2026 ECNP Congress

PS02-2224

Brain gene expression-informed analysis identifies retinoid receptor targets in suicidal thoughts and behaviours

Alfonso Martone — G. Fanelli, M. Arenella, L. Fortaner Uyà, A. Serretti, E.C.N.P. Network on Suicide Research and Prevention, B. Vai, F. Benedetti · Vita-Salute San Raffaele University, Faculty of Psychology, Milano, Italy

2026 ECNP Congress

Introduction

Suicidal thoughts and behaviours (STBs) comprise suicidal ideation (SI) and suicidal behaviours (SB), including suicide attempt (SA) and death by suicide. Collectively, STBs represent a major public health concern and cut across diagnostic boundaries in psychiatry. Although STBs exhibit a significant SNP-based heritability (h2SNP= 3-6.7%) [1] and are associated with alterations in brain morphology [2], little is known about how STB-associated genetic variants relate to brain tissue-specific gene regulation and molecular pathways. We investigated whether genetic liability for STBs converges on brain tissue-specific expression quantitative trait loci (eQTL)-informed gene-level associations and molecular pathways.

Methods

Publicly available summary statistics for SB, SA, and SI were obtained from the latest STB GWAS [1]. After quality control, eQTL-informed gene-based association testing was performed for each suicide phenotype using eMAGMA [3] and eQTL reference datasets for 13 brain tissues, including the amygdala, anterior cingulate cortex (BA24), basal ganglia, cerebellum, cerebellar hemisphere, cerebral cortex, frontal cortex (BA9), hippocampus, hypothalamus, nucleus accumbens, putamen, spinal cord (C1), and substantia nigra. MAGMA gene-set analysis (GSA) [4] was then performed on the mapped genes using the complete Molecular Signatures Database (MSigDB) of gene sets.

Results

After correcting for multiple testing, 2 to 43 significant differentially expressed genes (eGenes) were identified per trait–tissue analysis, resulting in 126 unique genes across the 13 brain tissues. SB and SA yielded the largest numbers of associated genes (SB, N=11–43; SA, N=7–37, across tissues), whereas SI showed fewer associated genes (N=2–13). Across STB phenotypes, the largest gene counts were observed for the cerebellum, cerebellar hemisphere, cerebral cortex, BA9, and nucleus accumbens. SB and SA also showed signals for striatal regions, including caudate and putamen, whereas SI showed additional mapping for the hippocampus and anterior cingulate cortex. GSA identified 14 significant gene sets, involving immune-related pathways, major histocompatibility complex region sets, and the chr3p21 cytoband. Notably, seven SA/SB trait–tissue pairs, involving SA/SB and the hypothalamus, BA9, BA24, and C1, were enriched for the FXR_IR1_Q6gene set, which collects genes harbouring an RXRA transcription factor binding site.

Conclusions

To our knowledge, this is the first study to test brain tissue-specific eQTL-informed gene-level associations for STBs and related molecular pathways. Results indicate that STBs are associated with brain-expressed genes across multiple tissues, with prominent signals in regions relevant to frontostriatal and corticolimbic systems and in the cerebellum, consistent with evidence from neuroimaging studies. GSA supported involvement of immune system-related molecular pathways and identified enrichment in Retinoid X Receptor (RXR) Alpha transcriptional targets. RXRs are nuclear receptors that mediate the biological effects of retinoids, acting as transcription factors for retinoic acid-mediated gene activation and mediating brain development, dendritic spine formation, and microglial activity. Altered retinoid signalling has previously been linked to psychiatric disorders, and exposure to retinoid drugs has been associated with reduced suicidal behaviours following prolonged exposure [5]. Although these findings do not establish causality, they nominate immune-related and retinoid signalling as candidate pathways for future mechanistic studies, warranting follow-up through colocalization, effector-gene prioritisation, integration with brain co-expression modules, and drug-target Mendelian randomisation.

References

  1. S. M. C. Colbert, the P. G. C. S. W. Group, D. Ruderfer, A. R. Docherty, and N. Mullins, ‘Genome-wide association studies identify 77 loci for suicidality and provide novel biological insights’, Oct. 23, 2025, medRxiv. doi: 10.1101/2025.10.22.25338076.
  2. L. S. van Velzen et al., ‘Structural brain alterations associated with suicidal thoughts and behaviors in young people: results from 21 international studies from the ENIGMA Suicidal Thoughts and Behaviours consortium’, Mol Psychiatry, vol. 27, no. 11, pp. 4550-4560, Nov. 2022, doi: 10.1038/s41380-022-01734-0.
  3. Z. F. Gerring, A. Mina-Vargas, E. R. Gamazon, and E. M. Derks, ‘E-MAGMA: an eQTL-informed method to identify risk genes using genome-wide association study summary statistics’, Bioinformatics, vol. 37, no. 16, pp. 2245-2249, Aug. 2021, doi: 10.1093/bioinformatics/btab115.
  4. C. A. de Leeuw, J. M. Mooij, T. Heskes, and D. Posthuma, ‘MAGMA: Generalized Gene-Set Analysis of GWAS Data’, PLOS Computational Biology, vol. 11, no. 4, p. e1004219, Apr. 2015, doi: 10.1371/journal.pcbi.1004219.
  5. N. K. W. Tan, A. Tang, N. C. Y. L. MacAlevey, B. K. J. Tan, and H. H. Oon, ‘Risk of Suicide and Psychiatric Disorders Among Isotretinoin Users: A Meta-Analysis’, JAMA Dermatol, vol. 160, no. 1, pp. 54-62, Jan. 2024, doi: 10.1001/jamadermatol.2023.4579.