses in schizophrenia and bipolar disorder. Because of this, it is possible that this approach may represent a new opportunity to accelerate clinical and preclinical development.

In the present study, we used locomotor activity along with EEG to examine ASSR in Fmr1 KO rats, with the dual goals of evaluating these animals as a translational model of FXS (with particular focus on abnormal oscillation bands) and progressing ASSR as a putative tool for establishing translational validity and accelerating preclinical development.

Methods: After recovery following implantation of surgical electrodes, Fmr1 KO rats and wild-type (WT) littermates were repeatedly examined for locomotor activity, ECoG sensory gating (SG) and ASSR measurements (10-80Hz). Locomotor activity was performed for 90 minutes in a novel environment under 300 lux using SM-024/w (Muromachi Kikai Co., Ltd) after one hour of test room acclimation. The data were categorized into three sections (0-30 min, 0-60 min, 0-90 min) and analyzed by CompACT AMSTM ver. 3.82. In SG paradigm, paired click sounds were produced by white noise generator. Both auditory ERPs were performed and recorded using a programing script on Spike2 with CED1401 system (CED). EEG data were analyzed using EEGLAB, a MATLAB (MathWorks) toolbox.

Results: Fmr1 KO rats showed hyper-locomotion, the most significant of which occurred within 30 minutes of introduction to the novel environment. In the SG paradigm, Fmr1 KO rats displayed a remarkable increase in S2/S1, suggesting these animals have sensory gating disruptions. Fmr1 KO rats also showed a significant decrease in overall theta power and increased overall gamma power when compared to WT littermates. In ASSR measurements, auditory click train stimuli in each frequency bands markedly induced augmented event-related spectral perturbations (ERSP) and synchronized inter trial coherence (ITC) in WT rats. It showed that ASSR was clearly observed. In contrast, Fmr1 KO rats showed disturbances in ERSP and ITC in gamma oscillation bands, but not in lower frequency oscillation bands.

Conclusions: Fmr1 KO rats recapitulate the hyperactivity, disruption of sensory gating and increased power of gamma frequency bands observed in FXS patients, suggesting they may serve as a useful disease model.

One of the most relevant clinical parameters of this animal model and FXS patients are the observed disruptions in gamma synchrony. These perturbations represent a fascinating shared feature, as both are readily measurable, comparable descriptors of functional state in both populations. As such, detection of network synchronization status by ASSR may be a promising translational tool for connecting animal model studies to patient disease conditions. Furthermore, it may be possible to extrapolate pathophysiological alterations from these data, as reduced evoked gamma band activity may be reflective of reduced inhibitory neuron activity and hyperexcitability in FXS patients.

In addition to providing a tool for linking disrupted gamma synchrony in Fmr1 KO model and FXS patients, ASSR-based measurements are a particularly efficient way to capture electrophysiological alterations in affected individuals when compared to other EEG-based approaches. This efficiency, combined with the translational validity of the approach, may allow for new therapeutic inroads to be tested and evaluated in this preclinical model, resulting in an acceleration in the advancement of drug development for patients suffering from FXS.

Keywords: Fragile X Syndrome, ASSR, Sensory Gating, Fmr1 KO Rat

Disclosure: Nothing to Disclose.

W193. Discovery of Non-Catechol Dopamine D1 Receptor Agonists With Impaired Beta-Arrestin Recruitment and Reduced Desensitization: Promising Novel Ligands for Psychiatry and Insights Into GPCR Functional Selectivity

John Allen*, David Gray, Scot Mente, Rebecca O’Connor, Patrick Tierney, George DeMarco, Dmitri Volfson, Edward Guilmette, Michelle Salafia, Rouba Kozak, Michael Ehlers

University of Texas Medical Branch, Galveston, Texas, United States

Background: Selective agonism of the dopamine D1 receptor (D1R) has been pursued for nearly 40 years as a therapeutic strategy for neurologic and psychiatric diseases due to the fundamental role of D1Rs in motor function, reward processing, and cognition. However, all known D1R-selective agonists are catechols, which are rapidly metabolized in vivo. Moreover, prolonged exposure to catechol agonists desensitizes the D1R, which reduces agonist response. As such, drug-like selective D1R agonists have remained elusive. Here we report a novel series of selective, potent non-catechol D1R agonists with excellent in vivo pharmacokinetics and prolonged agonism in vivo. These ligands stimulate adenylyl cyclase signaling, but exhibit distinct binding to the D1R orthosteric site that does not lead to desensitization or trigger recruitment of Beta-arrestin as observed with catechol ligands.

Methods: A high throughput screen measuring D1R-mediated cAMP production was performed followed by chemical optimization of lead molecules. The novel ligands were characterized for their orthosteric binding mode and receptor selectivity using a combination of radioligand binding, homology modeling and receptor mutagenesis. Desensitization of Gs-mediated cAMP signaling was assessed using primary rat striatal neurons while beta-arrestin recruitment to D1Rs was examined by imaging arrestin2-GFP using TIRF microscopy. To test for agonist-mediated efficacy and potential tachyphylaxis in vivo, novel non-catechol and catechol D1R agonists were assessed for their ability to increase eye-blink rates in non-human primates or rescue locomotor activity in 6-OHDA lesioned rats.

Results: Evaluation of nearly 3 million distinct compounds in a D1R functional screen yielded a single low affinity, non-catechol D1R agonist which was optimized through medicinal chemistry to create potent, highly selective, and orally bioavailable non-catechol agonists. Unlike catechol D1R agonists, non-catechol ligands did not trigger recruitment of beta-arrestin or receptor desensitization, despite strong activation of Gs/adenylyl cyclase. Structure-based modeling, receptor mutagenesis, and chemical biology probes further revealed that these unique agonist effects result from a novel binding interaction for non-catechols with key residues in extracellular loop 2. in vivo, unlike catechols, non-catechol agonists did not show tachyphylaxis of D1R-driven spontaneous eye blink rate in non-human primates. Furthermore, an exemplar non-catechol agonist also demonstrated continuous sustained locomotor activity with reduced desensitization after oral dosing over three days in the rat 6-OHDA unilateral lesion model of Parkinson's disease, compared to a matched catechol D1 agonist.

Conclusions: Taken together, these findings elucidate a unique orthosteric binding mode for selectively activating the D1 dopamine receptor, defines a molecular basis for ligand-specific recruitment of beta-arrestin and reveals structural underpinnings for agonist functional selectivity at a GPCR. In addition, discovery of highly selective non-catechol D1R agonists with favorable drug-like properties and limited desensitization provides therapeutic promise for treating diverse neurologic and psychiatric diseases involving reduced D1R/dopaminergic signaling (e.g. working memory deficits in schizophrenia; loss of motor function in Parkinson's disease).

Keywords: D1 Dopamine Receptors, Drug Discovery/Development, Functional Selectivity, Desensitization, β-arrestin

Disclosure: Part 1: Pfizer Global R&D, Patent, Part 3: Pfizer Global R&D, Employee.

W194. Genetic Isolation of Neurons in the Lateral Nucleus of the Cerebellum that Regulate Cognitive Behaviors

Erik Carlson*, Timothy Locke, Marta Soden, Samara Miller, Avery Hunker, Larry Zweifel

University of Washington, Seattle, Washington, United States

Background: Studies in humans and non-human primates have identified a region of the dentate nucleus of the cerebellum (DCN), or lateral nucleus in rodents (LCN), which is activated during performance of cognitive tasks involving complex spatial and sequential planning. Whether such a subdivision exists in the rodent is not known.

Methods: Utilizing viral and genetic strategies in mice, we examined cellular phenotypes of dopamine D1 receptor positive (D1R+) cells in the LCN with whole-cell patch clamp recordings, translational RNA profiling, and immunohistochemistry to examine D1R expression in the LCN of mice and DCN of humans. We used a chemogenetic strategy to inhibit D1R+ neurons, and examined behaviors including spatial navigation, social recognition memory, prepulse inhibition of the acoustic startle reflex, peak interval timing, and a working memory task to test the necessity of these neurons in these behaviors.

Results: We identified a population of neurons in the deep cerebellum expressing the D1R that are localized to an anatomically distinct region of the LCN. We also observed D1R+ neurons in the DCN of humans, suggesting an evolutionarily conserved population of dopamine-receptive neurons in this region. The genetic, electrophysiological, and anatomical profile of mouse D1R neurons is consistent with a heterogeneous population of GABAergic, and to a lesser extent glutamatergic cell types. Selective inhibition of D1R LCN neurons impairs spatial navigation memory, peak interval timing, working memory, social recognition memory, and prepulse inhibition of the acoustic startle reflex.

Conclusions: Collectively, these data demonstrate a functional link between genetically distinct neurons in the LCN and cognitive behaviors.

Keywords: Cerebellum, D1 Dopamine Receptors, Working Memory, Spatial Navigation, Timing

Disclosure: Nothing to Disclose.

