We propose a set of studies focused on the association of suicide with neuroinflammation and compromise of the blood-brain barrier, with the goal of identifying a pattern of quantifiable abnormalities that could serve as a biomarker for imminent suicidal risk in our Veterans. Autopsy studies are uniquely suited to do this, because they capture the state of the brain at the time of the suicidal act. Findings from our laboratories and others indicate that susceptibility to suicide includes inflammatory activation in the brain and systemically, accompanied by compromised integrity of the blood-brain barrier: (1) Most directly, we reported increased densities of microglia or other phagocytic cells associated with blood vessels in dorsal prefrontal white matter of people who died by suicide, Similar results are reported in cingulate white matter. (2) Studies of brains from individuals who died by suicide and studies of blood and CSF from live individuals who had previously attempted suicide found elevations of inflammatory cytokines. (3) Various infectious diseases are associated with increased risk of suicide, as is a history of hospitalization for any infection. (4) Laboratory animals exposed to stress show elevated levels of inflammatory cytokines, increased permeability of the blood-brain barrier, behavioral abnormalities, and activation of microglia. (5) We have reported an association of suicide with a polymorphism and decreased frontal and cingulate transcripts for CD44, which is involved in the normal function of the BBB. (6) Biochemical measures suggesting BBB impairment are reportedly associated with attempted suicide and with suicidal ideation. (7) In MDD subjects who died by suicide, compared with nonpsychiatric non-suicide cases, we found differential methylation of genes associated with cell death, both in whole cortical homogenates and in purified neuronal fractions. We also found significantly lower methylation in the promoter of the gene for CCL3, a powerful inflammatory cytokine synthesized by microglia and astrocytes and an attractant for microglia and white blood cells, but this difference was not present in the purified neuronal fraction. Taken together, these findings lead us to hypothesize a suicidal state characterized by impaired BBB function, elevation of pro-inflammatory cytokines, and abnormalities in DNA methylation of genes stimulating inflammation, all of which can be assessed in live individuals. To confirm this phenotype, we propose three specific aims, each employing the same set of 90 autopsy brains, already collected. In order to distinguish features of suicide from those of psychiatric illness, we employ a 3-group design with 30 cases of psychiatric disease and suicide, 30 cases of psychiatric disease without suicide, and 30 cases with neither psychiatric disease nor suicide, all from a well-characterized collection with a single collection protocol at a single autopsy service. To optimize our ability to distingu...