B6129SF1/J mice were purchased from Jax Labs (Bar Harbor, ME) and bred and housed in a vivarium at Arizona State University (ASU). These mice were kept on a 10 h light: 14 h dark schedule with ad libitum access to food and water. ASU is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AALAC). All procedures were carried out in compliance with the ASU institutional animal care and use committee and AALAC under an approved research protocol.

Primary muscle satellite cell cultures (MuSCs) were established from cells isolated from five 12 week old B6129SF1/J mice as previously described (Palade et al., 2019). Briefly, hind limb quadriceps femoris muscles were excised, trimmed of fat and connective tissue, and finely minced. The muscle tissue was digested with 1.25 mg of protease XIV (Sigma-Aldrich, St. Louis, MO) for 1 h at 37°C. The cell suspension was filtered, differentially centrifuged and pre-plated in DMEM (Corning, Corning NY), containing 2% donor horse serum (HS) (Atlanta Biologicals, Flowery Branch, GA), and 100 μg/mL Primocin (Invivogen, San Diego, CA). Fibroblasts were removed by three pre-plating rounds of 3 h. Satellite cells were grown in a humidified chamber at 37°C with 5% CO₂, on Matrigel Matrix Basement Membrane (BD Biosciences, Bedford, MA) coated plates, in growth medium, Hams F-10 (Corning), containing 20% FBS (Atlanta Biologicals), 10 ng/mL bFGF (BD Biosciences) and 100 μg/mL Primocin (Invivogen).

MuSCs were seeded onto Matrigel (BD Biosciences) coated 6 well plates at a density of 1 × 10⁵ cells/well in growth medium. Cells were then treated with LPS (1 mg/mL; Sigma-Aldrich) or phosphate buffered saline (PBS) for 1, 2, 4, 6, or 8 h. Cells were lysed in TRIzol (ThermoFisher Scientific, Waltham, MA) for RNA isolation, per the manufacturer’s protocol, and 2 μg of total RNA was used to synthesize cDNA using SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA) and random hexamer primers.

Quantitative Real Time PCR (RT-qPCR) was carried out using SYBRgreen (Eurogentec, Fremont, CA) on an ABI 7900 HT thermocycler using a 384 well format in 10 μL reactions. All samples were normalized to the Gapdh transcript and relative gene expression was calculated using ΔΔCt analysis (Haimes and Kelley, 2010). Primer sequences are listed in Supplementary Table S1. All data are the result of triplicates from 3 biological samples.

MuSCs were treated with LPS as above for 2 h. MuSCs were then trypsinized (0.05% trypsin 0.53 M EDTA, Mediatech, Manassas VA) to create a single cell suspension. Cells were then counted on an automated cell counter and RNA library preparation was performed by the ASU Genomics Core facility. Samples were processed using a 10X Chromium single cell 3′ GEM, Library and Gel Bead kit V3 with a Chromium single cell B chip. The quality of each library was determined using Agilent TapeStation automated electrophoresis. Samples were sequenced at a read depth of ∼31,000 reads per cell (BioFrontiers Sequencing Core, University of Colorado, Boulder). The 10X Genomics CellRanger v6.0.1 was used to align to the Mus musculus reference genome mm10-2020-A (GRCm39), quantify, and provide basic quality control metrics for the scRNA-seq data. Data was then analyzed utilizing the Seurat R package. Cells with < 2000 reads or >80,000 reads were discarded as well as cells with < 0.9% or >20% mitochondrial genes. Clusters of cells with similar expression patters were identified using the ‘FindClusters’ function which used a shared nearest neighbor modularity optimization-based clustering approach.

Acute muscle injury was generated in the quadriceps of 3 month old mice by intramuscular injection of 50 μL of 10 μM cardiotoxin (CTX) (Sigma-Aldrich) solution in DMSO (Garry et al., 2000). At 24 h post-injection, Evan’s blue dye (Sigma-Aldrich) was injected intraperitoneally at 1 mg/10 g body weight to allow for visualization of damaged cellular membranes. Mice were sacrificed at designated time points and tissue was harvested for analysis. The uninjured quadriceps muscle from the contralateral leg was also harvested for use as control tissue. Total RNA was isolated from muscle biopsies using TRIzol (ThermoFisher Scientific) as described in Majumdar et al. (2012).

