A Dopamine D2 Receptor-DISC1 Protein Complex may Contribute to Antipsychotic-Like Effects

Transcription

1 Neuron Article A Dopamine D2 Receptor-DISC1 Protein Complex may Contribute to Antipsychotic-Like Effects Ping Su, 1,6 Shupeng Li, 1,6 Sheng Chen, 1 Tatiana V. Lipina, 2 Min Wang, 1 Terence K.Y. Lai, 1 Frankie H.F. Lee, 1 Hailong Zhang, 1 Dongxu Zhai, 1 Stephen S.G. Ferguson, 3 José N. Nobrega, 1,4,5 Albert H.C. Wong, 1,5 John C. Roder, 2 Paul J. Fletcher, 1,4,5 and Fang Liu 1,5, * 1 Department of Neuroscience, Centre for Addiction and Mental Health, Toronto, ON M5T 1R8, Canada 2 Lunenfeld Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, ON M5G 1X5, Canada 3 Department of Physiology & Pharmacology, University of Western Ontario, London, ON N6A 5 K8, Canada 4 Departments of Psychology 5 Psychiatry, University of Toronto Toronto, ON M5S 2J7, Canada 6 Co-first author *Correspondence: f.liu.a@utoronto.ca SUMMARY Current antipsychotic drugs primarily target dopamine D2 receptors (D2Rs), in conjunction with other receptors such as those for serotonin. However, these drugs have serious side effects such as extrapyramidal symptoms (EPS) and diabetes. Identifying a specific D2R signaling pathway that could be targeted for antipsychotic effects, without inducing EPS, would be a significant improvement in the treatment of schizophrenia. We report here that the D2R forms a protein complex with Disrupted in Schizophrenia 1 (DISC1) that facilitates D2R-mediated glycogen synthase kinase (GSK)-3 signaling and inhibits agonist-induced D2R internalization. D2R- DISC1 complex levels are increased in conjunction with decreased GSK-3a/b (Ser21/9) phosphorylation in both postmortem brain tissue from schizophrenia patients and in Disc1-L100P mutant mice, an animal model with behavioral abnormalities related to schizophrenia. Administration of an interfering peptide that disrupts the D2R-DISC1 complex successfully reverses behaviors relevant to schizophrenia but does not induce catalepsy, a strong predictor of EPS in humans. INTRODUCTION Schizophrenia is a chronic mental illness, characterized by episodes of psychotic symptoms in addition to cognitive and social impairments that emerge in early adulthood. Dopamine D2 receptors (D2Rs) are the main target of antipsychotic medications, and the clinical potency of antipsychotic drugs is correlated with D2R binding affinity (Seeman et al., 1976). Some antipsychotics also block serotonin 2A receptors (5-HT2AR) in addition to D2R (Geddes et al., 2000). However, current antipsychotics are ineffective for controlling psychotic symptoms in many patients, and even with good symptom control, functional outcomes remain poor (Insel and Scolnick, 2006; Lieberman et al., 2005). In addition, current antipsychotic drugs cause serious side effects, including extrapyramidal symptoms (EPS), tardive dyskinesia, sexual dysfunction, weight gain, and diabetes (Leucht et al., 2009). Thus, understanding the details of D2R signaling is critical for the identification of novel therapeutic targets to develop more effective antipsychotic drugs with fewer side effects. D2Rs are G protein-coupled receptors (GPCRs) that activate both G protein-coupled signaling cascades and non-g protein pathways (Missale et al., 1998; Beaulieu et al., 2009). D2Rs couple with Gi/Go proteins to control adenylate cyclase activity, phosphatidylinositol turnover, arachidonic acid release, inwardly rectifying K + and Ca 2+ channels, and mitogen-activated protein kinases (Missale et al., 1998). D2Rs also signal through G protein-independent mechanisms such as b-arrestin2, which leads to formation of a complex that includes protein phosphatase 2A (PP2A), Akt, and activation of the GSK-3 pathway (Beaulieu et al., 2009). A primary mechanism for regulating D2Rs is kinase-mediated desensitization, endocytosis, and endosomal trafficking that is initiated by GPCR kinase phosphorylation. This in turn leads to the formation of a complex with b-arrestin, adaptor protein-2, and clathrin (Hanyaloglu and von Zastrow, 2008). D2R function is also regulated by various interacting proteins that modulate its signaling, trafficking, and stability (Fukunaga and Shioda, 2012; Fuxe et al., 2012). These D2R-associated protein complexes are therefore potential targets for new treatments. DISC1 is a susceptibility gene for schizophrenia and other psychiatric disorders, originally identified in a Scottish family with a high frequency of mental illness (Brandon and Sawa, 2011; Hodgkinson et al., 2004; Porteous et al., 2011; Taylor et al., 2003). DISC1 is a scaffold protein that interacts with many important signaling molecules including GSK-3 (Duan et al., 2007; Kvajo et al., 2008; Lee et al., 2011; Mao et al., 2009; Niwa et al., 2010; Pletnikov et al., 2008; Shen et al., 2008). As a result, 1302 Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

2 Neuron D2-DISC1 Complex and Antipsychotic Effects the role of DISC1 in mental illness involves a complex network of interacting proteins and pathways that has various functions depending on the developmental stage, signaling pathway, and brain region (Enomoto et al., 2009; Kim et al., 2009, 2012; Singh et al., 2010). There are several types of genetic mouse models for Disc1: the 129S6/SvEv strain, N-ethyl-N-nitrosourea (ENU)-induced mutants, sirna knockdown, and transgenics. The Disc1-L100P (334T/C) mutation produces abnormalities consistent with schizophrenia: enlarged lateral ventricles, abnormal cortical lamination, abnormal prepulse inhibition (PPI), disrupted latent inhibition, and decreased social interaction (Clapcote et al., 2007; Lee et al., 2011). The behavioral abnormalities are ameliorated by antipsychotic medications. In this study, we sought to identify D2R-interacting proteins that might be potential targets for new antipsychotic treatments. Since DISC1 is a scaffold protein implicated in schizophrenia, and since the D2R is the main target of antipsychotic drugs, we hypothesized that these two proteins interact to modulate the downstream GSK-3 signaling pathway. We also predicted an increased D2R-DISC1 interaction in postmortem brain from schizophrenia patients and in mouse models of schizophrenia. Having confirmed these predictions, we developed a peptide that disrupts the D2R-DISC1 interaction. This peptide exerts antipsychotic-like effects without inducing EPS and affects D2R signaling through GSK-3. Thus, we have demonstrated a strategy to achieve greater pharmacological specificity by targeting interactions between a neurotransmitter receptor and accessory proteins, rather than simply blocking the receptor itself. Our results could represent a more general approach to improve efficacy and reduce side effects for other medications that target neurotransmitter receptors. RESULTS D2R-DISC1 Complex Formation Is Enhanced in Postmortem Brain Tissue from Schizophrenia Patients We initially tested whether DISC1 forms a protein complex with D2R. As shown in Figure 1A (top panel), a D2R antibody, but not a dopamine D1 receptor (D1R) antibody, coimmunoprecipitated with DISC1 in proteins extracted from rat striatum, suggesting the existence of a D2R-DISC1 complex. The coimmunoprecipitation of D1R with D2R (Pei et al., 2010) but not with DISC1 verified both the efficiency and the specificity of the D2R antibody for immunoprecipitation. The specificity of the D2R antibody was also validated in D2R knockout mice (Figure S1A available online). Conversely, DISC1 antibody coimmunoprecipitated with D2R (Figure 1A, bottom panel), confirming the existence of a D2R-DISC1 complex in rat striatum. We hypothesized that D2R activation should alter D2R-DISC1 complex formation if the D2R-DISC1 interaction participates in D2R signaling. To investigate this possibility, we examined the D2R-DISC1 interaction using coimmunoprecipitation in rat primary striatal neuron cultures and in HEK293T cells expressing HA-D2R and GFP-DISC1, treated with/without a specific D2R agonist or antagonist. Activation of D2R with 10 mm quinpirole (30 min) significantly increased the D2R-DISC1 interaction in both striatal neurons (Figures 1B and 1C) and in transfected HEK293T cells (Figures S1B and S1C). The quinpirole effect could be blocked by haloperidol, a specific D2R antagonist. These data suggest that activation of D2R facilitates D2R- DISC1 complex formation. As a hyperdopaminergic state has been reported in schizophrenia (Seeman et al., 1976), we reasoned that the D2R- DISC1 complex may be increased in schizophrenia. We first tested our hypothesis by carrying out coimmunoprecipitation experiments on Disc1-L100P mice, as these mice display behaviors relevant to schizophrenia, including increased locomotor activity and disrupted PPI, all of which can be reversed by treatment with D2R antagonists such as haloperidol (Clapcote et al., 2007). As shown in Figures 1D and 1E, coimmunoprecipitation of DISC1 by the D2R antibody was significantly higher in Disc1-L100P mice versus wild-type (WT) mice (p < 0.01). There was no significant difference in direct immunoprecipitation of D2R between Disc1-L100P and WT mice (Figures 1D and 1F). To obtain direct evidence demonstrating that the L100P mutation alters the D2R-DISC1 interaction, we measured D2R-DISC1 complex levels in HEK293T cells coexpressing HA-D2R with either Disc1-L100P or WT DISC1. As shown in Figures S1D and S1E, coimmunoprecipitation results show that D2R-DISC1 interaction is significantly higher in D2R/Disc1-L100P coexpressing cells compared to D2R/WT- DISC1 cells. We then measured D2R-DISC1 complex levels in postmortem striatum samples from the Stanley Medical Research Institute, consisting of 15 schizophrenia and 15 unaffected control subjects matched for age, sex, race, postmortem interval, ph, side of brain, and mrna quality (Torrey et al., 2000). Demographic information including antipsychotic use is shown in Tables S1 and S2. Equal amounts of protein from each sample were incubated with anti-d2r antibody, and the precipitated proteins were immunoblotted with either DISC1 or D2R antibody. Each western blot included five samples from each group, and results for each sample are presented as the percentage of the mean of five control samples on the same blot. As shown in Figures 1G and 1H, coimmunoprecipitation of DISC1 by the D2R antibody was significantly increased in schizophrenia postmortem brain samples compared to control (n = 15, p < 0.05). The levels of directly immunoprecipitated D2R were not significantly different between the two groups (Figures 1G and 1I). One potential confounding factor in this postmortem analysis is antipsychotic treatment, since most (14/15) of the schizophrenia patients were treated with antipsychotics (Table S1). To address this issue, we tested whether haloperidol enhances the D2R-DISC1 interaction in Disc1-L100P mice. As shown in Figures S1F and S1G, both acute (1 mg/kg subcutaneously [s.c.]) and chronic (1 mg/kg/day s.c. 14 days) haloperidol treatment significantly reduced D2R-DISC1 interaction in Disc1- L100P mice. Thus, antipsychotic treatment is unlikely to be the cause of increased D2R-DISC1 complex levels. DISC1 Facilitated D2R-Induced Reduction of GSK-3 Phosphorylation To investigate the function of the D2R-DISC1 complex, we sought to determine whether DISC1 played a role in D2R-mediated GSK-3 signaling. Prior studies on D2R-mediated modulation of GSK-3 phosphorylation were carried out exclusively in Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1303