W195. Deficits in Sleep Spindles and Slow Waves in Early-Course Psychosis

Fabio Ferrarelli*, Rachel Kaskie, Bianca Graziano

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: Sleep spindles and slow waves are the two main rhythms occurring during NREM sleep. Spindles are waxing and waning, 12-16 Hz oscillations, which are generated within the thalamus and then relayed and amplified into the cortex, whereas slow waves are 0.8-1.2 Hz, large amplitude oscillations that are initiated and propagated primarily within the cortex. Marked deficits in sleep spindles have recently been reported in patients with schizophrenia (SCZ) compared to both healthy individuals and psychiatric non-schizophrenia patients. However, it remains to be establish when these deficits first occur, if they are indeed specific for SCZ, and if they are restricted to spindles, or rather extend to slow waves at illness onset. After reviewing previous findings on sleep spindle deficits in SCZ, I will present new results of an ongoing study in first-break psychosis patients.

Methods: Whole night sleep high density (hd)-EEG recordings, were performed in first-break psychosis (N=18) and healthy comparison subjects (HC, N=14). NREM sleep power spectra, power topography, as well as several parameters of sleep spindles -amplitude, duration, density, and Integrated Spindle Activity (ISA) - and slow waves -density, amplitude, up- and down-slope- were calculated and compared across groups.

Results: There were no differences in NREM power, both in the spindle or slow wave ranges, between groups; however, first-break psychosis patients had reduced slow wave density. Furthermore, only those patients who were eventually diagnosed with SCZ (N=11), showed a deficit in ISA, and more prominently in spindle density, compared to HC.

Conclusions: Altogether, these findings indicate that reduced slow wave density is a shared feature of early course psychosis, whereas deficits in sleep spindles appear to be specific for SCZ, and point to a critical involvement of the thalamus in the neurobiology of this disorder.

Keywords: Sleep Spindles, First Episode Psychosis, Slow Waves, Thalamus

Disclosure: Nothing to Disclose.

W196. The Role of the Endocannabinoid System in Schizophrenia and its Therapeutic Correlates

Bernardo Dell'Osso*, Beatrice Benatti, Laura Cremaschi, MariaCarlotta Palazzo, Vincenzo Micale, Mariangela Pucci, Daniela Galimberti, Elio Scarpini, A. Carlo Altamura, Mauro Maccarrone, Claudio D'Addario

Università degli studi di Milano, Fondazione IRCCS Cà Granda, Ospedale Maggiore Policlinico, Milano, Italy

Background: In the recent years, the endocannabinoid system (ECS) has received increasing attention and has been linked to psychosis vulnerability. Epigenetics plays a role in the modulation of the expression of genes critical for neuronal development and for synaptic plasticity and several studies have already highlighted its importance in psychiatric disorders. Pharmacological treatment of psychosis can show only partial efficacy in some patients and the mechanism of action of antipsychotic compounds is still not completely understood. Based on this background, the aim of the present study was to investigate the transcriptional regulation of genes coding for ECS components in patients with mental disorders and to analyze eventual differences across antipsychotic treatments.

Methods: 131 subjects, including 97 patients (24 suffering from Major Depressive Disorder, MDD; 24 from Bipolar Disorder type I, BDI; 24 from Bipolar Disorder type II, BDII; 25 from Schizophrenia, SZ), on stable pharmacological treatment, and 34 age-matched, healthy controls were recruited, and clinical, pharmacological and socio-demographic data, along with blood samples, were collected. Peripheral blood mononuclear cells were isolated and mRNA levels of ECS components, as well as DNA methylation at CNR1 (the gene coding for the type-1 cannabinoid receptor) promoter were analyzed.

Results: A significant decrease in DNA methylation at the promoter of CNR1 was observed in schizophrenic patients with no changes in any other disorder. Consistently, an up-regulation of the expression of CNR-1 gene was found only in SZ patients. Most of the SZ subjects were on a stable oral treatment with multiple compounds including antipsychotics (whole sample), mood stabilizers (24%, including lithium and anticonvulsants) and, in minor proportions, antidepressants (4%) and benzodiazepines (16%). With respect to antipsychotic treatments, no differences were found in relation to the CNR 1 promoter methylation in patients with SZ.

Conclusions: SZ patients showed a significant decrease in DNA methylation at the promoter of CNR1 and an up-regulation of the expression of CNR-1 gene. However, no differences were found in terms of antipsychotic treatments in SZ patients as regards CNR 1 promoter methylation. Further studies with larger samples are needed to additionally investigate whether available antipsychotics can modify DNA methylation and other epigenetic patterns of the ECS genes in patients with SZ.

Keywords: Endocannabinoids, Schizophrenia, Epigenetics

Disclosure: Nothing to Disclose.

W197. Impaired Plasticity in the Dorsolateral Prefrontal Cortex of Patients With Schizophrenia

Reza Zomorrodi, Mera S. Barr, Daniel M. Blumberger, Benoit Mulsant, Zafiris Daskalakis, Tarek Rajji*

University of Toronto, Toronto, Canada

Background: Several lines of evidence suggest that schizophrenia (SZ) is associated with disrupted plasticity in the dorsolateral prefrontal cortex (DLPFC). The N100 evoked potential is proposed as a sensory gating mechanism that optimizes frontal lobes function and that is supported by frontal lobes networks. Abnormal N100 has also been associated with prefrontal cortical structure abnormalities in patients with schizophrenia. Combining a Transcranial Magnetic Stimulation (TMS)-based protocol - Paired Associative Stimulation (PAS) - with electroencephalography (EEG), we assessed DLPFC plasticity in vivo in adults with SZ and compared it to DLPFC plasticity in healthy adults. We then examined the relationship between N100 and DLPFC plasticity.

Methods: 33 SZ participants were compared to 36 health control (HC) participants. Using PAS-EEG, DLPFC plasticity was measured using potentiation of cortical evoked activity post-PAS compared to pre-PAS. Pre-PAS TMS-evoked N100 was also measured to determine its relationship to PAS response in SZ and HC participants.

Results: SZ participants had a Mean (SD) Age of 45.7 (16.8) years; 9/33 were Female; HC participants had a Mean (SD) Age of 43.4 (20.5) years; 16/36 were Female. Other than on Education ((SZ Mean (SD) Education=13.2 (1.6); HC Mean (SD) Education=15.8 (2.3), p<0.001), the two groups did not differ significantly on any demographic variable.

Selecting time of maximum potentiation post-PAS, HC participants experienced significant potentiation of cortical evoked activity (t(35)=11.2, p<0.001; 52% potentiation) but not SZ participants. Compared to HC participants, SZ participants had impaired potentiation (SZ Mean (SD) Potentiation=1.04 (0.19); HC Mean (SD) Potentiation=1.52 (0.28), p<0.001). Further, pre-PAS N100 amplitude was also reduced in SZ compared to HC participants (SZ Mean (SD) N100 Amplitude (μV)=3.7 (2.0); HC Mean (SD) N100 Amplitude (μV)=5.0 (2.5), p=0.032). Finally, pre-PAS N100 amplitude was positively associated with potentiation of cortical evoked activity across both groups and after controlling for age, gender, education, attention, cortical excitability, and group effects.

Conclusions: Our findings demonstrating impaired in vivo DLPFC plasticity in patients with schizophrenia. They also suggest that impaired N100 related mechanisms, e.g. sensory gating, underlie DLPFC plasticity impairment in schizophrenia.

Keywords: Schizophrenia, Neuroplasticity, TMS EEG

Disclosure: Nothing to Disclose.

W198. Decreased ion Channel Gene Expression Levels in the Frontal Cortex Associated With Schizophrenia Diagnosis Revealed by Latent RNA Quality Modeling

Andrew Jaffe*, Richard Straub, Joo Heon Shin, Ran Tao, Yuan Gao, Leonardo Collado-Torres, Tony Kam-Thong, Hualin Xi, Jie Quan, Qiang Chen, Carlo Colantuoni, Bill Ulrich, Brady Maher, Amy Deep-Soboslay, BrainSeq Consortium, Alan Cross, Nicholas Brandon, Jeffrey Leek, Thomas Hyde, Joel Kleinman, Daniel Weinberger

Lieber Institute for Brain Development, Baltimore, Maryland, United States

Background: Transcriptomic studies of human postmortem brain tissue can better identify molecular correlates of psychiatric disorders like schizophrenia. However, these studies involve extracting RNA from tissue that has been susceptible to a wide variety of antemortem and postmortem factors that are unrelated to the biology of schizophrenia.

Methods: We performed polyA+ RNA sequencing (RNA-seq) from DLPFC tissue on 155 schizophrenia cases and 196 controls from the Lieber Institute for Brain Development (LIBD) and similarly processed publicly available DLPFC RiboZero RNA-seq data from 159 patients and 172 controls from the CommonMind Consortium (CMC). We additionally performed RNA-seq on DLPFC tissue aliquots from five donors left off ice for 0, 15, 30, and 60 minutes using polyA+ and RiboZero library preparations to develop a statistical model that corrects for RNA degradation.

