MicroRNA cargo of extracellular vesicles released by skeletal muscle fibro-adipogenic progenitor cells is significantly altered with disuse atrophy and IL-1β deficiency

Abstract

Fibro-adipogenic progenitor cells (FAPs) are a population of stem cells in skeletal muscle that play multiple roles in muscle repair and regeneration through their complex secretome; however, it is not well understood how the FAP secretome is altered with muscle disuse atrophy. Previous work suggests that the inflammatory cytokine IL-1β is increased in FAPs with disuse and denervation. Inflammasome activation and IL-1β secretion are also known to stimulate the release of extracellular vesicles (EVs). Here, we examined the microRNA (miRNA) cargo of FAP-derived, platelet-derived growth factor receptor A (PDGFRα⁺) EVs from hindlimb muscles of wild-type and IL-1β KO mice after 14 days of single-hindlimb immobilization. Hindlimb muscles were isolated from mice following the immobilization period, and PDGFRα⁺ extracellular vesicles were isolated using size-exclusion chromatography and immunoprecipitation. Microarrays were performed to detect changes in miRNAs with unloading and IL-1β deficiency. Results indicate that the PDGFRα⁺, FAP-derived EVs show a significant increase in miRNAs, such as miR-let-7c, miR-let-7b, miR-181a, and miR-124. These miRNAs have previously been demonstrated to play important roles in cellular senescence and muscle atrophy. Furthermore, the expression of these same miRNAs was not significantly altered in FAP-derived EVs isolated from the immobilized IL-1β KO. These data suggest that disuse-related activation of IL-1β can mediate the miRNA cargo of FAP-derived EVs, contributing directly to the release of senescence- and atrophy-related miRNAs. Therapies targeting FAPs in settings associated with muscle disuse atrophy may therefore have the potential to preserve muscle function and enhance muscle recovery.

INTRODUCTION

Skeletal muscle atrophy is associated with frailty and impaired mobility leading to a higher risk of falls, fractures, and hospitalization (1). A better understanding of the cellular and molecular mechanisms by which muscle atrophy occurs is critical in preventing muscle loss in conditions, such as bed rest, spinal cord injury, aging, and spaceflight (1). Fibro-adipogenic progenitor cells (FAPs) are a population of muscle stem cells that play a vital role in muscle repair and regeneration (2). FAPs can differentiate to form adipocytes or fibroblasts, and are positive for the surface marker platelet-derived growth factor receptor A (PDGFRα). FAPs comprise ∼5%–20% of the total muscle population, in contrast to satellite cells (SCs), which represent 5% or less of the muscle progenitor cell population (3). FAPs are normally quiescent but become activated in the setting of muscle injury (4–10). Once activated, FAPs will proliferate and interact with neighboring satellite cells and myotubes to stimulate their division and differentiation through paracrine signaling (10–14). These paracrine signals are collectively referred to as the FAP secretome.

The FAP secretome is a complex network of cytokines, myokines, and other factors that can promote myogenesis (2, 10). FAP-derived factors include myokines, such as IL-6 and IL-10, IL-33, WISP1, and follistatin all of which play a role in increasing myogenesis (2, 10). Recent data from our laboratory (15) and others (16) suggest that the proinflammatory cytokine IL-1β is also a key component of the FAP secretome. Specifically, IL-1β is secreted from FAPs in conditions associated with muscle atrophy, such as hindlimb immobilization (15) and muscle denervation (16). Inflammasome activation and IL-1 β secretion have been shown to stimulate the secretion of EVs, including exosomes and microvesicles, in a variety of cell types (17, 18). The cargo of FAP-derived EVs is, however, not well understood. Recently it was shown that small-molecule treatment could modify the microRNA cargo of FAP-derived EVs to promote muscle repair and regeneration (19). Here, we expand upon this body of research by characterizing the microRNA cargo of FAPs after hindlimb immobilization and disuse atrophy. We also explore the role of the inflammasome in this process by using the hindlimb immobilization model in mice lacking IL-1β. Our results show that hindlimb immobilization significantly alters the cargo of FAP-derived EVs, and that IL-1β deficiency, in turn, modifies the EV cargo released by FAPs.

