Poster 155: Muscle-Derived Beige Adipose Precursors Secrete Promyogenic Exosomes that Treat Rotator Cuff Muscle Degeneration in Mice and are Identified in Humans by Single-Cell RNA Sequencing

Abstract

Objectives:Muscle atrophy, fibrosis, and fatty infiltration are common to a variety of sports-related and degenerative conditions and are thought to be irreversible. Fibroadipogenic progenitors (FAPs) are multipotent resident muscle stem cells with the capacity to differentiate into fibrogenic as well as white- and beige adipose tissue (WAT; BAT). FAPs that have assumed a BAT differentiation state (FAP-BAT) have proven efficacious in treating muscle degeneration in a variety of injury models. The purpose of this study was to characterize the subpopulation of murine FAPs with FAP-BAT activity, determine whether their pro-myogenic effect is mediated via exosomes, and analyze human FAPs for an analogous pro-myogenic, exosome-rich subpopulation. Given their non-myogenic differentiation capacity, it was hypothesized that FAP-BAT exerts a pro-myogenic effect on the injured muscle environment through the secretion of exosomes.Methods:FAPs from UCP1-reporter mice were isolated via FACS and sorted according to the differential intensity of the UCP1 signal observed: negative for UCP1 (UCP1-), intermediate intensity (UCP1+), and high intensity (UCP1++). Bulk RNAseq was performed on UCP1-, UCP1+, and UCP1++ FAPs to evaluate distinct characteristics of each population. Exosomes were harvested from UCP1++ (Exo-FB) and UCP1- (Exo-nFB) cells using cushioned-density gradient ultracentrifugation and used to treat C2C12 cells and MEFs in vitro, and the myotube fusion index was assessed. Exo-FB and Exo-nFB were then used to treat WT C57B/L6J mice that had undergone massive rotator cuff tear. At 6 weeks mice were sacrificed and supraspinatus muscles were harvested and analyzed for muscle atrophy, fibrosis, fatty infiltration, and UCP1 expression. Single-cell RNA sequencing (scRNAseq) was then performed on FAPs isolated from human rotator cuff muscle that were treated with mirabegron or standard media to assess for the presence of a parallel promyogenic subpopulation of FAP-BAT cells in humans.Results:Flow cytometry analysis of sorted UCP1-reporter mouse FAPs revealed a trimodal distribution of UCP1 signal intensity, which correlated with three distinct transcriptomic profiles characterized with bulk RNAseq. UCP1++ cells were marked by high mitochondrial gene expression, BAT markers, and exosome surface makers (Table 1). UCP1- cells were marked by fibrogenic markers and UCP1+ cells were characterized differential enrichment of WAT markers (Table 1). Exo-FB treatment of C2C12 cells resulted in robust myotube fusion, with a myotube fusion index of 33.3 ± 6.3% compared to 7.1 ± 3.2% with Exo-nFB (p <0.001) and 8.8 ± 3.3% with exosome-free media (p = 0.0014), while treatment of MEFs resulted in transdifferentiation into myotubes (Fig. 1). Mice that were treated with Exo-FB at the time of rotator cuff injury demonstrated markedly reduced muscle atrophy and FI compared to treatment with Exo-nFB or PBS (Fig. 2). ScRNAseq of human FAPs from the rotator cuff revealed 6 distinct subpopulations of human FAPs, with differential expression of mitochondrial genes, certain BAT markers, and exosome surface markers noted within one of the subpopulations (Fig. 3).Conclusions:FAP-BAT cells comprise a subpopulation of FAPs with upregulated mitochondrial gene expression and exosome production that mediates promyogenic effects in vitro and in vivo, and are present as a transcriptomically similar subpopulation of FAPs in humans.Table 1.Highly Differentially Expressed Genes of UCP1++ FAPsFigure 1.C2C12 cells and MEFs cultured for 14 days with Exo-FB(2 x 109), Exo-nFB(2 x 109), or exosome-free media, and stained for myosin heavy chain (MF20)j scale bar = 50µm (A), Myotube fusion index was calculated as the percentage of nuclei per field contained within an MF20+ myotube. n-3 replicates per condition; ** indicates p<0.005, *** indicates p<0.001.Figure 3.Light microscoay and immunofluorescence of human FAPs grown in standard media (top row) or mirabegron treated media (bottom row); scale bar =50µm; red arrows indicates predominant perilipin expression, green arrow indicates predominant UCP1 expression, yellow arrow indicates expression of both UCP1 and perilipin (A) UMAP plot of human scRNAseq data demonstrating 5 distinct clusters of human FAPs identified with Seurat (B) UMAP plot demonstrating distribution of human FAPs by in vitro treatment conditions: standard media or mirabegron (c). Dot plot of top 20 differentially expressed genes by treatment condition based upon adjusted p-value; “Average Expression” denotes z-score from mean overaal gene expression (D). Dot plot demonstrating unifrom increase in mitochondrial gene expression with mirabegron (E). Dot plot showing 13 genes that are significantly differentially expressed in both murine UCP1++ FAPs and human FAP subpopulations (F). Dot plot demonstrating additional characterstic genes of interest stratified by cluster. Red arrows indicates populations favorted to contain FAP-BAT cells at varying differentiations stages; green arrow highlights Adrb2 receptor expression; blue arrow highlights exosome surface markers CD63 and CD81(G).