Schizophrenia and depression, two poles of endocannabinoid system deregulation

Abstract

The activity of certain G protein-coupled receptors (GPCRs) and of glutamate N-Methyl-D-aspartate receptors (NMDARs) is altered in both schizophrenia and depression. Using postmortem prefrontal cortex samples from subjects with schizophrenia or depression, we observed a series of opposite changes in the expression of signaling proteins that have been implicated in the cross-talk between GPCRs and NMDARs. Thus, the levels of HINT1 proteins and NMDAR NR1 subunits carrying the C1 cytosolic segment were increased in depressives and decreased in schizophrenics, respect to matched controls. The differences in NR1 C1 subunits were compensated for via altered expression of NR1 subunits lacking the C1 segment; thus, the total number of NR1 subunits was comparable among the three groups. GPCRs influence the function of NR1 C1-containing NMDARs via PKC/Src, and thus, the association of mu-opioid and dopamine 2 receptors with NR1 C1 subunits was augmented in depressives and decreased in schizophrenics. However, the association of cannabinoid 1 receptors (CB1Rs) with NR1 C1 remained nearly constant. Endocannabinoids, via CB1Rs, control the presence of NR1 C1 subunits in the neural membrane. Thus, an altered endocannabinoid system may contribute to the pathophysiology of schizophrenia and depression by modifying the HINT1-NR1 C1/GPCR ratio, thereby altering GPCR-NMDAR cross-regulation.

Fig. 1. HINT1 and NMDAR NR1 C1 subunit levels and their association with CB1Rs in postmortem human prefrontal cortices of schizophrenic and depressive subjects: a comparative study vs. control individuals

The analysis was first performed for each individual group, schizophrenics (S), depressives (D), and controls (C). The levels of (a) HINT1 and (b) NMDAR NR1 subunits containing C1 cytosolic segment were normalized when necessary to α-tubulin levels. Representative blots are shown. Left, S/C/D triplet 9; right S/C/D triplet 10 (see Supplementary Table S1 and Fig. S1). The assays were performed twice, and the average data for each individual were included in the subsequent matched analysis. The values of the schizophrenic and depressive subjects were compared to that of the corresponding matched control of the triplet (assigned an arbitrary value of 1), and for the S and D groups (n = 24) the means (bars), 95% CI (lines) and individual values (points) are shown. Analyses of covariance (ANCOVA) were performed with age, PMI and suicide as covariates. Inset: diagram indicating differences in HINT1 levels and ratios of the different NR1 subunits in the study groups. *p < 0.05 vs control in LSD post-hoc analyses. CB1R was immunoprecipitated, and CB1R-associated HINT1 (c) and NR1 C1 (d) protein levels were determined by western blotting. Data were normalized when necessary to the signals obtained probing the accompanying anti-CB1R IgGs used for immunoprecipitation with a secondary antibody (mouse anti-rabbit light chain HRP-conjugated monoclonal ab; Millipore #MAB201P). This antibody labels light chains on primary antibodies targeting the GPCR or a co-precipitated protein. Data expression and analyses as in (a) and (b). The CB1R-associated HINT1 and NR1 C1 were related to total HINT1 (e) and NR1 C1 (f) content, respectively. Inset: diagram showing the presence of CB1Rs and their association with HINT1 and NMDAR NR1 C1 subunits relative to the total content of these proteins in the study groups. ^Ф p < 0.05 vs control in LSD post-hoc analyses

Fig. 2. MOR association with HINT1 and NR1 C1 subunits in postmortem human prefrontal cortices of schizophrenic and depressive subjects: a comparative study vs. controls

The MOR was immunoprecipitated and its association with HINT1 (a) and NR1 C1 subunits (b) was determined by western blotting. The regions of the blotting membrane incubated with different antibodies are indicated. The levels of anti-MOR IgG light chain (IgG lc) were used as a loading control. Representative blots are shown. Left, S/C/D triplet 9; right S/C/D triplet 10 (see Supplementary Table S1 and Fig. S1). Data expression and analyses as in Fig. 1. The MOR-associated HINT1 and NR1 C1 were compared to total HINT1 (c) and NR1 C1 (d) content, respectively. Inset: diagram showing the presence of MORs and their association with HINT1 and NMDAR NR1 C1 subunits relative to the total content of these proteins in the study groups. *, ^Ф p < 0.05 vs control in LSD post-hoc analyses

Fig. 3. D2R association with HINT1 and NR1 C1 subunits in postmortem human prefrontal cortices of schizophrenic and depressive subjects: a comparative study vs. controls

The D2R was immunoprecipitated, and its association with HINT1 (a) and NR1 C1 subunits (b) was determined by western blotting. Representative blots are shown. Left, S/C/D triplet 9; right S/C/D triplet 10 (see Supplementary Table S1 and Fig. S1). Data expression and analyses as in Figs. 1 and 2. The D2R-associated HINT1 and NR1 C1 were referred to total HINT1 (c) and NR1 C1 (d) content, respectively. Inset: diagram showing the presence of D2Rs and their association with HINT1 and NMDAR NR1 C1 subunits relative to the total content of these proteins in the study groups. *, ^Ф p < 0.05 vs control in LSD post-hoc analyses

Fig. 4. Cannabinoids, via CB1Rs negatively control the presence of NMDAR NR1 C1 subunits in neural membranes

(a) In HINT1−/− and σ1R−/− mice, MOR/CB1R association with NR1 C1 subunits is impaired. IP immunoprecipitation, WB western blot. The immunoprecipitated levels of MORs and CB1Rs are indicated. (b) The cannabinoid agonist WIN55,212-2 promotes the co-internalization of CB1Rs and NR1 C1 subunits. Mice received three intracerebroventricular doses of WIN55,212-2 or saline spaced 90 min apart and were sacrificed 3 h after the last injection. Adapted from ref. . (c) The HINT1 protein restores the CB1R-NR1 C1 association. The HINT1 protein was introduced with lentiviral particles. Murine HINT1 was cloned in the pLVTHM vector downstream of the H1 promoter. Lentiviruses (pVLTHM-HINT1 cDNA, psPAX2, pMD2.G) were prepared in HEK-293T cells. Adapted from ref. . (d) Total NR1 and NR1 C1 subunits in the frontal cortices of HINT1−/−, σ1R−/− or CB1R−/− mice. For CB1R−/− mice, HINT1 levels are shown. Within each row, knockout (KO) values are compared with those of the respective wild-type (WT) mice (assigned an arbitrary value of 1). The * indicates a significant difference between the study and control group, ANOVA, followed by Dunnett’s multiple comparisons vs. control group, p < 0.05. Representative blots are shown. Details are given in the references indicated in Results. (e) Diagram showing the possible influence of HINT1/NR1 C1 levels on mental disorders such as schizophrenia and depression