Introduction

The Disrupted-in-Schizophrenia 1 (DISC1) gene was originally discovered at the breakpoint of a balanced translocation t(1;11) (q42;q14.3) in a Scottish family and later identified in a North American family with high rates of schizophrenia.1, 2, 3, 4 Since then, preclinical models have shown that DISC1 mutant animals exhibit behavioural, neurostructural and neurochemical features relevant to schizophrenia, 5, 6 although its significance for the human disease has been debated.7, 8, 9 DISC1 is described as a scaffold protein with multiple interactors involved in a wide range of cellular processes including neurotransmitter signalling.10, 11 In particular, DISC1 is known to interact with several proteins involved in dopamine signalling including fasciculation and elongation protein zeta 1, phosphodiesterase 4D9 and phosphodiesterase 4B, serine/threonine protein kinase Akt and glycogen synthase kinase-3 (GSK-3)12, 13, 14, 15, 16 as well as synaptic interactors such as kalirin-7 and the Traf2, Nck-interacting kinase,17, 18 and the microtubule/centrosomal proteins pericentriolar material 1 and Bardet–Biedl syndrome protein.19, 20 These multiple interactions have highlighted the potential of DISC1 as a therapeutic target.21, 22, 23

The neurotransmitter dopamine is widely thought to have a central role in the aetiology of psychotic disorders.24, 25, 26 The dopamine hypothesis of schizophrenia was initially based on the findings that the affinity of antipsychotic medications for dopamine receptors is closely related to their clinical potency,27, 28, 29 and that drugs that increase dopamine levels provoke psychotic symptoms in healthy people.30, 31 Molecular imaging studies since then have shown increased presynaptic dopamine synthesis capacity and release in schizophrenia32, 33, 34, 35 and in subjects with prodromal symptoms of schizophrenia.36, 37, 38, 39 Alterations in dopamine D1 and D2/3 receptors, tyrosine hydroxylase (TH) levels and baseline synaptic dopamine levels in schizophrenia have also been reported,40, 41 although with some inconsistency.42

These findings highlight why dopaminergic dysfunction has a pivotal role in schizophrenia. In view of this, we sought to review the evidence from animal models that DISC1 pathway alterations may impact on dopaminergic function, as it has not been comprehensively synthesised before. The aim of our review was therefore to summarise the impact of DISC1 on TH levels, dopamine transporter (DAT) levels, basal dopamine levels and after amphetamine administration, dopamine D2 receptor-binding potential (BP), dopamine D1 (D1R) and D2 receptor (D2R) levels, and locomotor activity after amphetamine administration for dopamine-related behaviour.43 We selected publications citing data collection in the midbrain, as this is the location of the majority of dopaminergic neuron cell bodies in the brain, and the frontal cortex, hippocampus and striatum as these are the target sites of the main dopaminergic pathways relevant to psychiatric disorders.44, 45

Materials and methods

Selection of studies

The entire PubMed database was searched to select publications. Studies were screened based on the terms (‘Disrupted-in-Schizophrenia-1’ OR ‘DISC1’) AND (‘dopamine’ OR ‘tyrosine hydroxylase’ OR ‘dopamine receptor’ OR ‘DAT’ OR ‘amphetamine’ OR ‘behavioral alterations’ OR ‘locomotor activity’ OR ‘Positron Emission Tomography’ OR ‘PET’ OR ‘Single Photon Emission Computed Tomography’ OR ‘SPECT’). Only articles meeting the following criteria were included: (1) original studies; (2) English language; (3) peer-review journals; (4) findings reporting TH levels, DAT levels, basal dopamine levels and/or dopamine levels after amphetamine administration, and/or dopamine receptor-binding potential, dopamine receptor levels and/or locomotion after amphetamine administration in a DISC1 model compared with a control group; and (5) in the frontal cortex, striatum, nucleus accumbens, midbrain and/or hippocampus, as these regions are major target sites of dopaminergic projections in the brain and are thought to be involved in the pathophysiology of schizophrenia.44, 45 The DISC1 models were selected based on gene mutation in DISC1 or alteration in the quantitative expression of DISC1 protein. Method and results sections of the eligible articles were screened to identify the measures of interest listed above.

