Multi-omic transcriptional, brain, and clinical variations in schizophrenia

Abstract

AbstractHow genetic risk variants may relate to brain abnormalities is crucial for understanding cross-scale pathophysiological mechanisms underlying schizophrenia. The present study identifies brain structural correlates of variation in gene expression in schizophrenia and its clinical significance. Of 43 patients with schizophrenia, RNA-seq data from blood samples, MRI, and clinical assessments were collected, together with data from 60 healthy controls. Gene expression differentiation between schizophrenia and health controls was assessed and cross-referenced to schizophrenia-related genomic variations (GWAS on 76,755 patients and 243,649 controls and GWAS on 22,778 East Asian patients) and brain gene expressions (samples from 559 patients and 175 individuals). Multivariate correlation analysis was employed to examine associations across gene expression, brain volume, and clinical assessments. Differentially expressed genes in blood samples from patients with schizophrenia were significantly enriched for genes previously reported in genome-wide association studies on schizophrenia (P= 0.002, false discovery rate corrected) and were associated with gene expression differentiation in the brain (P= 0.016, 5,000 permutations). Transcriptional levels of differentially expressed genes were found to significantly correlate with gray matter volume in the frontal and temporal regions of cognitive brain networks in schizophrenia (q< 0.05, false discovery rate corrected). A significant correlation was further observed between gene expression, gray matter volume, and performance in the Wechsler Adult Intelligence Scale test (P= 0.031). Our findings suggest that genomic variations in schizophrenia are associated with differentiation in the blood transcriptome, which further plays a role in individual variations in macroscale brain structure and cognition.