Synaptic-like coupling of macrophages to myofibers regulates muscle repair

Summary

Peripheral injury responses essential for muscle repair and nociception require complex interactions of target tissues, immune cells and primary sensory neurons. Nociceptors and myofibers both react robustly to signals generated from circulating immune cells, which promote repair, growth, and regeneration of muscle while simultaneously modulating peripheral sensitization. Here, we found that macrophages form a synaptic-like contact with myofibers to hasten repair after acute incision injury and to facilitate regeneration after major muscle damage. Transient chemogenetic activation of macrophages enhanced calcium dependent membrane repair, induced muscle calcium transients in vivo, elicited low level electrical activity in the muscles and enhanced myonuclear accretion. Under severe injury, macrophage activation could also modulate pain-like behaviors. This study identifies a novel mechanism by which synaptic-like functions of macrophages impact muscle repair after tissue damage.

eTOC:

Tripathi et al. investigate the effects of transiently activating macrophages within damaged skeletal muscle and their contribution to tissue repair versus pain. Macrophages make synaptic-like contacts with muscle fibers after injury to promote repair. These findings reveal new ways that circulating immune cells interact with muscle to modulate healing processes.

Graphical Abstract

Introduction

One mechanism for skeletal muscle injury is through disruption of the integrity of the myofiber plasma membrane and basal lamina, leading to altered intracellular calcium levels ^1,2. Functional restoration after muscle damage requires a coordinated response of multiple cellular systems including the myofibers, immune cells and sensory neurons ³. Muscle recovers through a variety of mechanisms that include restoring the disrupted basal lamina and myofiber membrane, or if necessary, myogenesis, ultimately resulting in re-establishment of muscle function ^4,5. Importantly, to allow proper repair of peripheral structures, an indispensable nociceptive signal is generated to inform the organism of the ongoing repair process ⁶. Nociceptors and myofibers both respond robustly to signals generated from circulating immune cells ^{7 8}. Given the architecture of skeletal muscle tissue and the wide range of injuries and diseases that prompt muscles to regenerate, it is important to obtain a better understanding of the time-course required and multi-cellular interactions that are established during various injuries to facilitate muscle repair ⁴.

Immune cells, including macrophages, infiltrate sites of injury to facilitate inflammation and repair processes ^9–11. In muscle, they can promote myofiber growth and regeneration ^12,13. Macrophages release a variety of molecules including cytokines and chemokines which can play a role in both muscle repair and nociception ^14–16. Vesicles are used to package these factors within macrophages and can contain other molecules such as various proteins or ions that are concurrently released with the cytokines/chemokines ^17,18. A recent study showed that in the heart, resident macrophages could also electrically couple to cardiomyocytes to modulate their function through direct delivery of calcium ions between cells ¹⁹. It has further been shown that macrophages can transport mitochondria to sensory neurons in the DRGs ²⁰ under injury conditions suggesting direct connections may be possible with neurons as well. However, it is not known whether a similar phenomenon with circulating immune cells could dually modulate skeletal muscle repair and/or nociception after injury. A better understanding of the complex interactions between these cellular systems would fill a major gap in our understanding of myalgia development and/or disorders of muscle repair/function. In this study, we hypothesized that infiltrating macrophages would couple to both myofibers and neurons to dually modulate muscle repair and pain. We performed a variety of assays assessing muscle repair and pain after acute or severe injury with chemogenetic or optogenetic manipulation of infiltrating macrophages and compared results to mice with synaptic transmission inhibited in macrophages. We found that macrophages form synaptic-like contacts with myofibers to hasten repair after both acute traumatic injury and major muscle damage. Interestingly however, muscle repair could be dissociated from pain-like behaviors during acute injuries indicating that these two distinct biological processes can be distinguished at the level of the infiltrating macrophage.

Results

Chemogenetic activation of macrophages restores membrane integrity after acute incision injury.

Reports suggested that macrophages may be able to couple to other cell types to modulate their functions ^21,22. We therefore utilized a chemogenetic strategy to induce calcium signaling selectively but transiently within macrophages (LysMCre;Chrm3 (GqDREADD)) in mice with incision injury to the flexor digitorum brevis (FDB) muscle. To tailor macrophage activation to shortened time frames, we only injected mice with Clozapine-N-Oxide (CNO) (0.4μg/μl) 1x daily intra -peritoneally beginning 1d after incision injury and immediately (within 1hr) prior to behavioral analyses. This prevented a sustained activation of the macrophage which could induce other signaling pathways or release of cytokines and chemokines which would have prevented us from uncovering a direct effect of macrophages on myofibers. To first determine if transient activation of macrophages could modulate muscle repair after injury, we assessed Evan’s Blue Dye (EBD) uptake in mice with incision. We found that incision of the muscles increased EBD uptake in the myofibers 2d after incision compared to uninjured contralateral muscles (Figures 1A–B). Surprisingly, we found that transient chemogenetic activation of macrophages was able to significantly block the uptake of EBD into the muscles and partially restore membrane integrity after incision (Figures 1C–D). Validation of calcium activation in macrophages containing the GqDREADD were performed in vitro and demonstrated that CNO can effectively induce calcium production in these immune cells (Figures S1 A–E).

Figure 1: Transient chemogenetic activation of macrophages regulates muscle repair.

EBD (red) is readily detectable in muscle in control (Chrm3) mice treated with CNO + incision at 2d (A) compared to contralateral control muscle (B), but this is inhibited in LysM;Chrm3 mice (C) with CNO+Incision. Quantification of EBD+ myofibers in the indicated groups (D). Purple=WGA. E-F: Areas of fibrosis (arrows) are detected in control Chrm3 mice 2d post incision as indicated by Sirius red (purple) staining, but this is significantly inhibited in CNO treated, LysM;Chrm3 mice. G-H: Myofiber damage (IgM+, red) is found in Chrm3 mice with incision at 2d but this is also significantly blunted in LysM;Chrm3 mice treated with CNO. *p<0.05, **p<0.01 vs. Chrm3. Unpaired t-test, n=3 mean±SEM. See also Figure S1, Figure S2, Table S1 and Data S1 A.

To more fully characterize the effects of incision on muscle repair processes, we also assessed the extent of fibrosis (Sirius red), myofiber damage (IgM+), muscle fiber maturation (eMHC) and myofiber cross-sectional area. We found that incision induced some fibrosis and muscle fiber damage 2d post injury, but chemogenetic activation of macrophages blunted these effects as well (Figures 1E–H). No changes in muscle fiber maturation were found between groups and no alterations in myofiber cross sectional area were detected between different conditions (Figures S1 F–H).

To then assess whether transient activation of macrophages could alter pain-like behaviors in mice with incision, we performed spontaneous and evoked assessments of hypersensitivity in LysMCre;Chrm3 mice treated with CNO. Surprisingly, we found that incision provoked spontaneous paw guarding and reduced muscle withdrawal thresholds by 1d after incision but transient chemogenetic manipulation of macrophages did not alter any of these behavioral changes induced by incision at any time point (Figures S1 I–J). This suggests that muscle repair processes can be dissociated from pain-like behaviors by the way macrophages interact with myofibers after incision injury.

Since LysM is also known to be expressed in neutrophils and dendritic cells (DCs), we stained muscles for these immune cells after incision. We found almost no neutrophils and few DCs present in the injured muscles at the time point (2d) in which an effect of DREADD dependent modulation of myofiber integrity was detected (Figures S2 A–D). Our previous work has shown that the muscles contain virtually no LysM+ cells prior to injury ²³. It is thus likely that the macrophages detected after incision are infiltrating and not resident. To confirm this, we stained muscles 2d post incision for CCR2 and found an abundance of CCR2+ cells (Figures S2 E, F). Therefore, effects are likely to be mainly attributed to infiltrating macrophages.