METHODS

Hindlimb Immobilization

Six (3 females, 3 males) C57BL/6 mice, 22 (16 females, 6 males) IL-1β KO mice, and 17 BALB/c (8 females, 9 males), all 4–5 mo of age, were randomized to cages and immobilized for 2 wk. C57BL/6 mice were obtained from Envigo and global IL-1β KO (Stock No. 034447) and BALB/c mice were obtained from the Jackson Laboratory. The IL-1β KO mice are on a BALB/c background and so the wild-type BALB/c mice were included as an additional control to assess loss of muscle mass with immobilization. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at Augusta University. The immobilization process was performed as described previously (15). In brief, the left hindlimb of the mice was immobilized using a casting procedure that involved a combination of tape, cast padding, and Vetwrap. This cast was replaced every 2 to 3 days as needed. After 2 wk, the mice were euthanized and the hindlimb muscles were isolated for further processing. The unloaded limb was used as a control comparison and the mortality rate of the mice was zero.

Quadriceps Histology

Quadriceps muscles were dissected, weighed, and placed in 10% buffered formalin. After 24 h fixation, the muscles were moved to 70% ethanol. The samples were paraffin embedded, cut in cross sections at 2–3 µm, and stained with hematoxylin and eosin (H&E). Fiber size was measured by taking images from each muscle section at ×20 magnification using a bright-field Leica DMCS microscope fitted with a micropublisher six camera (Qimaging). The technician was blinded to the treatments and groups. Ocular software (Qimaging) was then used to take images of each sample, which were uploaded to ImageJ to calculate individual muscle fiber areas. Thirty muscle fibers were measured for each sample to calculate an average fiber size per muscle.

Isolation of FAP-Derived EVs from Skeletal Muscle

Muscle-derived EVs were harvested using a modified protocol from (20–22). In brief, all hindlimb muscles (gastrocnemius, soleus, tibialis anterior, quadriceps, and extensor digitorum longus) were minced and placed into 5 mL of a mixture of Hibernate-E media, protease inhibitor, DNase I (40 U/mL), and collagenase D (2 mg/mL). The minced muscle was incubated for 30 min at 37°C and centrifuged in gradients to remove cellular debris. The supernatant was then filtered with a 0.22 µm membrane and the sample was placed into a 10-kDa Amicon Ultra filter (Millipore Sigma) to concentrate the sample. Once concentrated, 500 µL of the sample was added to the size exclusion chromatography (SmartSEC, System Biosciences) column, which was prepared per kit instructions. The samples were then incubated for 30 min with rotation and subsequently spun down. SmartSEC filtration was performed twice on the samples. Immunoprecipitation was performed as described previously (20) using PDGFRα antibody (Invitrogen, Cat. No. 14-1401-82; RRID:AB_467491) to isolate FAP-derived EVs from the muscle-derived EV population. In brief, streptavidin beads (Fisher Scientific, Cat. No. PI53117) were biotinylated with the PDGFRα antibody (Abcam, Cat. No. ab201796). The biotinylated antibody was then incubated with the EVs overnight at 4°C. Glycine-HCl (pH 3.0) was then used to separate the beads from PDGFRα⁺ exosomes. The solution was then neutralized with Tris-HCl (pH 8.0).

Characterization of Isolated EVs

EVs isolated before and after the immunoprecipitation procedure were analyzed using a Zetaview instrument from Particle Matrix to determine particle concentration and size (nm). The EVs were then imaged by transmission electron microscopy for characterization of EV morphology and stained with immunogold anti-CD63 and anti-CD81, both characteristic markers of exosomes. The samples were fixed in 4% paraformaldehyde in cacodylate buffer overnight. The fixed samples were then suspended and applied to a carbon-Formvar coated nickel grid. The grids were blocked for 1 h and then exposed to the primary antibody for an hour. The two antibodies used were 1:100 CD63 (Santa Cruz Biotechnology, H-193, SC-15363; RRID:AB_648179) and 1:100 CD81 (Santa Cruz Biotechnology, H-121, SC-9158; RRID:AB_638255). The grids were then placed in 1:1,000 anti-rabbit nanogold (Nanoprobes Inc) for 1 h, enhanced with HQ Silver (Nanoprobes, Inc), and negatively stained with 2% aqueous Uranyl Acetate. Grids were then examined using a JEM-1400 Flash transmission electron microscope (JEOL USA inc., Peabody, MA) at 110 kV and imaged with a Gatan One View Digital Camera (Gatan Inc., Pleasanton, CA).

