T.M. Bauer1, K.D. Mangum1, J. Shadiow3, A. Joshi3, S. Buckley1, R. Wasilowski2, B. Moore3, J. Gudjonnson2, F. Davis1, K.A. Gallagher1,3 1Michigan Medicine, Vascular Surgery, Ann Arbor, MI, USA 2Michigan Medicine, Dermatology, Ann Arbor, MI, USA 3Michigan Medicine, Immunology, Ann Arbor, MI, USA
Introduction: Early after tissue injury, fibroblasts transition to myofibroblasts to contract and close wounds. In pathologic conditions where wounds fail to heal (such as diabetes), fibroblasts do not increase myofibroblast gene expression. The epigenetic mechanisms that control myofibroblast transition are not defined.
Methods: Mice were subjected to either a normal diet or high fat diet resulting in diet-induced obesity (DIO), a murine model of diabetes. Fibroblasts were magnetically isolated from wounded mice on days seven following punch biopsy creation. Isolated fibroblasts were treated with TGF-β and subjected to reverse transcription-quantitative polymerase chain reaction (RT-qPCR) with a focus on myofibroblast genes. JMJD3 was inhibited with a small molecule inhibitor (GSK-J1) or knocked down with a fibroblast specific, tamoxifen induced jmjd3 knockout via a cre-recombinase driver unique to our laboratory. Important findings were translated to human tissue by utilizing single cell RNA sequencing (scRNAseq) from normal and T2D wounds.
Results: First, we identified that the cytokine TGF-β increases myofibroblast gene in isolated wound fibroblasts in vitro. We performed an unbiased epigenetic superarray on TGF-β stimulated dermal fibroblasts, where jmjd3, a chromatin modifying enzyme known to increase transcription of genes by modifying the DNA chromatin structure, was significantly upregulated. ScRNAseq of human wounds confirmed that jmjd3 is strongly upregulated in wound myofibroblasts. Analysis of murine wound fibroblasts showed that DIO mice failed to upregulate jmjd3 expression in wound fibroblasts compared to normal diet counterparts. Our lab has previously shown that injection of GSK-J1, a small molecule inhibitor of JMJD3 reduces wound healing, but the mechanism remains incompletely defined. Therefore, we hypothesized that JMJD3 regulated myofibroblast transition following tissue repair. Surprisingly, fibroblasts isolated from our mice with a fibroblast-specific knockdown of jmjd3 did not display reduced levels of traditional myofibroblast genes (e.g. smooth muscle actin, transgelin). To define a role for JMJD3, an unbaised superarray focused on remodelling and cell adhesion pathways identified that knockdown of JMJD3 resulted in reduced expression of many structural and remodelling genes, most notably the integrin-β subunit family, which is known to connect cells to the extracellular matrix and aid in wound healing. Small molecule inhibition of JMJD3 and siRNA knockdowns of jmjd3 expression confirmed JMJD3’s regulation of the integrin-β family in wound fibroblasts.
Conclusions: This work identifies JMJD3 as a TGFb-mediated regulator of structural genes necessary for wound healing. Further investigation of the epigenetic mechanisms by which myofibroblast transition is regulated may result in novel pathways to improve wound healing.