Activation of macrophages induces calcium transients in muscle tissue and rapidly modulates electrical activity in the muscles.

To then assess how the activation of macrophages could influence muscle function after injury, we performed intravital calcium imaging in the hind paw muscle of anesthetized LysMCre;Chrm3 mice with incision. Delivery of CNO directly to the muscles in the hind paw induced robust calcium florescence within a short timeframe after initiating delivery (~10–30s). CNO had virtually no effects on muscle calcium transients in vivo in control (Cre negative, Chrm3) mice with incision. (Figures 2A–D; Video S1). These data indicate that chemogenetic activation of macrophages is capable of inducing calcium transients in muscle tissue in injured mice on a rapid time scale.

Figure 2: Transient activation of macrophages induces calcium transients in muscle.

Example images of calcium induction (Rhod-2, red, arrows) in muscle in vivo in LysM;Chrm3 mice before (left) and 10–30s (two separate examples) after (right) delivery of CNO to the hind paw (A). Example traces of fluorescence intensity changes in muscle upon CNO delivery (~30–40s prior to onset of response) in control (B) and LysM;Chrm3 (C) mice with incision. Quantification of changes in peak fluorescence between conditions (D). *p<0.05 vs Chrm3, Unpaired t-test, n=3, mean±SEM. Chemogenetic activation of macrophages increases EMG activity in the muscles of LysM;Chrm3 mice compared to Chrm3 controls (E). Firing rates are significantly increased in the LysM;Chrm3 mice (F). **p<0.01 vs Chrm3. Unpaired t-test, n=4. Similar results are obtained in mice with optogenetic activation of macrophages. Immediately upon blue light delivery, EMG activity is detected in muscles in LysM;Chr2 mice with incision (G). This is not detected in controls (Chr2). Firing rates are significantly increased in the LysM;Chr2 mice (H). *p<0.05 vs Chr2. Mann-Whitney test, n=4, mean ± SEM. Representative images of calcium imaging of myotubes isolated from HSACre;Gcamp6BP mice co-cultured with macrophages obtained from Chrm3 or LysMCre;Chrm3 mice before (left) or after (right) delivery of CNO showing increased calcium in myofibers (arrows) after CNO only in LysM;Chrm3 co-cultures (I). Example traces of fluorescence intensity changes in myotubes upon CNO delivery in control (J) and LysMCre;Chrm3 (K) co-cultures. Quantification of changes in peak fluorescence between conditions (L). *p<0.05 vs Chrm3, Unpaired t-test, n=3 per condition, mean ± SEM. (n=5 cells/ well, 15 cells/ animal). See also Figure S3, Video S1, Video S2, Video S3, Table S1 and Data S1 B.

We then wanted to assess how DREADD dependent activation of macrophages may modulate muscle function. We therefore performed electromyographic (EMG) recordings in LysMCre;Chrm3 mice with incision and compared to controls. Similar to intravital calcium imaging experiments, CNO was delivered directly to the hind paw during EMG recording. Interestingly, within a relatively short time frame after CNO delivery (~10–30s), we detected bursts of activity within the muscles of LysMCre;Chrm3 mice with incision, but not in the Cre negative incised animals (Figures 2E–F). This suggests a direct correlation between calcium activation in macrophages and electrical activity in the muscles.

Although the results were promising and indicated that macrophage activation can directly and rapidly regulate muscle function during incision, we could not confirm whether this was due to slow release of vesicles from macrophages onto the muscle fibers or whether a more direct delivery was established. To begin to determine this, we performed electromyography on hind paw muscles from incised mice with channelrhodopsin expressed in macrophages (LysMCre;Chr2). Surprisingly, we found an almost immediate and prolonged induction of EMG activity upon delivery of blue light to the muscles that was not observed in Cre negative controls with incision (Figures 2G–H). This data strongly supports a direct coupling mechanism between macrophages and muscle fibers that modulates muscle function and repair processes after injury.

To determine if there was a direct effect of activating macrophages on muscle fibers, we then assessed calcium transients in myotubes isolated from Ai95-D mice co-cultured with macrophages isolated from LysM;Chrm3 and Cre negative control mice. After delivery of CNO to co-cultures in vitro, calcium activation was observed in myotubes which were co-cultured with LysMCre;Chrm3 macrophages. No major changes in fluorescence were detected in control cultures (Figures 2 I–L; Figure S3, Video S2 and S3). This data indicates that chemogenetic activation of macrophages can directly induce calcium transients in muscle fibers.

Macrophages make synaptic-like contacts with muscles to modulate repair.

As our data thus far suggested a rapid regulation of muscle repair upon activation of macrophages, we wanted to assess how this may occur. Macrophages are known to release a variety of cytokines and chemokines to modulate inflammation and repair ^9,22. We therefore assessed the levels of selected cytokines and chemokines in the muscles of LysMCre;Chrm3 mice with incision injection and treated with CNO. Interestingly, in mice with incision injury, no changes in genes such as IL6, TNFα, IL1β or MCP1 were observed in CNO treated LysMCre;Chmr3 mice compared to treated Chmr3 controls at 2d post injury (Figure S4A). Data suggests that transient activation of macrophages in vivo has no detectable effects on production of key cytokines/chemokines after incision injury.

Since previous reports suggested that resident macrophages can form gap junctions with cardiomyocytes to regulate their function ¹⁹, it is plausible that circulating macrophages formed similar contacts in an experimental model of peripheral injury (incision). We therefore first wanted to verify that the infiltrating macrophage population contained appropriate proteins for generation of gap junctions which could serve as “electrical synapses” between macrophages and myofibers. We analyzed our previously obtained RNA-Seq dataset on macrophages sorted using fluorescence activated cell sorting (FACS) methods obtained from macrophage reporter mice (LysMCre;tdTom) ²³. Surprisingly, we found that LysM+ cells do not contain gap junction proteins. They do, however, express readily detectable levels of synaptic genes including synaptobrevin 2 (VAMP2), extended synaptotagmin 1 (Esyt1), synaptophysin-like 1 (Sypl) and syntaxin 7 (Stx7) to name a few (Figure 3A).

Figure 3: LysM+ macrophage RNA-Seq and PSD95 staining in muscles from mice with incision.

RNA-Seq analysis for indicated genes from sorted LysMCre;tdTomato+ cells (A). Although gap junction mRNAs such as Gja1 (i.e. connexin 43) are not readily expressed in macrophages (TPM (Transcripts Per Million) values need to greater than 10 for reliable detection), synaptic proteins are highly expressed in these cells. An = animal. Post-synaptic density protein 95 (PSD95, green) is detectable in muscle marked by laminin (white) in naïve mice (B). Increased PSD95 is detected 2d post incision in the muscles and these are often colocalized with macrophages (C, F4/80, red). Higher magnification examples are also provided in C (lower panels). Quantification of PSD95 and F4/80 in muscles from mice with incision indicates an increase in PSD95 that colocalizes with F4/80+ macrophages (D). **p<0.01, ***p<0.001 vs naïve. Unpaired t-test, n=3 mean ± SEM, Scale Bars, 25μm. See also Figure S4, Table S1 and Data S1 C.