Mice (n = 3) were injected with CTX, quadriceps were harvested, and fixed in 4% paraformaldehyde in PBS overnight at 4°C. Tissue was cryopreserved in sucrose prior to embedding in Tissue-Tek OCT. Sections were then permeabilized and blocked in a buffer containing 0.1% Triton-X 100% and 5% normal donkey serum (NDS, Sigma-Aldrich) in PBS. Samples were then incubated in primary antibody solution containing antibodies recognizing a muscle marker, M-cadherin (Santa Cruz #sc-81471 1:250) and CCL5 or CCL2 (CCL5, R&D Systems AF478 1:300, CCL2 Novus Biologicals NBP 1-07035 1:500). Sections were then rinsed and incubated in secondary antibody solutions containing Alexa-Fluor conjugated antibodies against mouse (Thermofisher, A-21202, 1:1000), rabbit (Thermofisher, A-31573, 1:1000) or goat IgG (Thermofisher, A-21447, 1:1000) and DAPI (Sigma Aldrich 10,236,276,001, 1:1000) diluted in blocking solution. Tissue was rinsed in PBS and cover-slipped for microscopic analysis. Images were collected on a Zeiss LSM800 laser scanning confocal microscope and optimized for brightness and contrast. Blinded evaluators quantified M-cadherin positive (MCAD^+ve) and MCAD^+ve/CCL5^+ve cells in random 20X fields. Data were averaged for both CTX injured (n = 418 MCAD^+ve cells) and contralateral control quadriceps (n = 416 MCAD^+ve cells). Student’s t-test was used to determine statistical significance.

MuSCs treated with LPS (1 mg/mL) or PBS control for 2 h, were prepared for immunofluorescence (IF) by washing twice with cold PBS followed by fixation in 4% PFA for 20 min. Cells were then washed 3 times with PBS before antibody staining. Monoclonal mouse anti-MYOD (NB100-56511, Bio-techne Minneapolis, MN, 1:200) and polyclonal goat anti-CCL5 (AF478-SP, R&D Systems, 1:200) were used to detect cellular expression in culture. Anti-mouse-Alexa Fluor 488 (1:1000), anti-goat Alexa Fluor 647 (1:1000) and DAPI (1:1000) were then used to visualize MYOD, CCL5, and nuclei, respectively. Images were collected on a Zeiss LSM800 laser scanning confocal microscope at 20X magnification. Cells from 50 fields were then counted to determine the percent of total cells positive for these markers.

To assess the significance of the expression of cultured MuSCs post-treatment a one-way ANOVA was used. The ‘FindMarkers’ function in the Seurat package was used to determine the significance of the differential expression of marker genes from each cluster in the scRNA-Seq data. Marker genes were determined as having an adjusted p-value less than 0.05 and an average log 2 fold-change greater than 0.25.

Proinflammatory cytokines promote neutrophil and M1 macrophage infiltration to the damaged muscle via diapedesis and promote the activation and proliferation of MuSCs necessary for muscle regeneration (Howard et al., 2020). There are several studies that indicate the potential for MuSCs to contribute to the proinflammatory environment. For example, the pathogen-associated molecular pattern molecule (PAMP) LPS induced the expression of Tnfa, Il6, Ccl2, and Ccl5 in C2C12 and L6 muscle cell lines and human MuSCs (De Rossi et al., 2000; Frost et al., 2002; Frost et al., 2003). Further, expression of genes in the antiviral IFN pathway was elevated in the muscle of DM patients, predicting that muscle plays a role in the onset of inflammation (Moneta et al., 2019; Bivona et al., 2021). To examine the potential induction of an inflammatory response by MuSCs, we examined primary cell cultures derived from MuSCs isolated from the quadriceps of 3 month old B6129SF1/J mice treated with LPS (1 mg/mL) or a PBS control for 1, 2, 4, 6, or 8 h. Total RNA was isolated and transcription of Tnfa, Il6, and Ccl2 was assessed by RT-qPCR. We found that Tnfa was rapidly induced and reached its maximum at 1 h post-treatment (p < 0.0001) with a significant drop off between 2 and 4 h. In contrast, Il-6 and Ccl2 transcription were induced to maximal levels at 2 h (p < 0.0001) (Figure 1).

