Systems-Level Analysis of Genetic Variants Reveals Functional and Spatiotemporal Context in Treatment-resistant Schizophrenia

Treatment-resistant schizophrenia (TRS) occurs in one-third of the patients, but the molecular determinants of poor antipsychotic response remain unclear. We compared genetic data of patients with TRS ( n = 63) with non-TRS ( n = 111) by polygenic risk scores (PRS) calculated by PRSice software using PGC2_SCZ (Psychiatric Genomics Consortium – Schizophrenia) data. TRS criteria followed the International Psychopharmacology Algorithm Project SCZ algorithm. Statistical clustering and functional enrichment analyses of genes harboring TRS-linked variants were performed. Individuals on the top three deciles of schizophrenia PRS distribution exhibited higher odds of being refractory to antipsychotics than those on the bottom three deciles. Clusters of interacting variant-harboring genes were identified among the association signals. They are upregulated in the dorsolateral prefrontal, orbitofrontal, temporal, and inferior parietal areas during adolescence and early adulthood. Similar gene modules were found using transcriptional data from the same brain regions in individuals with schizophrenia. Genes were enriched among markers of cortical interneurons and somatosensory pyramidal cells. Finally, the enrichment of the clustered genes in drug-response expression signatures revealed compounds that could be employed to identify novel antipsychotic targets. In conclusion, we identified variant-harboring genes that may predispose SCZ patients to poor antipsychotic response and found statistically enriched clusters which provided functional and spatiotemporal context for TRS, suggesting that genotypic variation may converge to biological alterations at the interplay between actin dynamics and synaptic organization.