Early life stress, such as childhood trauma, is linked with the development of depression in adulthood, but the mechanisms that drive lasting changes in the brain are not well understood. In a study published March 15 in Nature Neuroscience, researchers found that early-life stress in mice induces epigenetic changes in a particular type of neuron, which in turn make the animals more prone to stress later in life. Using a drug that inhibits an enzyme that adds epigenetic marks to histones, they also show that the latent effects of early-life stress can be reversed.

“It is a wonderful paper because it is really advancing our ability to understand how events that happen early in life leave enduring signatures in the brain so that they influence what we do as adults,” says Tallie Z. Baram, a child neurologist and developmental neurobiologist at the University of California, Irvine, who wasn’t involved with the study.

Early life is an important period for brain development, and stress during this time can have long-lasting effects on the brain via epigenetic changes, modifications to the genome that influence gene expression. One region of the brain involved in depression is the nucleus accumbens, which regulates motivation- and reward-linked behaviors.

“We’ve previously seen that early-life stress changed gene expression in a number of different brain regions [in mice], including the nucleus accumbens, but we didn’t know how these long-lasting changes were regulated,” says Cate Peña, a neuroscientist at the Princeton Neuroscience Institute and a coauthor of the study who was formerly a postdoc in the lab of Eric Nestler at the Icahn School of Medicine at Mount Sinai. She and her colleagues suspected that early-life stress (ELS) influences gene expression in the nucleus accumbens through epigenetic regulation such as histone modification, whereby chemical groups are added to or removed from histones, the proteins that DNA winds around.

The lasting effects of stress on the epigenome

To investigate this hypothesis, the team first stressed mouse pups by separating them from their mothers for four hours per day and giving them limited bedding. Control pups stayed with their mothers and had ample bedding. In adulthood, these mice faced a second bout of stress involving a confrontation with a larger, more aggressive mouse. In response to the tense encounter, ELS mice exhibited more depression-like behaviors than control mice, including lower social interaction, decreased exploration, and greater immobility in a swim test. 

Peña and her colleagues then dug into ELS epigenetic changes by investigating the roles of two enzymes that add and remove methyl groups to histones. The particular tag they looked is called H3K79me2—dimethylation at the 79th lysine residue of the histone H3 protein—a mark that has been associated with the upregulation of gene activity.

The writer enzyme that creates this mark is called Dot1l, and the eraser enzyme that removes it is Kdm2b. The team used viral-mediated gene transfer to overexpress or knockdown Dot1l and Kdm2b in D2 neurons of the nucleus accumbens. These brain cells bear so-called D2 dopamine receptors, whose activation was known to promote animals’ susceptibility to stress. The overexpression of Dot1l or knockdown of Kdm2b caused control mice to exhibit more depression-like behaviors, similar to ELS mice, after the stressful encounter with an aggressive mouse. Conversely, knocking down Dot1l or overexpressing Kdm2b in ELS mice reversed depression-related behaviors in response to stress.

In contrast, manipulation of Dot1l in D2 neurons of the prefrontal cortex region of the brain or in nucleus accumbens neurons that produce a different type of dopamine receptor had no effect on the behavior of ELS or control mice, evidence that the effects of ELS are specific to D2 neurons of the nucleus accumbens.

“The brain is incredibly heterogeneous, and one brain region can have so many different cell subtypes. So one nice thing about the study is that they actually . . . found the specific neuron subtype that these changes were occurring in, and they did a lot of elegant cell subtype manipulations to look at these neurons in lasting behavioral effects of early life stress,” says Mary Kay Lobo, a neuroscientist at the University of Maryland School of Medicine who did her postdoc with Nestler but did not contribute to the study.

To investigate the downstream effects of epigenetic regulation on gene transcription, the authors performed RNA sequencing on D2 neurons. Overall patterns of gene upregulation and downregulation were very similar between ELS mice and control mice with overexpression of Dot1l, more evidence that this enzyme and H3K79me are responsible for transmitting the effects of stress over time.

“Cell-type specific changes in Dot1l not only mimics the effects of early-life stress on behavior, but also mimics the effects of early-life stress on gene expression across the genome” says Peña. “To mimic or rescue changes in gene expression across hundreds of genes was very exciting and proved to us . . . that Dot1l and H3K79me2 were important regulators of how the brain responds to early life stress.”

Unexpectedly, when the researchers looked at the overall abundance of H3K79me2 tags in the nucleus accumbens of ELS mice, they found it was lower than in control mice. “However, looking at where H3K79me2 was deposited across the genome seemed to resolve this,” says Peña. “There were both regions that were enriched for H3K79me2 and depleted for it [in ELS mice].” In other words, it was the pattern of changes that appears responsible for transmitting stress across an animal’s life, rather than the overall amount of H3K79me2.

Drug reverses the effects of early-life stress

Although the effects of ELS could be reversed with genetic manipulation in mice, such an intervention wouldn’t likely be feasible in people. So the researchers also investigated the effects of pinometostat, a small molecule inhibitor of the DOT1L enzyme that is currently in Phase 2 trials for the treatment of a type of leukemia. They injected the drug or saline control into the body cavity of adult ELS mice twice daily over a 10-day period of social stress. Compared with saline control mice, those receiving the drug had reduced H3K79me2 levels in the nucleus accumbens and had better social interaction scores after their time with an aggressor. 

“[This finding] gives hope for the future that although early-life stress might change epigenetic modifications in the brain, it is possible to reverse some of those effects and ameliorate the long-lasting impact on depression-like behavior,” says Peña.

According to Baram, this mechanism is likely to be just one of ELS’s effects on the brain. “The brain is so complex, and the effects of early-life stress are so profound and important. There are many, many different mechanisms,” says Baram. “[The researchers] pull out some very important mechanisms with wonderful resolution, but we need to realize that this is only part of the story, and there are many more aspects of that story.”

H. Kronman et al., Long-term behavioral and cell-type-specific molecular effects of early life stress are mediated by H3K79me2 dynamics in medium spiny neurons. Nat Neurosci. doi:10.1038/s41593-021-00814-8, 2021.