We therefore wanted to determine if infiltrating macrophages could form synaptic-like contacts with myofibers instead of forming electrical synapses using gap junction proteins. Muscle tissues were thus stained from mice with incision injury for the synaptic marker post-synaptic density protein 95 (PSD95) along with laminin and F4/80 (another macrophage marker). An increase in PSD95+ myofibers were observed in mice with incision injury and many of these co-localized in direct apposition with F4/80+ macrophages (Figures 3B–D). Since the neuromuscular junction also expresses PSD95, we further quantified α-BTX staining with PSD95 and found no differences in NMJ numbers two days post incision (Figures S4C, D). These results with those presented above collectively indicate that macrophages may be one cell capable of generating synaptic-like contacts to modulate repair.

To test the idea of synaptic-like coupling, we crossed LysMCre;Chrm3 mice with a Cre-inducible botulinum toxin mouse (iBot), which would disrupt synaptic machinery, including VAMP2 ²⁴. In these mice, macrophages are not able to release vesicles via these synaptic protein complexes even after chemogenetic activation of intracellular calcium using the GqDREADD approach. We therefore performed EBD uptake and EMG analyses on muscles from these groups and found that LysMCre;Chrm3;iBot mice treated with CNO or LysM;iBot mice alone after incision display increased EBD in the muscles 2d post incision and have reduced EMG activity (Figures 4A–G) similar to mice with no GqDREADD expressed in macrophages at this time point. To determine if calcium related repair processes could be at play in the observed hastening of myofiber repair after incision injury, we isolated muscles from CNO treated LysM;Chrm3 and LysM;Chrm3;iBot with incision injury on day two and stained muscles for the calcium dependent membrane repair protein, dysferlin. We found that compared to controls, incision-related reduction in dysferlin staining was restored by chemogenetic activation of macrophages in mice with incision and this is blocked in the iBot mice (Figures 4H–I).

Figure 4: Blocking synaptic vesicle release inhibited the effects of chemogenetically activating macrophages on muscle repair.

Examples of EBD labeling from Cre negative, Chrm3;iBot control (A), LysMCre;Chrm3 (B) and LysMCre;Chrm3;iBot (C) LysMCre;iBot (D) groups 2d after incision injury to the FDB muscles. Quantification of EBD positive fibers per section is provided (E). ***p<0.001, ****p<0.0001 vs indicated condition. 1-way ANOVA with Tukey’s post hoc test. LysM;Chrm3 mice containing Cre inducible botulinum toxin (iBot) in macrophages display little EMG activity upon CNO delivery unlike that observed in LysM;Chrm3 mice in vivo (F). Quantification of firing rates post CNO in three groups listed (G). *p<0.05, 1-way ANOVA with Tukey’s post hoc test, n=3, mean ± SEM. Scale Bars, 25μm. (H) Dysferlin (red) is readily detectable in uninjured muscles, and this is reduced at 2d post incision in control (Chrm3) mice. CNO treatment in LysM;Chrm3 mice with incision reverses this reduction and this is prevented in iBot mice. (I) Quantification of dysferlin intensity in the indicated conditions. DAPI=blue; **p<0.01 vs. uninjured. *p<0.05 vs uninjured and Chrm3+Inc. 1-way ANOVA with Tukey’s post hoc, n=3, mean ± SEM, Scale Bars, 25μm. See also Figure S5, Table S1 and Data S1 D.

To finally assess if calcium was one factor that could be released from macrophages to modulate these phenomena, we performed calcium imaging experiments in vitro on macrophages isolated from the LysM;Chmr3 mice and treated cultures with CNO. Cells were incubated in calcium free media to observe any release of these ions into the bath upon CNO treatment. We indeed found that chemogenetic activation of macrophages caused calcium to be detected in the bath readily upon CNO treatment while Cre negative macrophages showed little changes in bath levels of calcium. These effects were blocked in the iBot mice (Figures S5 A–G). To confirm whether more physiological-like stimuli could also activate macrophages and release calcium, we then tested whether lipopolysaccharide (LPS) treatment could elicit responses in macrophages in culture isolated from LysM;iBot mice and compared to controls. We found that LPS could indeed induce release of calcium from macrophages, and this is blunted in the LysM;iBot mice (Figures S5 H–L). Together these results indicate that infiltrating macrophages may synaptically couple to myofibers after damage in part to facilitate a calcium dependent repair process in the muscles after incision.

Macrophage activation enhances muscle regeneration and pain-like behaviors after cardiotoxin injury.

As there appeared to be direct coupling of macrophages to muscle tissue after acute traumatic injuries to regulate repair, we wanted to also determine if this phenomenon could be observed in other models of muscle injury. We thus performed an analysis of muscle regeneration from LysMCre;Chrm3 mice injected with cardiotoxin (CTX) into the tibialis anterior muscles which causes significant disruption of the muscles and allows for the assessment of muscle regeneration. We first found that a greater number of macrophages are present in the muscles after CTX injury compared to incision (Figure S6). Quantification of peripheral and central nuclei per myofiber in addition to fibers with more than 2 nuclei were determined in LysMCre:Chrm3 and Cre negative Chrm3 groups. We found that in control mice with CTX injury, we observed the expected increase in central nuclei present per myofiber, 10 days post injection. However, in LysMCre;Chrm3 mice injected with CTX and then CNO every other day beginning at 1d, we found an increase in fibers with more than 2 nuclei indicating either hastening or enhanced level of muscle regeneration 10 days post CTX (Figures 5A–D). No differences in peripheral or central nuclei were found between the different groups at 3d or 7d post CTX injection (Figure S6). To further characterize the effects of activating Gq in macrophages on muscle regeneration, we assessed stem cell activation marker, Pax7, over time post CTX injection in these mice. Although no changes in Pax7 cells were detected at 3d or 7d post CTX between groups (Figure S6), we did observe a significant increase in Pax7 cells in the muscles from LysMCre;Chrm3 mice with CNO delivery at 10d post CTX injection (Figures 5E–G). To determine if any pain-related behaviors could be modified in the CTX model, we performed similar guarding and mechanical hypersensitivity measurements in CNO treated LysMCre;Chrm3 mice injected with CTX. We found that control animals with CTX show modest guarding behaviors and a delayed onset mechanical hypersensitivity to muscle squeezing that lasted for the duration of our testing period (10d). Surprisingly, unlike that observed with incision injury (see Figure S1), LysMCre;Chrm3 mice injected with CTX and then CNO every other day display reduced guarding and mechanical hypersensitivity (Figures 5H, I).

Figure 5: Chemogenetic activation of macrophages alters muscle regeneration.

10d after cardiotoxin (CTX) injection into the tibialis anterior muscle of control mice treated with CNO (Chrm3), many myofibers (indicated by red arrows) surrounded by laminin (white) display central nuclei (blue; A). In LysMCre;Chrm3 mice treated with CNO after CTX injection, more fibers show multiple nuclei (B). Quantification of peripheral, central and multinucleated myofibers at 10d (C-D), unpaired t-test, n=3, mean±SEM. Control Chrm3 mice injected with CTX also show some Pax7 positive cells (red) in the muscles (white arrows) at 10d (E). CNO treated LysM;Chrm3 mice with CTX at 10d however display greater Pax7 staining (arrows) in the muscles (F). Quantification of Pax7 positive cells per 100 myofibers in CNO treated Chrm3 and LysM;Chrm3 mice injected with CTX at 10d, *p <0.05 vs Chrm3, unpaired t-test, n=3, mean±SEM, Scale Bars, 25μm. (G). Control (Chrm3) mice display significant guarding (H) and mechanical hypersensitivity to muscle squeezing (I) beginning 3d post CTX injection. These CTX-induced behaviors are significantly inhibited in LysMCre;Chrm3 mice treated with CNO. **p<0.01, ***p<0.001,****p<0.0001 vs baseline, #p<0.05, ##p<0.01 vs time matched controls. 2-way RM ANOVA with Tukey’s post hoc, n=16, mean ± SEM. See also Figure S6, Table S1, Data S1E and Data S1 F.