Data extraction

The main outcome measures were the differences between the DISC1 models and controls in (1) TH levels; (2) DAT levels; (3) basal dopamine levels; (4) dopamine levels after amphetamine administration; (5) dopamine receptor-binding potential; (6) D1R and D2R levels; and (7) locomotion after amphetamine administration. In addition, the following data were extracted: (8) authors; (9) year of publication; (10) the DISC1 model; (11) samples size; and (12) methods. The data were extracted by TD and checked by SVT. Findings related to the nucleus accumbens and olfactory tubercle were merged as both being part of the ventral striatum.46

Results

Fifty-one studies were excluded from a total of 65 studies screened (Figure 1). Fourteen studies were included of which two were of TH levels, three of DAT levels, nine of basal dopamine levels, six of induced dopamine release, four of dopamine receptor BP, four studies of D1R levels, four studies of D2R levels and thirteen of locomotion after amphetamine administration. Table 1 summarises all studies including the DISC1 model used, sample sizes and methods. It should be noted that we were not able to find evidence that dopaminergic function had been investigated in more recently disclosed DISC1 models, for example.47, 48

Figure 1
figure 1

Flow chart of identification, exclusion and inclusion of eligible studies. DISC1, disrupted-in-schizophrenia 1.

Full size image
Table 1 DISC1 models with available dopamine-related data
Full size table

DISC1 models

Five types of DISC1 models were identified across the studies as follows: (1) transgenic expression of truncated human Disc1 protein with dominant-negative (DN) effect; (2) DISC1 haploinsufficiency/silencing; (3) full-length human DISC1 overexpression; (4) artificial Disc1 mutation; and (5) wild-type model (Table 1). Data on locomotion after amphetamine administration from Su et al.62 were included despite the absence of a direct comparison between mutant and wild-type mice as they showed a functional relationship between Disc1 and the dopamine receptor. Both genotype effects (wild type versus transgenic) and genotype effect in a stress condition (isolated wild type versus isolated transgenics) were included from Niwa et al.51

TH levels

Two studies investigated TH levels in the hDISC1 and the Disc1 RNA interference (RNAi)/silencing models compared with controls.51, 54 These studies showed reduced TH levels in frontal cortical regions in isolated hDISC1 mice compared with isolated controls51 and in the Disc1 RNAi/silencing model compared with controls54 (Figure 2 and Table 3).

Figure 2
figure 2

The impact of DISC1 models on the dopamine system. AMPH, amphetamine; DA, dopamine; DAT, dopamine transporter; NAc, nucleus accumbens; SN, substantia nigra; TH, tyrosine hydroxylase; VTA, ventral tegmental area.

Full size image

One study showed no significant changes in TH levels between hDISC1 and controls, and between isolated hDISC1 mice and isolated controls in the nucleus accumbens.51

DAT levels

Three studies investigated DAT levels in the DISC1 model compared with controls.52, 56, 58 Two studies found increased DAT levels in the striatum of DN homozygous line 37 mice and tgDISC1 rats compared with controls Tables 2 and 3.52, 58

Table 2 Methods
Full size table
Table 3 Findings
Full size table

One study found no significant difference between the Disc1Δ2–3 mice compared with controls in the nucleus accumbens.56

Basal dopamine levels

Nine studies investigated basal dopamine levels in DISC1 models compared with controls.49, 50, 51, 52, 54, 56, 58, 59, 61 In vivo microdialysis and post-mortem high-performance liquid chromatography with electrochemical detection (HPLC-ED) were used, measuring extracellular and total dopamine levels, respectively.

Eight studies investigated basal dopamine levels in the frontal cortex/mPFC, six using HPLC-ED49, 50, 55, 58, 59, 61 and two using both post-mortem HPLC-ED and microdialysis.51, 54 One of the two studies using microdialysis showed decreased basal dopamine levels in the Disc1 RNAi/silencing model compared to controls,54 whereas the other study found no significant changes between the hDISC1 mice and controls and the isolated hDISC1 mice and isolated controls.51 For HPLC-ED, decreased basal dopamine levels were found at postnatal day 56 in the Disc1 RNAi/silencing model,54 and in males from the prenatal hDISC1 expression group (until embryonic day 17), the postnatal expression group (from embryonic day 17 on) groups and the pre- and postnatal hDISC1 expression (entire life) compared with controls.49 No significant differences were reported in the other studies.50, 51, 55, 58, 59, 61