3 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 1. Regulation of D2R-DISC1 Complex Formation (A) In rat brain striatal lysate, D2R antibody, but not D1R antibody or IgG (negative control), coimmunoprecipitated with DISC1 (top panel); DISC1 antibody coimmunoprecipitated with D2R (bottom panel). (B) Coimmunoprecipitation shows that activation of D2R facilitates D2R-DISC1 complex formation in cell lysate from striatal neurons pretreated with quinpirole (Quin) or haloperidol (halo)/quin. (C) Densitometric analysis of DISC1 coimmunoprecipitated by D2R from cultured rat striatal neurons pretreated with quinpirole or halopirodol/quinpirole. *p < 0.05 compared to controls. #p < 0.05 compared to the Quin group. One-way ANOVA. n = 3. (D) Coimmunoprecipitation shows higher D2R-DISC1 complex levels in Disc1-L100P mouse striatum compared to WT mice. (E and F) Densitometric analysis of the level of DISC1 and D2R. **p < 0.01, n = 5/6, t test. (G I) Coimmunoprecipitation shows significantly more D2R-DISC1 complex in striatal postmortem brain samples from schizophrenia (SZ) patients compared to controls (CTRL). Representative Western blot of DISC1 and D2R precipitated by D2R antibody (G). The intensity of each protein band for DISC1 (H) and D2R (I) was quantified by densitometry. Results for each sample are presented as the percentage of the mean of the control samples on the same blot. *p < 0.05, n = 15, t test. All data are shown as means ± SEM. See also Figure S1 and Tables S1 and S Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

4 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 2. DISC1 Modulates D2R-Mediated GSK-3 Phosphorylation (A) Western blot analysis shows that activation of D2R with quinpirole reduces GSK-3a/b (Ser21/9) phosphorylation in HEK293T cells expressing D2R and DISC1 but not in HEK293 cells expressing D2R with pcdna3. GSK-3a/b was used as a loading control. (B and C) Densitometric analysis of phosphorylated GSK-3a/b (Ser21/9). *p < 0.05, n = 6, t test. (D) Western blot analysis shows that activation of D2R reduces GSK-3a/b (Ser21/9) phosphorylation in rat striatal neurons. The D2R effect can be blocked by the D2R antagonist haloperidol. (E) Densitometric analysis of the intensity of phosphorylated GSK-3a/b (Ser21/9). *p < 0.05, n = 4, One-way ANOVA. (F) Western blot analysis shows that activation of D2R reduces AKT (Thr308) phosphorylation in rat striatal neurons. The D2R effect can be blocked by the D2R antagonist haloperidol. AKT was used as a loading control. (G) Densitometric analysis of the intensity of phosphorylated AKT (Thr308). *p < 0.05, n = 3, One-way ANOVA. (H) Representative western blot shows that GSK-3a/b (Ser21/9) phosphorylation was reduced in postmortem striatal tissue of schizophrenia patients (SZ) compared to control subjects (control). (I and J) Densitometric analysis of the intensity of phosphorylated GSK-3a/b. t test, n = 15, *p < All data are shown as means ± SEM. See also Figures S2 S4. brain tissue expressing both D2R and DISC1 (Beaulieu et al., 2007, 2009; Beaulieu, 2012). Thus, an initial step to investigate the role of DISC1 in D2R-mediated GSK-3 signaling is to compare D2R-induced changes in GSK-3 phosphorylation in HEK293T cells coexpressing DISC1 and D2R to HEK293T cells expressing D2R alone. As shown in Figures 2A 2C, preincubation with 10 mm quinpirole (30 min) significantly reduces GSK-3 a/b Ser-21/9 phosphorylation, but not GSK3a/b (Y-279/216) phosphorylation, in HEK293T cells coexpressing D2R and DISC1, while quinpirole has no such an effect in HEK293T cells Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1305

5 Neuron D2-DISC1 Complex and Antipsychotic Effects expressing D2R alone. The D2R-mediated reduction of GSK-3 a/b Ser-21/9 phosphorylation is also observed in rat striatal neurons (Figures 2D and 2E), which is consistent with previous observations in mouse striatum (Beaulieu et al., 2007, 2009; Beaulieu, 2012). In contrast, the D2R-mediated reduction of GSK-3 a/b Ser-21/9 phosphorylation is not observed in D2R knockout mice (Figures S2A and S2B). Furthermore, activation of D2R failed to modulate GSK-3 a/b Ser-21/9 phosphorylation in rat cortical neurons (Figures S2C and S2D), although coimmunoprecipitation of DISC1 by the D2R antibody is observed (Figure S2E). To further confirm the role of DISC1 in the modulation of D2R effects on GSK-3 phosphorylation, we knocked down the expression of DISC1 using DISC1-shRNA as previously described (DISC1-shRNA from Dr. Guo-li Ming at Johns Hopkins University) (Duan et al., 2007; Zhou et al., 2013). The ability of DISC1-shRNA to knock down the expression of DISC1 in rat striatal neurons was confirmed (Figure S3A). We then measured GSK-3 a/b Ser-21/9 phosphorylation. As shown in Figures S3B S3D, DISC1 shrna treatment significantly decreased the ability of D2R to reduce GSK-3 a/b Ser-21/9 phosphorylation in rat striatal neurons. More importantly, the effect of DISC1 shrna can be rescued by the expression of shrna-resistant DISC1 (DISC1 R from Dr. Guo-li Ming at Johns Hopkins University) (Figures S3E S3G). As Akt phosphorylation is a critical step in D2R-mediated GSK-3 signaling, we also tested whether activation of D2R would alter Akt phosphorylation in rat striatal neurons. As shown in Figures 2F and 2G, quinpirole stimulation decreased Akt phosphorylation at Thr-308 but not at Ser-473 site, consistent with previous studies (Beaulieu et al., 2007). In addition, DISC1 shrna treatment also decreased the ability of D2R to reduce Akt phosphorylation at Thr-308 site in rat striatal neurons (Figures S4A S4C). Furthermore, we tested whether DISC1 is required for D2R-mediated camp accumulation, the classic signaling pathway for D2R. HEK293T cells coexpressing DISC1 and D2R and HEK293T cells expressing D2R alone were treated with forskolin (1 mm, 10 min), with or without pretreatment with 10 mm quinpirole (30 min). As shown in Figure S4D, there is no difference in D2R-mediated camp accumulation between these two groups. Thus, DISC1 may selectively modulate D2R-mediated activation of the GSK-3 signaling pathway. GSK3a/b Phosphorylation Is Decreased in Postmortem Brain Tissue from Schizophrenia Patients Previous studies have shown that antipsychotics target GSK-3 signaling (Beaulieu et al., 2009), implicating GSK-3 in the treatment of schizophrenia. To investigate the functional consequence of elevated DISC1-D2R complex levels in schizophrenia, we measured GSK3a/b phosphorylation in these samples compared to controls. As shown in Figures 2H 2J, GSK3a/b Ser-21/9 phosphorylation is significantly decreased in postmortem striatal tissue from schizophrenia patients compared to control subjects (p < 0.05, n = 15). This suggest that enhanced DISC1-D2R complex formation may lead to increased GSK-3 signaling as part of the mechanism underlying the pathophysiology of schizophrenia. Development of an Interfering Peptide that Is Able to Disrupt DISC1-D2R Complex If the D2R-DISC1 interaction plays a role in D2R-mediated modulation of GSK-3 signaling, we predicted that disruption of the interaction should abolish D2R-mediated modulation of GSK-3 signaling. More importantly, if the enhanced D2R-DISC1 complex indeed plays a role in the pathophysiology of schizophrenia, disrupting the complex should have antipsychotic effects. To develop an interfering peptide, we identified the regions of the D2R that are essential for D2R-DISC1 complex formation. We initially used GST-fusion proteins encoding fragments of intracellular domains of D2R to affinity purify DISC1 in rat striatal extracts. GST-fusion proteins encoding the carboxyl tail (CT) of D2R (D2 CT :T 428 -C 443 ) and the third intracellular loop (IL3) of D2R (D2 IL3 :K 211 -Q 373 ) were made and used for affinity purification. As shown in Figure 3A, GST-D2 IL3, but not GST-D2 CT, precipitates DISC1, suggesting that D2 IL3 is involved in the DISC1-D2R interaction. Using the same strategy, we further dissected the D2 IL3 into smaller fragments (Figure S5A) and concluded that D2 IL3-1-1 [K 211 -T 225 ] is the region interacting with DISC1 (Figures 3B and 3C). Furthermore, we found that in-vitrotranslated full-length [ 35 S]-DISC1 probe specifically hybridized with the K 211 -T 225 fragment (Figure 3D), suggesting that the D2R and DISC1 interact with each other directly. Using the same experimental approach, we identified that the NT region of DISC1 is responsible for its interaction with D2R, because the [ 35 S]-DISC1 NT probe, but not the [ 35 S]-DISC1 CT probe, specifically hybridized with the D2 IL3-1-1 [K 211 -T 225 ] fragment (Figure 3E). Based on the sequence of K 211 -T 225, we synthesized a peptide: D2pep[K 211 -T 225 ] (KIYIVLRRRRKRVNT). If the K 211 -T 225 region is essential for D2R to interact with DISC1, application of D2pep[K 211 -T 225 ] should disrupt the D2R-DISC1 interaction by competing with D2R for DISC1. As expected, D2pep[K 211 -T 225 ] (10 mm), but not the scrambled control peptide (D2pep-sc) (VLRKTRIRRYKIRNV), inhibited D2R-DISC1 complex formation in rat striatum (Figure 3F) and agonist-induced increase in D2R-DISC1 complex formation in HEK293T cells coexpressing HA-D2R and GFP-DISC1 (Figure 3G). The ability of D2pep to inhibit D2R-DISC1 complex formation was also observed in rat striatal neurons pretreated with TAT-D2pep (10 mm, 1 hr) (Figure 3H). To allow D2pep[K 211 -T 225 ] to penetrate into cells, it was fused to the cell membrane transduction domain of the human immunodeficiency virus-type 1 TAT protein, as previously described (Lee et al., 2007; Pei et al., 2010). Thus, we concluded that the K 211 -T 225 region of D2R is critical for D2R- DISC1 complex formation and that the D2pep[K 211 -T 225 ] peptide disrupts the D2R-DISC1 complex. Disruption of DISC1-D2R Interaction Blocks D2R- Mediated Modulation of GSK-3a/b Phosphorylation Since D2pep[K 211 -T 225 ] disrupts the D2R-DISC1 interaction, we tested whether pretreatment with D2pep[K 211 -T 225 ] blocked D2R-mediated modulation of GSK-3a/b phosphorylation. The ability of D2R activation to modulate GSK-3a/b Ser-21/9 phosphorylation was significantly inhibited by preincubating with TAT-D2pep[K 211 -T 225 ] peptide (10 mm, 1 hr), but not with the scrambled peptide TAT-D2pep-sc in both HEK293T cells coexpressing D2R and DISC1 (Figures S5B and S5C), and in rat 1306 Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

