Sex-specific transcriptional and proteomic signatures in schizophrenia

Abstract

It has remained unclear why schizophrenia typically manifests after adolescence and which neurobiological mechanisms are underlying the cascade leading to the actual onset of the illness. Here we show that the use of induced pluripotent stem cell-derived neurons of monozygotic twins from pairs discordant for schizophrenia enhances disease-specific signal by minimizing genetic heterogeneity. In proteomic and pathway analyses, clinical illness is associated especially with altered glycosaminoglycan, GABAergic synapse, sialylation, and purine metabolism pathways. Although only 12% of all 19,462 genes are expressed differentially between healthy males and females, up to 61% of the illness-related genes are sex specific. These results on sex-specific genes are replicated in another dataset. This implies that the pathophysiology differs between males and females, and may explain why symptoms appear after adolescence when the expression of many sex-specific genes change, and suggests the need for sex-specific treatments.

Molecular signaling pathways underlying schizophrenia

“…We applied knowledge-based Ingenuity Pathway Analysis [9] to investigate two published DEG datasets comparing hiPSC-derived neurons from patients with schizophrenia to those of unrelated healthy individuals [8,10]. In addition, we aimed to include a comparison of monozygotic twin pairs discordant for schizophrenia, but this comparison yielded only one DEG surviving correction for multiple comparisons in ref [8], and no pathway analysis was possible.…”

“…We applied knowledge-based Ingenuity Pathway Analysis [9] to investigate two published DEG datasets comparing hiPSC-derived neurons from patients with schizophrenia to those of unrelated healthy individuals [8,10]. In addition, we aimed to include a comparison of monozygotic twin pairs discordant for schizophrenia, but this comparison yielded only one DEG surviving correction for multiple comparisons in ref [8], and no pathway analysis was possible. In order to have a larger number of DEGs for pathway analysis, we made a new comparison with neuronal cells dissociated for final maturation from neuroprogenitors maintained in culture for 12 weeks, including the same five twin pairs as earlier (in the previous studies, hiPSC-derived neuronal cells were maintained in culture for 6 [10] and 10 [8] weeks).…”

“…In addition, we aimed to include a comparison of monozygotic twin pairs discordant for schizophrenia, but this comparison yielded only one DEG surviving correction for multiple comparisons in ref [8], and no pathway analysis was possible. In order to have a larger number of DEGs for pathway analysis, we made a new comparison with neuronal cells dissociated for final maturation from neuroprogenitors maintained in culture for 12 weeks, including the same five twin pairs as earlier (in the previous studies, hiPSC-derived neuronal cells were maintained in culture for 6 [10] and 10 [8] weeks). In total, these datasets included results from 37 individuals (16 patients, 16 healthy unrelated controls, and 5 unaffected monozygotic twins).…”

“…In the US dataset, 9 of 11 patients were Caucasian and 2 Caucasian-Hispanic, and 5 of 11 controls were Caucasian, the rest 6 having mixed ethnicity (Supplementary data 1 [10]). From all three datasets, genes with a minimum of 2-fold change in expression and adjusted p<0.05 (232 DEGs/231 genes mapped by IPA [8] for affected twin vs. controls, and 356 DEGs/352 mapped by IPA for unaffected twin vs. controls; new hiPSC-derived neuron data made for this work, 69 DEGs/all mapped by IPA) or nominal p<0.05 (108 DEGs/100 mapped by IPA [10]) were used as input for IPA core analysis, and run with default settings against data content version 49932394 (dated 2019- [11][12][13][14][15]. In causal pathway analysis, the software used previous knowledge on molecular pathways to identify a master regulator, which regulated a cluster of genes via several upstream regulators.…”

“…However, the pathophysiology of neurological and psychiatric disorders can be modeled using human induced pluripotent stem cell (hiPSC)-derived neurons. So far, these studies [7,8] have tried to identify abnormalities related to illness by analyzing how differentially expressed genes (DEGs) interact in established pathways such as GO and KEGG, based on previous literature on gene functions, but no reproducible results identifying novel potential treatments have been published using these analyses. Ingenuity Pathway Analysis (IPA) is a data-driven method that may increase the possibilities to detect disease-specific abnormalities [9].…”