Muscle fibro-adipogenic progenitors from a single-cell perspective: Focus on their "virtual" secretome

Abstract

Skeletal muscle is a highly plastic tissue composed of a number of heterogeneous cell populations that, by interacting and communicating with each other, participate to the muscle homeostasis, and orchestrate regeneration and repair in healthy and diseased conditions. Although muscle regeneration relies on the activity of muscle stem cells (MuSCs), many other cellular players such as inflammatory, vascular and tissue-resident mesenchymal cells participate and communicate with MuSCs to sustain the regenerative process. Among them, Fibro-Adipogenic Progenitors (FAPs), a muscle interstitial stromal population, are crucial actors during muscle homeostasis and regeneration, interacting with MuSCs and other cellular players and dynamically producing and remodelling the extra-cellular matrix. Recent emerging single-cell omics technologies have resulted in the dissection of the heterogeneity of each cell populations within skeletal muscle. In this perspective we have reviewed the recent single-cell omics studies with a specific focus on FAPs in mouse and human muscle. More precisely, using the OutCyte prediction tool, we analysed the "virtual" secretome of FAPs, in resting and regenerating conditions, to highlight the potential of RNAseq data for the study of cellular communication.

Keywords: FAPs; cell-cell communication; extracellular matrix; in silico; scRNAseq; secretome; skeletal muscle.

FIGURE 1

Virtual secretome of murine and human FAPs (A) Workflow of the study from sc-RNA-seq datasets to virtual secretome. FAPs RNAseq data-set is an expression matrix that consists of FAPs cells barcodes, as given in each study, in columns and FAPs-markers gene-names in rows. SP = signal-peptide, UPS = unconventional peptide sequence. (B) Heatmap of k-means Clustering (with k = 6, produced with pheatmap () function in R) of the genes that are identified as FAPs markers and labelled as secreted (SP,UPS) from the virtual-secretome-tool “Outcyte”. The mean expression of FAPs cells was calculated for each time point per gene. Color values = z-score of expression values. Conditions = NI: non-injured, T0_5 = 0.5 days, T2 = 2 days, T3_5 = 3.5 days, T5 = 5 days, T10 = 10 days, T21 = 21 days after injury. (C) Gene names included in each cluster of the heatmap in (B), along with the top-2 Gene Ontology Terms, as given by clusterProfiler package with function “enrichGO ()” and parameters: pvalueCutoff = 0.1, qvalueCutoff = 0.1, pAdjustMethod = “BH”. The enrichment analysis highlights biological processes (BP) and molecular functions (MF) during mouse muscle regeneration. (D) Venn Diagram of the marker genes of FAPs subpopulations in human single-cell study, as identified via “FindMarkers ()” function of Seurat package. Values in parentheses are percentages of genes included in each subset. (E) List of the common secreted elements/109 genes between the three subpopulations of human FAPs. The proteins highlighted in bold are shared between the two studies. (F) Pie chart of human proteins identified as part of the matrisome (ECM and ECM-associated proteins) (Naba et al., 2016). (G) Venn diagram presenting the shared proteins between the two studies. The Venn Diagram shows the intersection of gene-names between FAPs-markers highly expressed at the “NI” time point from Oprescu et al and the 109 FAPs-markers in common from the 3 FAPs subpopulations in De Micheli et al. The genes highly expressed in “NI” time point were considered those from clusters 3 and 5 in the k-means clustering in 1B, colored in red. NI: non-injured. (H) Workflow of the study to identify downstream receptors and/or downstream regulated gene targets of interacting cells within skeletal muscle. (I) Expression of FN1 receptors in several resident cell types within skeletal muscle (extracted from De Micheli et al.).