Epigenetic suicide note

Szyf had discovered a way in which the early environment stably altered the genome for the rest of a rat’s life. “The question was,” he says, “is this relevant [for humans]? And the only way to know if it’s relevant is to look at human material.” Fortunately, his colleagues across town at the Douglas Mental Health Institute (DMHI) kept just the perfect material.

In 2005, Szyf and Meaney teamed up with Gustavo Turecki, director of the DMHI’s Quebec Suicide Brain Bank, a repository of brains from hundreds of suicide victims, complete with medical records and postmortem interviews with friends and family members. These detailed records allowed the researchers to reconstruct the victims’ personal histories, including whether they had been abused or neglected as children—the human equivalent of not getting licked and groomed. To test whether early childhood adversity left an epigenetic mark on the brain, the McGill team then compared cells from suicidal brains with those taken from people who died randomly in car crashes or other accidents.

First, the researchers investigated genes encoding ribosomal RNA because Szyf had already linked expression patterns with DNA methylation levels at these genes’ promoter regions in cell cultures (Mol Cell Biol, 27:4938–52, 2007). Szyf and his coauthors applied the exact same primers to the human brain tissues, and found that suicidal brains had highly methylated—and, thus, inactive—rRNA genes. What’s more, the epigenetic differences between suicidal and healthy brains were manifested only in the hippocampus, the epicenter of stress hormone regulation, but not in the cerebellum, which directs motor control (PLoS ONE, 3:e2085, 2008).

The results suggested that early childhood adversity was leaving a specific epigenetic mark on people who ultimately killed themselves, but the researchers couldn’t rule out other confounding life factors associated with suicide, such as depression or substance abuse. So, to tease apart the source of the environmental hardship, Szyf’s team then compared the brains of suicide victims who were abused as children with those who were not.

This time, instead of the rRNA locus, for logistical reasons the researchers returned to the GR gene, which is known to be hyperactive in people with major depression and which Szyf had already implicated in the rat studies of maternal care. “It’s the first suspect you would look at if you thought that epigenetics played a role in suicide,” Szyf says. Abused suicide victims indeed had significantly different methylation patterns in their hippocampi from both nonabused suicide victims and those who died suddenly in accidents, and these differences were linked with reduced GR expression (Nat Neurosci, 12:342–48, 2009). The other perfect control—brains from abused people who didn’t commit suicide—were not available because these people don’t generally die of unnatural causes. As for the nonabused suicide victims, “they committed suicide so there must be a reason, but we can’t pinpoint it,” says Szyf.

“Psychiatry has been in search of causative mechanisms of mental illness for many, many decades without finding any concrete clear evidence,” says Turecki. “[These findings] open the door to the development of treatment intervention and eventually prevention as well.”

These suicidal brain studies provide an “important signpost,” says Ezra Susser, an epidemiologist and psychiatrist at New York’s Columbia University. “This is exactly the direction we have to go.”

Szyf is now testing whether social adversity also leaves a chemical mark on T cells, since there is extensive crosstalk between the immune system and brain signaling. The goal is to identify markers that signal someone is at risk of suicide in nonbrain tissue (because once you have brain tissue, the patient is already dead). “Once this is possible it will change social studies forever,” Szyf says.