This protocol outlines a method to isolate Fibro-adipogenic progenitors (FAPs) and myogenic progenitors (MPs) from rat skeletal muscle. Utilization of the rat in muscle injury models provides increased tissue availability from atrophic muscle for the analysis and a larger repertoire of validated methods to assess muscle strength and gait in free-moving animals.
Fibro-adipogenic Progenitors (FAPs) are resident interstitial cells in skeletal muscle that, together with myogenic progenitors (MPs), play a key role in muscle homeostasis, injury, and repair. Current protocols for FAPs identification and isolation use flow cytometry/fluorescence-activated cell sorting (FACS) and studies evaluating their function in vivo to date have been undertaken exclusively in mice. The larger inherent size of the rat allows for a more comprehensive analysis of FAPs in skeletal muscle injury models, especially in severely atrophic muscle or when investigators require substantial tissue mass to conduct multiple downstream assays. The rat additionally provides a larger selection of muscle functional assays that do not require animal sedation or sacrifice, thus minimizing morbidity and animal use by enabling serial assessments. The flow cytometry/FACS protocols optimized for mice are species specific, notably restricted by the characteristics of commercially available antibodies. They have not been optimized for separating FAPs from rat or highly fibrotic muscle. A flow cytometry/FACS protocol for the identification and isolation of FAPs and MPs from both healthy and denervated rat skeletal muscle was developed, relying on the differential expression of surface markers CD31, CD45, Sca-1, and VCAM-1. As rat-specific, flow cytometry-validated primary antibodies are severely limited, in-house conjugation of the antibody targeting Sca-1 was performed. Using this protocol, successful Sca-1 conjugation was confirmed, and flow cytometric identification of FAPs and MPs was validated by cell culture and immunostaining of FACS-isolated FAPs and MPs. Finally, we report a novel FAPs time-course in a prolonged (14 week) rat denervation model. This method provides the investigators the ability to study FAPs in a novel animal model.
Fibro-adipogenic progenitor cells (FAPs) are a population of resident multipotent progenitor cells in skeletal muscle that play a critical role in muscle homeostasis, repair, and regeneration, and conversely, also mediate pathologic responses to muscle injury. As the name suggests, FAPs were originally identified as a progenitor population with the potential to differentiate into fibroblasts and adipocytes1 and were purported to be the key mediators of fibro-fatty infiltration of skeletal muscle in chronic injury and disease. Further study revealed that FAPs are additionally capable of osteogenesis and chondrogenesis2,3,4. Thus, they are more broadly notated in the literature as mesenchymal or stromal progenitors3,5,6,7,8. In acute skeletal muscle injury, FAPs indirectly aid in regenerative myogenesis by transiently proliferating to provide a favorable environment for activated muscle satellite cells and their downstream myogenic progenitor (MPs) counterparts1,9,10. In parallel with successful regeneration, FAPs undergo apoptosis, returning their numbers to baseline levels1,9,10,11. In contrast, in chronic muscle injury, FAPs override pro-apoptotic signals, which results in their persistence9,10,11 and abnormal muscle repair.
In vivo studies evaluating the cellular and molecular mechanisms by which FAPs mediate muscle responses have utilized murine animal models to date1,7,9,10,11,12,13,14. While genetically engineered mice are powerful tools for use in these analyses, the small size of the animal limits tissue availability for study in long-term localized injury models where muscle atrophy can be profound, such as traumatic denervation. Furthermore, measurement of muscle strength and physical function requires ex vivo or in situ measurements that necessitate termination of the mouse, or in vivo methods that require surgery and/or a general anaesthetic to permit evaluation of muscle contractile performance15,16,17,18,19,20. In rats, well validated and globally utilized muscle functional analyses, in addition to analyses for more complex motor behaviors such as gait analysis (e.g., Sciatic Function Index, CatWalk analysis) exist and are performed in awake and spontaneously moving animals21,22,23,24. This additionally optimizes the principles of minimal morbidity in animal experimentation, and numbers of research animals used. The rat thereby provides the FAPs investigator the added flexibility of greater injured muscle volume for protein and cellular analyses and the ability to undertake serial assessments of muscle complex static and dynamic functional activity and behaviors, in the alert animal.
FAPs have primarily been identified and isolated from whole muscle samples using flow cytometry and Fluorescence-activated cell sorting (FACS) respectively. These are laser-based assays that are able to identify multiple specific cell populations based on characteristic features such as size, granularity, and a specific combination of cell surface or intracellular markers25. This is highly advantageous in the study of an organ system such as skeletal muscle, as homeostasis and regeneration are complex, multifactorial processes coordinated by a plethora of cell types. A seminal study identified FAPs, as well as MPs, using flow cytometric methods in mouse skeletal muscle1. They demonstrated that FAPs are mesenchymal in nature, as they lacked surface antigens specific to cells from endothelial (CD31), hematopoietic (CD45), or myogenic (Integrin-α7 [ITGA7]) origins, but expressed the mesenchymal stem cell marker Sca-1 (Stem cell antigen 1)1 and differentiated into fibrogenic and adipogenic cells in culture. Other studies demonstrated successful isolation of mesenchymal progenitors in muscle based on the expression of an alternative stem cell marker, platelet-derived growth factor receptor alpha (PDGFRα)2,7,8 and further analysis revealed these likely to be the same cell population as FAPs3. FAPs are now commonly identified in flow cytometry using either Sca-1 or PDGFRα as a positive selection marker1,9,10,11,12,13,14,26,27,28,29,30,31. The use of PDGFRα is preferential for human tissue however, as a direct human homologue of murine Sca-1 has yet to be identified32. In addition, other cell surface proteins have been reported as markers of MPs (e.g., VCAM-1), providing a potential alternative to ITGA7 as an indicator of cells of myogenic lineage during FAPs isolation33.
While flow cytometry/FACS is a powerful methodology for studying the role and pathogenic potential of FAPs in skeletal muscle1,9,10,11,13,29, it is limited technically by the specificity and optimization of its required reagents. Since flow cytometric identification and isolation of FAPs has been developed and conducted in mouse animal models1,9,10,11,29, this poses challenges for researchers who wish to study FAPs in other model organisms. Many factors – such as optimal tissue size to be processed, as well as reagent and/or antibody specificity and availability – differ depending on the species used.