We also determined that DREADD dependent activation of macrophages increases the cross-sectional area of myofibers undergoing repair and reduces fibrosis 10d after CTX (Figure 6). We also observed similar effects on gene expression in the CTX treated mice with CNO as we observed after incision for TNFα, IL1β and MCP1 but we did observe an increase in IL6 in both LysMCre;Chrm3 mice and Chrm3 controls at 10d post CTX (Figure S4B). These data indicate that macrophage activation can enhance muscle regeneration in severe injury models and under these larger injury conditions, activation of these immune cells can also modulate pain-related behaviors over time.

Figure 6: Characterization of regenerated myofibers.

Quantification of myofiber cross sectional area (CSA) in muscles from mice with CTX. Increased in CSA is detected in LysmChrm3 group after CTX injury. *p<0.05 vs control. Unpaired t-test, n=3 mean ± SEM (A). Histological confirmation of muscle regeneration validates the increase of >2 central nuclei in LysM,Chrm3 mice (B). Quantification of >2 central nuclei in Chrm3 and LysM;Chrm3 mice treated with CNO at 10 days after injection with CTX. ***p<0.001 vs control. Unpaired t-test, n=3 mean ± SEM, Scale Bars, 25μm (C). Sirius red staining is decreased in muscles in LysmChrm3 mice after incision and CTX injury compared to Chrm3 controls (D). Quantification of Sirius red in muscles from mice with incision and CTX (E). *p<0.05 vs control. Unpaired t-test, n=3 mean ± SEM. Bars, 25μm. See also Figure S6, Table S1 and Data S1 G.

Discussion

This report is the first to show that infiltrating macrophages can form synaptic-like associations with myofibers (Figures 3–4) to rapidly regulate (Figure 2) calcium-dependent repair in the muscle tissue (Figures 1, 4). After severe muscle injury, transient macrophage activation could also modulate pain-like behaviors (Figures 5–6), however, during acute injuries, pain could be dissociated from muscle repair when macrophages were briefly activated (Figure S1). Results uncover a novel mechanism by which circulating immune cells contribute to repair processes in the skeletal muscles after damage.

Muscle injuries occur from a variety of causes, including direct trauma (e.g., lacerations, contusions, or strains) or indirect causes (e.g., ischemia or neuromuscular dysfunction) ^25,26. The general phases of healing occurring within the damaged muscle are mostly similar among the various types of muscle injuries, but the functional recovery of the injured muscle varies from one type of injury to another. It has become clear that the processes occurring in injured muscle (i.e., necrosis/degeneration, inflammation, repair, and scar-tissue formation [fibrosis]) are all interrelated and time-dependent ^1,27,28. In this study we considered hind paw incision injury as an acute trauma model ^29–31 and cardiotoxin injury ^32–34 as major injury condition resulting in muscle degradation. Skeletal muscle repair is a highly synchronized process involving the activation of various cellular and molecular responses, where the coordination between inflammation and regeneration is crucial for the beneficial outcome of the repair process following muscle damage ^35,36.

Data from several models of muscle injury (hindlimb ischemia, freeze-injury, unloading/reloading, and myotoxic agent injections) indicate that impairment of immune cell recruitment including macrophages into injured muscle results in delayed tissue regeneration in terms of appearance of regenerating central nucleated myofibers and persistence of intramuscular adipocytes and fibrosis ^37–39. Part of the contribution of macrophages to repair processes relies on their removal of debris after injury. However, several findings suggest that macrophages may play a more direct role in muscle repair and remodeling than merely removing tissue debris. For example, muscle regeneration by transplanted myogenic cells can be impaired if the recipients of whole muscle grafts are depleted of monocytes and macrophages by irradiation before transplantation, which has been interpreted as showing a role for macrophages in muscle repair and regeneration in vivo ⁴⁰. Channels known to be expressed on muscle fibers that can allow calcium influx (e.g. TRPC3) are capable of facilitating the influx of calcium from extracellular sources or in this case released by macrophages. This could then induce further intracellular calcium release ⁴¹, contributing to the transients observed in muscle (Figure 2) and likely to membrane repair via protein like dysferlin (Figure 4), which is itself a membrane repair protein dependent on calcium.

A unique feature of infiltrating immune cells is that in addition to releasing cytokines and chemokines ^{42 43}, they may physically couple to tissues through connexin 43 (Cx43) dependent gap junctions ⁴⁴. A recent report suggests that resident macrophages in particular have the capacity to electrically couple to cardiomyocytes in the heart and modulate atrio-ventricular (AV) conduction via newly formed Cx43 gap junctions ¹⁹. Interestingly, peripheral nociceptors can also form new Cx43 gap junctions with each other in the dorsal root ganglion (DRG) after injury to regulate nociceptive responses ^45–47. Thus, it was reasonable to hypothesize that infiltrating immune cells such as macrophages could have electrically coupled to damaged myofibers to facilitate functional recovery ^42,48. However, we found that infiltrating macrophages did not contain gap junction genes but highly expressed synaptic-like genes (Figure 3). This effectively permitted these circulating cells to rapidly communicate with muscle tissue to modulate both calcium transients in myofibers and EMG activity in the muscles (Figures 2, 4) without altering pain-like behaviors during acute muscle injury (Figure S1).

Evan’s blue dye, considered as a marker to study muscle fiber integrity ^49,50 along with vessel leakage, was used here to show that membrane disruption after injury could be restored in a hastened time frame upon chemogenetic activation of macrophages (Figures 1, 4). To confirm that calcium related repair processes were at play in the observed phenomena, we stained muscles from these conditions with dysferlin which is calcium dependent membrane repair protein ^51,52. Complementing EBD studies, dysferlin levels were also found to be restored upon chemogenetic activation of macrophages in mice with incision injury (Figure 4). Moreover, to check the effects of macrophage activation on muscle tissue, calcium imaging and EMG recordings were performed. Not only did chemogenetic activation of macrophages induce calcium transients in muscle in vivo and in myotubes in vitro, induction of calcium signaling in these cells also modulated EMG activity in the muscles (Figure 2). Although these studies indicated a possible direct effect of macrophages on muscle tissue, the notion of rapid vesicle release to modulate repair remained ⁵³. We therefore performed optogenetic activation of macrophages during EMG recordings and observed an immediate onset of EMG activity in the muscle during light dependent activation of macrophages (Figure 2). The amplitudes of the EMG waves were small in most cases both in the chemogenetic and optogenetic experiments, but in other cases were nevertheless seemingly capable of inducing a contraction (Video S1). Nevertheless, these experiments suggest that regardless of a direct behavioral effect, transient activation of macrophages was able to rapidly initiate low level EMG activity that also could facilitate calcium-dependent muscle repair and regeneration (Figures 1–4). There is obviously still a role for released factors from macrophages in the inflammatory and repair processes ⁹, but this work also uncovers a novel mechanism by which macrophages can modulate repair.