Results: We demonstrate differential expression analyses for schizophrenia are strongly biased by RNA quality differences between cases and controls, and that statistical adjustment using existing RNA quality measures largely fails to remove these effects. Using RNA-seq data from molecular degradation experiments of human primary tissues, we developed a method called quality surrogate variable analysis (qSVA) to estimate and remove the confounding effect of RNA quality in differential expression analysis. We show that this approach results in greatly improved replication rates (>3x) across the LIBD and CMC RNA-seq datasets, resulting in significant (at FDR<10%) and replicated differential expression signal in 237 genes. Gene ontology analysis of these genes implicated transporter- and channel-related signaling as significantly consistently downregulated in patients compared to controls.

Conclusions: Inefficient or disrupted signaling and tuning is thought to underlie the expression of illness in the adult brain, and the most successful therapeutics work through improving these processes. Consistent with this hypothesis, we find evidence for differences in the expression of genes coding for subunits of ion channels in the cortices of patients with schizophrenia compared to controls. We observed significant differential expression of both voltage-gated (KCNA1, KCNC3, KCNK1, KCNN1, SCN9A) and ligand gated ion channels (GRIN3A, GABRA5, GABRB3), transporters (SLC16A2, ALC25A33, SLC26A11, SLC35F2, SLC7A3), and ion channel auxiliary subunits (KCNIP3, SCN1B), supporting other evidence that the clinical phenomenology of schizophrenia is associated with altered neuronal excitability. While these findings implicating basic mechanisms of cortical circuit dynamics may underlie fundamental aspects of the clinical disorder, the possibility that they are driven by the effects of pharmacological treatment and are thus state dependent epiphenomena cannot be excluded.

Keywords: RNA Sequencing, Schizophrenia, Ion Channels, Postmortem Brain Tissue

Disclosure: Nothing to Disclose.

W199. Translational Development Strategies Utilized in the Development of an Inhibitor of PDE10A (TAK-063)

Thomas A. Macek, Kazunori Suzuki*, Karen Asin, Haruhide Kimura

Takeda Pharmaceutical Company, Ltd., Fujisawa, Japan

Background: TAK-063 is an inhibitor of phosphodiesterase 10A (PDE10A), which is highly expressed in the medium spiny neurons of the striatum. Although TAK-063 demonstrated antipsychotic-like effects in animal models, one of the challenges in the development of new treatments for schizophrenia is that the general relevance of these models to the clinical condition is questionable. In addition, compared with preclinical research using dopamine receptor antagonists or partial agonists, the meaningfulness of these preclinical models for novel targets (such as PDE10A) has not been established. The phase 1 clinical program for TAK-063 instituted a comprehensive translational development strategy in which the overarching objective was to demonstrate whether the antipsychotic-like pharmacodynamic effects observed in preclinical studies could be replicated in clinical studies. This strategy was used to determine whether further clinical development of TAK-063 for the treatment of schizophrenia was warranted and to select the dose for an initial Proof of Concept (POC) study. Although most preclinical and clinical results from this program have been presented elsewhere, the aim of this presentation is to illustrate the consistency and applicability of findings as translated from animal to human.

Methods: The main objectives of the program were to confirm target engagement and demonstrate pharmacodynamic effects in a translational model of schizophrenia and in people with schizophrenia. To evaluate these objectives, we developed a novel PET radiotracer for use in primates and humans. Four phase 1 studies were conducted, including a Single Rising Dose (SRD) study, a positron emission tomography (PET) study, a functional magnetic resonance imaging (fMRI) blood oxygen level-dependent (BOLD) study (ketamine-challenge), and a Multiple Rising Dose (MRD) study that included people with schizophrenia. In addition to PET and fMRI, assessments of cognition, EEG, prepulse inhibition, and other exploratory measures were included in the phase 1 program.

Results: Clinical results were generally consistent with preclinical studies. The main findings in humans included confirmation of target engagement in the PET study (as was observed in rats and monkeys), a replication in humans of the pharmacological effects of TAK-063 on fMRI BOLD (ketamine-challenge) observed in rats, and pharmacodynamic effects on EEG and cognition in subjects with schizophrenia, similar to the effects observed in rodents. These results and PK/PD modeling were utilized for dose selection in a small POC study of 20 mg TAK-063 vs. placebo.

Conclusions: The TAK-063 program used a variety of translational methods to replicate in humans the key findings in animal studies. Incorporation of this type of strategy is especially important in the development of antipsychotic agents for which the relevance of preclinical models of schizophrenia is questionable. The results of these studies, combined with PK/PD modeling, were utilized for dose selection in a POC study.

Keywords: Schizophrenia, Translational Medicine, New Drug Development, Antipsychotics, PDE10A

Disclosure: Part 1: Takeda Pharmaceutical Company Limited, Employee, Part 5: Takeda Pharmaceutical Company Limited, Employee.

W200. Multicenter Validation Study of Biomarkers for Glutamate-Targeted Drug Development in Psychotic Disorders: A Randomized Clinical Trial

Daniel Javitt*, Cameron Carter, John Krystal, Joshua Kantrowitz, Ragy Girgis, Lawrence Kegeles, John Ragland, Richard Maddock, Costin Tanase, Philip Corlett, Douglas Rothman, Graeme Mason, William Potter, Marlene Carlson, Melanie Wall, Tse Hwei Choo, Jack Grinband, Jeffrey Lieberman

Columbia University, New York, New York, United States

Background: Despite strong theoretical rationale and pre-clinical evidence, several glutamate-targeted treatments for schizophrenia have failed in recent pivotal trials, prompting questions as to target validity, compound inadequacy or lack of target engagement. A key limitation for glutamate-based treatment development is the lack of target-engagement biomarkers for translation between pre-clinical and early-stage clinical studies. Here, we evaluated the utility of three potential functional target engagement biomarkers - ketamine-evoked increases in functional magnetic imaging (fMRI) BOLD response (“pharmacoBOLD”); glutamate proton magnetic resonance spectroscopy (1H MRS); and task-based fMRI - for detecting ketamine-related alterations in brain glutamate. The objective was to identify one or more measure with sufficient effect-size and cross-site reliability to serve as glutamatergic target engagement biomarkers within early phase clinical studies.

Methods: This was a randomized clinical trial conducted between May 2014 and December 2016 as part of the NIMH-funded FAST-PS project. All raters were blinded to study group. The study was conducted in an academic research setting. Participants were healthy volunteer subjects age 18-55 or either gender and free of significant medical or psychiatric history were recruited from three sites. A total of 65 subjects were enrolled, of whom 59 had at least one valid scan. Subjects received either sequential ketamine (0.23 mg/kg bolus over 1-min followed by 0.58 mg/kg/hr over 30-min then 0.29 mg/kg/hr over 29-min) or placebo infusions There main outcome measures were ketamine-induced changes in phamacoBOLD, 1H MRS and task-based fMRI procedures, along with symptom ratings. Measures were pre-specified prior to data collection.

Results: In pharmacoBOLD, a highly robust (p<0.00001, d=5.4 SD units) increase in fMRI response was observed, with consistent response across sites. A smaller, but significant (p=0.039, d=0.64) signal was also observed in 1H MRS-determined levels of Glx (glutamate+glutamine) immediately following ketamine infusion. By contrast, no significant differences in task-activated fMRI responses were found between groups.

Conclusions: These findings demonstrate robust effects of ketamine on pharmacoBOLD across sites, supporting its utility for definitive assessment of functional target engagement. Other measures, while sensitive to ketamine effects, were not sufficiently robust for use as cross-site target engagement measures.

Keywords: Disorders of Glutamate, pharmacoBOLD, Functional MRI (fMRI), NMDA Antagonists, MR Spectroscopy

Disclosure: Part 1: Glytech, Stock / Equity, Amino Acid Solutions, Stock / Equity, NeuroRx, Stock / Equity, Takeda, Consultant, Lundbeck, Consultant, Autifony, Consultant, FORUM, Consultant, Pfizer, Honoraria, Part 2: Glytech, Stock / Equity, Part 3: Glytech, Stock / Equity.

W201. Stability of Behavioral and fMRI Measures of Cognitive Control Deficits in Early Psychosis: Evidence for Abnormal Neurodevelopment Rather Than Neurodegeneration

Jason Smucny*, Tyler Lesh, Tara Niendam, John Ragland, Cameron Carter

University of California, Davis, Sacramento, California, United States

Background: Although behavioral and neuronal deficits in cognitive control have been consistently demonstrated in schizophrenia (SZ), their longitudinal stability during the early course of psychosis has not been well-documented. Here we used functional magnetic resonance imaging (fMRI) during the AX-CPT, a cognitive control task, to examine the behavioral and functional stability of this cognitive process in SZ patients and control subjects.