6 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 3. DISC1 Forms a Complex with D2R via the K 211 -T 225 Region of the IL3 of D2R (A) Western blot showing that GST-D2 IL3, but not GST-D2 CT, can pull down DISC1 from rat striatal tissue. (B) Western blot showing that GST-D2 IL3-1, but not GST-D2 IL3-2 or GST-D2 IL3-3, can pull down DISC1 from rat striatal tissue. (C) Western blot showing that GST-D2 IL3-1-1, but not GST-D2 IL3-1-2, GST-D2 IL3-1-3 or GST-D2 IL3-1-4, can pull down DISC1 from rat striatal tissue. (D) GST-D2 IL3-1 and GST- D2 IL3-1-1 directly bind to DISC1 in vitro. GST-D2 IL3-1, GST-D2 IL3-1-1, GST-D2 IL3-1-2, GST-D2 IL3-1-3, and GST-D2 IL3-1-4 were incubated with [ 35 S]-DISC1, and the precipitated proteins were subjected to SDS-PAGE. (E) GST-D2 IL3-1-1 directly binds to DISC1 NT in vitro. GST-D2 IL3-1, GST-D2 IL3-1-1, GST-D2 IL3-1-2, GST-D2 IL3-1-3, and GST-D2 IL3-1-4 were incubated with [ 35 S]- DISC1 NT or [ 35 S]-DISC1 CT, and the precipitated proteins were subjected to SDS-PAGE. (F) Coimmunoprecipitation shows that D2pep[K 211 -T 225 ], but not D2pep-sc, is able to disrupt the DISC1-D2R complex in rat brain striatal lysate. (G) Coimmunoprecipitation shows that D2pep[K 211 -T 225 ] (D2), but not D2pep-sc (Sc), is able to block quinpirole facilitated DISC1-D2R complex formation in cell lysate from HEK293T cells expressing GFP-DISC1 and HA-D2R. (H) Coimmunoprecipitation shows that D2pep[K 211 -T 225 ], but not D2pep-sc, is able to disrupt the DISC1-D2R complex in striatal neurons. See also Figure S5A. striatal primary cultures (Figures 4A and 4B). These results suggest that the D2R-mediated modulation of GSK-3a/b phosphorylation is dependent on D2R- DISC1 interaction. As well, the D2R-mediated modulation of Akt phosphorylation can be blocked by the TAT-D2pep [K 211 -T 225 ], but not the scrambled peptide TAT-D2pep-sc in rat striatal neurons (Figures 4C and 4D). Furthermore, we found that the ability of D2R activation to inhibit forskolin-induced camp accumulation was not changed by preincubating with TAT-D2pep[K 211 -T 225 ] or TAT- D2pep-sc in HEK293T cells (Figure S5D). Similar results were also observed in striatal neurons with lentivirus-mediated overexpression of DISC1 (Figure S5E). These data provide further evidence that the D2R-DISC1 interaction is selectively involved in D2R-mediated activation of the GSK-3 signaling pathway, which is also specifically targeted by TAT-D2pep[K 211 -T 225 ]. We next tested whether GSK3a/b phosphorylation is altered in Disc1-L100P mice and, if so, whether disruption of the D2R-DISC1 complex can rescue the changes in GSK3a/b Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1307

7 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 4. Disruption of D2R-DISC1 Interaction Blocked D2R-Mediated Modulation of GSK-3 Signaling (A) Western blot showing that TAT-D2pep[K T 225 ], but not TAT-D2pep-sc, blocked quinpiroleinduced modulation of GSK-3 phosphorylation in rat striatal neurons. (B) Densitometric analysis of phosphorylated GSK-3a/b in striatal neurons. *p < 0.05, n = 3, twoway ANOVA. (C) Western blot showing that TAT-D2pep[K T 225 ], but not TAT-D2pep-sc, blocked quinpiroleinduced modulation of AKT phosphorylation in rat striatal neurons. (D) Densitometric analysis of phosphorylated AKT. *p < 0.05, n = 3, two-way ANOVA. (E) Western blot showing that TAT-D2pep[K T 225 ], but not TAT-D2pep-sc, blocked the decreased GSK-3a/b (Ser21/9) phosphorylation in Disc1-L100P mice. (F) Densitometric analysis of phosphorylated GSK-3a/b. n = 3, *p < 0.05, **p < 0.01 compared to WT, ##p < 0.01, ###p < compared to L100P+TAT-D2pep-sc, one-way ANOVA. All data are shown as means ± SEM. See also Figures S5B S5E and S6. phosphorylation observed in the striatum of these mice. As shown in Figures 4E and 4F, GSK3a/b phosphorylation is significantly decreased in Disc1-L100P mice. More importantly, intraperitoneal (i.p.) injection of TAT-D2pep[K 211 -T 225 ] (3 nmol/g), but not the scrambled TAT-D2pep (TAT-D2pep-sc), restored GSK3a/b phosphorylation in Disc1-L100P mice. Similar effects were also observed in Disc1-L100P mice with acute (1 mg/kg s.c.) (Figures S6A and S6B) and chronic (1 mg/kg s.c.14 days) (Figures S6C and S6D) haloperidol treatment, as well as chronic TAT-D2pep[K 211 -T 225 ] treatment (3 nmol/g i.p. daily for 7 days) (Figures S6E and S6F). D2R-DISC1 Interaction Inhibits Agonist-Induced D2R Internalization We have shown that D2R-DISC1 complex formation facilitates D2R-mediated GSK-3 signaling. As a GPCR, control of membrane expression is one of the major mechanisms regulating D2R function. Thus, we examined whether coexpression of DISC1 changed agonist-induced D2R internalization. We used a biotinylation assay to selectively label internalized D2R. HEK293T cells expressing D2R with or without DISC1 were first labeled with sulfo-nhs-ss-biotin and then treated with 10 mm quinpirole (30 min). Subsequently, all cell surface biotin was cleaved by using glutathione (GSH), leaving only the endocytosed proteins labeled with biotin. The ability of GSH to cleave all cell surface biotin was confirmed with a negative control. The proteins were extracted and purified using streptavidin beads to harvest internalized biotinylated proteins. As shown in Figures 5A and 5B, western blots of streptavidin-purified protein extracts show less biotinylated D2R in cells coexpressing DISC1 and pretreated with quinpirole compared to cells expressing D2R alone, suggesting that DISC1 may inhibit agonist-induced D2R internalization. There was no significant difference in total D2R expression. Furthermore, pretreatment with TAT-D2pep [K 211 -T 225 ], but not TAT-D2pep-sc, blocked DISC1-mediated inhibition of agonist-induced D2R internalization (Figures 5C and 5D), suggesting that DISC1 inhibits agonist-induced D2R internalization through the D2R-DISC1 interaction. Similar results were also obtained in rat striatal neurons (Figures 5E and 5F). D2R-DISC1 Interaction Blocks b-arrestin 2-clathrin-AP2 Complex Formation and Facilitates b-arrestin2-pp2a- Akt Complex Formation Agonist stimulation of the D2R can lead to both receptor internalization and activation of the GSK-3 pathway; both are potentially 1308 Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