We were able to confirm increases in synaptic-like contacts between macrophages and skeletal muscles after incision, that infiltrating macrophages do contain synaptic genes, and inhibition of synaptic-like transmission in macrophages blocked the chemogenetic effects of macrophages on myofiber repair and EMG activity (Figures 3–4). Thus, even though electrical coupling was not confirmed as was shown in the heart ¹⁹, synaptic-like coupling was observed between infiltrating macrophages and myofibers to regulate repair after incision. Specifically, postsynaptic density protein (PSD95) was colocalized to apposed macrophages ^54,55. This is the most abundant protein located at excitatory synapses in the nervous system and is involved in the stabilization, recruitment and trafficking of various receptors to the postsynaptic membrane ^56–58 and serves as a major component of neuromuscular junction (NMJ). Our data shows however that no changes in BTX/PSD95 are observed after incision injury, indicating that much of the increased PSD95 expression detected in the muscles are due to new contacts formed between myofiber and macrophages (Figure 3, Figure S4). Our data now includes a novel mechanism by which synapse-like contacts form between immune cells and myofibers to regulate repair.

We further showed a role for these novel contacts between macrophages and muscle in more severe injury models like CTX injury. Similar to that found after incision, activation of macrophages could hasten regeneration after damage (Figures 5–6) ^59–61. Normal skeletal muscles are composed of individual multinucleated myofibers with nuclei positioned at their periphery. Interestingly, some nuclei are positioned in the center of the myofiber, which is a marker of myofiber repair and has long been recognized as a sign of diseased muscle tissue ^{61 62}. In our severe injury model, LysMCre;Chrm3 mice treated with CNO injection after injury showed more fibers with multiple nuclei and increased Pax7 staining suggesting faster recovery than the control mice. Muscle injury typically initiates a rapid and sequential infiltration of muscle by inflammatory cell populations that can persist for hours to weeks, while muscle repair, regeneration, and growth occur¹⁵. This relationship between inflammation and muscle repair or regeneration has suggested that muscle inflammation after modified muscle use may be functionally beneficial.

Another interesting finding from these studies is the fact that pain-like behaviors were dissociated from tissue repair processes under acute injury conditions such as incision. This indicates that once initiated, sensitization mechanisms in neurons may continue their normal trajectory of sensitization and resolution regardless of the state of tissue repair. This would challenge the traditional dogma that pain manifests alongside tissue injury and resolves upon repair. We clearly showed here that through transient infiltrating macrophage activation, we can hasten muscle repair after incision injury at a time point when animals still display acute pain-like behaviors (Figures 1–4, Figure S1). Future experiments analyzing this relationship are therefore crucial. Nevertheless, these studies will offer a novel, mechanistic understanding of nociceptive processing and muscle repair after tissue damage in relation to the synaptic-like functions of macrophages. These studies are uniquely positioned to dramatically advance our mechanistic understanding of muscle pain development and myofiber repair and could identify a novel target for therapeutic intervention in several muscle injury states.

RESOURCE AVAILABILITY

Lead contact—

Further information and requests for resource and reagents should be directed to and will be fulfilled by the lead contact, Michael Jankowski (Michael.Jankowski@cchmc.org).

Materials availability—

Materials generated in this manuscript are available upon request

Data and Code Availability-

• The data have been deposited at Mendeley Data Repository at [DOI: 10.17632/4zbzb3rfg3.1] and are publicly available as of the date of publication.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.

STAR Methods

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Animals

Experiments were conducted with young adult (>P21) male and female mice. Several transgenic lines were used in various breeding pairs. We used the following lines in the current report: LysM-Cre (Jax stock #004781) mice express the Cre recombinase enzyme under the control of the lysozyme M (LysM) promoter. The LysM gene is expressed in myeloid cells, including monocytes, macrophages, and granulocytes, making this model useful for targeted gene manipulation within these cell types. Rosa26-LSL-hM3Dq-mCit (Jax #026220) allows Cre recombinase-inducible expression of a CAG promoter-driven HA-hM3Dq-pta-mCitrine. Rosa26-LSL-Chr2 (Jax #024109), express an improved channelrhodopsin-2/EYFP fusion protein following exposure to Cre recombinase. These mice can be used in optogenetic studies for rapid in vivo activation of excitable cells by illumination with blue light. Rosa26-LSL-iBot (Jax #018056), express a Clostridium botulinum neurotoxin serotype B, light chain (boNT/B) gene in a cre-dependent fashion allowing selective blockage of exocytosis. Ai95 (RCL-GCaMP6f)-D (Jax stock #028865) are a Cre-dependent, fluorescent, calcium-indicator strain. Ai95D has a floxed-STOP cassette preventing transcription of the GCaMP6 fast variant calcium indicator, these mice were crossed with a previously generated transgenic mouse which utilizes the human skeletal α-actin (HSA) promoter to drive expression of reverse tetracycline trans activator (rtTA) ⁶³. C57Bl/6 mice were also used for some control conditions. Mice were bred to generate LysMCre;Chrm3, LysMCre;Chr2 and LysMCre;Chrm3;iBot mice along with their littermate controls. All animals were housed in a barrier facility in group cages of no more than four mice, maintained on a 14/10 hour light/dark cycle with a temperature-controlled environment, and given food and water ad-libitum. All procedures were approved by the Cincinnati Children’s Hospital Medical Center Institutional Animal Care and Use Committee and adhered to National Institutes of Health Standards of Animal Care and Use under American Association for Accreditation of Laboratory Animal Care-approved practices. Animals were anesthetized with 2–3% isoflurane before induction of various treatments or with ketamine/xylazine (100 mg/kg / 10 mg/kg) prior to terminal procedures.

METHOD DETAILS

Hind paw Incision and Cardiotoxin (CTX) injury

Mice were anesthetized with 2–3% isoflurane and then an incision was performed from the hairy skin side of the hind paw between the bones through to the flexor digitorum brevis muscles. Blunt manipulation of the muscle was performed using #5 forceps, but the plantar skin was left untouched. Wounds closed with 6–0 sutures ⁷.

For CTX (Millipore Sigma, 217503) experiments (CTX aliquoted in sterile saline at a concentration of 10 μM), the agent was delivered directly into the tibialis anterior muscles. Mice were first anesthetized with isoflurane and legs shaved. Then 50 μL of CTX was injected using a 28-gauge needle ⁶⁴.

Pain-related Behavioral Paradigms

Behavioral examination of spontaneous paw guarding, and mechanical withdrawal thresholds using muscle squeezing assays were performed as described previously ⁶⁵. All behavioral assays were conducted by experimenters blind to conditions.

Spontaneous paw guarding

Before the commencement of behavioral studies, mice were randomly assigned to different groups into a rectangular transparent plastic box where they were acclimatized for approximately 10 minutes. Baseline pain scores were measured before the induction of injury. A cumulative pain score was used to assess spontaneous pain like behavior using a rating scale developed by Xu and Brennan ⁶⁵. Mice were allowed to move freely inside the transparent plastic box. The ipsilateral and the contralateral paws were closely monitored during a 1-minute period repeated every 5 minutes for 60 minutes. The testing scale scores movement of the hind-limb from where there is no observable hind paw movement off of the floor of the box (score = 0), through when a slight raise is observed (score = 1) to when the paw is completely raised off of the floor (score =2). Guarding was evaluated one animal at a time at 5-minute intervals for twelve readings and the average score is used as the mean value.

Mechanical withdrawal thresholds

The modified paw pressure device was used to quantify the nociceptive withdrawal reflex after muscle injury. This device was used to apply increasing mechanical force (up to 350 gram) on the plantar surface of both hind paws using a blunt/rounded tip which allows for stimulation of the muscles until a paw withdrawal is observed. The force at which the paw withdrew was determined to be threshold and this was repeated three times with a 5-minute interval in between trials. The average of the three measurements was determined per animal and measurements were averaged across groups for comparisons.