Methods: 45 patients with SZ (37 M, 8 F, mean age 20.8 years) and 62 healthy control subjects (37 M, 25 F, mean age 19.9 years) underwent fMR scanning during the AX-CPT task at two time points separated by an average of 360 days. In addition to behavior (d-prime context), cognitive control-associated activity (high control "B">low control "A" Trial) was examined using an apriori bilateral spherical dorsolateral prefrontal cortex (DLPFC) ROI at both time points. To determine if deficits in SZ improved, were stable, or worsened over time, two complementary analyses were conducted. First, repeated measures ANOVAs were performed on behavioral and neuroimaging data with time as a within-subjects factor and diagnosis as a between-subjects factor. Second, intraclass correlation coefficients (ICCs) were calculated for each diagnostic group. Data were normalized and adjusted for age and gender prior to analysis.

Results: Main effects of diagnosis were observed for both d-prime context (F(1,105)=11.10, p=0.001) and task-associated DLPFC activity (F(1,105)=10.16, p=0.002). No effects of time or time*diagnosis interactions were observed for d-prime context (Ftime(1,105)=0.005, p=0.95); Ftime*Dx(1,105)=0.19, p=0.66) or task-associated DLPFC activity (Ftime(1,105)=0.003, p=0.96); Ftime*Dx(1,105)=0.13, p=0.72). ICC's for d-prime context were 0.67 (controls) and 0.59 (patients). ICC's for DLPFC activity were 0.52 (controls) and 0.54 (patients).

Conclusions: Behavioral and functional deficits in cognitive control in SZ patients were quite stable throughout the first year of early psychosis, with no evidence of either deterioration or recovery. No evidence was observed for differential stability of behavioral and functional measures of cognitive control between SZ patients and healthy controls. These findings suggest that functional cognitive control deficits in SZ are non-degenerative in the early stage of the illness. The psychometric analyses also provide additional evidence to support the use of behavioral and fMRI measures related to cognitive control as an endophenotypic biomarker of the illness.

Keywords: Schizophrenia, Cognitive Control, fMRI, Longitudinal Imaging, Dorsolateral Prefrontal Cortex

Disclosure: Nothing to Disclose.

W202. New Insight in White Matter and Core Cognitive Deficits in Schizophrenia

Peter Kochunov*, Elliot Hong

Maryland Psychiatric Research Center, University of Maryland School of Medicine, Ellicott City, Maryland, United States

Background: Disconnections of cortical networks may underlie various cognitive deficits that take severe clinical tolls on patients with schizophrenia. Historically, neuropsychopharmacology of cognitive deficits is mostly conceptualized and studied in terms of neurons, neurotransmitters and synaptic receptors. We hypothesized that the dynamics of the extended lifetime development trajectory of the brain's white matter, and the consistency of connectivity deficits in schizophrenia, posit white matter as the key loci responsible for these cognitive deficits [1]. Using novel diffusion weighted imaging (DWI) techniques and a milestone development of identifying key white matter tracks most relevant to schizophrenia, we are now able to show that specific white matter pathways responsible for shared vs. unique contributions to some of the key cognitive deficits in schizophrenia.

Methods: First, an advance DWI protocol consisted of 16 b-shells (b=0 to 3500 s/mm2) and 32 diffusion directions per shell were applied in 74 patients, 113 healthy controls, and 41 non-ill siblings of the patients [2]. This protocol was prescribed to capture the integrity of commissural fibers in the corpus callosum. In the second study in the ENIGMA meta-analysis, each major white matter tract was ranked by their vulnerability to schizophrenia using the largest diffusion tensor imaging (DTI) sample in the world (n= 4375) assembled from 33 cohorts [3]. In the third study, a large and independent (n=388) cohort of schizophrenia patients (n=171) and controls (n=217) were assessed on DTI of each major white matter tracts and key cognitive deficits in schizophrenia: processing speed and working memory [4]. We examined possible neuroanatomical sources of two of the the primary cognitive deficit in schizophrenia – reduced processing speed of information processing and working memory.

Results: Relative to matched controls, schizophrenia patients showed significantly reduced processing speed (p=8.7•10-7 -10-25), reduced working memory (p=10-15), and significant microstructural abnormalities in the brain's white matter, assessed using DTI fractional anisotropy (FA) (p=5.5•10-5-10-40). Kurtosis Anisotropy (KA) (p=2.2•10-4) and permeability diffusivity index (PDI) (p=2.6•10-4) were also significantly reduced in patients compared with controls. We observed that diffusion parameters that showed significant patient-control differences (FA, KA and PDI) at the genu of the corpus callosum white matter explained over 90% of the variance of the patient-control differences in processing speed (p<0.001)[2]. This association was also found for the non-ill siblings of the patients [2]. Next, in the ENIGMA mega-analysis revealed that schizophrenia differentially impacts different white matter tracks, with the largest impacts seen in the genu which connects interhemispheric frontal cortex and the anterior corona radiata that connects the ipsilateral frontal areas [3]. Finally, we performed mediation analyses that showed a significant white matter->processing speed->working memory relationship pathway, particularly in white matter regions such as the genu and the anterior corona radiata. Importantly, the strength of this brain-to-cognition pathway in different white matter tracts was strongly associated with the level of vulnerability to schizophrenia in the corresponding white matter tracts as determined by the largest independent meta-analysis in schizophrenia (r=0.85-0.94, p<0.001), supporting a strong relationship between white matter anatomic specificity for cognition and white matter anatomic specificity for schizophrenia [4].

Conclusions: Our findings show that advanced diffusion MRI of the white matter may capture microstructural connectivity patterns and mechanisms that govern the association between core neurocognitive deficits and schizophrenia. White matter impairment in schizophrenia is regional and tract specific, particularly in tracts normally supporting processing speed performance. Our hope is that these findings may elevate research interest in the white matter directed pharmacological interventions for cognitive enhancement.


1. Kochunov, P. and L.E. Hong, Neurodevelopmental and neurodegenerative models of schizophrenia: white matter at the center stage. Schizophr Bull, 2014. 40(4): p. 721-8.

2.Kochunov, P., et al., Diffusion-weighted imaging uncovers likely sources of processing-speed deficits in schizophrenia. Proc Natl Acad Sci U S A, 2016. 113(47): p. 13504-13509.

3. Kelly, S., et al., Widespread white matter microstructural differences in schizophrenia across 4,375 individuals: results from the ENIGMA Schizophrenia DTI Working Group. Mol Psychiatry, 2017. In Press.

4. Kochunov, P., et al., White Matter and Core Cognitive Deficits in Schizophrenia. JAMA Psychiatry, 2017. In Press.

Keywords: DTI, White Matter Integrity, Cognition, Working Memory

Disclosure: Nothing to Disclose.

W203. Nicotine-Induced Activation of Caudate and Anterior Cingulate Cortex in Response to Errors in Schizophrenia

Lauren Moran*, Luke Stoeckel, Kristina Wang, Carolyn Caine, Vanessa Calderon, Rosemond Villafuerte, Justin Baker, Amy Janes, Dost Ongur, Anne Eden Evins, Diego Pizzagalli

Harvard Medical School, McLean Hospital, Belmont, Massachusetts, United States

Background: Nicotine improves attention and processing speed in individuals with schizophrenia. Few studies have investigated the effects of nicotine on cognitive control. Prior functional magnetic resonance imaging (fMRI) research demonstrates blunted activation of dorsal anterior cingulate cortex (dACC) and rostral anterior cingulate cortex (rACC) in response to error and decreased post-error slowing in schizophrenia. To our knowledge, there are no studies examining the effects of acute nicotine on error processing and performance adjustments following errors in healthy individuals or those with schizophrenia. We hypothesized that nicotine-related activation of ACC and associated regions may be associated with enhanced error processing in schizophrenia.

Methods: Participants with schizophrenia (n=13) and healthy controls (n=12) participated in a randomized, placebo-controlled, crossover study of the effects of transdermal nicotine on cognitive control. Participants received transdermal nicotine or placebo in two sessions approximately a week apart while undergoing functional magnetic resonance imaging (fMRI). For each drug condition, participants performed the Stop Signal Task where subjects attempt to inhibit pre-potent responses to “Go (motor activation)” signals when an occasional “Stop (motor inhibition)” signal appears. Neural regions involved in error processing were evaluated by comparing “Stop Error” trials (failed response inhibition) to “Go” trials. In addition, participants underwent resting state fMRI prior to the task to evaluate functional connectivity within error-related networks.