Six studies investigated basal dopamine levels in the striatum using HPLC-ED49, 50, 51, 55, 58, 61 and one using both HPLC-ED and microdialysis.59 One study found decreased total dopamine levels in the full-length DISC1-overexpressing rat model compared with controls in the dorsal striatum.58

Six studies investigated basal dopamine levels in the nucleus accumbens, one using in vivo microdialysis,52 four using HPLC-ED49, 58, 59, 61 and one using both techniques.54 Two studies using in vivo microdialysis showed decreased basal dopamine levels in the Disc1 RNAi/silencing model compared with controls54 and the hDISC1 heterozygous line 10 and 37 mice compared with controls.52 One study using HPLC-ED showed significant decreased basal dopamine levels in L100P ENU-generated missense mutation mice,61 whereas the others found no significant differences.49, 54, 58, 59

One study investigated basal dopamine levels in the midbrain and found no significant difference between the Disc1Δ2–3 haploinsufficiency model and controls.55

Seven studies investigated basal dopamine levels in the hippocampus using HPLC-ED.49, 50, 54, 55, 58, 59, 61 One study found decreased dopamine levels in females in the postnatal hDISC1 expression group compared with prenatal expression only and controls.49 The other studies found no significant differences.50, 54, 55, 58, 59, 61

Induced dopamine release

All the studies induced dopamine release by administrating amphetamine-related drugs. Two studies investigated induced dopamine release in the frontal cortex and found no significant differences, one using microdialysis51 and one using HPLC-ED.50

One study investigated induced dopamine release in the striatum using in vivo microdialysis59 and one study using HPLC-ED,50 both reporting no significant differences.

Four studies investigated induced dopamine release in the nucleus accumbens using microdialysis.51, 52, 54, 56 The four studies found significantly increased dopamine release. This was in the DISC1 knockdown compared with controls,54 in isolated hDISC1 compared with isolated controls,51 in heterozygous line 10 and 37 compared with controls 52 and female but not male Disc1Δ2–3 mice compared with controls.56

One study investigated induced dopamine release in the hippocampus using HPLC-ED and found no significant difference between the hDISC1 and controls.50

Dopamine D1 receptor

Two studies investigated D1R levels in the frontal cortex and found no significant differences between the hDISC1 and controls, and the Disc1Δ2–3 haploinsufficiency model and controls.51, 56

Three studies investigated D1R levels in the striatum.52, 56, 58 One study found increased levels in the hDISC1 model compared with controls,52 whereas the others found no significant differences.56, 58

Two studies investigated D1R levels in the nucleus accumbens.51, 56 One study found significant increased D1R levels in female and no significant changes in male and mixed Disc1Δ2–3 groups,56 whereas the other showed no significant differences.51

Dopamine D2 receptor

Three studies investigated D2R levels in the frontal cortex.51, 54, 56 One study found significant increased D2R levels in the hDISC1 mice compared with controls and isolated hDISC1 mice compared with isolated controls51 and the two other studies found no significant differences between the Disc1 RNAi/silencing/haploinsufficiency models and controls.54, 56

Two studies investigated D2R levels in the striatum.52, 56 The hDISC1 mice showed significant increased D2R levels,52 whereas the other showed no significant differences between the Disc1Δ2–3 models and controls.56

Four studies investigated dopamine receptor-binding potential in the striatum.50, 52, 58, 59 The dopamine D2 receptor is known to exist in two interconverting states, a low-affinity (μm) and a high-affinity (nm) state.63 Lipina et al.59 and Trossbach et al.58 found a significant increase in dopamine D2 high-affinity receptor levels using [3H]domperidone binding challenged with dopamine, but Trossbach et al. found no difference in [3H]raclopride binding by autoradiography. As raclopride does not distinguish low from high affinity or D2 from D3 receptors, taken together, these studies are consistent with a shift to the high-affinity state without a change in total D2/3 receptor levels. Jaaro-Peled et al.52 found significantly increased binding potential of D2/3 receptor availability in the striatum using [11C]raclopride PET and significantly increased levels of D2/3R in the medial part of the right rostral striatum using [3H]spiperone autoradiography but no significant differences in D2/3 levels in the total right rostral striatum and the lateral part of the right striatum in the hDISC1 compared with controls. Pogorelov et al.50 found no significant difference in the rostral part of the striatum using [11C]raclopride autoradiography in the hDISC1 mice compared with controls.