miRNA Microarray Analysis of FAP-Derived EVs

RNA from six C57BL/6 mice (3 females, 3 males) and six IL-1β KO mice (3 females, 3 males) were used for microarray analysis. RNA from the EVs was isolated using the miRNeasy kit (Qiagen) following manufacturer specifications. RNA purity and concentration were evaluated by spectrophotometry using NanoDrop ND-1000 (Thermo Fisher). The quality of the total and small RNA was assessed by the Agilent 2100 Bioanalyzer (Agilent Technologies). A total of 130 ng of total RNA was labeled with biotin using the FlashTag Biotin HSR RNA Labeling Kit (Thermo Fisher) according to the manufacturer’s procedure. The labeled samples were then hybridized to the GeneChip miRNA 4.0 array (Thermo Fisher) that contains 1,908 and 1,255 mouse mature and premature miRNAs, respectively. Array hybridization, washing, and scanning of the arrays were carried out according to the manufacturer’s recommendations. Data were obtained in the form of CEL files. Some samples within each sex, limb, and genotype were combined to reach the miRNA 4.0 array threshold limit. The final analysis included a total of five immobilized (L) and four nonimmobilized (R) wild-type samples and a total of five immobilized (L) and four nonimmobilized (R) IL-1β KO samples were analyzed. We did not test for sex differences in the miRNA data since we have found no sex differences in muscle loss with immobilization (15).

Protein Extraction and Digestion from Muscle Lysate for LC-MS Analysis

LC-MS analysis was used to characterize changes to protein with immobilization to identify targets of the miRNA contained in the EV cargo. Protein was isolated from immobilized (3) and nonimmobilized (3) gastrocnemius muscle samples using the Bio-Plex kit, following the manufacturer’s instructions (BioRad). In brief, samples were homogenized and incubated in cell lysis buffer and then spun for 20 min at 10,000 rpm. The supernatant was collected and the protein amount was measured using BCA Protein Assay Kit (Sigma). A total of 50 µg of extracted protein was reduced with dithiothreitol, alkylated with iodoacetamide, and diluted (10-fold) with 50 mM ammonium bicarbonate buffer. Samples were digested for 16 h using trypsin (Thermo Scientific #90057) at 37°C at a 1:20 ratio. The digested peptides were then cleaned using a C18 spin column (Harvard Apparatus #744101) and lyophilized. The digested samples were separated via an Ultimate 3000 nano-UPLC system (Thermo Scientific) and analyzed with an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). The peptide mixture was trapped and washed on a Pepmap100 C18 trap (5 μm, 0.3 × 5 mm) for 10 min at a rate of 20 μL/min using 2% acetonitrile in water with 0.1% formic acid. The peptides were then separated on a Pepman100 RSLC C18 column (2.0 μm, 75 μm × 150 mm) using a gradient of 2% to 40% acetonitrile with 0.1% formic acid over 120 min (flow rate: 300 nL/min; column temperature: 40°C). Eluted peptides were analyzed via data-dependent acquisition in positive mode using the following settings: Orbitrap MS analyzer for precursor scan at 120,000 FWHM from 300 to 1,500 m/z; Ion-trap MS analyzer for MS/MS scans in top speed mode (2-s cycle time) with dynamic exclusion settings (repeat count: 1; repeat duration: 15 s; exclusion duration: 30 s). The fragmentation method was higher-energy collision dissociation (HCD) with a normalized collision energy level of 30%. Raw MS peptide data were analyzed using Proteome Discoverer (v. 1.4; Thermo Scientific) before being searched against the Uniprot mouse protein database using TurboSequest with the following search parameters: 0.6 Da production tolerance, 10 ppm precursor size, 57.021 Da static carbamidomethylation for cysteine 15.995 Da dynamic oxidation for methionine. Proteins were grouped if they had comparable peptide characteristics. Peptide spectrum matching (PSM) was validated using the Percolator PSM validator algorithm. PSM counts were normalized by comparing total PSM counts between samples.

Bioinformatics and Statistical Analysis

The CEL files were imported to Partek Genomic Suite 7.19.1125 and processed through a robust multiarray average (RMA) process that includes background correction, quantile normalization, log (base 2) transformation, and median polishing. Principal component analysis (PCA) of the normalized miRNA expression was performed to visualize the partition among the groups and identify the major sources of variation within the experiment (Figs. 2 and 4) for KO and WT, respectively. One-way ANOVA was used to identify potentially differentially expressed miRNAs between the left and right limbs and miRNAs were filtered with a P value cutoff of 0.05 and an absolute fold-change cutoff of 1.0 as shown in Tables 2 and 3. Gene targets of the miRNAs were identified using miRsystem (48). For proteomic analysis, statistical analyses were performed using the R Project for Statistical Computing (v. 3.6.3). PSM values were log-transformed for normal distribution and LIMMA package in R to analyze the effect of treatment on protein levels. P values were adjusted using a false discovery rate (FDR) with threshold of <0.05. T tests and ANOVA analysis for quadriceps histology (weight and fiber size) were analyzed using GraphPad Prism software.

