NRN1 genetic variability and methylation changes as biomarkers for cognitive remediation therapy response in schizophrenia

Carmen Almodóvar-Payá,Irene París-Gómez,Mariona Latorre-Guardia,Maria Guardiola-Ripoll,Rosa Catalán,Bárbara Arias,Rafael Penadés,Mar Fatjó-Vilas
DOI: https://doi.org/10.1016/j.pnpbp.2024.111175
2024-10-18
Abstract:Cognitive remediation therapy (CRT) demonstrates potential in enhancing cognitive function in schizophrenia (SZ), though the identification of molecular biomarkers remains challenging. The Neuritin-1 gene (NRN1) emerges as a promising candidate gene due to its association with SZ, cognitive performance and response to neurotherapeutic treatments. We aimed to investigate whether NRN1 genetic variability and methylation changes following CRT are related to cognitive improvements. Twenty-five SZ patients were randomly assigned to CRT or treatment-as-usual (TAU) groups, with cognitive function and NRN1 methylation assessed pre- and post-intervention using the MATRICS Consensus Cognitive Battery and EpiTYPER. Besides, eleven NRN1 polymorphisms were genotyped. Methylation changes (Δm = post - pre) were analyzed via sparse Partial Least Square Discriminant Analysis (sPLS-DA) to identify latent components (LCs) distinguishing CRT from TAU. To further explore methylation patterns of these LCs, CpG units were grouped into two subsets, yielding Δm means for those with increased and decreased methylation. Cognitive changes (Δcog = post - pre) were used to identify CRT improvers (CRT-I, Δcog ≥ 1), and the association between methylation changes and cognitive improvements post-therapy was also tested. We identified two LCs that differentiated CRT from TAU with a classification error rate of 0.28. The main component, LC1, included 25 CpG units. The subsets of CpG units with increased and decreased post-therapy methylation differed significantly between the two treatment arms, suggesting that differences were not merely data-driven but reflected meaningful biological variation. Additionally, CpG units linked to therapy were also associated with cognitive improvement, with LC1 and the subset of CpG units showing increased methylation post-therapy distinguishing CRT-I from the rest of the patients across multiple cognitive domains. Furthermore, the effect of LC1 on speed processing improvement after CRT was enhanced by considering the NRN1-rs9405890 polymorphism. Notably, these CpG units, particularly those with increased methylation after CRT, overlapped with key gene regulatory elements. Our model, integrating genetics and epigenetics, boosts the understanding of CRT response variability and highlights this multi-level approach as a promising strategy for identifying potential NRN1-related biomarkers of CRT effects, though further studies with larger samples are needed.
What problem does this paper attempt to address?