Immunohistochemistry

In order to evaluate the integrity of the skeletal muscle fibers, we injected mice (i.p.) with 200 μl of Evans-blue dye (EBD: 1% in sterile saline solution). EBD has been used extensively to evaluate the integrity and permeability of the membrane of muscle fibers ⁶⁶. Wheat germ agglutinin (WGA) conjugated with FITC (Invitrogen, Catalog # 32466) was used to co-stain the tissue to visualize the membranes in the skeletal muscle, as previously described ⁶⁷. Briefly, muscle tissue was embedded in Tissue-Tek O.C.T. compound (Sakura Finetek USA Inc.), flash frozen in liquid nitrogen and sectioned at 12 μM on a cryostat and mounted on slides. Tissue was fixed on slide using 4% paraformaldehyde in 0.1 M PBS. The samples were subsequently washed, blocked in 0.01 M PBS containing 5% horse serum, 1% bovine serum albumin, and 0.2% Triton X-100 for 10 min. Sections were stained with WGA-FITC (1:100), incubated for 1 h, washed and cover slipped.

In other experiments, muscle samples were collected from LysM;Chrm3 and control mice in the same way and stained with dysferlin antibodies (rabbit anti-dysferlin 1:200; Abcam, catalog #ab124684) using an overnight incubation. Sections were then washed with PBS and incubated with labeled secondary antibodies (Alexa Fluor 488, dilution 1:400; Thermo Scientific, catalog# A11034) for 35 mins at room temperature and cover slipped after PBS washing.

A separate set of muscle samples from mice were cut and stained with PSD95 (rabbit anti PSD95 1:1000 Gentex, catalog # GTX133091), F4/80 (Rat anti F4/80 1:200 Abcam, catalog #ab 6640), and Laminin (rabbit anti Laminin 1: 200 Abcam, catalog #ab 11575), incubated overnight, washed and labeled with appropriate secondary antibodies (Alexa Fluor Fab 488, Jackson ImmunoResearch, catalog # 111–547-003), 594 Donkey anti Rat, Thermo Scientific, Catalog # A21209), 647 goat Anti Rabbit, Thermo Scientific, A21244) at the dilution of 1:400 respectively for 35 mins and cover slipped after PBS washing.

In order to quantify nuclei in the CTX treated mice, muscle samples were collected as mentioned above after 3, 7 and10 days of CTX injection which were stained with laminin (rabbit anti Laminin 1: 200, Abcam, catalog #ab 15277), incubated overnight, washed and labeled with secondary antibody 647 goat Anti-Rabbit(Thermo Scientific, catalog # A21244) for 35 mins. The slides were then rinsed in PBS and cover slipped using Fluro media with DAPI (Electron microscope Sciences, Cat# 17985–50) to stain nuclei. Nuclei were determined in three non-consecutive sections and quantified as containing either one, two or multiple nuclei per myofiber and reported as mean ± SEM.

For anti- Pax7 staining an additional antigen retrieval step was included in addition to above protocol, boiling slides in 1× Antigen Retrieval Citra Plus Solution (Invitrogen,# 00500) for 30 minutes before blocking with M.O.M. mouse IgG blocking reagent (MKB-2213-NB),followed by incubation with 1:10 dilution Pax 7 antibodies (DSHB, #AB528428 ) overnight. The following day, secondary (IgG1) (Thermo scientific, 555, A-21127) was incubated for 30 min similar to that described above followed by cover slipping.

Other muscle samples were also stained with anti-eMHC (DSHB, catalog # F1.652), anti CD11c (1:250 Novus biologicals, catalog # NBP2–81066), or anti CCR2 (Bioss Antibodies, catalogue # BS23026R), incubated overnight. Sections were then washed with PBS and incubated with labeled secondary antibody (donkey anti mouse, dilution 1:400; Abcam, catalog# ab150101 or donkey anti rabbit, dilution 1:400; Jackson ImmunoResearch, catalog # 711–585-152) for 35 mins at room temperature and cover slipped after PBS washing. For IgM staining, primary antibodies conjugated to Texas Red (1:1000, Millipore Sigma, catalog# SAB3701210) were used with an overnight incubation followed by washing and cover slipping. The same protocol was used for Alpha-Bungarotoxin (BTX) (1:1000: Invitrogen: catalog# B13423) immunocytochemistry.

H&E processing was performed in the CCHMC pathology core laboratory for histology and neutrophil quantification. Sirius Red staining was performed (to identify Fibrosis) in PFA fixed slides and then stained with picro Sirius red solution for 1 hour. The cross-sectional area (CSA) of individual myofibers were measured using NIS Elements software (Nikon), which can quantify the area within laminin-labeled myofibers. After all CSA measurements were collected, CSA distribution graphs were generated by binning all data to produce a parametric distribution in control groups and then comparing experimental distributions using the same binning.

Distribution of fluorescent staining was determined with a Nikon confocal microscope with sequential scanning to avoid bleed-through of the fluorophores. Three nonconsecutive sections, separated at least by four sections, from three different animals per condition were used to quantify the images. Exposure times during microscopic analysis for each image was performed at the same intensity level to confirm staining above background. All the muscle cells in a section were labeled using ImageJ and muscle cells that were observed to contain red staining were considered positive. The percentage of positive cells obtained from each animal was used for comparisons.

Intravital calcium imaging

Mice were anesthetized under isoflurane anesthesia, had the hind paw muscles exposed and the hind paw isolated in a clay mold to allow superfusion of Krebs-Heinslfelt buffer (KH buffer) with 5% CO₂ for 10 mins. The hind paw muscles were then loaded with 5% Rhodamine-2 AM dye in KH buffer, for 30 mins before imaging. Muscles of the anesthetized mice were then analyzed under a two-photon confocal microscope (Nikon FN1 upright Mutliphoton) after rinsing with KH buffer for 2–3mins. CNO (2mL of 0.4 μg/μl) was delivered directly to the hind paw during imaging to assess changes in calcium transients upon chemogenetic activation of macrophages. Videos were taken for up to 5 minutes.

3 random ROIs (region of interest) at red emission filter (around 580nm) using NIS software were chosen from each condition for time dependent analysis of fluorescence intensity. Changes in fluorescence intensity were then analyzed and the change in intensity was calculated as ΔF/F=(Fmax-FO)/FO. Fmax= maximum intensity during stimulation. FO= intensity measured in the ROI immediately prior to delivery of CNO. Data was averaged within a condition and compared across indicated conditions.

In vitro calcium imaging

For macrophage cultures analyzed in calcium free media, macrophages from LysMCre;Chrm3 and control mice were collected from peritoneal cavity by injecting isolation media (3–5 mL of cold sterile PBS with 3%FBS) into the peritoneal cavity of mice. Subsequently, the cell-containing fluid of the peritoneal cavity and isolation media was gently collected in falcon tube, then centrifuged and resuspended in complete MEM-alpha medium (Sigma, M4655–1). After 24 hours, non-adherent cells were discarded and replaced with fresh medium. This allows for isolation of peritoneal macrophages as > 90% of cells remaining in dish will be macrophages. For imaging the following day, wells were loaded with 5% Rhodamine-2 AM dye in KH buffer (without calcium component) for 30 mins at 37°C/5% CO₂ and washed with KH buffer. 100μl(0.4 μg/μl) of CNO was delivered directly to the wells during imaging to assess changes in calcium transients upon chemogenetic activation of macrophages. During analysis, macrophages were encircled as ROIs (region of interest) at red emission filter (around 560nm) using NIS software for time dependent analysis of fluorescence intensity. Changes in fluorescence intensity were then analyzed, the average of the cells per well were calculated across each condition, then the average of the three wells were determined per condition and the change in intensity was calculated for that group and reported. Specific cell numbers are indicated in the figure legends for the respective experiment. Outer Green circles (in Figure S6) are the regions of interest around the macrophages to record the florescent intensity of the calcium signal outside the macrophage in the media indicating calcium release upon CNO delivery.

