BackgroundThe synaptic dysfunction hypothesis of schizophrenia is supported by multiple lines of evidence. However, in vivo evidence is lacking. Moreover, it is not known if antipsychotics alter synaptic protein levels. We addressed this in a combined clinical study using [11C]UCB-J, a positron emission tomography (PET) radioligand specific for synaptic vesicle glycoprotein 2A (SV2A), and rodent study of clinically relevant antipsychotic drug exposure.MethodsWe scanned 18 subjects with schizophrenia (SCZ) and 18 healthy volunteers (HV) with [11C]UCB-J PET and T1-weighted MRI, estimating grey matter volumes of distribution (VT) and corrected grey matter volumes (GMV) for the frontal cortex (FC), anterior cingulate cortex (ACC) and hippocampus. In addition, we estimated the [11C]UCB-J distribution volume ratio (DVR) in these regions, using the centrum semiovale (CS) as a pseudoreference region. We collected clinical data including PANSS score, chlorpromazine-equivalent antipsychotic dose (CPZ) and duration of illness (DOI). In parallel, we investigated the effects of olanzapine and haloperidol administration on SV2A levels in the Sprague-Dawley rat frontal cortex using western blotting, [3H]UCB-J autoradiography and immunostaining with confocal microscopy. We used two-way analysis of variance (ANOVA) to test effects of group, ROI and group-by-ROI interaction on VT and DVR. We used planned post hoc t-tests to test group effects at each ROI, with false discovery-rate adjustment for multiple comparisons. We used two-way ANOVA with Bonferroni’s post-hoc test to analyse autoradiography and SV2A immunostaining data, and the Kruskal-Wallis test, with alpha=0.05, to analyse western blot data.ResultsClinical study: We found significant group-by-region interaction (F2, 68=7.472, p=0.001), group (F1, 34=6.170, p=0.02) and ROI (F2, 68=426.0, p<0.0001) effects on VT, with significant reductions in mean [SEM] VT of large effect size in the SCZ group in the FC (SCZ=16.93 [0.80]; HV=19.50 [0.64]; t=2.51, df=34.0, p=0.03, Cohen’s d=0.8) and ACC (SCZ=19.55 [0.75]; HV=22.49 [0.72]; t=2.83, df=34.0, p=0.02, Cohen’s d=0.9), but not in the hippocampus (SCZ=14.09 [0.59]; HV=15.44 [0.50]; t=1.75, df=34.0, p=0.09, Cohen’s d=0.6). Furthermore, we found significant group-by-region interaction (F2, 68=7.97, p=0.0008), group (F1, 34=8.1, p=0.007) and ROI (F2, 68=510.9, p<0.0001) effects on DVR, with significant reductions of large effect size in the FC (SCZ=2.93 [0.17]; HV=3.48 [0.09]; t=2.89, df=34.0, p=0.01, Cohen’s d=1.0), ACC (SCZ=3.39 [0.17]; HV=3.99 [0.09]; t=3.05, df=34.0, p=0.01, Cohen’s d=1.0) and, additionally, hippocampus (SCZ=2.40 [0.12]; HV=2.74 [0.07]; t=2.32, df=34.0, p=0.03, Cohen’s d=0.8). There were no significant relationships (p>0.05) between VT and GMV, PANSS score, DOI and CPZ.Preclinical study: Chronic olanzapine or haloperidol exposure did not significantly affect total SV2A immunostaining intensity (p=0.97 and p=0.71 respectively) in the rat frontal cortex as compared to vehicle-exposed controls. Moreover, chronic haloperidol exposure did not significantly affect [3H]UCB-J specific binding (p=0.27) or SV2A protein levels (relative to GAPDH, p=0.71).DiscussionThese findings provide evidence for the first time that presynaptic marker protein levels are reduced in schizophrenia in vivo and that antipsychotic drug exposure is unlikely to account for this, consistent with the synaptic dysfunction hypothesis.
Read full abstract