Results: Stop Signal Task – Behavioral Data: We replicated prior findings of decreased post-error slowing in schizophrenia and found that nicotine was associated with more adaptive (i.e., increased) post-error reaction time (RT). Stop Signal Task - fMRI Data: Participants with schizophrenia had increased nicotine-induced activation of right caudate in response to errors compared to controls (DRUG x GROUP effect: pcorrected<0.05). Both groups had significant nicotine-induced activation of dACC and rACC in response to errors. Resting State fMRI Data: Using right caudate activation to errors as a seed for resting state functional connectivity analysis, relative to controls, participants with schizophrenia had significantly decreased connectivity between right caudate and dACC/bilateral dorsolateral prefrontal cortices (DLPFC). fMRI Correlations: We found a significant positive correlation between post-error RT and dACC ROI activation for nicotine (r=0.45, p=0.04), replicating previous findings. The correlation between dACC and post-error RT for the nicotine condition was significantly greater than for the placebo condition (z=2.57, p=0.01). We also found significant negative correlations between right caudate activation to errors and resting state connectivity between the right caudate and both the right (r=-0.50, p=0.01) and left dorsolateral prefrontal cortex (DLPFC; r=-0.45, p=0.03) and a trend for connectivity between right caudate and dACC (r=-0.34, p=0.09) for the nicotine condition but no significant correlations for the placebo condition (right DLPFC: r= 0.26, p=0.21; left DLPFC: r=-0.07, p=0.73; dACC: r=-0.11, p=0.61). This constellation of findings suggests that nicotine-induced activation of the right caudate in response to errors may overcome deficits in functional connectivity in schizophrenia within a network of regions associated with cognitive control (right caudate, dACC, DLPFC); greater nicotine-induced caudate activation was found in patients with the greatest decrements in connectivity.

Conclusions: In this study, we report the novel findings that a) nicotine administration is associated with increased activation of the right caudate in response to errors in participants with schizophrenia relative to controls; b) nicotine administration is associated with increased activation of both rACC and dACC in response to errors in participants with and without schizophrenia. We also replicated prior research demonstrating decreased post-error RT in schizophrenia and demonstrate the novel finding that nicotine administration is associated with increased post-error RT. This proof-of-concept pilot study suggests a role for nicotinic agents in targeting cognitive control deficits in schizophrenia.

Keywords: Schizophrenia, Nicotine, Caudate, Anterior Cingulate, Errors

Disclosure: Nothing to Disclose.

W204. 20-Year Nationwide Follow-Up Study on Antipsychotic Treatment, Re-Hospitalization, and Mortality in First-Episode Schizophrenia (FIN20 Study)

Jari Tiihonen*, Antti Tanskanen, Heidi Taipale

Karolinska Institutet, Stockholm, Sweden

Background: It is generally believed that 15–20% of patients with schizophrenia do not need antipsychotic treatment after their first episode, because they do not relapse without antipsychotic. Many treatment guidelines recommend that after stabilization, antipsychotic treatment should be continued for 1–5 years, and longer exposure should be avoided if possible. However, there is no published evidence to substantiate any of these views.

Methods: We used prospectively gathered nationwide databases to study the risk of re-hospitalization and death after discontinuation of antipsychotic treatment with multivariate Cox regression among all patients without previous use of antipsychotics who were hospitalized for the first time with schizophrenia diagnosis during 1996–2014 in Finland (N=8,738).

Results: In 10-year follow-up, the lowest risk of re-hospitalization or death was observed for patients using antipsychotic continuously (HR 1.00), followed by patients discontinuing the medication within 1 year (1.88, 95% CI 1.57–2.24), within 1–2 years (2.12, 1.43–3.14), within 2–5 years (3.26, 2.07–5.13), and after 5 years (7.28, 2.78–19.05) (p<0.01 for trend, Spearman correlation 2-tailed). During 20-year follow-up among patients without re-hospitalization, the risk of death was 214% higher (HR 3.14, 1.29–7.68) among those patients having not used antipsychotic compared with those with continuous use.

Conclusions: The results indicate that in first-episode schizophrenia, the longer the duration of antipsychotic treatment has been, the higher the relative risk of treatment failure becomes after discontinuation of medication, possibly due to supersensitivity of D2 receptors. Regardless of underlying mechanisms, these results provide definite evidence that, contrary to the general belief, relapse risk does not decrease as function of time during the first 10 years of illness. The risk of death is substantially higher among relapse-free patients having not used antipsychotic (i.e., “patients who do not need antipsychotic”) compared with those having used antipsychotic continuously up to 20 years. Antipsychotic treatment should be continued at least for 10 years after the first episode to prevent re-hospitalization and death.

Keywords: Pharmacology, Antipsychotic, Relapse and Hospitalization, First Episode Schizophrenia

Disclosure: Part 1: The Finnish Medicines Agency (Fimea), Consultant, Eli Lilly, Consultant, Honoraria, Board Member, F. Hoffman-La Roche, Consultant, Janssen-Cilag, Consultant, Honoraria, Board Member, Lundbeck, Consultant, Honoraria, Otsuka, Honoraria, Board Member. Part 2: Eli Lilly, Honoraria, Janssen-Cilag, Honoraria, Part 4: Stanley Foundation, Grant, Sigrid Jusélius Foundation, Grant, Eli Lilly, Grant, Janssen-Cilag, Grant.

W205. Blood Folate Levels Predict Ventrolateral Prefrontal Thickness, Activation, and Underlying White Matter Microstructure

Anais Rodriguez-Thompson, Kevin Dowling, Franklin Huntington, Hamdi Eryilmaz, Jordan Smoller, Joshua Roffman*

Massachusetts General Hospital, Charlestown, Massachusetts, United States

Background: Folate is a B-vitamin that supports methylation-dependent processes in the brain, including gene expression and neurotransmitter regulation. Reduced blood folate has been linked to several psychiatric disorders including schizophrenia. Further, several common, low-functioning genetic variants in the one-carbon pathway augment risk for negative symptoms and related cognitive impairment. Three recent clinical trials of folate supplementation in schizophrenia have demonstrated modest benefits for negative symptoms, although the neural mechanisms underlying these effects remain uncertain. In this study of healthy young adults, we investigated basic relationships among blood folate, its key underlying dietary and genetic determinants, and MRI phenotypes salient to psychiatric illness.

Methods: Healthy young adult volunteers (n=152), age 17-35, provided blood for serum folate levels and underwent standardized 3T MRI scans (Siemens Skyra) for measures of cortical thickness (sMRI), working memory-related activation (fMRI), and fractional anisotropy (DTI) within a single study visit. fMRI was conducted during the Sternberg Item Recognition Paradigm (SIRP), which assessed working memory across multiple loads (1, 3, 5, or 7 consonants). Scans with excessive motion were removed, as were SIRP runs with below-chance performance. MRI analyses were conducted with Freesurfer and covaried by age and sex. Surface-wide regression of cortical thickness on serum folate was used to identify significant clusters, with extent defined by voxelwise p<.001 and clusterwise p<.05 after 10,000 Monte Carlo simulations. Subsequent analysis of SIRP activation (all loads versus fixation) and its relationship to blood folate level were assessed using both ROI-based (i.e., significant clusters from the folate-thickness analysis) and exploratory whole-brain approaches. DTI images were acquired with diffusion weighting along 64 directions (b=1,000 s/m2) and processed with Tracula to reconstruct white matter pathways for fractional anisotropy (FA) measurements. Subjects also completed detailed questionnaires regarding natural and synthetic folate intake (DHQ-2), and were genotyped for three common, functional variants in the folate pathway (MTHFR rs1801133, MTR rs1805087, FOLH1 rs202676).

Results: Serum folate levels directly predicted cortical thickness in four regions: bilateral ventrolateral prefrontal cortex (vmPFC), right middle temporal gyrus (MTG), and right post-central gyrus (PoCG) (p<.05, whole brain corrected). Each of these regions demonstrated significant (p<.01) working memory load-dependent activation (bilateral vmPFC) or deactivation (MTG, PoCG). Within the right vlPFC, which engaged primarily at higher working memory loads, serum folate level also predicted activation during SIRP performance (p<.05, corrected for 4 regions). DTI analysis revealed a specific correlation between serum folate and FA within the right uncinate fasciculus (p<.05, corrected for 18 tracts), which directly underlies vlPFC and connects it to medial temporal lobe regions. While folate intake correlated significantly with serum folate level, surprisingly variation in neither folate intake nor folate-related genotypes influenced cortical thickness, activation, or FA measures.

Conclusions: Blood folate level associates with a confluence of structural and functional markers within vlPFC, a region associated with impaired higher-order cognition in psychosis. Folate may strengthen cortico-limbic circuits that facilitate working memory under more challenging conditions. However, in contrast to previous studies of schizophrenia, prefrontal structure and function among healthy young adults appear insensitive to important genetic and environmental modifiers of folate status. Improved “buffering” of one-carbon metabolism through folate supplementation may potentially help restore vlPFC structure and function in schizophrenia, and ameliorate related symptoms.

Keywords: Folate, Ventrolateral Prefrontal Cortex, Working Memory, Cortical Thickness, Schizophrenia

Disclosure: Part 1: Pamlab, Consultant.