For macrophage/myotube cocultures, myotubes were developed from HSA-Gcamp6 mice as described previously ⁶⁸. Briefly, the hind limb muscles were isolated from mice, excess connective tissue and fat were removed in sterile PBS. Muscle tissues were then minced and enzymatically digested at 37 °C for 1 h by adding 400 IU/ml collagenase II (Worthington). The homogenate was spun, pelleted, and triturated multiple times, and then sequentially passed through a 70-μm and then 30-μm cell strainer (BD Falcon). The filtrate was spun at 1000 × g and suspended in myoblast growth medium (Ham’s F-10 medium with 20% FBS supplemented with 10 ng/ml of basic fibroblast growth factor (bFGF). Cells had media replenished after 3 days of initial plating in Matrigel. Cells were pre-plated for 15–30 min for the first few passages to select for a pure myoblast population (cells in suspension). The cells were cultured in their corresponding culturing medium until reaching at least 80% confluence. Purified cells were then seeded 50000 cell/well into Ibidi chambers (catalog #80821) containing myoblast growth medium for 24hr. The myoblasts cells then treated with 10μM tamoxifen (Sigma-Aldrich, T5648) dissolved in 100% ethanol once every 24h for 2 days. To induce differentiation, the cells were incubated in differentiation medium (DM; 2% horse serum in DMEM). Then macrophages from LysMCre;Chrm3 mice and Chrm3 mice were isolated from peritoneal cavity, and cocultured with myotubes for 3 hours. 100μl (0.4 μg/μl) of CNO was delivered directly to the wells during imaging to assess changes in calcium transients upon chemogenetic activation of macrophages. For analysis, green emission (around 488nm) using NIS software was chosen for each condition for time dependent analysis of fluorescence intensity. Changes in fluorescence intensity were then analyzed and the change in intensity was calculated.

In order to verify macrophage activation in macrophage/myotube cocultures, myotubes were differentiated from primary myoblasts isolated from C57 mice as described above. To visualize the myotubes, these purified primary myoblasts were transduced with retrovirus carrying pBabe X-GFP vector ⁶⁹. Once GFP expression was confirmed we seed 50000 cell/well into a 8 well Ibidi chamber containing myoblast growth medium for 24hr. The cells were incubated in differentiation media for 72 hours to induce differentiation to myotubes. The cells in the wells were then loaded with 5% Rhodamine-2 AM dye (Invitrogen, R1244) in Krebs-Heinslfelt buffer ( KH buffer) for 30 mins at 37°C/5% CO₂ and washed with KH buffer. Then we followed the macrophage isolation and imaging process as described above. Changes in fluorescence intensity (at ~560 nm) were then analyzed and the change in intensity was calculated.

Electromyography (EMG) on hind paw muscle

Electromyography was performed as previously described ⁷⁰. Briefly, mice were first anesthetized with 2% isoflurane. The sciatic nerve was exposed near the biceps femoris muscle. Then the hind paw muscles including the flexor digitorum brevis muscles were exposed. Mylar-coated steel recording wires (California Fine Wire) were implanted into the flexor digitorum brevis muscles, and reference wires were inserted under the skin near the base of the tail. A concentric bipolar stimulating electrode was placed on the sciatic nerve and used for electrical activation to confirm connectivity.

CMAPs were amplified using an Axoclamp 900a, recorded with a Micro 1401 data acquisition unit, and analyzed offline with Spike2 software (Cambridge Electronic Design, Cambridge, UK). A 2mA electrical stimulation of the sciatic nerve immediately proximal to the tibial, sural, and common peroneal branches was employed via a stimulus isolation unit (World Precession Instruments) connected to the Micro 1401. In the indicated instance, either blue light (10mW) or CNO was delivered directly to the hind paw muscles. Activity was recorded for 2min. After recording, the sciatic nerve was axotomized. The proximal end of the sciatic nerve was stimulated to ensure that CMAPs were generated from direct nerve stimulation, blue light or CNO delivery. CMAP amplitude, and duration were calculated from each stimulation paradigm. The average stimulation of the sciatic nerve for each paradigm was obtained and averaged across animals.

RNA isolation and real-time PCR

Muscle tissue was collected from mice at different time points. Tissue RNA was isolated using the Qiagen RNeasy kit for fibrous tissues, according to the manufacturer’s protocol (QIAGEN stock #74704). For real-time PCR, 500 ng of total RNA was DNase I treated (Invitrogen) and reverse transcribed using Superscript II (Invitrogen) reverse transcriptase. A total of 20 ng of cDNA were used in SYBR Green real-time PCR reactions that were performed in duplicate and analyzed on a Step- One real-time PCR machine (Applied Biosystems).

Primer sequences for GAPDH, IL6, TNFα, IL1β, MCP1 were obtained from previously published work ^7,71. Cycle time (Ct) values for all targets were all normalized to GAPDH as internal control. Differences in expression are determined from the normalized ΔΔCt values and standard error of the difference in means is determined. This was used to calculate fold change between conditions and values are then converted to a percentage where 2-fold = 100% change.

Quantification and Statistical analysis

Data were analyzed using GraphPad prism or SigmaPlot software. All values are presented as mean ± SEM unless stated differently. Comparisons of imaging, RT-qPCR data, and electrophysiological responses were tested with a one-way ANOVA, or a two-way repeated measures (RM) ANOVA with Tukey’s post hoc test when appropriate. For behavioral data containing the same animals treated with an intervention over time, a two-way repeated measures (RM) ANOVA was used. Two group comparisons that failed normality tests were tested with a Mann–Whitney U test. The critical significance level was set at p<0.05.

Supplementary Material

1

Data S1. Raw data from main figures. Related to Figures 1–6.

A) Description of Data S1 A- Quantification of EBD+ myofibers (Figure 1D), area of Sirius red (Figure 1F) and IgM+ fibers (Figure 1H) in muscle post incision in Chrm3 vs LysM;Chrm3 groups.

B) Description of Data S1 B- Quantification of CNO induced changes in peak fluorescence in muscle in vivo between Chrm3 vs LysM;Chrm3 groups (Figure 2 D), quantification of CNO induced firing rates between Chrm3 vs LysM;Chrm3 groups (Figure 2F), quantification of blue light induced firing rates between ChR2 vs LysM;ChR2 groups (Figure 2H), and quantification of CNO induced changes in peak fluorescence (co-cultures) in myotubes from Chrm3 vs LysM;Chrm3 groups (Figure 2L).

C) Description of Data S1 C- Quantification of PSD95, F4/80 and colocalization of PSD95 with F4/80 in muscle from naïve and incised mice (Figure 3D).

D) Description of Data S1 D- Quantification of EBD+ myofibers in Chrm3;iBot, LysMCre;Chrm3, LysMCre;Chrm3;iBot and LysMCre;iBot groups 2d after incision injury (Figure 4E), bin firing rates in Chrm3;iBot, LysMCre;Chrm3 and LysMCre;Chrm3;iBot groups (Figure 4G) and dysferlin staining in uninjured, Chrm3, LysMCre;Chrm3, Chrm3;iBot, and LysMCre;Chrm3;iBot groups post CNO (Figure 4I).

E) Description of Data S1 E- Quantification of peripheral nuclei (Figure 5C), 1 central nucleus, 2 central nuclei, and >2 central nuclei (Figure 5D) in muscle, and Pax7 positive cells per 100 myofibers at 10 days after CTX injection and CNO treatment (Figure 5G).