8 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 5. DISC1 Inhibits Agonist-Induced D2R Internalization (A) Western blot using anti-ha showing that quinpirole-induced internalized HA-D2R proteins (biotinylated cell membrane proteins) are decreased in HEK293T cells expressing HA-D2R with DISC1 compared to cells expressing HA-D2R alone. ( ) control: biotin is cleaved before agonist stimulation. (+) control: biotin is not cleaved after biotinylation. (B) Densitometric analysis of D2R biotinylation assay (relative to untreated cells, mean ± SEM twoway ANOVA *p < 0.05, n = 3). (C) Western blot using anti-ha showing TAT-D2pep [K 211 -T 225 ] (D2), but not TAT-D2pep-sc (Sc), blocks DISC1-mediated inhibition of quinpirole-induced D2R internalization in HEK293T cells expressing HA-D2R with or without DISC1. (D) Densitometric analysis of D2R biotinylation assay (relative to the untreated cells, mean ± SEM, two-way ANOVA *p < 0.05, n = 3). (E) Western blot using anti-d2r showing TAT- D2pep[K 211 -T 225 ] (D2), but not TAT-D2pep-sc (Sc), blocks DISC1-mediated inhibition of quinpiroleinduced D2R internalization in striatal neurons. (F) Densitometric analysis of D2R biotinylation assay (relative to the group treated with scrambled peptide without quinpirole, mean ± SEM, two-way ANOVA **p < 0.01, ***p < 0.001, n = 3). initiated by D2R-b-arrestin2 complex formation. We found that the D2R-DISC1 interaction facilitated activation of the GSK-3 pathway and downregulates D2R internalization. Thus, we hypothesized that the D2R-DISC1 interaction facilitates formation of the b-arrestin2-pp2a-akt protein complex and inhibits formation of the b-arrestin2-clathrin-ap2 complex. The b-arrestin2-pp2a-akt complex activates GSK-3 signaling, while b-arrestin2-clathrin-ap2 is required for D2R endocytosis in response to agonist activation of D2R. We initially assessed by coimmunoprecipitation whether the presence of DISC1 affects b-arrestin2 recruitment to D2R after agonist stimulation. As shown in Figure 6A, we found that the presence of DISC1 did not alter the D2R-b-arrestin2 interaction in HEK293T cells. We obtained similar results in striatal neurons; agonist stimulation enhanced D2R-b-arrestin2 interaction, which was unaffected by TAT-D2pep[K 211 -T 225 ] (Figure 6B). We then examined the effect of DISC1 on the formation of the b-arrestin2-clathrin-ap2 and b-arrestin2-pp2a-akt protein complexes. As shown in Figure 6C (top panel), activation of D2R enhanced D2R coupling with clathrin, and the effect was blocked by DISC1. In contrast, coexpression of DISC1 dramatically enhanced b-arrestin2 coupling with PP2A, which was further enhanced by D2R activation (Figure 6D, top panel). These DISC1-associated effects were consistently blocked by TAT-D2pep[K 211 -T 225 ], but not the control TAT-D2pep-sc (Figure 6C, bottom panel; Figure 6D, middle and bottom panel). Similar results were also obtained in rat striatal slices (Figures 6E and 6F). Therefore, it is likely that the formation of the DISC1-D2R complex selectively favors D2R-mediated GSK-3 signaling, which would lead to upregulation of the GSK3 pathway. Disruption of DISC1-D2R Interaction Reverses Hyperactivity and the Disruption of PPI in Disc1-L100P Mice Disc1-L100P mutant mice have relevant behavioral abnormalities including increased locomotor activity and disrupted PPI that can be reversed by treatment with antipsychotic medications (Clapcote et al., 2007). If the higher D2R-DISC1 complex levels in Disc1-L100P mice are part of the pathophysiology of their abnormal behavior, then disruption of the DISC1-D2R complex with TAT-D2pep[K 211 -T 225 ] should reverse these abnormalities. Prior to the behavior tests, we first confirmed the ability of TAT-D2pep[K 211 -T 225 ] to disrupt the D2R- DISC1 complex in Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1309

9 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 6. DISC1 Inhibits D2R-Clathrin Complex Formation but Promotes D2R-PP2A Complex Formation (A) Coimmunoprecipitation using anti-ha (for D2R) shows that DISC1 has no effect on quinpirolefacilitated FLAG-b-arrestin 2-HA-D2R complex formation in HEK293T cells transfected with FLAGb-arrestin 2 and HA-D2R with/without GFP-DISC1. (B) Coimmunoprecipitation shows that TAT-D2pep [K 211 -T 225 ] (D2) and TAT-D2pep-sc (Sc) has no effect on quinpirole-facilitated b-arrestin 2-D2R complex formation in rat striatal neurons. (C) Coimmunoprecipitation using anti-ha shows that the presence of DISC1 inhibits quinpirolefacilitated clathrin-ha-d2r complex formation (top panel), while TAT-D2pep[K 211 -T 225 ] (D2) but not TAT-D2pep-sc (sc) blocks the inhibitory effect of DISC1 (bottom panel) in HEK293T cells expressing HA-D2R and Flag-b-arrestin 2 with or without DISC1. (D) Coimmunoprecipitation using anti-flag (for b-arrestin 2) shows that the presence of DISC1 dramatically enhanced b-arrestin 2 coupling with PP2A-B and activation of D2R further enhanced b-arrestin 2 coupling with PP2A-B (top panel), while TAT-D2pep[K 211 -T 225 ] (D2), but not TAT- D2pep-sc (sc), blocks the enhancement induced by DSIC1 and quinpirole (middle and bottom panel) in HEK293T cells expressing HA-D2R and FLAG-b-arrestin 2 with or without DISC1. (E) Coimmunoprecipitation using anti-b-arrestin 2 shows that TAT-D2pep[K 211 -T 225 ] (D2), but not TAT-D2pep-sc (Sc), facilitates b-arrestin 2-clathrin complex formation in striatal neurons. (F) Coimmunoprecipitation using anti-b-arrestin 2 shows that TAT-D2pep[K 211 -T 225 ] (D2), but not TAT-D2pep-sc (Sc), inhibits b-arrestin 2- PP2A-B complex formation with or without agonist stimulation in striatal neurons. Disc1-L100P mice (Figures S7A S7D). We also measured the bioavailability of TAT-D2pep[K 211 -T 225 ] in plasma and brain of Disc1-L100P mice at 30, 90, and 1 min after the i.p. injection (Table S3). As shown in Figure 7A, Disc1-L100P mice treated with saline or TAT-D2pep-sc were hyperactive compared to WT mice. This hyperactivity was not observed in Disc1-L100P mice treated with TAT-D2pep[K 211 -T 225 ]. Furthermore, Disc1- L100P mice showed reduced PPI at all prepulse intensities compared to WT mice. This reduced PPI was restored by TAT- D2pep[K 211 -T 225 ] but not TAT-D2pep-sc (Figure 7B). TAT- D2pep[K 211 -T 225 ] had no effect on baseline startle response in Disc1-L100P and WT mice (Table S4). These data suggest that TAT-D2pep[K 211 -T 225 ] is able to rescue abnormal locomotor activity and PPI, when it is observed, in Disc1-L100P mutant mice. Disruption of the D2R-DISC1 Interaction Reverses Amphetamine-Induced Hyperactivity and the Disruption of PPI Induced by Apomorphine in Both Mice and Rats Amphetamine induces hyperactivity primarily by releasing DA and activating the D2R, and this is often used as a behavioral test relevant to schizophrenia (Featherstone et al., 2007). Thus, we examined the effects of TAT-D2pep[K 211 -T 225 ] on amphetamine-induced hyperactivity and apomorphine-induced PPI deficits. As shown in Figure 7C, TAT-D2pep[K 211 -T 225 ] (1.5 nmol/g i.p.) and haloperidol blocked amphetamine-induced hyperactivity in mice. The scrambled peptide TAT-D2pep-sc was inactive. Neither TAT-D2pep[K 211 -T 225 ] nor TAT-D2pep-sc affected baseline locomotor activity in mice (Figure S8A). PPI can be disrupted by the D1R/D2R agonist apomorphine. Antipsychotics that have affinity for D2R such as haloperidol prevent the apomorphine-induced deficits (Geyer et al., 2001; Ralph and Caine, 2005). As shown in Figure 7D, apomorphine disrupted PPI in comparison with saline (p < 0.05), and this effect was blocked by haloperidol. TAT-D2pep[K 211 -T 225 ] rescued apomorphine-disrupted PPI in mice while TAT-D2pep-sc had no effect. Neither peptide altered baseline startle response (Table S5). TAT-D2pep[K 211 -T 225 ] (intracerebroventricular [i.c.v.] 4 nmol) also blocked amphetamine-induced hyperactivity and apomorphine-induced disruption of PPI in rats (Figures 7E and 7F). Neither TAT-D2pep[K 211 -T 225 ] nor TAT-D2pep-sc altered baseline locomotor activity or baseline startle response in rats (Figure S8B; Table S6) Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