Figure 2.

PCA and heatmap analysis of miRNA changes occurring in wild-type exosomes with immobilization. A: principal components analysis of the immobilized (left, blue dots) and nonimmobilized (right, red dots) samples with clear separation of the treatment groups. B: heatmap shows clear clustering of the differentially expressed miRNAs with downregulation of genes indicated in green, and upregulation of genes indicated in red. The genes that are downregulated with immobilization (left) are upregulated comparatively in the nonimmobilized group (right) and vice versa (n = 9; 5 left, 4 right). miRNA, microRNA; PCA, principal components analysis.

Figure 4.

PCA and heatmap analysis of miRNA changes occurring in IL-1β KO exosomes with immobilization. A: principal component analysis of the immobilized (left, blue dots) and nonimmobilized (right, red dots) samples of the IL-1β KO mice. There is no clear separation of treatment groups based on immobilization vs. nonimmobilization. B: heatmap shows clustering of the differentially expressed miRNAs with downregulation of genes indicated in green, and upregulation of genes indicated in red. The genes that are downregulated with immobilization (left) are upregulated comparatively in the nonimmobilized group (right) and vice versa (n = 9; 5 left, 4 right). miRNA, microRNA; PCA, principal components analysis.

Table 2.

The five most up- and downregulated miRNAs from FAP-derived EVs isolated from the immobilized and nonimmobilized hindlimbs of wild-type C57BL/6 mice and their role(s) in muscle biology and senescence

miRNA	P Value	Fold Change	Effect on Muscle	Senescence?
mmu-let-7c-5p	0.016327	3.83704	Inhibits utrophin expression (23,24)	Yes (25–29)
mmu-let-7b-5p	0.010675	2.32484	Suppresses muscle repair and vitamin D accumulation (30)	Yes (27–29)
mmu-miR-667-5p	0.014745	2.006	N/A	No
mmu-miR-181a-5p	0.002832	1.89312	Overexpressed with muscle atrophy and muscular dystrophies (31,32); increases in fasting conditions (33); promotes skeletal muscle development (34)	Yes (29)
mmu-miR-124-3p	0.026632	1.88281	Suppresses myogenic differentiation (35); decreases fibrosis by inhibiting TGFB1 (36,37)	Yes (27)
mmu-miR-6933-5p	0.00233	−1.41691	N/A	N/A
mmu-mir-208a	0.014375	−1.42461	Increases with exercise (38); induces slow muscle genes by skeletal muscle and adult heart (39)	No
mmu-miR-3059-3p	0.006264	−1.46635	N/A	N/A
mmu-miR-20b-5p	0.006531	−1.48134	Glucose metabolism (40); promote myoblast differentiation (41)	Yes (42)*
mmu-miR-667-3p	0.000291	−1.78568	N/A	No

N/A, no literature found in that area for that miRNA. *Inhibits senescence, the upregulated miRNAs increase senescence. EVs, extracellular vesicles; FAPs, fibro-adipogenic progenitor cells; miRNA, microRNA.

Table 3.

List of five most up- and downregulated miRNAs from FAP-derived EVs isolated from the immobilized and nonimmobilized hindlimbs of IL-1β KO mice and their role(s) in muscle biology and senescence

miRNA	P Value	Fold Change	Effect on Muscle	Senescence?
mmu-mir-130a	0.00534837	1.48014	Angiogenesis in muscle (43)	N/A
mmu-mir-6923	0.00852097	1.45646	N/A	N/A
mmu-miR-181a-2-3p	0.00906754	1.43673	N/A	N/A
mmu-miR-190a-3p	0.0281067	1.42858	N/A	N/A
mmu-mir-210	0.0269355	1.39369	Promotes vascularization (44) and myogenesis (45)	No
mmu-miR-1905	0.00656974	−1.417	N/A	No
mmu-miR-291b-3p	0.0118063	−1.42002	N/A	N/A
mmu-miR-106b-3p	0.0130243	−1.42159	N/A	Yes (46, 47)
mmu-miR-5128	0.0444912	−1.44276	N/A	No
mmu-miR-684	0.000183845	−1.54663	N/A	No

N/A, no literature found in that area for that miRNA. EVs, extracellular vesicles; FAPs, fibro-adipogenic progenitor cells; miRNA, microRNA.