It is becoming increasingly clear that MuSCs are heterogeneous based on the expression of specific markers, cell function, and their ability to contribute to muscle regeneration (Biressi and Rando, 2010; Rodriguez-Outeiriño et al., 2021). Thus, it is likely that the chemokine and cytokine profiles of individual MuSCs will vary across different subpopulations. To examine this, scRNA-Seq was performed on actively proliferating MuSCs treated with LPS or PBS. Based on the results in Figure 1, cells were exposed to LPS for 2 h for maximal induction of chemokines and cytokines. Single cells were isolated and profiled by scRNA-Seq. Following quality control for mitochondrial RNA content and reads per cell, this resulted in a PBS-treated library of 2,751 cells with 20,235 mean reads/cell and an LPS-treated library of 2,802 cells with 19,330 mean reads/cell.

Unsupervised clustering and UMAP embedding was performed using the Seurat R package and identified seven unique clusters in PBS-treated MuSCs based on differentially expressed genes (Figures 2A,B; Supplementary Table S2). All clusters were determined to consist of myogenic lineage cells based on the expression of one or more of the muscle-specific transcription factors, Pax7, Myod1, Myog, and Myf5 (Figure 2C), known to have dynamic expression in satellite cells as they progress along the continuum from quiescent satellite cell through activation, proliferation, and differentiation (Figure 3A). The pattern of transcription of these four factors and other differentially expressed genes was used to identify the cell clusters within the PBS-treated MuSC culture (Figure 3B). Activated satellite cells (ASC) were identified based on transcription of Pax7 and the satellite cell activation and proliferation markers H19, EIF2S3Y, and HGF (Sheehan et al., 2000; Li et al., 2016; Martinet et al., 2016). The vast majority of the cells sorted into 3 cell clusters, PSC1—PSC3, with a similar pattern of reduced Pax7, and increased Myod1, transcription. PSC1 had a marked increase in ribosomal protein expression (Rspa, Rps2, Rps26, and Rps29), necessary for ribosome biogenesis and post-transcriptional regulation. Expression of these genes is associated with the increase in protein production needed for the transition of ASCs to proliferating myoblasts (Chaillou et al., 2017; Gayraud-Morel et al., 2018). The second and largest cell cluster, PSC2, was enriched for markers of proliferating myoblasts, including Peg3, Gas6, and C1qtnf3 (Correra et al., 2018; Mervis et al., 2020; Otani et al., 2020). The third PSC cluster (PSC3) had a higher expression level of glutathione peroxidase 3 (Gpx3) and the RNA helicase Mtrex, implying that these cells were responding to local environmental stressors that may not be related to the normal progression of satellite cells (El Haddad et al., 2012). There was a small cell population related to the PSC cell clusters, except for reduced transcription of Myod1 and Myog and enrichment for genes in the antiviral IFN pathway and downstream interferon stimulated genes (ISG), including Oasl1, Ifit1, Ifit3, and Rsad2 (Figure 3B). Thus, we have designated this cluster, IFN stimulated cells or ISCs. Finally, two clusters of differentiating myocytes (DMC1 and 2) were present in the culture, based on the expression of Myog and Myod1, as well as sarcomeric proteins including Myl1, Acta1, Tnnc2, and Myh1. Thus, the PBS treated primary satellite cell culture was comprised of cells along the continuum from activated satellite cells to differentiating myocytes with several subsets of myoblasts with distinct patterns of gene transcription.

