Abstract 189: Postnatal Engrailed-1 Expression Activates A Pro-Fibrotic Transcriptional Program In Wound Fibroblasts

Abstract

Purpose: Skin scars represent a massive biomedical burden on patients, but the cellular mediators of scarring remain poorly understood. We previously showed that embryonic expression of Engrailed-1 (En-1) defines a lineage of fibroblasts (En-1-positive fibroblasts; eEPFs) responsible for the deposition of fibrotic scar tissue in dorsal skin. More recently, we demonstrated that a subpopulation of En-1-negative fibroblasts (ENFs) activates En-1 during adult wound healing and contributes to scar formation as postnatally-derived EPFs (pEPFs). However, it was not known if such postnatal En-1 expression is accompanied by the acquisition of a pro-fibrotic phenotype. We thus sought to determine if pEPFs are a viable therapeutic target to minimize fibrosis in postnatal wound healing. Methods: En-1Cre-ERT;Ai6 (En-1-positive cells GFP+, En-1-negative cells no reporter) mice were systemically induced with tamoxifen prior to dorsal excisional wounding. This novel transgenic mouse model reliably distinguishes ENFs (GFP-CD26-), embryonically-derived EPFs (GFP-CD26+), and postnatally-derived EPFs (GFP+). The scars and surrounding unwounded tissue were harvested upon complete wound healing (day 14), enzymatically-digested, and then sorted by fluorescence-activated cell sorting. Five sorted fibroblast populations (pEPFs from wounded skin; eEPFs from unwounded and wounded skin; and ENFs from unwounded and wounded skin) were then analyzed by bulk RNA-sequencing (experimental schematic in n = 2 biological replicates, 6 pooled mice each). Results: Hierarchical clustering and principal components analysis of differentially expressed genes after wounding revealed that pEPFs clustered more closely with eEPFs than with ENFs. Both postnatally- and embryonically-derived EPFs showed increased expression of fibrosis-related genes in response to wounding, including Dpp4 (CD26). In contrast, ENFs showed increased expression of mechanotransduction signaling-related genes (Notch ligands Jag1, Dll1), suggesting that they are responsive to wound mechanical cues. Supporting these findings, gene set enrichment analysis of ranked whole genomes revealed that scar ENFs enriched for terms related to ECM adhesion and Notch signaling, while postnatal EPFs enriched for terms related to ECM production. Finally, we compared transcriptional activity of genes known to differentiate ENFs and eEPFs. Once again, pEPFs diverged from ENFs, exhibiting a gene expression profile more closely resembling that of eEPFs. Conclusion: Postnatal En-1 activation in ENFs during wound healing is accompanied by the acquisition of a pro-fibrotic transcriptional profile similar to that of embryonically-derived EPFs. These RNA-seq data also support our recent finding that ENFs activate En-1 through a canonical mechanotransduction mechanism involving YAP and Notch, before transitioning to an ECM-producing phenotype (pEPF). Thus, inhibition of mechanotransduction signaling may mitigate scarring by blocking this phenotypic switch. In future studies, we will compare the chromatin profiles of ENFs, eEPFs, and pEPFs to determine whether postnatal En-1 activation recapitulates the epigenomic shift that occurs during embryogenesis.