Abstract
We performed a multivariate gene-by-environment genome-wide interaction study (GEWIS) of suicidality in 123,633 individuals using a covariance matrix based on 26 environments related to traumatic experiences, posttraumatic stress, social support, and socioeconomic status. We discovered sex-specific risk loci exhibiting interaction effects driven by PTSD, including the male-specific rs2367967 (CWC22), which replicated in an independent cohort. Sex-specific cell-type transcriptome enrichment, gene-based analyses, and locus annotation point to extracellular matrix biology and synaptic plasticity as biological mediators of the effects of lifetime traumatic experiences on genetic risk for suicidality.
Introduction
Suicidality is a major global health crisis with a complex etiology. While its heritability is estimated between 25% and 48%, traditional Genome-Wide Association Studies (GWAS) have often failed to identify robust genetic signals. This discrepancy is likely due to the "stress-diathesis" nature of suicidality, where genetic risk (diathesis) remains dormant until activated by environmental stressors.
This study moves beyond the standard "nature vs. nurture" dichotomy by performing a massive, multivariate gene-by-environment genome-wide interaction study (GEWIS). By accounting for the interplay between DNA and life experiences, the researchers aimed to uncover genetic vulnerabilities that are specifically "unlocked" by trauma and posttraumatic stress disorder (PTSD).
Methodology and Datasets
The researchers utilized a sophisticated statistical framework called StructLMM (Structured Linear Mixed Model). Unlike traditional methods that analyze one environmental factor at a time, StructLMM allows for the simultaneous testing of interactions between genetic variants and a "covariance matrix" of multiple environments.
Cohorts
- Discovery Cohort: The UK Biobank (UKB), comprising data from 123,633 individuals of European ancestry.
- Replication Cohort: The Yale-Penn cohort ($N=3,099$), used to validate findings in an independent group.
Modeling
Suicidality was modeled on an ordinal scale (0–4), ranging from no suicidality to attempted suicide. The environmental model incorporated 26 factors, including childhood trauma, adult violence, social support, and socioeconomic status. Given the known differences in trauma response between sexes, analyses were performed on the full cohort as well as sex-stratified groups.
Key Results
The GEWIS identified five significant risk loci associated with suicidality through environmental interactions. Notably, these loci were not significant in standard GWAS, confirming that these genetic risks are purely interactive. PTSD and its associated symptom scores (PCL-6) were identified as the strongest environmental mediators for nearly all identified genetic risks.
As shown in the table above, the genetic architecture of suicidality appears highly sex-dependent. For instance, the locus rs2367967 (near the CWC22 gene) was specific to males and successfully replicated in the independent Yale-Penn cohort. Conversely, rs118118557 (near CHST14) was identified as a female-specific risk factor.
Biological Insights and Discussion
A central finding of the study is the convergence of these genetic risks on the Extracellular Matrix (ECM) and synaptic plasticity. The ECM is a dynamic structure in the brain that regulates memory and the ability of neurons to rewire themselves. These results suggest that trauma may physically alter the brain's "scaffolding" in individuals with specific genetic backgrounds, thereby increasing the risk for suicidality.
Furthermore, the researchers identified "qualitative interactions" or crossover effects. In these instances, a genetic variant that appears protective in a low-stress environment can become a significant risk factor in a high-stress environment. Beyond individual variants, gene-based analyses highlighted CNTN5, PSMD14, and HEPACAM as key biological players, with CNTN5 specifically linked to neurite outgrowth.
Limitations and Future Directions
While this study represents a major advancement, it is limited by its focus on individuals of European ancestry. Future research must include more diverse ancestral groups to ensure findings are broadly applicable. Additionally, while the study points toward the ECM, the exact molecular mechanisms by which PTSD triggers these genetic pathways require further investigation in animal models or human post-mortem tissue.
Ultimately, these findings suggest that "Environmental Enrichment" (EE)—a concept involving sensory and social stimulation—might one day be adapted as a personalized clinical intervention to mitigate genetic risk in trauma survivors.
Summary updated