In addition to the technical barriers to studying FAPs in a novel animal model, they have largely been studied in an acute, toxic setting – usually via intramuscular chemical injection or cardiotoxin. Evaluation of the long-term dynamics of FAPs is limited primarily to assessment in Duchenne's muscular dystrophy, using the mdx mouse model9,10,11, and models of combination muscle injury such as massive rotator cuff tear where concurrent tendon transection and denervation is performed on shoulder musculature26,27,28. The response of FAPs to the sole insult of chronic traumatic denervation, a common occurrence in work-place accidents in heavy industry, agriculture, and in birth traumas (brachial plexus injury)34,35,36,37 with significant morbidity, has not been as well characterized, often limited to a short-term time frame11,38.
We describe a method for identifying and isolating FAPs and MPs from healthy as well as severely atrophic and fibrotic skeletal muscle in the rat. First, identification of CD31-/CD45-/Sca-1+/VCAM-1- FAPs and CD31-/CD45-/Sca-1-/VCAM-1+ MPs using a tissue digestion and flow cytometry staining protocol is demonstrated and subsequent validation of our findings is performed through culture and immunocytochemical staining of FACS-isolated cells. Using this method, we also report a novel FAPs time-course in a long-term isolated denervation injury model in the rat.
Investigators conducting this protocol must receive permission from their local animal ethics board/care committee. All animal work was approved by the St. Michael's Hospital Unity Health Toronto Animal Care Committee (ACC #918) and was conducted in accordance with the guidelines set forth by the Canadian Council on Animal Care (CCAC). A schematic of the flow cytometry protocol is shown in Figure 1. If the downstream application is FACS and subsequent cell culture, all steps should be completed with proper aseptic technique.
1. Muscle harvesting
2. Muscle digestion
3. Generation of single cell suspension
4. Antibody staining for flow cytometry
NOTE: The Sca-1 antibody must be conjugated to APC prior to flow cytometry/FACS experiments, as per the manufacturer's instructions. Performance must be validated for each batch of conjugates (Figure 2). Final conjugations can be stored in 20 µL aliquots at -20 °C and are stable for three weeks. Refer to the Supplementary File for full conjugation protocol.
5. Flow cytometry and fluorescence-activated cell sorting (FACS)
6. Immunocytochemistry of cultured FAPs and MPs
7. Oil Red O (ORO) staining of cultured FAPs and MPs
8. Tissue staining of contralateral and denervated rat gastrocnemius sections
Identifying FAPs and MPs via flow cytometry using a novel antibody panel including Sca-1 and VCAM-1
The gating strategy for identifying FAPs in rat muscle is based upon flow cytometry protocols in the mouse29, which gate on CD31 (endothelial) and CD45 (hematopoietic) positive cells (termed the lineage [Lin]) and examines the fluorescent profile of FAPs marker Sca-1 and MPs marker ITGA7 from the linage-negative (Lin-) population. In the absence of commercially available, fluorophore-conjugated, and flow cytometry-validated antibodies for rat Sca-1 and ITGA7, we self-conjugated a rat Sca-1::APC antibody, and undertook an alternative strategy to identify the MPs. As VCAM-1 was recently shown to be an efficient single positive selection marker for the isolation of rat MPs33 and a rat-specific, conjugated, flow cytometry-validated antibody targeting VCAM-1 is available, we utilized VCAM-1 to identify MPs, as opposed to ITGA7.
We confirmed successful conjugation and performance of the Sca-1:APC antibody with single staining of compensation beads and cell suspensions generated from healthy rat gastrocnemius (Figure 2A,B). A five-point titration of Sca-1::APC (Figure 2C) was performed in addition to all other antibodies (CD31::FITC, CD45::FITC, VCAM-1:PE) used in the protocol (Supplementary Figure 1), to identify the optimum concentrations. Using this novel panel design, putative FAPs and MPs were simultaneously identified from healthy gastrocnemius (Figure 3), whereby cells single-positive for Sca-1 (Lin-/Sca-1+/VCAM-1-) were designated FAPs (red box) and cells single-positive for VCAM-1 (Lin-/Sca-1-/VCAM-1+) were designated MPs (blue box). We also identified a population of cells double positive for Sca-1 and VCAM-1 (Lin-/Sca-1+/VCAM-1+) (Figure 3F; upper right quadrant).
Validation of identification of FAPs and MPs by FACS and cell culture
To validate the protocol presented for flow cytometric identification of FAPs and MPs in rat skeletal muscle, we sought to isolate live cells for culture in vitro. Using FACS, viable FAPs and MPs were isolated using the same gating strategy as in the flow cytometric analysis. Approximately 20,000-40,000 FAPs and 30,000-50,000 MPs were collected for cell culture from a single gastrocnemius muscle from a 230 g rat.
To confirm the purity of each population, we co-immunostained freshly sorted FAPs and MPs for PDGFRα and Pax7. PDGFRα is the second mesenchymal progenitor marker, besides Sca-1, commonly used to identify FAPs2,7,8. Pax-7 is a well-recognized and widely used marker of muscle satellite (stem) cells41. The sorted population of FAPs displayed positive staining for PDGFRα with no contamination by Pax7 positive cells (Figure 4A; top row). Conversely, the sorted population of MPs stained positive for Pax7 with an absence of PDGFRα positive cells (Figure 4A; bottom row), validating the ability of Lin-/Sca-1+/VCAM-1- and Lin-/Sca-1-/VCAM-1+ cell surface antigen profiles to isolate pure populations of FAPs and MPs respectively.
