Approaching the Molecular Pathology of Suicide

Abstract

uicide risk and personality traits associated with suicidal behavior are often related to childhood maltreatment and other “environmental” stressors (1). Therefore, the pathohysiology of suicide is likely to involve molecular mechanisms hat are not necessarily related to genetic factors (defined here as hanges in DNA sequence), albeit this hypothesis is difficult to est conclusively. Somatic cells (including neurons) employ a wide array of olecular mechanisms to alter and regulate gene expression nd function without changing the sequence of the genomic NA itself. These include (but are not limited to) “epigenetic” echanisms, which are commonly defined by a chemical odification of the DNA and/or the histone proteins associted with it. For example, the methylation of DNA cytosine esidues at the sites of CpG dinucleotides within gene prooters is an epigenetic mark typically associated with represion or downregulation of RNA transcription (2). In addition, ene expression is also regulated by a rich set of postranslational modifications of specific histone residues; these nclude lysine acetylation, methylation, SUMOylation and biquitinylation, arginine methylation, serine phosphorylaion, and proline isomerization (2). Now, two provocative articles, one by Poulter et al. in this ournal (3) and the other one by McGowan et al. in PLoS ONE 4), provide evidence for DNA methylation changes in the frontal ortex (3) and hippocampus of suicide victims (4). Poulter et al. (3) focused on the promoter of the ionotropic -aminobutyric acid (GABA)-A a1 receptor subunit, which acording to their previous work is expressed at decreased levels in he anterior prefrontal cortex of suicide victims (5). However, it hould be noted that another study, conducted in anterior ingulate—which is in close proximity to the prefrontal cortex— eported a consistent increase in GABA-A a1 expression in ompleted suicides (6). The clinical cohorts in both studies were efined by an underlying diagnosis of mood disorder, and herefore the discrepancy remains to be resolved. If not exlained by demographic differences, medication history, postortem variables, and other confounds, one interesting hypothsis would be that levels of GABA-A a1 transcript show opposing lterations in different subdivisions of the frontal lobe of suicidal ubjects. In any case, Poulter et al. (3) now report that a small ubset of CpG dinucleotides at the proximal GABA-A a1 prooter are hypermethylated in prefrontal cortex of 10 male uicide victims diagnosed with major depression, compared with on-psychiatric control subjects who died from other causes. The ethylation levels in control subjects ranged from 6% to 10% but ere as high as 38% in suicide victims. When taken together with heir previous finding of decreased GABA-A a1 messenger RNA