O11.1. MULTIPLE GENOMIC MEASURES OF SCHIZOPHRENIA RISK ARE RELATED TO CORTICAL THICKNESS IN TYPICALLY DEVELOPING YOUTH AND YOUNG ADULTS

BACKGROUND: The incidence of psychotic disorders increases in adolescence and young adulthood. Transition to a psychotic disorder is associated with atypical development of brain structures, specifically protracted developmental course. It is unknown how polygenic risk for schizophrenia and gene expression profiles of schizophrenia risk genes affect typical brain development. The goal of the current study is to examine relationships multiple genomic measures associated with schizophrenia risk and structural neuroimaging measures thickness in typically developing youth. METHODS: We combined structural neuroimaging and genetic data from three different cohorts of typically developing youth (N=994, 5–30 years old): the Philadelphia Neurodevelopmental Cohort, Pediatric Imaging Neurocognition and Genetics Study, and a locally collected sample at the University of Pittsburgh. All youth were free from psychiatric disorders and not taking psychiatric medications. We used Freesurfer to process the T1-weighted structural scans and calculate subcortical volumes, cortical thickness, and surface area measurements. After regressing out study, sex, ancestry eigenvectors, and grey matter signal-to-noise ratio, we ran principal components analysis on all neuroimaging measures (N=156). We calculated a schizophrenia polygenic risk score using genome-wide summary statistics from the Psychiatric Genome Consortium. Using a generalized linear model, each of the top five principal components was evaluated in relation to the risk score. We then used a computational method, Predixcan, to calculate expected gene expression profiles from the genotype data. We selected 125 genes that were associated with schizophrenia in a previous case-control comparison. Elastic net regression was used to determine significant associations between individual gene expression and the principal components. RESULTS: Schizophrenia polygenic risk was statistically associated with the 5th principal component (b=-0.10, p=0.001), which consisted of contributions from multiple measures of cortical thickness. Reduced cortical thickness in frontal and temporal regions was associated with increased genetic liability for schizophrenia. Increased cortical thickness in sensory-motor areas was associated with higher schizophrenia polygenic risk scores. This relationship remained when age was included as a predictor of interest and there were no statistically significant interactions between schizophrenia polygenic risk and age. Sixteen unique gene expression profiles were also associated with this principal component, significantly increasing the proportion of variance explained in this measure (from ~1% with the schizophrenia polygenic risk only to ~6% when including the additional gene expression measures). Many of the genes significantly associated with this principal component have important roles during early fetal brain development, including neuronal migration (e.g., SDCCAG8) and DNA repair (e.g., MLH1). DISCUSSION: These results suggest that that genetic risk for schizophrenia has a consistent influence on subtle, individual differences in a distinct spatial pattern of cortical thickness across typical development. This spatial pattern of cortical thickness is also associated with schizophrenia risk genes that have important functions during early brain development. Taken together, these findings suggest that increased genetic risk for schizophrenia is related to early subtle alterations during early brain development, setting up individuals with higher risk profiles to have a small biological vulnerability for later developing the illness.