We next cultured sorted FAPs and MPs over a time course of 10-12 days in conditions to induce adipogenic, fibrogenic, and myogenic differentiation. By day 12, FAPs cultures subjected to adipogenic conditions contained cells with either a fibroblast-like morphology or a multilocular morphology similar to that of white pre-adipocytes with fat droplets (data not shown). Immunostaining for fibroblast specific protein-1 (FSP-1) confirmed fibroblast differentiation, while staining for Plin-1 (Perilipin-1; an adipocyte marker)42 and Oil red O confirmed the differentiation of adipocytes and the presence of neutral triglycerides and lipids, respectively (Figure 4B). The absence of contaminating cells from a myogenic lineage in the FAPs cultures was confirmed by co-immunostaining for myosin heavy chain (MHC), a marker of differentiated myocytes. FAPs cultures subjected to fibrogenic differentiation (FD) were assessed for Collagen type 1 (Col1a1) expression, one of the main FAPs-derived collagens8. Co-immunostaining for Col1a1 and MHC at Day 11 revealed robust Collagen type 1 expression with no contaminating MHC positive cells (Figure 4B). Final confirmation of purity of the FAPs population was undertaken by culturing FAPs in myogenic media, to encourage the outgrowth of any contaminating MPs. No MHC positive cells were observed (Supplementary Figure 2A).
MP cultures subjected to myogenic conditions demonstrated MHC-expressing mature myocytes and multi-nucleated myotubes 12 days post plating (Figure 4C). Co-immunostaining of the MP cultures with FSP-1, and Plin-1 demonstrated the absence of contaminating fibroblasts or adipocytes, respectively. Similarly, ORO staining did not demonstrate lipid contamination (Fig. 4C). Col1a1, which is highly expressed by fibroblasts, has also been reported to be produced to a lesser extent by myogenic precursors and myoblasts, although there are contradictory data in the literature in this regard8,43,44. While co-immunostaining of MPs with Col1a1 and MHC revealed myocyte differentiation of our MP culture, no evidence of Col1a1 immunopositivity was found (Figure 4C). To further confirm the purity of the population, MP cultures were subjected to adipogenic or fibrogenic conditions to encourage the growth of adipocytes and fibroblasts (Supplementary Figure 2B). No contaminating cells from the FAPs lineage were observed.
While we had demonstrated the ability to identify and sort pure populations of FAPs and MPs from rat muscle, we next sought to determine the identity of the Lin-/Sca-1+/VCAM-1+ (double positive) cells. Since Sca-1 expression has been reported on a very small proportion of MPs (approximately 3%) in healthy muscle45 and similarly few FAPs (approx. 4%) have been reported to express VCAM-1 in healthy muscle10, immunostaining of freshly sorted Lin-/Sca-1+/VCAM-1+ cells was performed for PDGFRα and Pax7 along with culturing them in myogenic, adipogenic, and fibrogenic conditions to induce myogenesis, adipogenesis and fibrogenesis respectively. It was found that the double-positive cells were a mixed population of MPs and FAPs, with freshly sorted cells immunostaining positive for either PDGFRα or Pax7 (Supplementary Figure 3A) and cultured cells differentiating into mature myocytes, adipocytes or fibroblasts (Supplementary Figure 3B,C).
ITGA7 vs VCAM-1 to identify MPs during flow cytometric FAPs identification
While a successful protocol for flow cytometric identification of rat FAPs and MPs using Sca-1 and VCAM-1 respectively was established, we sought to determine if a self-conjugated ITGA7 antibody could similarly be used to identify MPs instead of VCAM-1, as is standard in the mouse. A rat-specific ITGA7 antibody was conjugated to PE-Cy7 (see Supplementary File) and the antibody's performance was validated on commercial compensation beads and single cell suspensions generated from rat gastrocnemius (Supplementary Figure 4A-C). The ITGA7::PE-Cy7 antibody performed adequately on single staining and FMO experiments (Supplementary Figure 4D). However, when full staining gastrocnemius cell suspensions with CD31::FITC, CD45::FITC, Sca-1::APC, and ITGA7::PE-Cy7, an interaction became evident between the Sca-1::APC and ITGA7::PE-Cy-7 antibodies. Subsequent culture of both the FACS sorted Lin-/Sca-1+/ITGA7- cells (purported FAPs) and Lin-/Sca-1-/IGTA7+ cells (purported MPs) (Supplementary Figure 4E) yielded predominantly FAPs and with few MPs (Supplementary Figure 4F), indicating the interaction between Sca-1::APC and ITGA7::PE-Cy-7 antibodies negatively impacted the specificity of cell identification.
A novel FAPs time-course in long-term denervated skeletal muscle
As FAPs dynamics have been assessed in the context of short-term traumatic denervation, in murine models11, 38, we sought to validate the performance of our method for FAPs isolation from both healthy and severely atrophic, fibrotic muscle. Rats were subjected to traumatic long-term denervation injury using the well-validated unilateral tibial nerve transection model46 with gastrocnemius muscle harvested at four serial timepoints over 14 weeks post-denervation from the denervated limb and contralateral innervated limb (to serve as an internal control). As has been previously reported, denervated muscle demonstrated progressive atrophy (Figure 5A,B), with increasing fibrosis and fat deposition (Figure 5C-F) over time47.
Our protocol generated an adequate number of cells for flow cytometric analysis, even though 12 and 14 week-denervated gastrocnemius weighed approximately 0.2-0.3 g (15-20% of respective contralateral control muscle) (Figure 5B). We observed up-regulation of FAPs in the denervated gastrocnemius muscle compared to the innervated contralateral control muscle, maintained for the 14 week duration of the experiment (Figure 6A,B), concordant with the progressive muscle fibrosis and fat deposition. Sca-1 immunostaining of muscle histologic cross sections confirmed localization of Sca-1 expressing cells to areas of fibro-fatty change in 14 week-denervated muscle (Figure 5G), in addition to baseline expression in the interstitium between myofibers in healthy muscle. A distinct subset of cells emerged in the FAPs population over time post-denervation characterized by a robust increase in Sca-1 signal (Sca-1 high; Figure 6A, red box) on flow cytometric analysis, compared to the FAPs basal Sca-1 expression (Sca-1 Med/Low; Figure 6A, green box). Quantification of these subpopulations revealed differential dynamics over time; Sca-1 High FAPs increased significantly at 12- and 14-weeks post-denervation (Figure 6B), comprising approximately half of the total FAPs population (Figure 6C). In contrast, Sca-1 Med/Low FAPs were the dominant subpopulation at 2- and 5-weeks post-denervation (Figure 6C) and showed only a transient increase in levels early post-denervation (Figure 6B).
