Abstract

Background Schizophrenia (SCZ) is associated with genetic factors, and specific risk loci have recently been identified. Still, the biology and functional or clinical translation of SCZ risk genes remain largely unknown. At least some of the Psychiatric Genomic Consortium (PGC) genetic variants control gene expression [1] . Since the expression of individual genes is co-regulated and results in the co-expression of gene sets, we hypothesized that SCZ risk genes may converge into co-expression pathways which, in turn, may be associated with clinical phenotypes in patients with SCZ. Previous work in microarray post-mortem brain gene expression data shows that co-expression networks have biological plausibility for SCZ [2] . Methods Here, we used RNA sequencing data from the Lieber Institute for Brain Development (LIBD, N=343) [1] and the Common Mind Consortium (CMC, N=345) [3] to identify gene co-expression networks through Weighted Gene Co-expression Analysis. To detect gene modules in which PGC SCZ risk protein coding genes (n=300) were overrepresented, we computed hypergeometric tests and corrected the results for multiple comparisons (Bonferroni, number of modules=43). Then, we generated a Polygenic Co-expression Index (PCI) predicting gene co-expression based on a meta-analysis of LIBD and CMC. We replicated the PCI-co-expression association in BrainEAC, an independent post mortem microarray dataset (N=38). Finally, we studied a clinical cohort of drug-free patients with SCZ treated with olanzapine in monotherapy (N=49). We correlated patients’ PCI both with symptoms severity at the baseline assessment and treatment response (endpoint – baseline) in terms of positive, negative, and general PANSS domains. We Bonferroni-corrected results for the number of PANSS subscales (n=3). Results In LIBD, we discovered a single gene set enriched for SCZ risk loci (157 protein coding genes; PGC hits=12; corrected p=.0007), which was successfully replicated in CMC (empirical p Discussion These results corroborate the hypothesis that co-expression pathways are relevant to the neurobiology of SCZ and that a significant proportion of the risk genes is co-expressed in the dorsolateral prefrontal cortex. Furthermore, these findings provide preliminary evidence for the clinical relevance of this co-expression pathway and of the genetic variants associated with it.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call