Comparative analysis of single-cell and spatial data from multiple fibrotic skin conditions

Summary: Abstract Body: Cutaneous fibrotic conditions are characterized by over-production of extracellular matrix (ECM) components, inflammation, and vascular dysregulation. However, whether these share unique or overlapping mechanisms is unknown. We performed single-cell RNA sequencing of skin biopsies from 4 healthy controls, 8 hypertrophic scars (HS), 8 keloid scars (KS), 5 morphea, 4 post-radiation morphea, 2 lichen sclerosus (LS), and 4 early systemic sclerosis (SSc) patients, along with single-cell resolution spatial profiling (Xenium). A total of 206,882 cells were annotated into 12 major cell types. This included 67,347 fibroblasts (FB) in eight clusters, with four clusters showing a profibrotic signature. These included two distinct myofibroblast (MF) subsets – COL8A1+ and TNC+ MF – that segregated along separate pseudotime trajectories. KS, HS, and morphea showed expansion in both MF subsets, but SSc showed an increase only for COL8A1+ MF. Variable immune infiltrate was seen across fibrotic diseases, with notable T cell expansion in SSc, morphea, and LS but not KS or HS. Nine upregulated genes in FBs were common to all fibrotic conditions: JUNB, FOS, RUNX1, EGR1, CCN1, CCN2 (CTGF), TNC, PRSS23, and MCL1, also validated on spatial profiling. Distinct mechanisms included enrichment of Wnt signaling terms in KS and HS but TGFb signaling terms in SSc, morphea, and LS. Of the five endothelial cell subtypes, only the pre-venular capillary (PVC) subtype also overexpressed mesenchymal cell markers. Highest ECM production by COL8A1+ MFs and the PVC endothelial cell subtype was observed in SSc, with cellular crosstalk between COL8A1+ FBs and PVCs being strong only in SSc. Our data identify a defined number of shared pathways as well as distinct disease mechanisms involving FBs and endothelial cells in different fibrotic skin conditions, with immune contribution being highly variable among different fibrotic states. Poulami Dey¹, Rachael Bogle¹, Paul Harms¹, Allison C. Billi¹, Joseph Kirma¹, Jennifer Fox¹, Olesya Plazyo¹, Mehrnaz Gharaee-Kermani¹, Michelle Kahlenberg¹, Alex Tsoi¹, John Varga¹, Dinesh Khanna¹, Johann E. Gudjonsson¹ 1. University of Michigan Michigan Medicine, Ann Arbor, MI, United States. Bioinformatics, Computational Biology, and Imaging