10 Neuron D2-DISC1 Complex and Antipsychotic Effects TAT-D2pep[K 211 -T 225 ] Does Not Induce Catalepsy Although the interfering peptide we developed is able to ameliorate the schizophrenia-like behaviors, it is not of great clinical utility if it also has side effects similar to current antipsychotics. Because catalepsy is a strong predictor of EPS from antipsychotics, we tested whether disruption of the DISC1-D2R interaction induces catalepsy in rats. As shown in Figure 7G, haloperidol (0.5 mg/kg s.c.) induced catalepsy, but TAT-D2pep[K 211 -T 225 ] (40 nmol, i.c.v.) did not, even at doses ten times higher than doses that induced the behavioral responses presented above. These data suggest that disruption of the DISC1-D2R interaction may be a way to modulate D2R activity and exert antipsychotic effects without causing acute EPS. DISCUSSION In summary, we have identified a possible target for the development of antipsychotic treatment. This conclusion is based on: (1) finding a protein complex composed of D2R and DISC1 that facilitates D2R-mediated GSK-3 signaling, (2) an enhanced D2R-DISC1 complex and decreased GSK-3 phosphorylation in postmortem brain tissue from schizophrenia patients and Disc1-L100P mice, and (3) developing an interfering peptide that disrupts the D2R-DISC1 complex and reduces schizophrenia-related behaviors without inducing catalepsy. Our proposed model to integrate all these experimental observations is shown in Figure 8. We posit that activation of D2R recruits b-arrestins to D2R, which under physiological conditions has two outputs that are in equilibrium: (1) in the presence of DISC1, a b-arrestin-pp2a-akt signaling complex forms, leading to the activation of GSK-3 signaling, or (2) in the absence of DISC1, a b-arrestin-clathrin-ap2 internalization complex forms, leading to D2R internalization. In schizophrenia, however, excessive dopamine release (Laruelle et al., 1996) combined with elevated D2R-DISC1 complex levels would lead to overactivation of the GSK3 pathway and a relative decrease in D2R internalization. Thus, disrupting the DISC1-D2R complex with our peptide may exert antipsychotic effects by restoring this equilibrium. We found increased D2R-DISC1 interaction and decreased GSK-3 phosphorylation in postmortem brain from schizophrenia patients. A common confounding factor in such postmortem brain samples is that of antipsychotic treatment. In our sample, 14 of the 15 schizophrenia patients received antipsychotic treatment (Table S1). However, we argue that the higher DISC1-D2R complex levels and lower GSK-3 phosphorylation are not consequences of long-term antipsychotic treatment, because the antipsychotic haloperidol blocks the increase in DISC1-D2R complex formation (Figures S1F and S1G) and the decrease in GSK-3 phosphorylation (Figures S6A S6D) in Disc1-L100P mice, where a hyperdopaminergic state is reported. Thus, antipsychotic medication has the opposite effect to the abnormalities seen in the schizophrenia patients. We also found that DISC1 facilitates D2R-mediated GSK-3 signaling through the D2R-DISC1 interaction. Previous studies have shown that increased dopamine signaling and GSK3 activity are associated with psychosis (Laruelle et al., 1996; Beaulieu, 2012). Thus, an enhanced D2R-DISC1 interaction, leading to increased GSK-3 activity, could contribute to psychotic symptoms in schizophrenia. We are not claiming that the D2R- DISC1 interaction itself is pathological, since it is also present in unaffected human control subjects. Instead, we propose that the hyperdopaminergic state seen in schizophrenia leads to excessive D2R-DISC1 complex formation and abnormally increased GSK-3 activity. Both could contribute to psychotic symptoms. We chose the striatum to characterize the functional role of the D2R-DISC1 interaction because of the following: (1) all the studies on the D2R-mediated GSK-3 pathway from Dr. Caron s group were conducted in striatum, including the effect of antipsychotics; (2) the vast majority of neurons (95%) within this brain structure are medium spiny neurons (MSNs) (Nicola et al., 2000; Surmeier et al., 2007), which are a principal site of action of antipsychotic drugs because they express the highest levels of D2R (Kellendonk et al., 2006); (3) striatal D2R occupancy is strongly correlated with antipsychotic response (Agid et al., 2007); and (4) MSNs participate in many of the cognitive functions and behaviors that are altered in schizophrenia (Simpson et al., 2010). DISC1 is a scaffold protein that can modulate neurodevelopment and neurotransmitter signaling in a number of brain regions, through a large interactome (Duan et al., 2007; Kang et al., 2011; Kim et al., 2009, 2012; Mao et al., 2009; Millar et al., 2005; Singh et al., 2010). For example, DISC1 knockdown reduces interactions with NDEL1 to result in premature cell cycle exit and mispositioning of adult neurons in the adult hippocampus. Similarly, in the embryonic cerebral cortex, DISC1 knockdown reduces GSK3b, Wnt, and b-catenin interactions, also resulting in premature cell-cycle exit and abnormal neuron positioning. However, neuron migration is regulated by DISC1 interactions with Dixdc1 and NDEL, not GSK3. Furthermore, development of the cortex is also regulated by phosphorylation of DISC1, which regulates the transition from neuronal progenitor proliferation to neuron migration (Ishizuka et al., 2011). These examples illustrate that DISC1 has multiple effects in different pathways. Thus, the D2R-DISC1 interaction may be just one of many pathways through which DISC1 affects psychiatric illness. Disc1-L100P mutant mice are hyperactive, with reduced PPI compared to WT mice (Clapcote et al., 2007). In the present study we have shown that disrupting the D2R-DISC1 complex with TAT-D2pep[K 211 -T 225 ] reversed these behavioral deficits and blocked the reduction of GSK-3 phosphorylation. We are aware of the fact that TAT-D2pep[K 211 -T 225 ] increases the basal phosphorylation of GSK3 (Figures 4E and 4F), which suggests that D2pep[K 211 -T 225 ] may inhibit endogenous dopamine tone. In addition, hyperactivity induced by amphetamine, or disruption of PPI induced by apomorphine, were reversed by TAT-D2pep [K 211 -T 225 ] in normal mice and rats. The effects of TAT-D2pep [K 211 -T 225 ] were specific to the particular amino acid sequence, since the scrambled form of the peptide was inactive in all tests. In these behavioral experiments the effects of TAT- D2pep[K 211 -T 225 ] were similar to those observed with the antipsychotic haloperidol, except that TAT-D2pep[K 211 -T 225 ] did not induce catalepsy. Overall, these results indicate that interfering with the D2R-DISC1 complex using TAT-D2pep[K T 225 ] produces an antipsychotic-like response comparable to haloperidol but devoid of EPS. Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1311

11 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 7. Disruption of D2R-DISC1 Interaction Blocks Locomotor Hyperactivity and Restores PPI (A) Effect of TAT-D2pep[K 211 -T 225 ] on distance traveled (cm) in 30 min in the open-field. Disc1-L100P / mutant mice (n = 9/16) and WT littermates (n = 8/8) were treated with TAT-D2pep[K 211 -T 225 ] or TAT-D2pep-sc (3 nmol/g i.p.) 30 min before testing. ANOVA detected a main effect of genotype (F(1, 49) = 6.35; p < 0.05), peptide (F(1, 49) = 18.95; p < 0.001), testing interval (F(5, 245) = 63.07; p < 0.001), and their interactions (p s < ). # p < 0.05; ### p < compared to TAT-D2pep-sc-treated WT mice; *p < 0.05; ***p < compared to TAT-D2pep-sc-treated mice within each genotype. (B) Effect of TAT-D2pep[K 211 -T 225 ] (3 nmol/g i.p.) on PPI in Disc1-L100P (n = 8/8) and WT (n = 8/8) mice. The TAT-D2pep[K 211 -T 225 ] reversed PPI deficits in Disc1- L100P mutants at all four prepulses. ANOVA detected a main effect of genotype (F(1, 42) = 21.9; p < 0.001), peptide (F(1, 42) = 40.3; p < 0.001), prepulses (F(3, 126) = 24.6, p < 0.001), and gene x peptide interactions (F(2, 42) = 22.3, p < 0.001) on PPI. # p < compared to TAT-D2pep-sc-treated WT mice; ***p < compared to TAT-D2pep-sc-treated Disc1-L100P mutant mice. (legend continued on next page) 1312 Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

