Project 1: Repressive Histone and DNA Methylation in Rodent Depression Models (Eric Nestler, Mount Sinai)
Project 1 focuses on dimethylation of Lys 9 on histone H3 (H3K9me2), a major form of repressive histone methylation implicated in gene regulation in many systems. While H3K9me2 is just one of many histone modifications likely involved in depression-related phenomena, our initial focus on H3K9me2 is supported by several observations. 1) Chronic social defeat stress (CSDS) decreases global levels of H3K9me2 in NAc. 2) This change is associated with, and presumably mediated by, decreased expression of G9a and GLP (G9a-like protein), two histone methyltransferases (KMTs) that catalyze dimethylation of H3K9. 3) We have strong evidence that such repression of H3K9me2 in NAc controls susceptibility vs. resilience to CSDS. Conversely, enhancing H3K9me2 in NAc exerts an antidepressant-like effect. 4) Our recent ChIP-chip study provided a preliminary, but partly non-specific and incomplete, genome-wide map of H3K9me2 binding in the NAc after CSDS (see Figure). This genome-wide map, while imperfect, revealed novel biochemical pathways in NAc that control susceptibility vs. resilience to CSDS. Interestingly, we uncovered overlap between mechanisms of resilience and antidepressant action, suggesting fundamentally novel approaches for future drug discovery efforts. 5) Downregulation of H3K9me2 and of G9a/GLP, and associated regulation of several key target genes, have been validated in NAc of depressed humans postmortem. 6) We have more preliminary evidence for regulation of H3K9me2 and G9a/GLP in the NAc in several other chronic stress models, including CUS and IC/EC. We are now following up on this extensive preliminary work in several key ways. First, we are defining, using ChIP-Seq with more specific antibodies, the precise genes and non-genic regions that show stress-induced changes in H3K9me2 binding in NAc and thereby provide an unprecedented roadmap by which this histone modification controls depression- and antidepressant-like responses in CSDS. Second, we are comparing this role for H3K9me2 in CSDS with that in other key depression models over the life cycle of the animal. Third, we are studying the mechanisms by which these different forms of stress alter H3K9me2 binding at key target genes in the NAc, with a focus on the transcription factor, CREB (cAMP response element binding protein). We have evidence that H3K9me2/G9a and CREB are part of a pathological molecular circuit in NAc that drives susceptibility to CSDS. Fourth, we are focusing on stress regulation of H3K9me2 at a small number of selected target genes identified in our genome-wide analyses. One target of initial focus is DNMT3a (a DNA methyltransferase), which is induced in NAc by CSDS, mediated in part by reduced binding of H3K9me2. Such induction of DNMT3a, also seen in maternal separation stress and chronic unpredictable stress, are being compared to genome-wide DNA methylation patterns in NAc in conjunction with the Chromatin and Gene Analysis Core, with the goal of identifying genes whose methylation is regulated by chronic stress, something accomplished for only a small handful of genes to date. Finally, we are initiating comparable studies of these same chromatin mechanisms in PFC, where many equivalent changes are seen. Parallel molecular endpoints are being characterized in human brain via Project 4. Together, this work represents a uniquely penetrating analysis of the role played by repressive histone and DNA methylation in depression-related phenomena. The work also provides a template for how other major forms of histone modifications, as they are identified over the years, can be explored in depression models.
FIGURE: Regulation of dimethylation of lysine 9 of histone 3 (H3K9me2) and phospho-CREB (pCREB) binding in nucleus accumbens (NAc) in chronic social defeat stress (CSDS). Mice were subjected to 10 days of CSDS. Susceptible and resilient mice were identified on Day 11. Half of each group was then given imipramine (20 mg/kg ip) or vehicle daily for 28 days, at which time NAc was subjected to ChIP-chip for H3K9me2 and pCREB. A. The heatmaps show that, for both marks, the large majority of changes (red, up; green, down) seen in susceptible mice are not seen after chronic imipramine; as well, most changes observed in susceptible mice are not seen in resilient mice and vice versa. Note the lower bars in each heatmap are normalized to the upper bars, meaning that green would depict reversal of a red change, not a change in the opposite direction. Note also that the heatmaps to the right show results from vehicle-treated mice. Imipramine produced many fewer changes in resilient mice (not shown). B. The Venn diagrams show the appreciable overlap between apparent mechanisms of imipramine and of resilience. From Wilkinson et al., The Journal of Neuroscience, 29:7820-7832 (2009).