Dissecting Schizophrenia Biology Using Pleiotropy With Cognitive Genomics

Upasana Bhattacharyya; Jibin John; Todd Lencz; Max Lam

doi:10.1016/j.biopsych.2025.02.890

Dissecting Schizophrenia Biology Using Pleiotropy With Cognitive Genomics

Biol Psychiatry. 2025 Feb 22:S0006-3223(25)00989-8. doi: 10.1016/j.biopsych.2025.02.890. Online ahead of print.

Authors

Upasana Bhattacharyya¹, Jibin John¹, Todd Lencz², Max Lam³

Affiliations

¹ Institute of Behavioral Science, Feinstein Institutes for Medical Research, Manhasset, New York; Division of Psychiatry Research, Zucker Hillside Hospital, Glen Oaks, New York.
² Institute of Behavioral Science, Feinstein Institutes for Medical Research, Manhasset, New York; Division of Psychiatry Research, Zucker Hillside Hospital, Glen Oaks, New York; Departments of Psychiatry and Molecular Medicine, Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York. Electronic address: tlencz@northwell.edu.
³ Institute of Behavioral Science, Feinstein Institutes for Medical Research, Manhasset, New York; Institute of Mental Health, Singapore; Department of Population and Global Health, Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore. Electronic address: max.lam@ntu.edu.sg.

PMID: 39993652
DOI: 10.1016/j.biopsych.2025.02.890

Abstract

Background: Given the increasingly large number of loci discovered by psychiatric genome-wide association studies (GWASs), specification of the key biological pathways that underlie these loci has become a priority for the field. We have previously leveraged the pleiotropic genetic relationships between schizophrenia (SCZ) and 2 cognitive phenotypes (educational attainment and cognitive task performance) to differentiate 2 subsets of illness-relevant single nucleotide polymorphisms (SNPs): 1) those with concordant alleles, which are associated with reduced cognitive performance and educational attainment and increased SCZ risk, and 2) those with discordant alleles, which are linked to reduced educational and/or cognitive levels but lower SCZ susceptibility.

Methods: In the current study, we extended our prior work, utilizing larger input GWAS datasets and a more powerful statistical approach to pleiotropic meta-analysis, the pleiotropic locus exploration and interpretation using optimal test (PLEIO).

Results: Our pleiotropic meta-analysis of SCZ and the 2 cognitive phenotypes revealed 768 significant pleiotropic loci (166 novel). Among these, 347 loci harbored concordant SNPs, 270 encompassed discordant SNPs, and 151 dual loci contained concordant and discordant SNPs. Competitive gene-set analysis using MAGMA linked concordant SNP loci with neurodevelopmental pathways (e.g., neurogenesis), whereas discordant loci were associated with mature neuronal synaptic functions. These distinctions were also observed in BrainSpan analysis of temporal enrichment patterns across developmental periods, with concordant loci containing more prenatally expressed genes than discordant loci. Dual loci were enriched for genes related to messenger RNA translation initiation, which represents a novel finding in the SCZ literature.

Conclusions: Pleiotropic analysis permits not only enhanced statistical power for locus discovery but also the ability to parse distinct biological processes associated with endophenotypes.

Keywords: Cognition; Educational attainment; GWAS; Pleiotropy; Schizophrenia.