12 Neuron D2-DISC1 Complex and Antipsychotic Effects Figure 8. Schematic of Interactions between D2R and DISC1 Stimulation of D2R leads to recruitment of b-arrestins, which provide a balanced platform under physiological conditions for either the formation of the b-arrestin-pp2a-akt signaling complex in the presence of DISC1 or the b-arrestin-clathrin-ap2 internalization complex in the absence of DISC1. Under pathological conditions, such as excess stimulation of D2R in schizophrenia, the GSK3 pathway is overactive, while the D2R internalization is downregulated. Antipsychotic drugs have often been divided into typical and atypical categories, with clozapine sometimes considered distinct due to efficacy in refractory schizophrenia (Geddes et al., 2000). Typical antipsychotics are D2R antagonists, while atypical antipsychotics block D2R in combination with 5HT2A receptors. D2R signaling is the primary target for antipsychotics to exert their therapeutic effects, since no antipsychotics exist that do not interact with D2Rs. However, D2Rs have a very complex pattern of signal transduction, including both G proteindependent and G protein-independent Akt-GSK-3 pathways (b-arrestin 2-dependent) (Missale et al., 1998; Beaulieu et al., 2009). Thus, blockade of D2R with existing antipsychotics will shut down all of these downstream signaling pathways, only some of which may mediate the desired antipsychotic effect, while others may be essential in maintaining normal brain function. Our strategy is to identify antipsychotic drug targets by disrupting only some aspects of D2R signaling. Thus, antipsychotic effects could result from a more focused alteration of D2R signaling, which would therefore have fewer side effects. Antipsychotic drugs (including haloperidol, clozapine, and olanzapine) strongly block agonist-dependent recruitment of b-arrestin 2 to the D2R, thus blocking D2R-mediated modulation of GSK-3 signaling (Alimohamad et al., 2005). Therefore, it is possible that antipsychotics affect GSK-3 signaling downstream of the D2R. Our results demonstrate that the D2R-DISC1 complex is involved in D2R-mediated GSK-3 signaling, but not signaling via the G protein-dependent PKA pathway. This may explain why our interfering peptide exerts antipsychotic effects without inducing EPS. However, our data do not address the question of whether the PKA pathway is associated with EPS. If this were so, a PKA inhibitor would prevent EPS, but again, since the PKA pathway has multiple physiological functions, complete blockage of PKA pathway could lead to side effects as well. For the same reason, a strong GSK-3 inhibitor might not be suitable (C) Effects of TAT-D2pep[K 211 -T 225 ] (1.5 nmol/g i.p.) and haloperidol on hyperlocomotion induced by D-amphetamine (AMPH: 1 mg/kg i.p.). All values represent mean ± SEM. For all groups, n = TAT-D2pep[K 211 -T 225 ] and HAL significantly reduced locomotion (F(3, 30) = 46.68, p < 0.001, two-way ANOVA; post hoc Tukey s test p < 0.01). (D) Effect of TAT-D2pep[K 211 -T 225 ] (3 nmol/g i.p.) and haloperidol on apomorphine (APO)-induced PPI deficits in mice. TAT-D2pep[K 211 -T 225 ] did not affect baseline PPI but restored APO-disrupted PPI (pretreatment: F(3, 61) = 11.55, p < 0.001; treatment F(1, 61) = 37.93, p < 0.01; interaction treatment 3 pretreatment F(3, 61) = 6.24, p < 0.01). Post hoc comparisons revealed that both TAT-D2pep[K 211 -T 225 ] and HAL rescued PPI deficits (p < for comparisons saline+apo versus TAT-D2pep[K 211 -T 225 ]+APO; p < for HAL+APO versus saline+apo; Turkey s). # p < 0.01 compared to saline +APO; **p < compared to non-apo within each group. For all groups, n = (E) Effects of TAT-D2pep[K 211 -T 225 ] (4 nmol, i.c.v.) and haloperidol on AMPH-induced hyperlocomotion in rats. One-way ANOVA analysis revealed a significant effect treatment (F3, 21 = 9.641, p < ). For all groups, n = 6 8. Post hoc testing indicated that both haloperidol and TAT-D2pep significantly reduced amphetamine-induced hyperactivity compared to saline. **p < 0.01; *p < (F) Effect of TAT-D2pep[K 211 -T 225 ] (4 nmol, i.c.v.) and haloperidol on apomorphine (APO)-induced PPI deficits in rats. The peptide reversed PPI deficits at all prepulses (70 db, 75 db, and 80 db). A three-way ANOVA showed no effect oftat-d2pep[k 211 -T 225 ] on baseline PPI but showed a rescue of apomorphinedisrupted PPI (main effect of pretreatment: F(3, 72) = 4.24, p < 0.01; main effect of treatment F(1, 72) = 58.3, p < 0.001; interaction of treatment 3 pretreatment F(3, 72) = 3.29, p < 0.05). Post hoc comparisons revealed that both TAT-D2pep[K 211 -T 225 ] and HAL significantly reversed the PPI deficit induced by APO. #p < 0.05 for comparisons HAL+APO versus saline+apo; ## p < 0.01 for comparisons saline+apo versustat-d2pep[k 211 -T 225 ] +APO; ***p < in comparison with respective non-apo controls. For all groups, n = (G) Haloperidol (0.5 mg/kg s.c.) induced catalepsy but the TAT-D2pep[K 211 -T 225 ] (40nmol, ICV) did not. HAL induced catalepsy in comparison to other groups (F(1,23) = 46.05; p < 0.001). For all groups, n = All data are shown as means ± SEM. See also Figures S7 and S8 and Tables S3 S6. Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1313

13 Neuron D2-DISC1 Complex and Antipsychotic Effects for the treatment of schizophrenia. In addition, although our current data suggest that D2pep[K 211 -T 225 ] has no effect on D2Rmediated camp accumulation, we were not able to rule out the possibility that other D2R signaling components might be involved. For example, previous studies have suggested that D2R interacts with prostate apoptosis response 4 (Par-4) and calmodulin through the region of D2R (Bofill-Cardona et al., 2000; Park et al., 2005). The Par-4/D2R interaction is involved in the D2R-mediated PKA pathway. In conclusion, we have identified a specific target with antipsychotic-like effects that is associated with D2R-mediated GSK-3 signaling, but not other D2R-mediated signaling pathways or other GSK-3-mediated pathways. This provides an opportunity to develop antipsychotics with fewer side effects. In particular, the lack of catalepsy strongly suggests that interfering with the D2R-DISC1 interaction should not cause EPS. The overall strategy of targeting interactions between receptors and other proteins could represent a general approach to improve efficacy and reduce side effects for other medications. EXPERIMENTAL PROCEDURES A brief summary is provided here, with more details in the Supplemental Information accompanying this article. Coimmunoprecipitation, Protein Affinity Purification, In Vitro Binding, and Western Blot Protein affinity purification, in vitro binding, coimmunoprecipitation, and western blot analysis were performed as previously described (Lee et al., 2007; Pei et al., 2010). For antibody details, see the Supplemental Experimental Procedures. GST Fusion Protein Constructs GST-fusion proteins encoding D2R and DISC1 fragments were amplified by PCR from full-length human cdna clones. All constructs were sequenced to confirm the absence of spurious PCR-generated nucleotide errors. GST fusion proteins were prepared from bacterial lysates as described by the manufacturer (Amersham). camp Accumulation Assay camp accumulation assays were performed in HEK293T cells transiently transfected with the indicated cdna constructs and in striatal neurons infected with DISC1 in lentiviral vector using the Parameter camp Assay Kit as directed by the manufacturer (R&D systems, Minneapolis). Lentivirus Production Lentivirus was produced as previously described in our lab (Zhai et al., 2013). DISC1 shrna and shrna-resistant DISC1 (DISC1 R ) in lentiviral vector were gifts from Dr. Guo-li Ming at Johns Hopkins University. The lentivirus packaging was done using X-tremegene HP transfection reagent. Receptor Internalization with Surface Biotinylation Biotinylation assay were performed as previously described (Lee et al., 2007). Anti-HA or anti-d2r was used to detect the internalized D2Rs. Negative control: biotin is cleaved before agonist stimulation (right after biotinylation). Positive control: biotin is not cleaved after biotinylation. Animals Each experiment was performed on separate groups of 12- to 16-week-old male mice or 10- to 12-week-old male Sprague-Dawley rats ( g). Disc1-L100P homozygous mutant mice and their WT littermates were generated as previously described. For drug and peptide doses and administration routes, see the Supplemental Experimental Procedures. Behavioral Experiments Locomotor activity testing in mice and rats was performed according to previously published methods (Lee et al., 2007; Pei et al., 2010). PPI in mice was slightly modified from previously published descriptions (Clapcote et al., 2007). Rat PPI was assessed using six SR-Lab Startle Response systems (San Diego Instruments, San Diego) as previously described (Sills, 1999). Catalepsy was measured in rats as described previously (Creed-Carson et al., 2011). Statistical Analyses Data were analyzed by one-, two-, or three-way ANOVA, followed where appropriate by post hoc Fishers LSD test. Bonferroni corrections were applied in cases of multiple comparisons (Statistica, Statsoft, Tulsa). SUPPLEMENTAL INFORMATION Supplemental Information includes eight figures, six tables, and Supplemental Experimental Procedures and can be found with this article online at doi.org/ /j.neuron ACKNOWLEDGMENTS We gratefully acknowledge Dr. Guo-li Ming at Johns Hopkins University for providing us the DISC1-shRNA, DISC1 R. This work is supported by an operating grant from Canadian Institutes for Health Research (F.L.). We thank Dr. Meaghan Creed for technical support in the Catalepsy test. Accepted: November 4, 2014 Published: November 26, 2014 REFERENCES Agid, O., Mamo, D., Ginovart, N., Vitcu, I., Wilson, A.A., Zipursky, R.B., and Kapur, S. (2007). Striatal vs extrastriatal dopamine D2 receptors in antipsychotic response a double-blind PET study in schizophrenia. Neuropsychopharmacology 32, Alimohamad, H., Rajakumar, N., Seah, Y.H., and Rushlow, W. (2005). Antipsychotics alter the protein expression levels of beta-catenin and GSK-3 in the rat medial prefrontal cortex and striatum. Biol. Psychiatry 57, Beaulieu, J.M. (2012). A role for Akt and glycogen synthase kinase-3 as integrators of dopamine and serotonin neurotransmission in mental health. J. Psychiatry Neurosci. 37, Beaulieu, J.M., Gainetdinov, R.R., and Caron, M.G. (2007). The Akt-GSK-3 signaling cascade in the actions of dopamine. Trends Pharmacol. Sci. 28, Beaulieu, J.M., Gainetdinov, R.R., and Caron, M.G. (2009). Akt/GSK3 signaling in the action of psychotropic drugs. Annu. Rev. Pharmacol. Toxicol. 49, Bofill-Cardona, E., Kudlacek, O., Yang, Q., Ahorn, H., Freissmuth, M., and Nanoff, C. (2000). Binding of calmodulin to the D2-dopamine receptor reduces receptor signaling by arresting the G protein activation switch. J. Biol. Chem. 275, Brandon, N.J., and Sawa, A. (2011). Linking neurodevelopmental and synaptic theories of mental illness through DISC1. Nat. Rev. Neurosci. 12, Clapcote, S.J., Lipina, T.V., Millar, J.K., Mackie, S., Christie, S., Ogawa, F., Lerch, J.P., Trimble, K., Uchiyama, M., Sakuraba, Y., et al. (2007). Behavioral phenotypes of Disc1 missense mutations in mice. Neuron 54, Creed-Carson, M., Oraha, A., and Nobrega, J.N. (2011). Effects of 5-HT(2A) and 5-HT(2C) receptor antagonists on acute and chronic dyskinetic effects induced by haloperidol in rats. Behav. Brain Res. 219, Duan, X., Chang, J.H., Ge, S., Faulkner, R.L., Kim, J.Y., Kitabatake, Y., Liu, X.B., Yang, C.H., Jordan, J.D., Ma, D.K., et al. (2007). Disrupted-In- Schizophrenia 1 regulates integration of newly generated neurons in the adult brain. Cell 130, Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