W206. Effects of Amphetamine on Sensorimotor Gating and Neurocognition in Antipsychotic-Medicated Schizophrenia Patients

Neal Swerdlow*, Savita Bhakta, Jo Talledo, Daniel Franz, Brinda Rana, Gregory Light

University of California, San Diego, La Jolla, California, United States

Background: Drug effects on laboratory measures of brain function provide evidence for target engagement that can be used in an experimental medicine strategy to identify novel therapeutics. Sensorimotor gating of startle, measured by prepulse inhibition (PPI), is regulated by forebrain circuitry, and is impaired in specific brain disorders, including schizophrenia (SZ); it is also sensitive to acute drug effects, can provide evidence for target engagement, and is a stable, reliable, quantitative measure suitable for repeated testing designs. We recently reported that amphetamine (AMPH: 10 mg, po) enhanced simple auditory learning in both healthy subjects (HS) and antipsychotic (AP)-medicated SZ patients, using a frequency modulation task from a targeted cognitive training (TCT) suite for SZ; enhanced TCT learning among SZ patients after acute AMPH was retained for at least one week. We now report the effects of AMPH on PPI and MATRICS Comprehensive Cognitive Battery (MCCB) performance in an extended sample of the HS and AP-medicated SZ patients tested in the TCT suite. Past findings suggest that AMPH elevates PPI in subjects with low PPI baselines, and elevates MCCB scores in subjects with low baseline MCCB performance; thus, we predicted that AMPH would elevate PPI and MCCB performance in AP-medicated SZ patients.

Methods: Subjects (HS=44; SZ=38) were carefully screened. Patients were on stable AP doses for>1 month; all were taking atypical APs. Three laboratory sessions (screen day+2 test days) were each separated by 1 week. PPI and MCCB were assessed at each visit. Test days began with administration of placebo or AMPH (10 mg po) in a double-blind, cross-over design; PPI, MCCB, autonomic and subjective measures followed, along with measures of TCT learning, reported previously. The rs4680 polymorphism of catechol-O-methyltransferase gene was determined for all subjects. Parallel PPI testing in Long Evans rats was used to assess potential interactions of AMPH with AP medications.

Results: All testing was well tolerated; bioactivity was confirmed via autonomic and subjective changes, both of which were blunted in patients vs. HS. Startle magnitude and habituation did not differ based on diagnosis or drug. After placebo, SZ patients exhibited PPI deficits with 60 ms prepulse intervals (diagnosis x interval: p<0.004; effect of diagnosis at 60 ms: p=0.05, d=0.45); these deficits were absent in patients after AMPH (F<1). The magnitude of AMPH-enhanced PPI was significantly greater in patients than in HS (p<0.032), and was associated with positive symptoms (r=0.52; p<0.007), AP load (r=0.47; p<0.015) and amphetamine-increased "happy" ratings (r=-0.59, p<0.003) but not other demographic or clinical variables. There was a significant main effect of genotype (F=3.50, df 2,56, p<0.04), independent of diagnosis. PPI AMPH sensitivity was greatest among “AA” individuals (p<0.006) and was greater among individuals carrying at least one methionine allele vs. “GG” subjects (p<0.008). Rodent PPI testing demonstrated dose-dependent interactions of AMPH with haloperidol, quetiapine and the D1 antagonist, SCH 23390, resulting in both increases (haloperidol, quetiapine) and decreases (SCH 23390) in PPI, along with robust changes in startle magnitude. Patients exhibited significant MCCB deficits; no significant main effects of AMPH on MCCB scores were noted in either patients or HS, even after correcting for significant order (practice) effects. The MCCB did not detect pro-attentional effects of AMPH; nonetheless, AMPH-induced attentional changes correlated significantly (p<0.05) with AMPH-induced gains in TCT learning.

Conclusions: AMPH acutely normalized PPI in AP-medicated SZ patients; despite evidence of enhanced PPI and TCT learning in these patients, no concurrent neurocognitive changes were detected by the MCCB. Thus, in the context of stable AP medication, a low dose of AMPH enhances brain processes associated with higher function in SZ patients, without accompanying changes in MCCB performance. Both rs4680 and AP medications may be important moderators of AMPH effects on PPI. These findings do not support the clinical use of AMPH to treat neurocognitive deficits in SZ patients, but do provide evidence that: 1) substantial neuroplasticity exists in brain mechanisms regulating sensorimotor gating in SZ patients; 2) pro-attentional drugs can acutely normalize deficient brain mechanisms (sensorimotor gating) and enhance learning in AP-medicated SZ patients, and 3) MCCB may not be sensitive at detecting such acute pro-attentional drug effects.

Keywords: Schizophrenia, Amphetamine, Prepulse Inhibition, Neurocognition

Disclosure: Nothing to Disclose.

W207. A Potential Pathogenic Role of Hyperphosphorylated MAP2 in Schizophrenia

Megan Garver, Matthew MacDonald, Xiaolin Sun, Melanie Grubisha, Robert Sweet*

University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States

Background: Reduced neuron somal volume, dendrite length, complexity, and spine density have been reported in multiple brain regions, including primary auditory cortex, in schizophrenia. The structural features of dendrites are critical for signal processing. Not surprisingly, then, sensory processing within auditory cortex is impaired in schizophrenia and contribute to disability. We previously reported that the critical dendritic cytoskeletal protein, microtubule-associated protein 2 (MAP2), had profoundly reduced immunoreactivity in schizophrenia that was not due to loss of protein itself or of the neurons containing it. We hypothesized instead that MAP2 phosphorylation, known to regulate its immunoreactivity, is altered in schizophrenia.

Methods: Total protein extracted from auditory cortex gray matter homogenates of schizophrenia and matched control subjects was enriched for phosphopeptides using titanium dioxide and analyzed via LC-MS/MS. Cortical gray matter from antipsychotic- and sham-exposed monkey were similarly processed and analyzed, as were samples from a mouse model of postmortem interval. Computational modeling of the impact of phosphorylated residues on MAP2 structure was conducted using all atom Molecular Dynamics simulations. Phosphomimetic and phosphonull MAP2c constructs were generated by site-directed mutagenesis and transiently expressed in HEK 293 cells.

Results: Multiple phosphorylated residues in MAP2 that are abnormally increased (up to ~1000%) in schizophrenia were identified. MAP2 hyperphosphorylation was not an artifact of factors such as postmortem interval or long-term antipsychotic exposure. Levels of several phosphopeptides correlated inversely with dendritic spine density, suggesting that hyperphosphorylation of MAP2 may cause neurotoxicity. To further identify how hyperphosphorylation of MAP2 may alter MAP2 function, we modeled phosphorylation of MAP2 in silico and in vitro, focusing initially on pS426 (numbering as per the MAP2c sequence). pS426 was one of the most elevated residues and is homologous to pS396 in MAPTau, which is known to confer neurotoxicity. In silico simulations demonstrated that pS426 MAP2c induces helical motifs in MAP2 to become extended and exposed to solvent facilitating novel protein interactions. in vitro, expression of phosphomimetic S426E, resulted in increased protein levels compared to WT and to a phosphonull S426A mutant in transiently transfected HEK293 cells. Data contrasting the protein interactome of S426E and S426A MAP2 will also be presented.

Conclusions: Our findings demonstrate that abnormal MAP2 phosphorylation is closely linked to dendritic spine pathology in schizophrenia. Because MAP2 shares substantial sequence, regulatory, and functional homology with MAPTau, the deep knowledge of MAPTau biology and neurotoxicity provide resources that can be further leveraged to enhance translation of findings regarding MAP2.

Keywords: Schizophrenia, Synaptic Aberrations, Microtubule, Dendritic Spine, Tau

Disclosure: Nothing to Disclose.

W208. Prenatal One-Carbon Dysregulation Programs Cognitive Deficits

Amal Alachkar*, Ryan Yoshimura, Lien Wang, Geoffrey Abbott, Olivier Civelli, Xiangmin Xu

University of California, Irvine, Irvine, California, United States

Background: The one-carbon metabolism, which relies on the methionine-folate cycles, plays a critical role in the integration of genetic and environmental factors. This universal pathway cycles methyl groups from methionine to various products. Methionine, supplied through the diet, is converted to S-adenosylmethionine (SAM) the main methyl donor, and is used by over 100 methyltransferases. The one-carbon metabolism has been implicated in the pathophysiology of a number of psychiatric disorders such as autism, Alzheimer's, and schizophrenia. Here we examined the effects on mouse development and behavior of perturbation of one-carbon metabolism evoked by daily methionine administration to dams in the last week of gestation.