RESULTS

Muscle-Derived EVs Show Characteristic Features of Exosomes

EVs isolated using the filtration and size-exclusion chromatography approach were on average within the known size range for small EVs (<200 nm; Table 1 and Fig. 1A). The average particle size of EVs isolated after immunoprecipitation for PDGFRα was significantly smaller (∼15%, P < 0.001) than the average size of particles isolated from muscle before immunoprecipitation (Table 1), possibly indicating a larger population of exosomes versus microvesicles. Transmission electron microscopy (TEM) images show that the EVs have a cup-shaped morphology, characteristic of exosomes (Fig. 1B). Positive staining for the exosome markers CD63 and CD81 is also observed in these EVs (Fig. 1C).

Table 1.

Size of extracellular vesicles isolated from wild-type muscle tissue

Sample	Mean Size Before AB Pulldown, nm	Mean Size After Pulldown, nm
10R	165.5	181.6
8R	191.4	152.3
9R	181.2	164.3
6L	174.8	162.1
7L	187.2	157.7
8L	196.6	150.8
6R	183.4	123.2
7R	192.8	165.7
9L	192.9	133
10L	186.9	193.1
Average	185 ± 5.57	158 ± 12.12

Zetaview size measurements of the extracellular vesicles isolated from tissue before PDGFRα antibody pulldown and after the antibody pulldown. Size distribution after pulldown differs significantly from the size distribution before immunoprecipitation. Each sample was measured once. L, immobilized; PDGFRα, platelet-derived growth factor receptor A; R, nonimmobilized. (P < 0.001, ANOVA, n = 10.)

Figure 1.

Characterization of EVs isolated from skeletal muscle. A: representative image of the ZetaView results with the average particle size <200 nm. B: representative unstained TEM images of the EVs. Note the cup-shape morphology characteristic of exosomes (arrows). C: representative images of immunogold labeling of EVs positive for CD63 (left) and for CD81 (right). EVs, extracellular vesicles; TEM, transmission electron microscopy.

Atrophy- and Senescence-Associated miRNAs Are Upregulated with Immobilization in FAP-Derived, PDGFRα⁺ EVs, from Wild-Type Mice

Single hindlimb immobilization induced significant muscle atrophy in 2 wk (Supplemental Figs. S1 and S2; see https://doi.org/10.6084/m9.figshare.17430698.v4). Microarray was performed on miRNAs isolated from FAP-derived, PDGFRα⁺ EVs. ANOVA revealed 364 differentially expressed miRNAs: 214 upregulated with immobilization and 150 downregulated with immobilization (Supplemental Table S1; see https://doi.org/10.6084/m9.figshare.17430701.v2). PCA plot shows a clear separation between the immobilized and nonimmobilized EV miRNAs (Fig. 2A). Heatmaps show clustering of the differentially expressed miRNAs isolated from EVs from the immobilized versus nonimmobilized limbs (Fig. 2B). Mass spectrometry analysis of immobilized muscle shows that two proteins, SLC25A3 and SLC25A4, were decreased with immobilization with 21 proteins upregulated (Fig. 3 and Supplemental Table S2; see https://doi.org/10.6084/m9.figshare.19874227.v1). Target prediction analysis indicates that SLC25A3 is a target of miR-124-3p and miR-181-5p, whereas SLC25A4 is a target of both let-7c and let-7b (Supplemental Figs. S1 and S2). In addition, let-7c, let-7b, miR-181a, and miR-124-3p all share various roles in cellular senescence (Table 2). Upregulated proteins include embryonic and perinatal myosins, such as Myh3 and Myh8, which are associated with muscle regeneration in adults. Among the five most downregulated miRNAs, only mir-20b-5p and mir-208a-5p have a known role in muscle biology (Table 2).

Figure 3.

Heatmap showing changes in skeletal muscle proteins with single hindlimb immobilization. The mitochondrial protein transporters, Slc25a3 and Slc25a4, are downregulated (green) with immobilization, whereas embryonic and perinatal myosins associated with muscle regeneration (e.g., Myh3, Myh8) are increased (red) in muscle with immobilization.