14 Neuron D2-DISC1 Complex and Antipsychotic Effects Enomoto, A., Asai, N., Namba, T., Wang, Y., Kato, T., Tanaka, M., Tatsumi, H., Taya, S., Tsuboi, D., Kuroda, K., et al. (2009). Roles of disrupted-in-schizophrenia 1-interacting protein girdin in postnatal development of the dentate gyrus. Neuron 63, Featherstone, R.E., Kapur, S., and Fletcher, P.J. (2007). The amphetamineinduced sensitized state as a model of schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry 31, Fukunaga, K., and Shioda, N. (2012). Novel dopamine D2 receptor signaling through proteins interacting with the third cytoplasmic loop. Mol. Neurobiol. 45, Fuxe, K., Borroto-Escuela, D.O., Marcellino, D., Romero-Fernandez, W., Frankowska, M., Guidolin, D., Filip, M., Ferraro, L., Woods, A.S., Tarakanov, A., et al. (2012). GPCR heteromers and their allosteric receptor-receptor interactions. Curr. Med. Chem. 19, Geddes, J., Freemantle, N., Harrison, P., and Bebbington, P. (2000). Atypical antipsychotics in the treatment of schizophrenia: systematic overview and meta-regression analysis. BMJ 321, Geyer, M.A., Krebs-Thomson, K., Braff, D.L., and Swerdlow, N.R. (2001). Pharmacological studies of prepulse inhibition models of sensorimotor gating deficits in schizophrenia: a decade in review. Psychopharmacology (Berl.) 156, Hanyaloglu, A.C., and von Zastrow, M. (2008). Regulation of GPCRs by endocytic membrane trafficking and its potential implications. Annu. Rev. Pharmacol. Toxicol. 48, Hodgkinson, C.A., Goldman, D., Jaeger, J., Persaud, S., Kane, J.M., Lipsky, R.H., and Malhotra, A.K. (2004). Disrupted in schizophrenia 1 (DISC1): association with schizophrenia, schizoaffective disorder, and bipolar disorder. Am. J. Hum. Genet. 75, Insel, T.R., and Scolnick, E.M. (2006). Cure therapeutics and strategic prevention: raising the bar for mental health research. Mol. Psychiatry 11, Ishizuka, K., Kamiya, A., Oh, E.C., Kanki, H., Seshadri, S., Robinson, J.F., Murdoch, H., Dunlop, A.J., Kubo, K., Furukori, K., et al. (2011). DISC1-dependent switch from progenitor proliferation to migration in the developing cortex. Nature 473, Kang, E., Burdick, K.E., Kim, J.Y., Duan, X., Guo, J.U., Sailor, K.A., Jung, D.E., Ganesan, S., Choi, S., Pradhan, D., et al. (2011). Interaction between FEZ1 and DISC1 in regulation of neuronal development and risk for schizophrenia. Neuron 72, Kellendonk, C., Simpson, E.H., Polan, H.J., Malleret, G., Vronskaya, S., Winiger, V., Moore, H., and Kandel, E.R. (2006). Transient and selective overexpression of dopamine D2 receptors in the striatum causes persistent abnormalities in prefrontal cortex functioning. Neuron 49, Kim, J.Y., Duan, X., Liu, C.Y., Jang, M.H., Guo, J.U., Pow-anpongkul, N., Kang, E., Song, H., and Ming, G.L. (2009). DISC1 regulates new neuron development in the adult brain via modulation of AKT-mTOR signaling through KIAA1212. Neuron 63, Kim, J.Y., Liu, C.Y., Zhang, F., Duan, X., Wen, Z., Song, J., Feighery, E., Lu, B., Rujescu, D., St Clair, D., et al. (2012). Interplay between DISC1 and GABA signaling regulates neurogenesis in mice and risk for schizophrenia. Cell 148, Kvajo, M., McKellar, H., Arguello, P.A., Drew, L.J., Moore, H., MacDermott, A.B., Karayiorgou, M., and Gogos, J.A. (2008). A mutation in mouse Disc1 that models a schizophrenia risk allele leads to specific alterations in neuronal architecture and cognition. Proc. Natl. Acad. Sci. USA 105, Laruelle, M., Abi-Dargham, A., van Dyck, C.H., Gil, R., D Souza, C.D., Erdos, J., McCance, E., Rosenblatt, W., Fingado, C., Zoghbi, S.S., et al. (1996). Single photon emission computerized tomography imaging of amphetamine-induced dopamine release in drug-free schizophrenic subjects. Proc. Natl. Acad. Sci. USA 93, Lee, F.J., Pei, L., Moszczynska, A., Vukusic, B., Fletcher, P.J., and Liu, F. (2007). Dopamine transporter cell surface localization facilitated by a direct interaction with the dopamine D2 receptor. EMBO J. 26, Lee, F.H., Fadel, M.P., Preston-Maher, K., Cordes, S.P., Clapcote, S.J., Price, D.J., Roder, J.C., and Wong, A.H. (2011). Disc1 point mutations in mice affect development of the cerebral cortex. J. Neurosci. 31, Leucht, S., Corves, C., Arbter, D., Engel, R.R., Li, C., and Davis, J.M. (2009). Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. Lancet 373, Lieberman, J.A., Stroup, T.S., McEvoy, J.P., Swartz, M.S., Rosenheck, R.A., Perkins, D.O., Keefe, R.S., Davis, S.M., Davis, C.E., Lebowitz, B.D., et al.; Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Investigators (2005). Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N. Engl. J. Med. 353, Mao, Y., Ge, X., Frank, C.L., Madison, J.M., Koehler, A.N., Doud, M.K., Tassa, C., Berry, E.M., Soda, T., Singh, K.K., et al. (2009). Disrupted in schizophrenia 1 regulates neuronal progenitor proliferation via modulation of GSK3beta/betacatenin signaling. Cell 136, Millar, J.K., Pickard, B.S., Mackie, S., James, R., Christie, S., Buchanan, S.R., Malloy, M.P., Chubb, J.E., Huston, E., Baillie, G.S., et al. (2005). DISC1 and PDE4B are interacting genetic factors in schizophrenia that regulate camp signaling. Science 310, Missale, C., Nash, S.R., Robinson, S.W., Jaber, M., and Caron, M.G. (1998). Dopamine receptors: from structure to function. Physiol. Rev. 78, Nicola, S.M., Surmeier, J., and Malenka, R.C. (2000). Dopaminergic modulation of neuronal excitability in the striatum and nucleus accumbens. Annu. Rev. Neurosci. 23, Niwa, M., Kamiya, A., Murai, R., Kubo, K., Gruber, A.J., Tomita, K., Lu, L., Tomisato, S., Jaaro-Peled, H., Seshadri, S., et al. (2010). Knockdown of DISC1 by in utero gene transfer disturbs postnatal dopaminergic maturation in the frontal cortex and leads to adult behavioral deficits. Neuron 65, Park, S.K., Nguyen, M.D., Fischer, A., Luke, M.P., Affar, B., Dieffenbach, P.B., Tseng, H.C., Shi, Y., and Tsai, L.H. (2005). Par-4 links dopamine signaling and depression. Cell 122, Pei, L., Li, S., Wang, M., Diwan, M., Anisman, H., Fletcher, P.J., Nobrega, J.N., and Liu, F. (2010). Uncoupling the dopamine D1-D2 receptor complex exerts antidepressant-like effects. Nat. Med. 16, Pletnikov, M.V., Ayhan, Y., Nikolskaia, O., Xu, Y., Ovanesov, M.V., Huang, H., Mori, S., Moran, T.H., and Ross, C.A. (2008). Inducible expression of mutant human Dros. Inf. Serv.C1 in mice is associated with brain and behavioral abnormalities reminiscent of schizophrenia. Mol. Psychiatry 13, Porteous, D.J., Millar, J.K., Brandon, N.J., and Sawa, A. (2011). DISC1 at 10: connecting psychiatric genetics and neuroscience. Trends Mol. Med. 17, Ralph, R.J., and Caine, S.B. (2005). Dopamine D1 and D2 agonist effects on prepulse inhibition and locomotion: comparison of Sprague-Dawley rats to Swiss-Webster, 129X1/SvJ, C57BL/6J, and DBA/2J mice. J. Pharmacol. Exp. Ther. 312, Seeman, P., Lee, T., Chau-Wong, M., and Wong, K. (1976). Antipsychotic drug doses and neuroleptic/dopamine receptors. Nature 261, Shen, S., Lang, B., Nakamoto, C., Zhang, F., Pu, J., Kuan, S.L., Chatzi, C., He, S., Mackie, I., Brandon, N.J., et al. (2008). Schizophrenia-related neural and behavioral phenotypes in transgenic mice expressing truncated Disc1. J. Neurosci. 28, Sills, T.L. (1999). Amphetamine dose dependently disrupts prepulse inhibition of the acoustic startle response in rats within a narrow time window. Brain Res. Bull. 48, Simpson, E.H., Kellendonk, C., and Kandel, E. (2010). A possible role for the striatum in the pathogenesis of the cognitive symptoms of schizophrenia. Neuron 65, Singh, K.K., Ge, X., Mao, Y., Drane, L., Meletis, K., Samuels, B.A., and Tsai, L.H. (2010). Dixdc1 is a critical regulator of DISC1 and embryonic cortical development. Neuron 67, Neuron 84, , December 17, 2014 ª2014 Elsevier Inc. 1315

