A reversible epigenetic memory of inflammatory injury controls lineage plasticity and tumor initiation in the mouse pancreas.

Inflammation is essential to the disruption of tissue homeostasis and can destabilize the identity of lineage-committed epithelial cells. Here, we employ lineage-traced mouse models, single-cell transcriptomic and chromatin analyses, and CUT&TAG to identify an epigenetic memory of inflammatory injury in the pancreatic acinar cell compartment. Despite resolution of pancreatitis, our data show that acinar cells fail to return to their molecular baseline, with retention of elevated chromatin accessibility and H3K4me1 at metaplasia genes, such that memory represents an incomplete cell fate decision. In vivo, we find this epigenetic memory controls lineage plasticity, with diminished metaplasia in response to a second insult but increased tumorigenesis with an oncogenic Kras mutation. The lowered threshold for oncogenic transformation, in turn, can be restored by blockade of MAPK signaling. Together, we define the chromatin dynamics, molecular encoding, and recall of a prolonged epigenetic memory of inflammatory injury that impacts future responses but remains reversible.

PRC2-Inactivating Mutations in Cancer Enhance Cytotoxic Response to DNMT1-Targeted Therapy via Enhanced Viral Mimicry.

Polycomb repressive complex 2 (PRC2) has oncogenic and tumor-suppressive roles in cancer. There is clinical success of targeting this complex in PRC2-dependent cancers, but an unmet therapeutic need exists in PRC2-loss cancer. PRC2-inactivating mutations are a hallmark feature of high-grade malignant peripheral nerve sheath tumor (MPNST), an aggressive sarcoma with poor prognosis and no effective targeted therapy. Through RNAi screening in MPNST, we found that PRC2 inactivation increases sensitivity to genetic or small-molecule inhibition of DNA methyltransferase 1 (DNMT1), which results in enhanced cytotoxicity and antitumor response. Mechanistically, PRC2 inactivation amplifies DNMT inhibitor-mediated expression of retrotransposons, subsequent viral mimicry response, and robust cell death in part through a protein kinase R (PKR)-dependent double-stranded RNA sensor. Collectively, our observations posit DNA methylation as a safeguard against antitumorigenic cell-fate decisions in PRC2-loss cancer to promote cancer pathogenesis, which can be therapeutically exploited by DNMT1-targeted therapy. SIGNIFICANCE: PRC2 inactivation drives oncogenesis in various cancers, but therapeutically targeting PRC2 loss has remained challenging. Here we show that PRC2-inactivating mutations set up a tumor context-specific liability for therapeutic intervention via DNMT1 inhibitors, which leads to innate immune signaling mediated by sensing of derepressed retrotransposons and accompanied by enhanced cytotoxicity. See related commentary by Guil and Esteller, p. 2020. This article is highlighted in the In This Issue feature, p. 2007.

A cell type-specific expression map of NCoR1 and SMRT transcriptional co-repressors in the mouse brain.

The ability to rapidly change gene expression patterns is essential for differentiation, development, and functioning of the brain. Throughout development, or in response to environmental stimuli, gene expression patterns are tightly regulated by the dynamic interplay between transcription activators and repressors. Nuclear receptor corepressor 1 (NCoR1) and silencing mediator for retinoid or thyroid-hormone receptors (SMRT) are the best characterized transcriptional co-repressors from a molecular point of view. They mediate epigenetic silencing of gene expression in a wide range of developmental and homeostatic processes in many tissues, including the brain. For instance, NCoR1 and SMRT regulate neuronal stem cell proliferation and differentiation during brain development and they have been implicated in learning and memory. However, we still have a limited understanding of their regional and cell type-specific expression in the brain. In this study, we used fluorescent immunohistochemistry to map their expression patterns throughout the adult mouse brain. Our findings reveal that NCoR1 and SMRT share an overall neuroanatomical distribution, and are detected in both excitatory and inhibitory neurons. However, we observed striking differences in their cell type-specific expression in glial cells. Specifically, all oligodendrocytes express NCoR1, but only a subset express SMRT. In addition, NCoR1, but not SMRT, was detected in a subset of astrocytes and in the microglia. These novel observations are corroborated by single cell transcriptomics and emphasize how NCoR1 and SMRT may contribute to distinct biological functions, suggesting an exclusive role of NCoR1 in innate immune responses in the brain.