In contrast to the sustained presence of FAPs in long-term denervated muscle, MPs demonstrated an expected biphasic response to denervation. Initially the MPs increased in denervated muscle above that seen in the control limb, but at 5 weeks post-denervation the population began to decline (Figure 6D). In keeping with the well-known exhaustion of the muscle satellite cell population in long-term denervated muscle47, by 12 weeks MPs were either at or below baseline levels post tibial nerve transection.
We also analyzed the dynamics of the Lin-/Sca-1+/VCAM-1+ population (Supplementary Figure 3D-E). While the frequency of double positive cells relative to the Lin- population progressively decreased in denervated muscle over time, when the absolute cell count was determined per gram of muscle a temporary increase in the double positive population was observed, prior to falling to or below baseline levels by 14 weeks post-denervation.
Figure 1: Graphical schematic for FAPs and MPs identification, isolation, and culture: Graphical schematic depicting FAPs and MPs identification from rat gastrocnemius. Samples are harvested and mechanically minced before undergoing two sequential enzymatic digestions. Tissue preparations are then processed and filtered to generate a single cell suspension. A cocktail of fluorescently-conjugated antibodies is added to each sample which is then run through a flow cytometer or cell sorter to respectively identify or isolate FAPs and MPs. Please click here to view a larger version of this figure.
Figure 2: Sca-1::APC self-conjugated antibody validation and titration. Validation of Sca-1::APC antibody conjugation by single-staining on commercial compensation beads (A) and single cell suspensions generated from healthy rat gastrocnemius muscle (B). Distinct populations of both Sca-1 labeled beads and cells are evident (black boxes). Antibody titration was performed by testing five different concentrations of Sca-1::APC on single cell suspensions (C); the optimal concentration was chosen based on greatest fluorescence intensity with minimal background staining and found to be batch dependent. A representative titration is shown with the batch specific optimal concentration indicated by the red box. Please click here to view a larger version of this figure.
Figure 3: Flow cytometry identification of FAPs and MPs in rat gastrocnemius. (A-F) Gating strategy for flow cytometric analysis of FAPs and MPs. (A) Samples are first gated to exclude debris (black box) as well as to gate for counting beads (red box). Cells are then gated to exclude doublets both by front scatter (FSC) (B) and side scatter (SSC) characteristics (C). The viability of resulting single cells is assessed by staining with SYTOX Blue (D). SYTOX Blue negative (live) singlets are then assessed for FITC signal identifying CD31 and CD45 (Lineage; Lin) in order to exclude Lin+ fractions from further analysis (black box Lin- cells) (E). The Lin- population is then assessed for Sca-1::APC versus VCAM-1::PE signal (F). FAPs are identified as Lin-/Sca-1+/VCAM-1- (F; red box) while MPs are identified as Lin-/Sca-1-/VCAM-1+ events (F; blue box). A Lin-/Sca-1+/VCAM-1+ double positive population is also evident (upper right quadrant). (G) FMO (Fluorescence Minus One) controls for CD31::FITC and CD45::FITC demonstrate proper compensation and gating for Lin- cells. (H-I) Sca-1::APC versus VCAM-1::PE plots of FMO controls demonstrate proper compensation and gating for FAPs (Sca-1 FMO) and MPs (VCAM-1 FMO). Please click here to view a larger version of this figure.
Figure 4: Sorted FAPs and MPs cell culture and immunostaining. (A) Co-immunostaining for PDGFRα (green) and Pax7 (red) on freshly sorted FAPs (top row) and MPs (bottom row). Nuclei stain blue with DAPI. FAPs solely express PDGFRα and no Pax7 positive cells are evident, while MPs exclusively express Pax7 with no PDGFRα positive cell contamination, demonstrating the purity of both sorted populations. (B) FAPs at Day 12 in adipogenic differentiation media (AD) stain positive for Fibroblast-specific Protein 1 (FSP-1; green) and Perilipin-1 (Plin-1; green), demonstrating differentiated fibroblasts and adipocytes, respectively. Nuclei are stained with DAPI (blue). Oil Red O (ORO) staining (red) indicates the presence of neutral triglycerides and lipids arising from mature adipocytes by Day 12. FAPs subjected to fibrogenic differentiation media (FD) demonstrate expression of FAPs-derived Collagen type 1 (Col1a1; green). FAPs cultured in either adipogenic or fibrogenic media do not co-immunostain for myosin heavy chain (MHC, red), indicating an absence of contaminating myocytes. (C) MPs grown in myogenic differentiation media (MD) on Day 12 display MHC positive (red) staining demonstrating the presence of mature myocytes and fused multinucleated myotubes. Nuclei are stained with DAPI (blue). MP cultures were clear of fibroblast and adipocyte contamination as indicated by the absence of co-immunostaining for FSP-1 (green), Plin-1 (green) and ORO staining. MPs grown on laminin to accommodate Co1a1 immunostaining do not display the same degree of myotube fusion by Day 12 as occurs on collagen. Col1a1 (green) is absent from MP cultures. Please click here to view a larger version of this figure.
