679. DISSECTING EPIGENETIC HETEROGENEITY IN SCHIZOPHRENIA

Abstract Background Clinical management of schizophrenia is complicated by vast heterogeneity in terms of onset, progression, severity and treatment response. While varying combinations of genetic and environmental risk factors are thought to underpin this heterogeneity, recent epigenetic studies suggest DNA methylation may also contribute. Consequently, there is a growing need to identify DNA methylation changes pertinent to phenotypic heterogeneity in schizophrenia to better understand drivers of individual variability and explore precision medicine. Aims & Objectives We hypothesised that DNA methylation at CpG dinucleotides contributes to clinical variability in schizophrenia. To address this, we: 1) characterised methylation patterns associated with measures of progression and severity, and 2) identified highly variable CpG sites potentially related to phenotypic heterogeneity. Method Blood DNA methylation was quantified using Illumina MethylationEPIC arrays in 381 individuals with schizophrenia from the Australian Schizophrenia Research Bank. Patterns of methylation were explored in association with age of onset, cognitive deficits, global assessment of functioning scores, treatment resistance and common variant polygenic risk scores. In a follow-up study, we conducted the largest ever meta-analysis of DNA methylation variance in schizophrenia using three publicly available cohorts (Nschizophrenia = 1036, Nnon-psychiatric control = 954). Results We identified 662 CpG sites suggestively associated with (P < 6.72 x 10–5) clinical features of schizophrenia. Interestingly, these sites overlap several genes related to neuronal function (e.g. CACNA1C, NGFR) and converge on methylation patterns previously observed for known comorbidities and risk factors (e.g. myocardial infarction, type-2 diabetes). The meta-analysis of DNA methylation variance uncovered 213 loci with significant variability among individuals with schizophrenia. Despite being sampled from blood, these methylation events were overrepresented amongst brain-enriched genes, and several showed evidence for concordant changes in post-mortem brain. Strikingly, 48 CpG sites with high variability also exhibited evidence for a mean effect in association with clinical features. Discussion & Conclusions Our findings support a role for DNA methylation in schizophrenia-associated clinical variability, wherein some CpG sites exhibit high inter-individual variability and/or association with progression and severity. These results emphasise DNA methylation as an exciting prospect for reconciling molecular and phenotypic heterogeneity in schizophrenia, therein providing a promising avenue for precision medicine.