Prrx1 Marks Ventral Fibroblasts With Increased Fibrogenic Potential
Mimi R. Borrelli, MBBS, MSc, Tripp Leavitt, MD, Michael S. Hu, MD, Julia T. Garcia, Michael Januszyk, MD, PhD, Alessandra L. Moore, MD, PhD, Shamik Mascharak, BS, Michelle Griffin, MD, PhD, Derrick C. Wan, MD, Hermann P. Lorenz, MD, Geoffrey C. Gurtner, MD, Howard Y. Chang, PhD, Michael T. Longaker, MD, MA.
Stanford, menlo park, CA, USA.
Purpose: Skin fibrosis amounts to significant morbidity due to the prevalence of trauma and burn injuries. Fibroblasts are chiefly responsible for extracellular matrix (ECM) deposition in the skin and are increasingly recognized to be a heterogeneous population comprised of multiple subpopulations with distinct functions and scarring propensity. While anatomic and embryonic origin are known to determine fibroblast heterogeneity, most prior work in mice has focused on the dorsal dermis. Here, we explore how embryonic lineage and dermal spatial location define fibroblast heterogeneity in the mouse ventral dermis in both homeostasis and in acute and chronic fibroses. Methods: Prrx1Cre;R26mTmG mice were used to identify two ventral dermal fibroblast lineages; Prrx1-positive fibroblasts (GFP+ PPFs), and Prrx1-negative (RFP+ PNFs) (Fig.1A). Fibrogenic potential was explored by comparing fibroblast abundance using FACS and histology: 1) throughout 6 developmental timepoints; embryonic day(E)16.5 (non-scarring), E18.5, postnatal days(P) 1, 30, and 60 (scarring) (Fig.1B); 2) in wounds (acute fibrosis, Fig.1Ci); 3) following irradiation (chronic fibrosis, Fig.1Di); and 4) after implantation of melanoma cells (chronic fibrosis, Fig.1Ei). Reciprocal transplantation into the (non-scarring) oral dermis and co-localization with collagen type 1 by Imaris 24 hours after transplantation was used to assess whether fibrogenic potential was cell-intrinsic (Fig.1Fi&iii). Single-cell sequencing of PPFs and PNFs from unwounded and scarred ventral skin was performed using 10X Genomics. Manifold-based dimensionality reduction was used to delineate transcriptionally-distinct fibroblast clusters (subpopulations) and non-linear discriminant analyses were applied to determine cluster expression profiles and surface markers and thus identify papillary and dermal subpopulations (Fig.1Gi). Gene-set enrichment analysis was performed to confirm cluster profibrotic potential, and immunofluorescence and FACS were performed to confirm findings at the protein level. Results: PPFs progressively increased as a proportion of the total ventral fibroblasts over development, with the largest difference seen E16.5 to E18.5 (Fig.1Bi-ii). In adult ventral skin, there were significantly more PPFs in wounded (Fig.1Cii) and irradiated skin (Fig.1Dii), as well as in tumor stroma (Fig.1Eii). PPFs transplanted into the oral mucosa exhibited 23.89% collagen I co-localization, whereas Wnt-1positive (non-scarring) fibroblasts transplanted into the ventral dermis exhibited low collagen co-localization (1.49%), demonstrating that the fibrogenic potential of PPFs is cell-intrinsic (Fig.1Fii&iv). Single cell analysis revealed unique PPF and PNF subpopulations which responded differently to wounding; the papillary (CD26+) fibroblasts were largely PPFs and expanded with wounding (Fig.Gii-iii), whereas reticular (Dlk1+) fibroblasts were PNFs and minimally responded to wounding (Fig.Giv-vi).
Conclusion: For the first time, we highlight the functional heterogeneity of ventral dermal fibroblasts. We show how embryonic expression of paired related homeobox 1 (Prrx1) gives rise to two distinct fibroblast lineages comprised of unique fibroblast subpopulations with distinct fibrogenic potential. These data highlight that PPFs are responsible for the majority of connective tissue deposition in ventral skin and targeted modulation of PPF papillary subtypes may reduce the acute fibrosis during wound repair and the chronic fibrosis of tumor stroma formation.
Back to 2020 Abstracts