Figure 5: Atrophy, fibrosis, and fatty infiltration of long-term denervated muscle. (A) Representative harvested gastrocnemius muscles at four serial timepoints post-denervation. Gastrocnemius denoted CONTRA is a representative contralateral limb sample from a two-week denervated animal. (B) Quantification of muscle atrophy post-denervation expressed as the ratio of denervated (DEN) gastrocnemius wet weight to that of the contralateral (CONTRA) limb. (C) Picrosirius Red (PSR) stained histologic cross-sections of gastrocnemius harvested 2 or 14 weeks post-denervation demonstrate progressive fibrosis. Muscle stains yellow, collagen stains red. (D) Fibrosis quantified by determining area of PSR positive tissue relative to total tissue area. (E) Oil Red O (ORO) stained cross-sections of gastrocnemius harvested 2- or 14-weeks post-denervation, counter-stained with hematoxylin. Lipids stain red. (F) Fatty infiltration quantified by determining area of ORO stained tissue relative to total tissue area. (G) Gastrocnemius histologic cross-sections harvested 2- or 14-weeks post-denervation, immunostained for laminin (green) and Sca-1 (red). Laminin positively stains the basal lamina that surrounds individual myofibers. Sca-1 identifies multiple progenitor cell types. Nuclei are stained with DAPI (blue). Inset shows localization of Sca-1+ cells outside the basal lamina in healthy muscle; yellow arrows denote presence of Sca-1+ cells within fibrotic areas. (Data are mean +/- S.D; n=3 for 2 and 14 week timepoints. Data in panel B analyzed by one-way ANOVA, data in panels D and F analyzed by two-way ANOVA. Post-hoc Sidak's test performed to correct for multiple comparisons. * = P<0.05 between CONTRA and DEN; # = P<0.05 between samples) Please click here to view a larger version of this figure.
Figure 6: FAPs and MPs dynamics in long-term denervated rat gastrocnemius. (A) Representative Sca-1::APC versus VCAM-1::PE panels of the Lin- (CD31-/CD45-) population from muscle samples harvested 2-, 5-, 12-, or 14-weeks post-denervation. Top row shows plots from the denervated (DEN) gastrocnemius, while the bottom row shows those from the contralateral, unoperated (CONTRA) gastrocnemius. Lin-/Sca-1+/VCAM-1- FAPs (Total) are subdivided into Sca-1 High (red box) and Sca-1 Med/Low (green box) fractions, while Lin-/Sca-1-/VCAM-1+ MPs are shown in the blue box. (B) Quantification of FAPs (Total, Sca-1 High, Sca-1 Med/Low) at each timepoint post-denervation, reported as the frequency (%) of Lin- cells (top row) and the number of cells per gram muscle (bottom row) normalized to the contralateral control gastrocnemius. (C) Relative proportions of Sca-1 High and Sca-1 Med/Low subpopulations of the total FAPs population. (D) Quantification of MPs at each timepoint post-denervation reported as the frequency (%) of Lin- cells (left graph) and as the number of cells per gram muscle (right graph) normalized to the contralateral control. (Data are mean +/- S.D; n=4 for 2 and 12 week timepoints; n=5 for 5 week timepoint; n=2 for 14 week timepoint. Data analyzed by one-way ANOVA; Post-hoc Sidak's test to correct for multiple comparisons. # = P<0.05.) Please click here to view a larger version of this figure.
Supplementary Figure 1: Antibody validation and titration. Validation of CD31::FITC (A-C), CD45::FITC (D-F), and VCAM-1::PE (G-I) commercially-conjugated antibodies on compensation beads (A,D,G) and single-cell suspensions from healthy rat gastrocnemius muscle (B,E,H). Each antibody was titrated at 5 different concentrations (C,F,I). The optimal concentration was chosen based on greatest fluorescence intensity with minimal background staining (red boxes). Please click here to download this File.
Supplementary Figure 2: Sorted FAPs and MPs in reverse differentiation culture conditions. (A) FAPs cultured in myogenic differentiation media (MD) at Day 12 co-immunostained for FSP-1 (green), Plin-1 (green) or Col1a1 (green) and MHC (red) do not demonstrate myocyte contamination. Positive FSP-1, and Col1a1 staining reveal FAPs differentiation to fibroblasts. ORO staining (red) reveals lipid production although Plin-1 positive adipocytes are not readily visible. Nuclei stained with DAPI (blue). (B) MPs subjected to adipogenic differentiation media died. MPs cultured in FAP Growth Media (FAP GM) for 12 days and co-immunostained for FSP-1 (green) or Plin-1 (green) and MHC (red) demonstrate differentiated myocytes and myotubes with an absence of FSP-1 or Plin-1 positive cells. The absence of ORO staining further confirms the lack of adipogenic cells in culture. MPs grown in fibrogenic differentiation media (FD) and co-immunostained for Col1a1 and MHC show mature myocytes and an absence of Col1a1 positive cells. MPs grown on laminin to accommodate Col1a1 immunostaining do not display the same degree of fusion to myotubes at 12 days in culture as occurs on collagen. Please click here to download this File.
Supplementary Figure 3: Lin-/Sca-1+/VCAM-1+ cells are a mixed population of FAPs and MPs. (A) PDGFRα and Pax7 co-immunostaining of freshly isolated Lin-/Sca-1+/VCAM-1+ cells show a mixed population of PDGFRα single positive (white arrows) and Pax7 single positive (yellow arrow) cells identifying FAPs and MPs, respectively. (B) Lin-/Sca-1+/VCAM-1+ cultures on Day 12 grown in myogenic differentiation media (MD) contain both FSP-1 single positive (green) and MHC single positive (red) cells identifying fibroblasts and myocytes/tubes, respectively. An absence of Plin-1 (green) and ORO (red) staining indicate the absence of mature adipocytes. Co-immunostaining for Col1a1 (green) and MHC (red) show mature myocytes and fibroblasts. (C) Lin-/Sca-1+/VCAM-1+ cells on Day 12 grown in adipogenic differentiation media (AD) similarly show a mix of FSP-1 (green) single positive and MHC (red) single positive cells, but also with Plin-1 (green) single positive cells and ORO staining present, confirming the differentiation of fibroblasts, myocytes and adipocytes. Cultures grown in fibrogenic differentiation media (FD) show mature myocytes and Col1a1 single positive cells (fibroblasts). (D) Representative Sca-1::APC versus VCAM-1::PE panels of the Lin- (CD31-/CD45-) population from denervated muscle samples harvested 2-, 5-, 12-, or 14-weeks post-denervation; Lin-/Sca-1+/VCAM-1+ population are indicated in the purple box. (E) Quantification of Lin-/Sca-1+/VCAM-1+ cells across four serial time-points post-denervation, reported as the frequency of Lin- cells (left graph) and cells per gram muscle (right graph) normalized to the contralateral gastrocnemius. (Data are mean +/- S.D; n=4 for 2 and 12 week timepoints; n=5 for 5 week timepoint; n=2 for 14 week timepoint. Data were analyzed by one-way ANOVA; post-hoc Sidak's test to correct for multiple comparisons; # = P<0.05.) Please click here to download this File.