IL-1β Knockout Disrupts the miRNA Network Observed with Immobilization in FAP-Derived, PDGFRα⁺ EVs

Analysis of miRNAs isolated from FAP-derived, PDGFRα⁺ EVs of the IL-1β KO mice revealed a total of 105 differentially expressed miRNAs between the immobilized and nonimmobilized limbs (Supplemental Table S3; see https://doi.org/10.6084/m9.figshare.17430707.v2). PCA of the samples showed that immobilized and nonimmobilized samples did not have clear separation (Fig. 4A). Thirty of these miRNAs were upregulated with immobilization, whereas 75 of these miRNAs were downregulated with immobilization as shown in the heatmap (Fig. 4B). Although many of the miRNAs altered with immobilization in wild-type mice have established functions related to muscle biology and senescence (Table 2), only miR-210 (upregulated) and miR-106b-2-3p (downregulated) among the IL-1β KO mice are thought to be involved in muscle biology or senescence (Table 3).

IL-1β KO Mice Experience Less Muscle Atrophy Compared with Wild-Type Mice

Analysis of quadriceps muscles isolated after immobilization reveals that IL-1β KO mice exhibit less muscle atrophy compared with their BALB/c wild-type counterparts (Fig. 5). On average, the IL-1β KO mice lost 16.5% of their quadriceps muscle mass, compared with 22.6% of muscle mass lost in the BALB/c mice (P = 0.0036). The IL-1β KO mice lost on average 375.7 µm area fiber size with immobilization compared with the 522.6 µm lost with immobilization in the wild-type mice (P = 0.0196).

Figure 5.

Loss of muscle mass and fiber size with immobilization are attenuated in IL-1β KO mice. A: percent change in quadriceps muscle mass with immobilization of BalbC wild-type and IL-1β KO mice. Loss of muscle mass in the IL-1β KO mice is 16.5% loss (n = 22) compared with an average of 22.6% lost in the wild type (P = 0.0036, n = 17, unpaired t test). B: change in quadriceps fiber size with immobilization. A significant decrease is seen with immobilization (L) compared with the nonimmobilized muscle (R) in both the IL-1β KO (P = 0.0021) and the BalbC wild type (P < 0.0001). However, there is a significant difference between the fiber size of the BALB/c wildtype and the IL-1β KO muscle (P = 0.0196). (For BALB/c WT n = 17, for IL-1β KO n = 12.)

DISCUSSION

Extracellular vesicles, including exosomes and microvesicles, are known to play important roles in cell-cell communication in a variety of cell types. Skeletal muscle is an important source of EVs (49), and we have previously shown that the miRNA cargo of muscle-derived EVs is significantly altered with age (20). The role(s) of FAP-derived EVs in muscle atrophy is, however, not well understood. It is recognized that FAPs are key to stimulating muscle regeneration mediated by satellite cell activation (2, 7, 11), and that the miRNA cargo of FAPs can be “tuned” using histone deacetylase (HDAC) inhibitors to reduce fibrosis (19). Yet, the miRNA cargo of FAPs in both normal and atrophy-associated settings has not been explored. To our knowledge, our study is the first to isolate and characterize FAP-derived, PDGFRα⁺ EVs from skeletal muscle ex vivo. This approach has enabled us to profile FAP-derived miRNAs carried by EVs in normal muscle and muscle impacted by disuse. Results show that a number of the miRNAs upregulated in FAPs with immobilization have previously been shown to play various roles in muscle biology. For example, miR-181a-5p is overexpressed in muscular dystrophies (31, 32), is upregulated with fasting (33), and plays important roles in regulating muscle development and hypertrophy by targeting factors, such as MyoD and Sirt1 (34).

Our analysis also showed that let-7 family members let-7c-5p and let-7b-5p were upregulated in FAP-derived EVs with immobilization. The let-7 family are well-known tumor suppressors and increase with aging (23–25). They have also been shown to increase in skeletal muscle after 21 days of bed rest (50). Furthermore, these miRNAs are associated with cellular senescence. For example, let-7c is a well-established inducer of cellular senescence in a number of different cell types (26–28). Likewise, let-7b shares a number of targets with let-7c related to senescence (26, 27). The let-7 family, like miR-181a, can both modulate mitochondrial activity and contribute directly to senescence, forming a group of senescence-associated “mitomiRs” (29). Proteomic analysis of immobilized muscle revealed downregulation of two mitochondrial transport proteins, SLC25A3 and SLC25A4 (Fig. 3). SLC25A3, which decreased with immobilization, is a target of miR-124-3p and miR-181-5p, both of which increased with immobilization in FAP-derived EVs. Decreased levels of SLC25A3 are observed in sarcopenia patients with hip fracture (51) and SLC25A3 mutations are associated with myopathies (52, 53). SLC25A4, which also decreased with immobilization, is a target of let-7c and let-7b, which increased with immobilization. Mutations of SLC25A4 are also associated with myopathies (54).