Two studies investigated D2R levels in the nucleus accumbens.51, 56 One study found significantly increased D2R levels in female but not male and mixed Disc1Δ2–3 groups,56 whereas the other showed no significant differences.51

One study investigated D2/3R-binding potential in the nucleus accumbens using [11C]raclopride autoradiography PET and found no significant differences in the nucleus accumbens but significantly decreased levels in the right olfactory tubercle of female hDISC1 mice compared with controls.50 They used the same approach to investigate D2/3R-binding potential using [11C]raclopride autoradiography in the midbrain (substantia nigra/VTA) and found no significant difference between the hDISC1 and controls.50

One study investigated D2R levels in the hippocampus and found no significant difference between the Disc1Δ2–3 haploinsufficiency and controls.56

Locomotion after amphetamine administration

Thirteen studies investigated locomotion after amphetamine administration.49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 62 Ten studies found increased locomotion after amphetamine administration in the DISC1 models compared with control animals, in the pre- and postnatal hDISC1 expression groups,49 hDISC1 mice,52 the Disc1 RNAi/silencing model,54 female but not male Disc1Δ2–3 mice,55 male but not female Disc1Δ2–3 mice,56 full-length hDISC1-overexpressing rats,57, 58 Disc1-L100P mice,59 isolated hDISC1 mice compared with isolated controls51 and wild-type Disc1 mice with no DISC1–D2R disruption.62 Two studies found decreased locomotion after amphetamine administration, in female but not male hDISC1 mice after escalating dose of methamphetamine treatment compared with controls in Pogorelov et al.50 and wild-type mice with Disc1–D2R disruption in Su et al.62 No significant changes were found in the hDISC1 mice and Disc1-L100P/L100P mice compared with controls in three studies.51, 53, 60

Discussion

The main findings show that compared with controls, the DISC1 models exhibit reasonably consistent (1) increased locomotion after amphetamine administration (2) increased dopamine levels after amphetamine administration in the nucleus accumbens but (3) inconsistent alterations in basal dopamine levels and dopamine receptor levels and binding potentials. These findings extend other studies showing increased methamphetamine-induced dopamine release in the nucleus accumbens and locomotor hyperactivity in mice lacking DISC1-interacting proteins, such as fasciculation and elongation protein zeta 164 and PDE4,65 to indicate that the DISC1 pathway affects specific aspects of dopaminergic function.

Limitations

The findings presented in this systematic review must be considered in the light of the following limitations. First, the number of studies was modest for some aspects of dopaminergic function, such as transporter levels, and some regions. This limits the conclusions that can be drawn about these aspects, and highlights the needs for further studies. Second, the studies used a heterogeneous set of DISC1 models (Table 1), which could contribute to variability in results. Third, the evidence comes from a relatively small number of research groups. Thus, replication would be useful to determine generalisability. And fourth, alterations in other neurotransmitter system such as noradrenaline might also contribute to the locomotor hyperactivity phenotype observed. However, several reports indicate that locomotor hyperactivity after amphetamine is specifically mediated through dopamine and not noradrenergic transmission in the nucleus accumbens.66, 67, 68

Potential mechanisms underlying locomotor hyperactivity

The majority of the DISC1 models used showed locomotor hyperactivity following amphetamine challenge. This shows a relatively conserved phenotype of the DISC1 models that might be explained by (1) the presynaptic effects of DISC1 on dopamine release in the nucleus accumbens or (2) a direct impact of the DISC1 models on postsynaptic dopaminergic signal transduction, such as the protein serine/threonine protein kinase (Akt)–glycogen synthase kinase-3 (GSK-3) pathway. In support of the first hypothesis, the nucleus accumbens is thought to have an important role in regulating locomotor activity.69, 70 Local administration of dopamine and amphetamine has been shown to induce hyperactivity similar to systemic administration,66, 70, 71, 72 and our review has identified reasonably consistent evidence that DISC1 models are associated with greater dopamine release to amphetamine. With regards to the second hypothesis, Akt and GSK-3 are two proteins regulated by DISC1 with respectively indirect and direct interactions.15, 16, 73 The Akt–GSK-3 pathway modulates dopamine neurotransmission and amphetamine-mediated locomotor activity.74, 75, 76 Amphetamine/methamphetamine-induced dopamine release decreases Akt activation (phosphorylation state77), which activates GSK-3 by dephosphorylating the Serine 9 site78 to modulate dopamine-dependent behaviours.74 Although Disc1 wild-type protein decreases Akt and GSK-3 activation,15, 73, 79 the impact of mutant DISC1 on Akt and GSK-3 is less clear. Evidence shows increased and decreased Akt activation in DISC1 knockdown,15, 62, 80 no effects on Akt and GSK-3 levels in hDISC1 mice50 and consistently increased GSK-3 activation in DISC1 knockdown and Disc1 point mutation Q31L.62, 81, 82 Interestingly, mice overexpressing GSK-3 develop locomotor hyperactivity,83 GSK-3 knockdown mice express reduced locomotor activity84 and administration of GSK-3 inhibitor decreases amphetamine-induced hyperactivity.85