Supplementary Figure 4: ITGA7 as a marker of myogenic cells in flow cytometry of rat skeletal muscle. Validation of ITGA7::PE-Cy7 antibody conjugation by single-staining on commercial compensation beads (A) and single cell suspensions generated from healthy rat gastrocnemius muscle (B). Populations of both ITGA7 labeled beads and cells are evident (black boxes) (C) Antibody titration was performed by testing five different concentrations on single cell suspensions. Red box indicates ideal concentration. (D) Plots of Fluorescence Minus One (FMO) controls demonstrate the absence of fluorescence spillover across the antibody conjugates, but full stain of cell suspensions reveals a non-specific interaction between Sca-1::APC and ITGA7::PE-Cy7 (green box). Gating (E) to separate Lin-/Sca-1+/ITGA7- cells (purported "FAPs") and Lin-/Sca-1-/IGTA7+ cells (purported "MPs"). (F) Co-immunostaining of FACS-sorted cell cultures at 12 days post plating with Perilipin-1 (green) and MHC (red) demonstrates both groups of sorted cells matured predominantly into adipocytes with few myocytes present. Please click here to download this File.
1. Unstained |
2. Viability Control |
3. CD31 + CD45 FMO |
4. Sca-1 FMO |
5. VCAM-1 FMO |
6. CD31 + CD45 Single Stain (Beads) |
7. CD31 + CD45 Single Stain (Cells) |
8. Sca-1 Single Stain (Beads) |
9. Sca-1 Single Stain (Cells) |
10. VCAM-1 Single Stain (Beads) |
11. VCAM-1 Single Stain (Cells) |
Table 1: Flow cytometry antibody staining controls. Full complement of antibody staining controls. If the experiment is being performed for the first time, single stained controls should be run on both compensation beads and single cell suspensions. For all subsequent experiments, single stained controls only need to be run on compensation beads.
CD31::FITC (µg) | CD45::FITC (µg) | Sca-1::APC (µg)* | VCAM-1::PE (µg) | SYTOX Blue (µM; Final Concentration) |
|
Unstained Control | |||||
— | — | — | — | — | |
Single-Stained Controls | |||||
SYTOX Blue (Viability Control) | — | — | — | — | 1 |
CD31 & CD45 | 0.5 | 0.25 | — | — | 1 |
Sca-1 | — | — | 0.25 | — | 1 |
VCAM-1 | — | — | — | 0.25 | 1 |
Fluorescence Minus One (FMO) | |||||
CD31 & CD45 | — | — | 0.25 | 0.25 | 1 |
Sca-1 | 0.5 | 0.25 | — | 0.25 | 1 |
VCAM-1 | 0.5 | 0.25 | 0.25 | — | 1 |
Experimental Samples | |||||
Full Stain | 0.5 | 0.25 | 0.25 | 0.25 | 1 |
Table 2: Flow cytometry antibody staining matrix. The antibody amount for staining of experimental and control samples for flow cytometry is shown. All staining is performed in a total volume of 100 µL of wash buffer. *The optimal amount of self-conjugated Sca-1::APC antibody may vary depending on batch used. All freshly conjugated Sca-1::APC batches should be first validated by single-staining both compensation beads and single cell suspensions.
Supplementary file: Reagent recipes. Please click here to download this File.
An optimized, validated FAPs isolation protocol for rat muscle is essential for researchers who wish to study injury models that are not feasible in the mouse for biologic or technical reasons. For example, mice are not an optimal animal model in which to study chronic local, or neurodegenerative injuries such as long-term denervation. Biologically, the short lifespan and rapid aging of mice make it difficult to accurately delineate the muscle sequalae due to denervation from the confounding factor of aging. From a technical standpoint, the drastically reduced muscle mass due to severe atrophy would be insufficient for effective flow cytometric analysis. In fact, mouse FAPs isolation protocols suggest grouping healthy muscles together to better increase yield of sorted FAPs29. While this solves the technical barrier of obtaining adequate muscle tissue for analysis, it simultaneously limits researchers from assessing FAPs on a muscle-specific level. Since specific muscle groups vary inherently in their fiber type composition, degree of vascularization, and mitochondrial numbers47 for example, combining multiple different muscles for flow cytometric analysis prevents muscle-specific FAPs characterization. In contrast, we show that both healthy and long-term denervated, atrophic and fibrotic rat gastrocnemius provide an adequate amount of starting material for effective flow cytometry. Moreover, in healthy samples, the surplus of muscle can be further subdivided for multiple downstream assays, allowing researchers to concomitantly analyze the cellular, molecular, and histological features of the same tissue. Finally, for experiments manipulating FAPs in injury models with therapeutics, well validated analyses of physical function and gait in alert and active rats are available21,22,23,24, enabling serial assessment of muscle function without the need for termination of the animal.
A key obstacle in creating an optimized FAPs isolation protocol for the rat was antibody availability. Our initial approach sought to copy the antibody panel and gating strategy protocols widely employed in the mouse1,7,8,9,10,11,12,13,14. While commercially available, conjugated, flow cytometry-tested, and validated antibodies that recognize rat were available for the lineage markers CD31 and CD45, none existed for positive FAPs identifying markers Sca-1 or PDGFRα, as well as for the negative selection marker ITGA7. Unconjugated, primary antibodies specific to these markers and purported to be effective in flow cytometry were available, but of the FAPs markers Sca-1 and PDGFRα, only the Sca-1 antibody had been validated in the published literature in flow cytometric analysis of rat cells48. We, therefore, selected Sca-1 as the positive selection marker for FAPs. The prospect of using secondary antibodies to delineate Sca-1 and ITGA7 markers was not feasible for several reasons, but primarily because the only rat-specific antibodies for Sca-1 and ITGA7 validated for flow cytometry were generated in the same host species. Therefore, the remaining option was self-conjugation of both antibodies using commercially-available kits.