To extend our evaluation of LPS induction of cytokines and chemokines in muscle (Figure 1), MuSC cultures were treated with LPS (1 mg/mL) for 2 h and profiled by scRNA-Seq. Unsupervised clustering and UMAP embedding was performed using the Seurat R package under the same conditions as the control culture (Figure 4, Supplementary Table S3). A similar set of cell clusters, as described above, were identified based on differential gene expression. However, there were exceptions, including a fourth PSC cluster (PSC4) and the absence of the second DMC cluster. The PSC4 cluster was distinguished by decreased expression of more than 20 ribosomal proteins. The significance of this reduction of the translational machinery is not clear, though there is evidence of LPS reducing protein synthesis in muscle tissue (Gordon et al., 2013).

The level of transcription of the myogenic markers, Pax7 and Myod1, was examined in the cell clusters of PBS and LPS treated cells (Figure 4C). Pax7 transcription was reduced in PSC1, PSC3 and ISC clusters and reduced in the ASC cluster. In contrast, Myod1 transcription levels are unaffected by LPS treatment. Alternatively, the distribution of cells along the satellite cell continuum was evaluated independent of cell clusters by comparing the number of cells co-expressing the myogenic markers, Pax7, Myod1, and Myog (Figure 4D). While the number of Pax7 ^+ve -only satellite cells and Myod1 ^+ve /Myog ^+ve differentiating myocytes remained relatively constant, there was an expansion of the Myod1 ^+ve myoblasts cells at the cost of the Pax7 ^+ve /Myod1 ^+ve activated satellite cells. This is consistent with the reduction of Pax7 transcription after LPS treatment. It remains to be determined if this is due to direct regulation by LPS signaling, or the indirect result of promoting proliferation of myoblasts through cytokine production.

An evaluation of chemokine production in response to LPS by scRNA-Seq showed that 13 members of the C-C and C-X-C subfamilies (Ccl2, Ccl5, Ccl7, Ccl17, Ccl20, Ccl27a, Cxcl1, Cxcl2, Cxcl4, Cxcl10, Cxcl12, Cxcl16, and Cx3cl1) were expressed in MuSCs. A comparison of transcription of these factors across the cell clusters in the LPS and PBS control conditions revealed two distinct patterns (Figure 5). The first pattern included chemokines not expressed in the control cells and broadly transcribed after treatment with LPS (e.g., Ccl2, Cxcl1, Ccl7, and Cx3c1). This is in contrast to chemokines that were induced in a single cell cluster. This was observed for Ccl5 and Cxcl10, which were induced in the ISC cluster, and Ccl17 and Cxcl16 in the ASC cluster (Figure 5B). Ccl5 and Cxcl10 are induced by type I IFNs (Kelly-Scumpia et al., 2010; Nakano et al., 2012) consistent with our designation of this cluster as IFN stimulated.

LPS induction of chemokines was validated in independently isolated primary MuSCs by RT-qPCR of total RNA using gene specific primers. Transcripts for the chemokines Ccl2, Ccl5, Ccl7, Cxcl1, and Cxcl10 were significantly higher in LPS-treated cells as compared to the PBS-treated controls (Supplementary Table S4). Overall, these studies revealed that MuSCs can be induced to express many C-C and C-X-C chemokines with LPS and they are differentially expressed across the cell clusters. ASC and ISC cell clusters express separate sets of chemokines, raising the possibility that they serve unique functions during muscle inflammation response.

An examination of proinflammatory cytokine transcription in response to LPS treatment revealed a modest induction of a limited number of factors when compared to the robust induction of chemokines (Figure 5). Among the TGFβ superfamily, known to positively regulate cells of the innate immune system (Chen and ten Dijke, 2016), Tgfb1 was constitutively transcribed at low levels in the ASC, PSC2, PSC3, and ISC clusters. With LPS treatment, Tgfb1 was induced in PSC1 cells (Figure 6A). Similarly, Tgfb2 was induced in PSC1 and PSC2 cells, but constitutively expressed in DMCs (data not shown). In contrast, Bmp2, a known antiviral ISG in bone marrow macrophages (Liu et al., 2012), was induced selectively in the ISC cluster (Figure 6A). Interestingly, Il18 and Il33, members of the IL-1 family of proinflammatory cytokines (Chan et al., 2019; Martinon et al., 2009), were weakly induced in ISCs (Figure 6B). Overall, the MuSCs expressed cytokines and chemokines in a heterogeneous pattern in response to inflammatory signals.