Potential mechanisms underlying increased dopamine release to amphetamine

The studies reporting increased dopamine levels following amphetamine administration in the nucleus accumbens used a Disc1Δ2–3 haploinsufficiency,56 a DN hDISC1 model in combination with adolescent isolation stress,51 a transient knockdown in prefrontal cortex 54 and a DN hDISC1 model targeting specifically pyramidal neurons of the cortex and hippocampus.52 This raises the questions of (1) the time course of changes in dopamine and whether there are developmental periods that are particularly vulnerable to DISC1 alterations, (2) the brain regions minimally required to lead to increased dopamine release to amphetamine, and in particular, the role of the cortical regions in regulating the nucleus accumbens dopamine levels.

With regards to the first point, recent studies suggest that DISC1 alterations interact with stress to impact on dopaminergic neurons during adolescence.51, 86 These findings are in line with evidence showing that adolescence is a critical time life for the development of psychotic disorders including schizophrenia.87 With regards to the second point, a possible mechanism underlying increased dopamine levels in the nucleus accumbens could be a reduction in cortical parvalbumin-positive interneurons. Supporting this, studies have shown a decreased number of parvalbumin-positive interneurons in the cortex of DN DISC1 models.49, 88, 89, 90 Parvalbumin-positive interneurons are GABAergic inhibitory neurons thought to regulate the dopaminergic activity in the nucleus accumbens and to have a role in schizophrenia through the modulation of cortical glutamate excitatory pyramidal neurons.91, 92, 93 Finally, the specific localisation of the findings in the nucleus accumbens might be related to an increased sensitivity of this region to stimulants, as it has been shown to release more dopamine following amphetamine administration compared with other striatal subdivisions.94

Inconsistent basal dopamine levels and dopamine receptor-binding potential and levels

We summarise here a series of inconsistent findings on basal dopamine levels and dopamine receptor-binding potentials and levels in the frontal cortex, striatum, nucleus accumbens and hippocampus. These findings might be due the heterogeneity of the DISC1 models used (Table 1). Among these, only the short interfering RNA knockdown or knockout models should have loss of function phenotypes whereas all others could have either loss of function, or gain of function, or combined phenotypes at the same time. However, no more consistency is observed when looking only at the loss of function models. It should also be noted that the tgDISC1 rat was conceived as a model for protein pathology related to DISC1 rather than a model for mutant DISC1.58, 95 Another possible explanation could be that these are adaptive changes not always seen following the core dopamine release alteration.

Implications

The effects of DISC1 on dopamine release and the behavioural effects of amphetamine are in line with evidence showing increased amphetamine-induced dopamine release in schizophrenia, and that this positively correlates with amphetamine-induced positive psychotic symptoms.34, 35, 96, 97 The absence of clear receptor changes is also consistent with the lack of changes in dopamine D2/3 receptors alterations seen in a meta-analysis of in vivo findings in schizophrenia.98 However, the inconsistent findings in striatal basal dopamine levels do not agree with the in vivo evidence showing increased basal dopamine levels in schizophrenia.33, 99 Taken together, these findings indicate that DISC1 alterations may increase the risk of schizophrenia by dysregulating the presynaptic regulation of dopamine but they do not result in the full dopaminergic phenotype, suggesting other factors must interact with DISC1. Stress is one likely candidate factor100 that has been shown increase dopamine release in psychosis.38

