Research articleGenetic association analysis of serotonin and signal transduction pathways in suicide attempters from an Italian sample of psychiatric patients

- A polygenic risk score detected for three genes HTR2A (A-1438G), TPH1 and TPH2 increased the prediction of SA risk.
- A gender stratification evidenced nominal associations with HTR1A, TPH1 and GNB3 genes.
- A phenotypic dissection evidenced nominal associations with SLC6A4, HTR1A, HTR2A (-1438A/G) and GNB3 genes.
- Although some limitations, this could represent a further study for future meta-analyses in larger samples.

Genetic factors have been reported to contribute to the liability of suicide. We aimed to investigate functional polymorphisms in eight genes (serotonin transporter, SLC6A4; receptors, 5HTR1A, 1B, 5HTR2A; Tryptophan Hydroxylase, TPH1, TPH2; Monoamine Oxidase, MAOA and G Protein Subunit Beta 3, GNB3) to investigate their predictive value for suicide. The possible confounding effects of gender and phenotypic patients dissection were also valued. A sample of 111 consecutive psychiatric inpatients was recruited and assessed using specific psychometric instruments. Genomic DNA was isolated from peripheral white blood cell samples and polymorphisms were genotyped by pyrosequencing technology. Although no differences were observed between allele and genotype frequencies for all polymorphisms and suicide attempt (SA), a polygenic risk score was detected for three genes HTR2A (A-1438G), TPH1 and TPH2 increasing the prediction of SA risk (Thresh = 0.43, p = 0.038, R2 = 0.053). Moreover some nominal associations were obtained after gender and phenotypic dissection stratification (TEMPS-A, TEMPs-H, GSMD, SHSS, GAF, CGI) for SLC6A4 (5-HTTLPR), HTR1A (C-1019G), HTR2A (A-1438G), TPH1 (A799C) and GNB3 (C825T) genes, that were lost after Bonferroni correction. This is a first evidence that specific additive combinations of genes could increase the prediction of SA risk and that gender and phenotypic dissection could influence the association of the genes with SA. This could represent a further study also for future meta-analyses on larger samples.