Methods: Pregnant mice that mated with the same male mouse were administered with MET (750 mg/kg) or saline (SAL) twice a day for seven to eight consecutive days from the fourteenth day of pregnancy (gestational day 14) until delivery. The dose of methionine is equivalent to double their normal daily dietary intake. Initially, male mice were tested with a battery of behavioral paradigms: social interaction and novelty, spontaneous T maze alternation, novel object/location recognition, prepulse inhibition, contextual fear conditioning. Neurogenesis was evaluated in the hippocampus using immunohistochemistry. RNA microarray, followed by qtPCR, was used to determine the changes in genes expressions. We investigated whether the MET prenatal treatment affects synaptic connectivity in the hippocampus, in particular the CA1 region, by applying the Laser scanning photostimulation (LSPS) approach.

Results: MET mice exhibited deficits in cognitive functions and memory as shown in T-maze, social recognition, novel object recognition, novel location recognition, and fear conditioning. MET mouse brains also exhibited decreased neurogenesis and synaptic plasticity, increased gliogenesis, and abnormally reduced local excitatory synaptic connections in CA1 neurons. Strikingly, neural transcript expression of only one gene, encoding the Npas4 transcription factor, was >2-fold altered (downregulated) by prenatal methionine administration.

Conclusions: Our data support a role for prenatal one-carbon pathway in memory, and suggest methionine metabolism as a potential therapeutic target for psychiatric disorders associated with cognitive deficits.

Keywords: One-Carbon Metabolism, Methionine, Memory, Gene Expression

Disclosure: Nothing to Disclose.

W209. Dose Effects of Varenicline on Cognition as a Function of Smoking Status in Schizophrenia and Non-Psychiatric Controls

Karolina Kozak*, Tony George

University of Toronto, Toronto, Canada

Background: In comparison to the general population (<18%), patients with schizophrenia have substantially higher rates of tobacco smoking (58-88%), as well as greater cognitive deficits. Interestingly, smoking may have procognitive effects in this disorder due to downstream modulatory dopaminergic effects associated with dysregulated nicotinic acetylcholine receptors (nAChR). Moreover, effective smoking cessation aids may also target cognitive deficits. Whether such procognitive effects vary as a function of smoking status and psychiatric diagnosis remain unclear. We studied the dose-dependent effects of the nicotinic partial agonist varenicline in both SZ and non-psychiatric smokers and non-smokers on cognitive outcomes.

Methods: In a randomized, placebo-controlled cross-over human laboratory study, varenicline (0, 1, 2 mg/day) was administered over 3 days with a 1-week washout period between test weeks. Fifteen non-smokers with SZ, 15 non-psychiatric non-smokers, 14 schizophrenia tobacco smokers, and 14 non-psychiatric smokers completed the study. A cognitive battery was administered; including effects on VSWM, verbal memory, executive function, attention, and impulsivity were evaluated.

Results: A 2 (diagnosis) x 2 (smoking status) x 3 (varenicline dose) mixed factorial ANOVA with dose as the within-subjects factor and diagnosis and smoking status as the between-subject factors was performed on all cognitive measures. There was a nearly significant smoking x diagnosis interaction within doses on VSWM (F(1,54)=2.871, p=0.096). A significant diagnosis x smoking status interaction was found on the HVLT-R sub-variables including: delayed recall (F(1,54)=6.319, p=0.015), retention (F(1,54)=4.819, p=0.032), discrimination index (F(1,54)=6.319, p=0.030) and total score (F(1,54)=5.439, p=0.023), as well as CPT Omissions (F(1,54)=4.384, p=0.041). Subsequent one-way ANOVA's generally found diagnostic and smoking status differences in either the placebo or high dose (p's<0.05). On CPT Hit rate (msec) and Digit Span Forwards, there was a significant dose x smoking status within-subjects interaction which was not moderated by diagnosis (F(2,108)=4.505, p=0.013; F(2,108)=3.449, p=0.035, respectively).There were no three-way within subject interactions across any other cognitive outcomes. Graphically, performance on VSWM, verbal memory displayed as such: control non-smokers>control smoker>schizophrenia smoker>schizophrenia non-smokers.

Conclusions: Our preliminary results suggest complex effects of diagnosis and smoking status on several cognitive domains as a function of varenicline dose. These findings may have implications for understanding factors that may contribute to the cognitive effects of nicotinic partial agonists in people with schizophrenia.

Keywords: Schizophrenia, Cognition, Tobacco Smoking, Varenicline, Smoking Status

Disclosure: Nothing to Disclose.

W210. Intranasal Oxytocin Modulates Social Cognitive Errors in the Psychosis Spectrum

M. Mercedes Perez-Rodriguez*, Amanda Fisher, Sarah Rutter, Derek Alexander Smith, Caridad Benavides, Margaret McClure, Daniel Rosell, Erin A. Hazlett, Harold Koenigsberg, Antonia S. New, Scott J. Moeller, Prantik Kundu

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

Background: Social cognition is crucial for developing and maintaining social relationships. Abnormal social cognition is characteristic of schizophrenia spectrum disorders, and causes significant disability and impaired social functioning.

Schizotypal personality disorder (SPD) is a milder, non-psychotic disorder within the schizophrenia spectrum. SPD is characterized by attenuated, schizophrenia-like traits without overt psychosis. It shares many of the genetic, psychophysiological and neural abnormalities found in schizophrenia. As a model of schizophrenia, SPD represents a cleaner unit of analysis, free from confounders such as medication, global neurocognitive deficits and institutionalization.

In schizophrenia, multiple domains of social cognition are often abnormal, including mentalizing, which maps onto the Research Domain Criteria (RDoC) construct “Perception and Understanding of Others: Understanding Mental States”. Mentalizing is the capacity to understand the mental states of other individuals, such as their intentions, thoughts, feelings and beliefs. Patients with schizophrenia make mentalizing errors characterized by simplistic, incomplete interpretations of social cues (termed “hypomentalizing” or “no mentalizing” errors). “Hypomentalizing” and “no mentalizing” errors are correlated with negative symptoms, while “hypermentalizing” errors -distorted misinterpretations of social cues- are associated with positive symptoms.

Oxytocin has been proposed as a modulator of social cognition. However, oxytocin's effect on social cognitive errors remains unexplored.

We aimed to: 1) characterize mentalizing errors in SPD patients and test their relationship with positive and negative symptoms of psychosis; 2) test the effect of intranasal oxytocin on mentalizing errors in SPD patients.

Methods: Subjects: 15 unmedicated SPD patients, 15 healthy controls [HC], and 15 psychiatric controls with other personality disorders. Intervention: intranasal oxytocin 24/40IU/placebo. Measures: Mentalizing was assessed with the Movie for the Assessment of Social Cognition (MASC). The MASC is a naturalistic video task simulating a real-life interaction among 4 characters. The MASC was designed to assess mentalizing accuracy and errors. Multiple-choice questions about the feelings, thoughts and intentions of the 4 characters yield 4 outcome measures: mentalizing accuracy, “no mentalizing” errors, “hypomentalizing” errors and “hypermentalizing” errors. The “hyper-hypomentalizing ratio” can be computed to capture the predominant mentalizing tendency. Those with ratios above 1 tend to hypermentalize, those with ratios below 1 tend to hypomentalize.

The Positive and negative symptom scale (PANSS) assessed positive and negative symptoms. The Schizotypal Personality Questionnaire (SPQ) assessed dimensional psychosis traits. Mentalizing measures were compared across groups (BPD,SPD,HC), and treatments (oxytocin 24IU/40IU vs placebo) using ANOVA.

We used Pearson correlations to assess the relationship between mentalizing accuracy, “hypermentalizing”, “hypomentalizing” and “no mentalizing” errors and symptom measures.

Results: SPD patients had significantly lower mentalizing accuracy (F=10.11;df=1;p=0.003,Cohen's d=0.91) and made significantly more “No mentalizing” or “hypomentalizing” errors than HC (F=12.92;df=1;p=0.001,Cohen's d=1.00). SPD patients had lower hyper-hypomentalizing ratios compared to HCs (F=2.84; df=1;p=0.099,trend level, Cohen's d=0.78). In a subset of patients –including 8 SPD-, a single dose of intranasal oxytocin significantly increased the hyper-hypomentalizing ratio (F=6.84, df=1,p=0.019,partial eta2=0.30, large effect size). Oxytocin also caused increases in hypermentalizing and decreases in hypomentalizing errors, tending to reverse the abnormalities found at baseline.

“No mentalizing” and “hypomentalizing” errors were significantly correlated with the PANSS negative symptoms subscale (r=0.25,p=0.013; r=0.37,p<0.001 respectively). “Hypermentalizing” errors were significantly correlated with the PANSS positive symptoms subscale (r=0.20,p=0.044) and the “ideas of reference” (r=0.37,p=0.033) and “suspiciousness” SPQ subscales (r=0.41,p=0.017). MASC accuracy was significantly inversely correlated with PANSS positive and negative symptom subscales (r=-0.30,p=0.003; r=-0.48,p<0.001 respectively) and with the “ideas of reference” SPQ subscale (r=-0.39,p=0.024).