It should be noted that DISC1 is also associated with affective disorders including major depression.101, 102 The implications of the findings for this association remains unclear, as human PET studies have shown decreased dopamine synthesis capacity in patients with major depression particularly in individuals with reduced affect or psychomotor slowing symptoms103, 104, 105 and some endophenotypes such as anhedonia are thought to be underlined by dopamine function.106, 107

Our conclusions drawn from the preclinical research reviewed here may have interesting implications for clinical research and hence translational value at pointing to the necessity of identifying a biomarker to identify illness subtypes related to DISC1 dysfunction, to guide treatment choice and as a lead for the development of novel therapies. Determining whether DISC1 function is aberrant in a given individual could be a useful to subtype patients. Given that aberrant DISC1 function modulates aspects of dopaminergic function, this may help identify patients who may be responsive to drugs that act on the dopaminergic system, in line with emerging evidence on dopaminergic and non-dopaminergic subtypes of schizophrenia.108 What directions could the search for identifying biomarkers for aberrant DISC1 function take? Screening for DISC1 polymorphisms may be one way to assess this as some polymorphisms have been associated with different neuronal functions and with treatment-resistant schizophrenia.109, 110, 111 As it has been demonstrated that single-nucleotide polymorphisms of DISC1-interacting genes are overrepresented in schizophrenia,112 the use of a DISC1-interactome polygenic risk score might also be a complementary approach to stratify the risk associated with a specific signalling pathway or response to treatment. However, it should be recognised that genetic diagnostics alone may not provide sufficient information because DISC1 levels also depend on other factors, for example, BACE1-dependent cleavage of neuregulin 1.113 Large cohort studies of patients are needed to determine whether DISC1 polymorphisms and/or DISC1 protein levels in peripheral cells do identify subsets of patients with distinct illness characteristics or treatment response.114 This may require the combinatorial analysis of blood-based, imaging and/or neurophysiological factors, to both identify those patients with both aberrant DISC1 and neuronal function. Another key implication is that understanding how DISC1 alterations lead to dopamine dysregulation could identify new treatment approaches to address the dopamine dysfunction seen in schizophrenia and people at risk of schizophrenia in a broader sense. Pharmacological targeting of aberrant DISC1 function may be able to correct dopamine dysfunction without directly interfering with dopamine receptors themselves, providing an alternative to existing antipsychotic drugs, which are all D2/3 receptor blockers. In that sense, clinical development of diagnostics and pharmacotherapy of DISC1-related disorders may go hand in hand95 to support the development of precision medicine in psychiatry.

Future directions

We identify four key lines of direction for future studies based on the findings: first, as results to date come from a relative small number of studies, it would be useful to investigate dopamine function in DISC1 models recently developed.47, 48 Second, the mechanism by which DISC1 leads to increased dopamine release to amphetamine needs further investigation, in particular to determine whether this could be mediated by disinhibition of parvalbumin-positive interneurons or the Akt–GSK-3 pathway. Interestingly, a DISC1 model has been recently developed with selective knockdown of interneuronal DISC1 in parvalbumin neurons,47 which might provide insightful knowledge on the mechanisms linking DISC1 and dopamine regulations. In that context, it is also remarkable that DISC1 as a single factor is able to both regulate dopamine neuroanatomy as well as parvalbumin-positive interneuron placement in cortical layers.115 Third, elevated dopamine synthesis capacity is the other aspect of presynaptic dopamine dysregulation widely linked to schizophrenia and people at risk of schizophrenia.36, 37, 116 Thus, future work should test if DISC1 alterations affect this aspect of presynaptic dopamine function in humans. Fourth, it would be useful to examine further the impact of environmental stress on dopamine release and dopamine levels in DISC1 models as proposed by some authors.51, 117

Conclusions

Compared with controls, the majority of the DISC1 models but not all exhibits (1) increased locomotion after amphetamine administration and (2) increased dopamine levels after amphetamine administration in the nucleus accumbens but (3) inconsistent basal dopamine levels and dopamine receptor levels and binding potentials. This suggests that presynaptic dopamine dysregulation is a potential mechanism for the increased rates of psychotic disorders seen in the original DISC1 families and DISC1 variant carriers, and identifies a number of potential therapeutic targets for treating or even preventing schizophrenia based on the DISC1 pathway.