15 Neuron D2-DISC1 Complex and Antipsychotic Effects Surmeier, D.J., Ding, J., Day, M., Wang, Z., and Shen, W. (2007). D1 and D2 dopamine-receptor modulation of striatal glutamatergic signaling in striatal medium spiny neurons. Trends Neurosci. 30, Taylor, M.S., Devon, R.S., Millar, J.K., and Porteous, D.J. (2003). Evolutionary constraints on the Disrupted in Schizophrenia locus. Genomics 81, Torrey, E.F., Webster, M., Knable, M., Johnston, N., and Yolken, R.H. (2000). The stanley foundation brain collection and neuropathology consortium. Schizophr. Res. 44, Zhai, D., Li, S., Wang, M., Chin, K., and Liu, F. (2013). Disruption of the GluR2/ GAPDH complex protects against ischemia-induced neuronal damage. Neurobiol. Dis. 54, Zhou, M., Li, W., Huang, S., Song, J., Kim, J.Y., Tian, X., Kang, E., Sano, Y., Liu, C., Balaji, J., et al. (2013). mtor Inhibition ameliorates cognitive and affective deficits caused by Disc1 knockdown in adult-born dentate granule neurons. Neuron 77, Neuron 84, , December 17, 2014 ª2014 Elsevier Inc.

16 Neuron, Volume 84 Supplemental Information A Dopamine D2 Receptor-DISC1 Protein Complex may Contribute to Antipsychotic-Like Effects Ping Su, Shupeng Li, Sheng Chen, Tatiana V. Lipina, Min Wang, Terence K.Y. Lai, Frankie H.F. Lee, Hailong Zhang, Dongxu Zhai, Stephen S.G. Ferguson, José N. Nobrega, Albert H.C. Wong, John C. Roder, Paul J. Fletcher, and Fang Liu

17 DISC1 Co-IP/Total DISC1 (Percent of L100P Group) DISC1 Co-IP/Total DISC1 (Percent of L100P Group) DISC1 Co-IP/Total DISC1 (Percent of WT Group) GFP-DISC1 Binding to HA-D2R ( Percent of Control Group) A Mouseanti-D2R Goatanti-D2R B HEK-293T cells IP: HA-D2R C Anti-GFP -DISC1 Anti-HA -D2R ** # D2-KO WT D E HEK-293T cells IP ** 100 IgG Extract L100P WT HA-D2R Anti-DISC1 L100P WT F Anti-DISC1 Anti-D2R L100P L100P mice IP: anti-d2r L100P+ Haloperidol (acute) 100 G WT *** L100P IP: anti-d2r L100P L100P+Halo (acute) Anti-DISC1 Anti-D2R L100P+ Haloperidol (chronic) L100P ** 0.0 L100P L100P+Halo (chronic) Su et al., 2014 Figure S1

18 P-GSK3/GSK3 P-GSK3/GSK3 (Percent of Control) P-GSK3/GSK3 (Percent of Control) A. B 1.0 WT GSK 3a (Ser21) GSK 3b (Ser 9) 0.8 * p-gsk 3α (Ser 21) p-gsk 3β (Ser 9) Mouse striatal tissue WT D2R KO ** GSK 3α GSK 3β CTRL Quin Halo+Quin D2R KO GSK 3a (Ser21) GSK 3b (Ser 9) C Rat cortical neurons D 0.0 Control Quin Halo/Quin CTRL Quin Halo+Quin E p-gsk 3α (Ser 21) p-gsk 3β (Ser 9) IgG IP: D2R Anti-DISC1 GSK 3α GSK 3β Extract Cor. Stri GSK 3a (Ser21) GSK 3b (Ser 9) Su et al., 2014 Figure S2

19 P-GSK3/GSK3 (Percent of Control) P-GSK3/GSK3 (Percent of Control) P-GSK3/GSK3 (Percent of Control) P-GSK3/GSK3 (Percent of Control) A B Anti-DISC1 100 p-gsk3α p-gsk3β DISC1-shRNA NS-shRNA Anti-α-Tubulin GSK3α GSK3β Quinpirole C 1.0 DISC1-shRNA Control Quin D 1.0 NS-shRNA Control Quin ** ** GSK 3a (Ser21) GSK 3b (Ser 9) 0.0 GSK 3a (Ser21) GSK 3b (Ser 9) E F G p-gsk 3α (Ser 21) p-gsk 3β (Ser 9) Vector DISC1-shRNA DISC1 R GSK 3a (Ser21) GSK 3b (Ser 9) GSK 3a (Ser21) GSK 3b (Ser 9) ** * GSK 3α GSK 3β Control Quin 0.0 Control Quin DISC1-shRNA+Vector DISC1-shRNA+DISC1 R Su et al., 2014 Figure S3

20 P-Akt (T308)/Akt (Percent of Control Group) camp (fold change relative to the control) P-Akt (T308)/Akt (Percent of Control Group) A B 1.2 NS-shRNA NS-shRNA DISC1-shRNA 1.0 p-akt (Thr 308) p-akt (Ser 473) Akt ** Control Quin C DISC1-shRNA D 6 Control Foskolin Foskolin + Quinpirole * * Control Quin 0 D2R+pcDNA3 D2R+ DISC1 Su et al., 2014 Figure S4

21 camp (fold change relative to the control) camp (fold change relative to the control of Vector) P-GSK3/GSK3 (Percent of control) A K 211 D2R IL3 Q 373 K 211 D2R IL3-1 V 270 E 271 D2R IL3-2 H 330 H 312 D2R IL3-3 Q 373 K 211 D2R IL3-1-1 T 225 K 226 D2R IL3-1-2 L 240 K 241 V 255 D2R IL3-1-3 I 256 D2R IL3-1-4 V 270 B C 1.4 Control Quinpirole TAT-D2pep/Quinpirole TAT-D2pep-sc/Quinpirole 1.2 p-gsk3α (S21) p-gsk3β (S9) Quinpirole TAT-Peptide - - D2 sc GSK3α GSK3β * * * * 0 GSK3α (S21) GSK3β (S9) D E *** *** Vector DISC1 2 * * * Forskolin Quinpirole TAT-Peptide D2pep D2pep-sc Forskolin - + Quinpirole - - ## ## + + Su et al., 2014 Figure S5

22 P-GSK3/GSK3 (Percent of L100P+D2sc Group) P-GSK3/GSK3 (Percent of L100P Group) P-GSK3/GSK3 (Percent of L100P Group) A p-gsk3α (Ser21) p-gsk3b (Ser 9) GSK3α GSK3b B GSK 3a (Ser21) GSK 3b (Ser 9) * ** ** *** L100P WT L100P +Halo (acute) C D 0.0 L100P WT L100P+halo (acute) p-gsk3α (Ser21) p-gsk3b (Ser 9) GSK3α GSK3b GSK 3a (Ser21) GSK 3b (Ser 9) ** *** L100P +Halo (chronic) L100P E F 0.0 L100P L100P+Halo (chronic) 2.5 GSK 3a (Ser21) GSK 3b (Ser 9) p-gsk3α (Ser21) p-gsk3b (Ser 9) 2.0 * ** GSK3α GSK3b L100P +TATD2pep-sc (chronic) L100P +TATD2pep (chronic) L100P+D2pep-sc L100P+D2pep Su et al., 2014 Figure S6