The process of choosing an optimal kit for antibody conjugation was extensive, as many kits are completely or partially intolerant to various buffer diluents (e.g., Glycine, Glycerol, BSA, Sodium Azide) present in primary antibodies. In addition, the fluorophore provided must be compatible with the lasers equipped within the flow cytometer/cell sorter available to the investigator, and not emit at a similar wavelength to other fluorophores used to identify other cell types in the sample. The presence of diluent components in the rat-specific PDGFRα primary antibodies that were poorly compatible with the conjugation kits, was an additional consideration in the choice of Sca-1 as the positive FAPs selection marker in this protocol. While these results demonstrate that both Sca-1::APC and ITGA7::PE-Cy7 conjugations resulted in identification of distinct positive-stained populations on both compensation beads and cells, the former were found to exhibit decay in fluorescence sooner than was advertised by the manufacturer. It is highly recommended that researchers conjugate Sca-1::APC according to the manufacturer's instructions, immediately prior to the first time use (e.g., the day before the experiment) to ensure robust signal, plan experiments accordingly such that a single batch of conjugated antibody can be used within the antibody's window of effectiveness, and validate each batch by single-staining compensation beads and titrating on single cell suspensions. More importantly, however, the critical interaction noted between Sca-1::APC and ITGA7::PE-Cy7 prevented the accurate delineation of populations of FAPs vs MPs, necessitating an alternative strategy be employed.
Boscolo Sesillo et al. recently demonstrated the successful flow cytometric isolation of rat muscle stem cells using VCAM-1 as a single positive selection marker33 in CD31, CD45 and CD11b negative cells. With VCAM-1 as the MP selection marker, we utilized a novel antibody panel to identify FAPs (Lin-/Sca-1+/VCAM-1-) and MPs (Lin-/Sca-1-/VCAM-1+) and validated the approach with in vitro culture of FACS sorted cells from healthy gastrocnemius muscle. Freshly isolated FAPs and MPs immunostained solely for their respective alternative markers PDGFRα and Pax7. Cultured FAPs differentiated into populations of mature fibroblasts and adipocytes, and MPs into mature myocytes/myotubes. No cross contamination of FAPs and MPs within the cultures occurred. Immunostaining of histologic cross-sections confirmed infiltration of areas of fibro-fatty degradation in denervated muscle with Sca-1 positive cells.
While we undertook flow cytometric analysis of long-term denervated muscle to validate our protocol in severely atrophic, fibrotic and fat-infiltrated muscle, we incidentally observed the emergence of a FAPs sub-population with increased Sca-1 signal (denoted Sca-1 High) as compared to the baseline Sca-1 expression (denoted Sca-1 Med/Low) over time. Sca-1 High FAPs increased significantly late post-denervation (12 weeks plus), while Sca-1 Med/Low FAPs peaked early at 5 weeks before declining back to baseline levels. Heterogeneity in the FAPs phenotype has been reported10,30. FAPs with higher Sca-1 expression were shown to differentiate more readily into adipocytes, and when exposed to fibrogenic stimulation increased expression of Col1a130,49. The dynamics of the FAPs sub-populations observed here appear to be in-line with the time-course of denervation-induced sequelae and muscle's regenerative capacity47 and thus may characterize FAPs sub-populations with differing cellular programs. Sca-1 High FAPs become up-regulated at late time-points concomitant with fibro/fatty infiltration and a decline in regenerative potential, while Sca-1 Med/Low FAPs become up-regulated during muscle's regenerative window and may aid in effective regenerative myogenesis. The FAPs isolation protocol presented here provides investigators the ability to identify and isolate these various populations for future study.
We identified a population of cells staining positive for both Sca-1 and VCAM-1 (Lin-/Sca-1+/VCAM-1+) and ascertained this double positive population to be a mixture of FAPs and MPs. Separation of this population using a third marker – ITGA7 – was not possible due to the interaction between Sca-1 and ITGA7 primary antibodies. Malecova and colleagues reported a sub-population of VCAM-1 expressing FAPs that is nearly absent in healthy muscle, transiently increased with acute inflammation, and their persistence in mdx mice (model of muscular dystrophy) was associated with chronic muscle inflammation and fibrosis10. In contrast, and although not a pure FAPs population, we found that the Lin-/Sca-1+/VCAM-1+ population decreased to or below baseline levels with chronic muscle fibrosis at 12- and 14-weeks post-denervation. Denervation does not induce the same inflammatory reaction and cycling regeneration attempts experienced by mdx mice, which may explain the differences in our results. Our identification of MPs in the Lin-/Sca-1+/ VCAM-1+ cell population is in keeping with the report of Sca-1 expression on a very small proportion of MPs in healthy muscle, with a transient increase following injury during myoblast proliferation and subsequent withdrawal from the cell cycle45. Thus, while our antibody labeling and FACS gating strategy successfully isolates pure populations of FAPs and MPs for study, experiments requiring flow cytometry-based quantification of these populations may slightly underestimate their numbers due to the small percentage of cells in both progenitor lines that co-express Sca-1 and VCAM-1. Regardless, the current protocol clearly demonstrates the classic biphasic response of MPs to denervation injury, with short-term upregulation and subsequent depletion of the population in long-term denervated muscle. Similarly, the increase in FAPs reported in isolated shorter-term denervation injury is recapitulated here, and shown to be further sustained using the long-term tibial nerve transection model.
In summary, this protocol provides researchers with a previously unexplored animal, the rat, in which to use flow cytometric methods to simultaneously study FAPs and MPs. Experiments in mice limited by quantity of skeletal muscle, and the frequent need to undertake terminal experiments to assess muscle strength and function, can be readily conducted in the larger rat and with a more extensive range of non-lethal strength and functional assessment methods. Overall, FAPs studies in the rat may uncover novel roles of this key progenitor population in acute and chronic muscle and peripheral nerve trauma and disease, subsequently increasing the potential for developing cell-specific therapies.
The authors have nothing to disclose.