Conclusions: As hypothesized, SPD patients had impaired, less accurate social cognition, and made more “no mentalizing” and “hypomentalizing” errors, which were correlated with negative symptoms. Conversely, “hypermentalizing errors” were correlated with positive symptoms and delusional and paranoid traits. Mentalizing accuracy was inversely correlated with positive and negative symptoms, and with delusional traits. Oxytocin increased the tendency to hypermentalize. This effect may normalize the low hyper/hypomentalizing ratios found at baseline in SPD patients.

These results support the role of social cognitive impairments as an underlying factor of positive and negative symptoms of psychosis, with specific associations with paranoid and delusional traits. Our results also suggest that intranasal oxytocin modulates social cognitive errors in the psychosis spectrum.

Keywords: Oxytocin, Social Cognition, Mentalizing, Schizotypy, Schizophrenia

Disclosure: Nothing to Disclose.

W211. Relationship of Prolonged Acoustic Startle Latency to Diagnosis and Biotype in the Bipolar-Schizophrenia Network on Intermediate Phenotypes (B-SNIP) Cohort

Nicholas Massa, Arpita Bhattacharya, Andrew V. Owens, Sarah Keedy, Wesley Harmon, John A. Sweeney, Godfrey Pearlson, Matcheri S. Keshavan, Elena Ivleva, Carol Tamminga, Brett Clementz, Erica Duncan*

Emory University School of Medicine, Decatur, Georgia, United States

Background: The acoustic startle reflex is mediated by a three-synapse subcortical neural circuit. Prepulse inhibition of startle (PPI), an operational measure of sensorimotor gating, is extensively studied in schizophrenia (SCZ). Less often studied is startle latency, which is the time required for the startle reflex to occur after a startling stimulus. Latency provides a putative index of neural processing speed. Significant findings regarding latency have emerged: SCZ subjects have slower latency compared to healthy controls (CON), and latency is highly heritable (up to 90% heritability). Additionally, slower latency at baseline is predictive of conversion to schizophrenia during a two-year follow-up period in young subjects at high risk for development of schizophrenia.

The Bipolar-Schizophrenia Network on Intermediate Phenotypes (B-SNIP) Consortium has discovered three clustered intermediate phenotypes (Biotypes-1, -2, and -3) derived from a battery of neurocognitive and biological measures (Clementz et al., Am J Psychiatry 2016). These Biotypes are seen in subjects with psychotic disorders (schizophrenia, schizoaffective disorder, bipolar disorder) and occur spanning across these clinical diagnoses. We investigated whether a slowing of latency occurs in this population of SCZ, bipolar subjects (BP), and their families. Furthermore, we tested associations of the B-SNIP Biotypes with latency and PPI.

Methods: From the B-SNIP cohort 1149 subjects were included: SCZ, n=270; SCZ-relatives (SCZ-Fam), n=331; BP, n=139; BP-relatives (BP-Fam), n=183; and controls (CON), n=226. Subjects in the Biotype analysis included both SCZ or BP probands and their relatives: 150 in Biotype-1, 313 in Biotype-2, and 343 in Biotype-3. A Biopac system recorded the eyeblink component of the startle reflex to assay baseline startle magnitude, latency, and PPI. The paradigm contained pulse-alone trials to evaluate baseline startle, and prepulse+pulse trials with inter-stimulus intervals of 120 or 4500ms. Repeated measures analyses of variance (ANOVAs) were used to compare startle outcome measures in subject groups, with covariates included where appropriate (age, race, sex, and baseline startle magnitude).

Results: Latency was slower in SCZ than CON (F(1,489)=7.28, p=0.007). When SCZ-Fam subjects were added to this model, the SCZ-Fam group was intermediate between SCZ and CON in pulse-alone trials. Likewise, latency was slower in BP than CON in pulse-alone trials in post hoc tests (p=0.01), however not significant for other trial types (F(1,360)=0.79, p=.37). When BP-Fam subjects were added to this analysis, the BP-Fam group was intermediate in pulse-alone trials. A model comparing latency in SCZ, BP and CON was significant for diagnosis (F(2,629)=5.24, p=0.006), with post hoc tests indicating that both SCZ (p=0.001) and BP (p=0.03) were slower than CON in pulse-alone trials, and SCZ slower than BP in 4500ms trials (p=0.013). PPI, however, was not significant in models comparing SCZ or BP to CON. In comparisons of these probands to one another and to their families, there were no significant differences in PPI (p>0.4).

In ANOVAs on Biotype that included SCZ, SCZ-Fam, BP, and BP-Fam subjects, Biotype-1 was associated with the slowest startle latency compared to Biotype-2 and Biotype-3 (F(2,800)=3.58, p=0.028). Similarly, PPI was most impaired in Biotype-1 (F(2,801)=3.80, p=0.02).

Conclusions: These analyses confirm that startle latency is slowed in SCZ. Both slow latency and impaired PPI associate with Biotype-1 compared to the other two Biotypes. These data thus add to prior work supporting the Biotype approach to uncovering the neurobiology of psychosis that is in part orthogonal to diagnostic boundaries.

Keywords: Latency of Acoustic Startle, Prepulse Inhibition, Schizophrenia, Bipolar, Biotype

Disclosure: Part 4: Auspex Pharmaceuticals, Grant, Teva Pharmaceuticals, Grant, Posit Science, Inc., Grant.

W212. Serum Metabolomics of Schizophrenia Participants by Fasting Insulin Reveals a Broad Metabolic Range Not Associated With Current Atypical Antipsychotic Therapy

Kristen Ward*, Cora McHugh, Larisa Yeomans, Zarina Kraal, Stephanie Flowers, Alla Karnovsky, Vicki Ellingrod, Kathleen Stringer

University of Michigan, Ann Arbor, Michigan, United States

Background: It is not fully understood why patients with schizophrenia are at exceptionally high risk of developing metabolic syndrome (MetS). One potential contributing factor is the widespread use of atypical antipsychotics (AAPs), which increase the risk of developing dyslipidemias and diabetes. The aim of this pilot study was to improve understanding of MetS in schizophrenia by comparing metabolic profiles of AAP-treated patients across a broad spectrum of fasting insulin levels.

Methods: Serum samples from 93 schizophrenia patients were grouped by fasting insulin concentration into quartiles. Groups were matched for age, gender, and total daily calorie intake. Serum samples were extracted into aqueous and lipophilic fractions. Nuclear magnetic resonance (NMR) and gas chromatography were used to generate quantified aqueous metabolite profiles from the aqueous fraction, and free fatty acid metabolite profiles from the lipophilic fraction. Data was normalized and transformed, then metabolite concentrations were compared between quartiles using ANOVA. To correct for multiple comparisons, differentiating metabolites were considered significant if they met a false discovery rate of <20%.

Results: The median (IQR) age of all participants was 47.0 (47.0-52.0), and the median time in years since schizophrenia diagnosis was 19.0 (12.8-29.3). The median fasting insulin concentrations (mU/mL) were: 9.4 (8.4-10.1), 14.6 (13.7-15.7), 20.6 (18.7-22.7), and 34.7 (28.1-42.9). Significant metabolites included acetate, glutamate, and several other amino acids, as well as saturated long-chain fatty acids. All results were independent of current AAP therapy when compared by percent users taking clozapine or olanzapine, and when total AAP dose was normalized using chlorpromazine equivalents and compared between quartiles.

Conclusions: These preliminary results suggest that current AAP therapy may not be a significant contributing factor to metabolic risk in middle-age schizophrenia patients. This study also highlights the broad metabolic range of MetS and suggests that current diagnostic criteria may be too simplistic to capture the breadth of risk across this patient population.

Keywords: Schizophrenia, Metabolomics, Insulin Resistance, Atypical Antipsychotics

Disclosure: Nothing to Disclose.

W213. Neuron-Specific Deficits of Neuroenergetic Processes in Schizophrenia

Courtney Sullivan, Sinead Odonovan, Amy Ramsey, Robert McCullumsmith*

University of Cincinnati, Cincinnati, Ohio, United States

Background: Schizophrenia is a devastating illness that affects over 2 million people in the U.S. and costs society billions of dollars annually. Patients with schizophrenia experience a wide range of psychotic symptoms, as well as cognitive deficits and profound negative symptoms that are often treatment resistant.

A growing body of evidence suggests abnormal bioenergetic function in chronic schizophrenia, including deficits in energy storage and usage in the brain. Microarray studies found significant decreases in expression of genes encoding proteins involving the malate shuttle, transcarboxylic acid (TCA) cycle, as well as the ornithine–polyamine, aspartate–alanine, and ubiquitin metabolism groups in the dorsolateral prefrontal cortex (DLPFC). Several postmortem studies have also found abnormalities in metabolic enzyme activity, including dysregulation of TCA cycle enzyme activity in the DLPFC and decreased specific activity of mitochondrial respiratory chain enzymes in the frontal cortex. in vivo studies also implicat