We previously found that single hindlimb immobilization not only increased IL-1β in FAPs but also increased expression of the senescence-related gene Cdkn2a (p16 INK4A) (15). Increased IL-1β and Cdkn2a expression in FAPs with immobilization, and increased let-7b and -7c expression in FAP-derived EVs, are all consistent with immobilization inducing a senescence-associated phenotype in FAPs (15). Importantly, the miRNA profile we observed in FAP-derived EVs with immobilization is markedly altered in mice lacking IL-1β. The network analysis reveals few muscle-specific targets among the miRNAs isolated from the FAP-derived EVs of IL-1β KO mice, and only one senescence-associated miRNA (miR-106b-3p) was detected in these EVs from IL-1β KO mice. Of note, this miRNA was actually downregulated in FAP-derived EVs from the IL-1β KO mice, whereas other senescence-related miRNAs were upregulated in EVs of the wild-type mice. In addition, immobilized IL-1β KO mice experience less muscle atrophy compared with wild-type mice. These findings suggest that IL-1β expression may have a critical function in FAPs by stimulating the expression and release of miRNAs driving senescence and muscle atrophy. Together these results point to a working model in which disuse increases IL-1β expression FAPs, leading to the secretion of EVs carrying pro-senescent and atrophy-related miRNAs that impact surrounding myofibers leading to reductions in muscle mass.

Prolonged inactivity due to a sedentary lifestyle, bed rest, or exposure to microgravity with spaceflight are all associated with muscle atrophy and fracture risk (1). Recently, it has been suggested that prolonged disuse in the form of bed rest may actually represent a setting of accelerated aging (55). This concept is based on the observation that a number of frailty biomarkers, including inflammatory cytokines, are increased with bed rest (55). These cytokines include members of the senescence-associated secretory phenotype (SASP), such as IL-1β and IL-6. Our data provide some support for this model in showing that FAP-derived EVs carry miRNAs that are also associated with cellular senescence, and that depletion of the SASP factor IL-1β may reverse many of the miRNA changes that are seen in FAP-derived EVs with disuse and immobilization. These data in turn suggest that senolytic therapies may suppress some of the changes in FAPs and other skeletal muscle cells in conditions associated with disuse-related muscle atrophy.

This study has several limitations. The first is that the IL-1β KO mice were global knockouts and not tissue specific, and so systemic effects cannot be ruled out. The second is that we did not examine recovery following reloading, and it is certainly possible that many of the observed changes may be reversed with normal ambulation. Future studies should be directed at determining to what extent the FAP-derived EV cargo may be modified with reloading, and what changes in muscle with immobilization are more difficult to reverse. Finally, our study focused on miRNAs, which are known to be abundant in EVs. It is likely that proteins and various lipids are also transported from FAPs via EVs, and future studies should be directed at further characterizing these components of the FAP secretome with disuse atrophy.

DATA AVAILABILITY

The data that support the findings of this study will be made available upon reasonable request from the corresponding author.

SUPPLEMENTAL DATA

Supplemental Figs. S1 and S2: https://doi.org/10.6084/m9.figshare.17430698.v4.

Supplemental Table S1: https://doi.org/10.6084/m9.figshare.17430701.v2.

Supplemental Table S2: https://doi.org/10.6084/m9.figshare.19874227.v1.

Supplemental Table S3: https://doi.org/10.6084/m9.figshare.17430707.v2.

GRANTS

This work was supported by the National Institute on Aging (AG036675).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E.P., B.M., L.R., B.M., and W.Z. performed experiments; E.P., Y.L.T., T.J.L., A.S., M.J., and J.C. analyzed data; E.P., Y.L., S.F., Y.L.T., M.M.-L., J.C., and M.W.H. interpreted results of experiments; E.P. prepared figures; E.P. drafted manuscript; M.W.H. edited and revised manuscript; M.W.H. approved final version of manuscript.