We would like to thank the Flow cytometry Core Facilities at the University of Ottawa and the Keenan Research Centre for Biomedical Sciences (KRC), St Michaels Hospital Unity Health Toronto for their expertise and guidance in optimization of the flow cytometry/FACS protocol presented in this manuscript. This work was funded by Medicine by Design New Ideas 2018 Fund (MbDNI-2018-01) to JB.
5 mL Polypropylene Round-Bottom Tube | Falcon | 352063 | |
5 mL Polystyrene Round-Bottom Tube with Cell-Strainer Cap | Falcon | 352235 | |
10 cm cell culture dishes | Sarstedt | 83.3902 | |
12-well cell culture plate | ThermoFisher | 130185 | |
12 mm glass coverslips, No.2 | VWR | 89015-724 | |
10 mL Syringe | Beckton Dickenson | 302995 | |
15 mL centrifuge tubes | FroggaBio | 91014 | |
20 gauge needle | Beckton Dickenson | 305176 | |
25mL Serological pipette | Sarstedt | 86.1685.001 | |
40µm cell strainer | Fisher Scientific | 22363547 | |
50mL centrifuge tubes | FroggaBio | TB50 | |
AbC Total Antibody Compensation Beads | ThermoFisher | A10497 | |
Ammonium Chloride, Reagent Grade | Bioshop | AMC303.500 | |
APC Conjugation Kit, 50-100µg | Biotium | 92307 | |
Aquatex Aqueous Mounting Medium | Merck | 108562 | |
Biolaminin 411 LN | Biolamina | LN411 | |
Bovine Serum Albumin (BSA) | Bioshop | ALB001 | |
Calcium Chloride | Bioshop | CCL444.500 | |
Collagenase Type II | Gibco | 17101015 | |
CountBright Plus Absolute Counting Beads | ThermoFisher | C36995 | |
Dexamethasone | Millipore Sigma | D4902 | |
Dispase | Gibco | 17105041 | |
Dulbecco’s Modified Eagle Medium (DMEM) (1X) | Gibco | 11995-065 | (+)4.5 g/L D-Glucose (+)L-Glutamine (+)110 mg/L Sodium Pyruvate |
EDTA | FisherScientific | S311 | |
FACSClean Solution | Beckton Dickenson | 340345 | |
FACSDiva Software | Beckton Dickenson | — | |
FACSRinse Solution | Beckton Dickenson | 340346 | |
Fetal Bovine Serum | Sigma | F1051 | |
Flow Cytometry Sheath Fluid | Beckton Dickenson | 342003 | |
FlowJo Software | Beckton Dickenson | — | |
Fluorescent Mounting Medium | Dako | S302380-2 | |
Goat anti-mouse Alexa Fluor 555 secondary antibody | Invitrogen | A21424 | |
Goat anti-rabbit Alexa Fluor 488 secondary antibody | Invitrogen | A11008 | |
Goat anti-rabbit Alexa Fluor 555 secondary antibody | Invitrogen | A21429 | |
Goat Serum | Gibco | 16210-064 | |
Ham's F10 Media | ThermoFisher | 11550043 | (+) Phenol Red (+) L-Glutamine (-) HEPES |
Hank’s Balanced Salt Solution (HBSS) (1X) | Multicell | 311-513-CL | |
Heat Inactivated Horse Serum | Gibco | 26050-088 | |
Hemocytometer | Reichert | N/A | |
HEPES, minimum 99.5% titration | Sigma | H3375 | |
Horse Serum | ThermoFisher | 16050130 | |
Human Transforming Growth Factor β1 (hTGF-β1) | Cell Signaling | 8915LF | |
Humulin R | Lilly | HI0210 | |
IBMX | Millipore Sigma | I5879 | Also known as 3-Isobutyl-1-methylxanthine |
Isopropanol | Sigma | I9516 | Also known as 2-propanol |
Lewis Rat, Female | Charles River Kingston | 004 (Strain Code) | 200-250 grams used |
LSRFortessa X-20 Benchtop Cytometer | Beckton Dickenson | — | |
Microcentrifuge | Eppendorf | EP-5417R | |
MoFlo XDP Cell Sorter | Beckman Coulter | — | |
Mouse Anti-CD31::FITC Antibody | Abcam | ab33858 | Clone TLD-3A12 |
Mouse Anti-CD45::FITC Antibody | Biolegend | 202205 | Clone OX-1 |
Mouse Anti-CD106::PE Antibody | Biolegend | 200403 | Also known as VCAM-1 |
Mouse Anti-MHC Antibody | Developmental Studies Hybridoma Bank (DSHB) | N/A | Also known as MF20 |
Mouse Anti-Pax7 Antibody | Developmental Studies Hybridoma Bank (DSHB) | N/A | |
Neutral Buffered Formalin, 10 % | Sigma | HT501128 | |
Oil Red O | Millipore Sigma | O0625 | |
PE-Cy7 Conjugation Kit | Abcam | ab102903 | |
Penicillin-Streptomycin | Sigma | P4333 | |
Phosphate Buffered Saline, pH 7.4 (1X) | Gibco | 10010-023 | (-)Calcium Chloride (-)Magnesium Chloride |
Potassium Bicarbonate, Reagent Grade | Bioshop | PBC401.250 | |
Rabbit Anti-Fibroblast Specific Protein 1 (FSP-1) Antibody | Invitrogen | MA5-32347 | FSP-1 also known as S100A4 |
Rabbit Anti-Integrin-a7 Antibody | Abcam | ab203254 | |
Rabbit Anti-Laminin Antibody | Sigma | L9393 | |
Rabbit Anti-Perilipin-1 Antibody | Abcam | ab3526 | |
Rabbit Anti-Sca-1 Antibody | Millipore Sigma | AB4336 | |
Rabbit Recombinant Anti-Collagen Type I Antibody | Abcam | ab260043 | Also known as Col1a1 |
Rabbit Recombinant Anti-PDGFR Alpha Antibody | Abcam | ab203491 | |
Recombinant Human FGF-basic | Gibco | PHG0266 | |
Sodium Azide | Sigma | S2002 | |
Triton-X-100 | Fisher Scientific | BP151 | |
Troglitazone | Millipore Sigma | T2573 | |
Tween-20 | Bioshop | TWN510 |