Single cell sequencing of the PBS control cells revealed a low level expression of components of the IFN signaling pathway in the ISC cluster (Figure 2). LPS induced the expression of Ccl5, Cxcl10, Il18, and Il33 in these cells (Figures 5, 6), supporting a proinflammatory role for this cluster. Since LPS acts through the TLR4 receptor to induce IFN α and γ expression (Richez et al., 2010; Wang et al., 2017), transcription of members of the antiviral IFN pathway were examined (Table 1). Of note, two cytosolic sensors of viral DNA and RNA, Ifih1/MDA-5 and Ddx58/Rig-1 (Loo and Gale, 2011), as well as their downstream transcription factor, Irf7, were selectively induced in the ISC cells. Further, twelve other IFN induced signaling factors associated with antiviral activity were also induced (Table 1). It should be noted that IFN-α, -β, or -γ transcription was not detected, raising the possibility that the antiviral IFN pathway was activated through the cytosolic receptors independent of the IFNs.

The expression of the IFN pathway marker, CCL5, in a small subset of MuSCs was confirmed by IF done using antibodies specific to both MYOD and CCL5. CCL5^+ve/MYOD^+ve cells were detected in newly isolated MuSCs from the quadriceps femoris of 3 month old mice treated with LPS or PBS (Figure 7). Double positive cells represented 2.3% of the LPS treated MuSCs, which was comparable to the distribution (2.08%) of Myod1 positive cells expressing ISG genes identified in the scRNA-Seq analysis (Table 2). These genes were also detectable in PBS treated control cultures (Figure 7), indicating that activation of the ISG gene expression was independent of LPS activation, but the expression level was increased in response to this inflammatory signal.

Activation of the antiviral IFN pathway is associated with the chronic muscle inflammation noted in DM patients and IFN-γ has been linked to muscle regeneration (Greenberg et al., 2005; Cheng et al., 2008; Suárez-Calvet et al., 2014). Since we observed that the IFN pathway was activated in a small percentage of MuSCs isolated from healthy mice, it was possible that it has a role in skeletal muscle repair as well. We further examined if we could detect activation of a subset of the ISG in response to acute injury. CTX was injected into the quadriceps femoris of Bl10 mice and total RNA isolated from the site of injury at 1, 2, 3, and 5 days post-injury (DPI) and from the uninjured contralateral leg. RT-qPCR was done using gene specific primers. When compared to the uninjured muscle, both Ifih1 and Isg15 transcription were rapidly induced at 1 DPI followed by a gradual decline. In contrast, expression of the IFN-regulated chemokines, Ccl5 and Cxcl10, peaked at 3 DPI before sharply declining (Figure 8). This is distinctly different from the broadly transcribed chemokines, Ccl2 and Cxcl1, that peaked earlier at 1 DPI. Thus, these data demonstrate that components of the antiviral IFN pathway originally observed with LPS treatment are expressed in vivo in response to injury also.

The distribution of satellite cells that had the antiviral IFN pathway activated in muscle was evaluated by IF with antibodies specific for CCL5 and M-Cadherin (MCAD), a cell adhesion protein expressed on satellite cells in their niche and proliferating myoblasts. MCAD expression is used to distinguish the mononuclear satellite cells from the adjacent muscle fiber (Yin et al., 2013). IF was performed on sections of CTX injured and contralateral control quadriceps. Analysis demonstrated that CCL5^+ve/MCAD^+ve cells represented 10.18% of total MCAD^+ve satellite cells of uninjured muscle and 10.52% of the total MCAD^+ve cells in CTX injured muscle at 3 DPI (Figure 9A). In comparison, we found CCL2, which was transcribed at high levels in LPS-induced MuSCs, was broadly expressed in both satellite cells and intact muscle fibers at 1 DPI (Figure 9B).