F) Description of Data S1 F- Data from behavioral assays including guarding scores (Figure 5H) and mechanical hypersensitivity to muscle squeezing (Figure 5I) between Chrm3 and LysM;Chrm3 group after CTX injection and CNO treatment.

G) Description of Data S1 G- Quantification of cross-sectional area of myofibers post CTX and CNO (Figure 6A), quantification of myofibers with >2 nuclei at 10 days (H&E in CTX) (Figure 6C), and quantification of Sirius red in muscle from Chrm3 and LysM;Chrm3 groups post CTX and CNO (Figure 6E).

2

Data S2. Raw data from supplementary figures. Related to Figures S1–S6.

A) Description of Data S2 A- Quantification of changes in peak fluorescence in Chrm3 and LysMCre;Chrm3 macrophages post CNO using calcium imaging (related to Figure S1 E).

B) Description of Data S2 B- Quantification of eMHC (related to Figure S1 G) and cross-sectional area of myofibers (related to Figure S1H) between Chrm3 and LysM;Chrm3 post incision and CNO.

C) Description of Data S2 C- Data from behavioral assays including guarding scores (related to Figure S1 I) and mechanical hypersensitivity to muscle squeezing (related to Figure S1 J) between Chrm3 and LysM;Chrm3 groups after incision treated with CNO.

D) Description of Data S2 D- Quantification of neutrophils (related to Figure S2 B), dendritic cells (related to Figure S2 D) and CCR2 (related to Figure S2 F) in muscle post incision and CNO treatment.

E) Description of Data S2 E- Quantification of changes in peak fluorescence in Cre-negative control and LysMCre;Chrm3 macrophages and myotubes post CNO (related to Figure S3 E).

F) Description of Data S2 F- Percentage change in mRNA of IL6, TNF, IL1B, and MCP1, two days after incision with CNO treatment (related to Figure S4 A).

G) Description of Data S2 G- Percentage change in mRNA of IL6, TNF, IL1B, and MCP1,10 days after CTX injection and CNO treatment (related to Figure S4 B).

H) Description of Data S2 G- Quantification of BTX relative to PSD95 expression in muscles from naïve and incised mice (related to Figure S4 D).

I) Description of Data S2 I- Quantification of changes in peak fluorescence of indicated outer ring of macrophages cultured in vitro isolated from Chrm3, LysMCre;Chrm3 and LysMCre;Chrm3;iBot mice during calcium imaging after CNO delivery. (related to Figure S5 G).

J) Description of Data S2 J- Quantification of changes in peak fluorescence of the indicated outer ring in Cre negative; iBot and LysM;iBot macrophages upon LPS delivery (related to Figure S5 L).

K) Description of Data S2 K- Quantification of F4/80 post incision or CTX injection (related to Figure S6 B).

L) Description of Data S2 L- Quantification of myofibers with peripheral nuclei (related to Figure S6 E), 1 central nucleus, 2 central nuclei, or >2 central nuclei (related to Figure S6 F) and Pax7 positive cells per 100 myofibers at 3 days after CTX injection with CNO treatment (related to Figure S6 G).

M) Description of Data S2 M- Quantification of myofibers with peripheral nuclei (related to Figure S6 J), 1 central nucleus, 2 central nuclei, or >2 central nuclei (related to Figure S6 K) and Pax7 positive cells per 100 myofibers at 7 days after CTX injection with CNO treatment (related to Figure S6 L).

3

Table S1. Detailed information on statistical parameters for all comparisons in main figures and supplemental figures. Related to Figures 1–6 and Figures S1–S6.

4

Document S1. Figures S1–S6.

5

Video S1. Intravital calcium imaging shows DREADD dependent activation of macrophages induces calcium transients in muscle tissue. Related to Figure 2.

6

Video S2. In vitro calcium imaging shows no effect of CNO on myotubes in co-cultures of macrophages isolated from Cre negative, Chrm3 mice. Related to Figure 2.

7

Video S3. In vitro calcium imaging of co-cultures of myotubes and macrophages shows a robust effect on myotubes after CNO activation of macrophages isolated from LysMCre;Chrm3 mice. Related to Figure 2.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
WGA	Invitrogen	W32466
Dysferlin	Abcam	Ab124684;
PSD95	Gentex	GTX133091
F4/80	Abcam	Ab6640; RRID:AB_1140040
Laminin	Abcam	Ab11575; RRID:AB_298179
Alexa Fluor 488	Thermo Scientific	A11034; RRID:AB_2576217
Alexa Fluor Fab 488	Jackson ImmunoResearch	111-547-003; RRID:AB_2338058
594 Donkey anti Rat	Thermo Scientific	A21209; RRID:AB_2535795
647 Goat Anti Rabbit	Thermo Scientific	A21244; RRID:AB_2535812
Pax 7	DSHB	AB528428;
Secondary (IgG1)	Thermo Scientific	A-21127; RRID:AB_2535769
eMHC	DSHB	F1.652; RRID:AB_528358
CD11c	Novus biologicals	NBP2-81066; RRID:AB_3411335
CCR2	Bioss Antibodies	BS23026R
IgM	Millipore Sigma	SAB3701210
Alpha-Bungarotoxin (BTX)	Invitrogen	B13423
Chemicals, peptides, and recombinant proteins
Evan’s Blue Dye	Sigma	E2129–10
Cardiotoxin (CTX)	Millipore Sigma	217503
Antigen Retrieval Citra Plus Solution	Invitrogen	00500
M.O.M. mouse IgG blocking reagent	Vector Laboratories	MKB-2213-NB
Rhodamine-2 AM dye	Invitrogen	R1244
MEM-alpha medium	Sigma	M4655-1
Collagenase II	Worthington	N/A
Ham’s F-10	Gibco	11550043
basic Fibroblast Growth Factor (bFGF)	Stemcell technology	78134.1
DMEM (Dulbecco’s Modified Eagle Medium)	Gibco	11960-044
FBS (Fetal Bovin Serum)	Gibco	26140-079
CNO (Clozapin-N-Oxide)	Tocris Bioscience	4936
Tamoxifen	Sigma-Aldrich	T5648
Critical commercial assays
RNeasy kit for fibrous tissues	Qiagen	74704
Deposited data
Other data	This paper	Mendeley Data: 10.17632/4zbzb3rfg3.1
Experimental models: Organisms/strains
LysM-Cre	Jax	004781
Rosa26-LSL-hM3Dq-mCit	Jax	026220
Rosa26-LSL-Chr2	Jax	024109
Rosa26-LSL-iBot	Jax	018056
Ai95 (RCL-GCaMP6f)-D	Jax	028865
C57Bl/6	In house	N/A
Oligonucleotides
IL6 FWD	IDT	ACTGATGCTGGTGACAAC
IL6 REV	IDT	CCGACTTGTGAAGTGGTATAG
TNFα FWD	IDT	CCTATGTCTCAGCCTCTTCT
TNFα REV	IDT	GGGAACTTCTCATCCCTTTG
IL1β FWD	IDT	TACAAGGAGAACCAAGCAAC
IL1β REV	IDT	GGTGTGCCGTCTTTCATTA
MCP1FWD	IDT	CACCTGCTGCTACTCATTC
MCP1 REV	IDT	CTACAGCTTCTTTGGGACAC
Software and algorithms
Spike2 software	Cambridge Electronic Design, Cambridge, UK	N/A
NIS -Element AR6-2002	Nikon	N/A
Other

Highlights.

• Synaptic-like contacts are found between macrophages and injured muscle fibers.

• Transient activation of macrophages hastens muscle repair after injury.

• Calcium and EMG activity in skeletal muscle is modulated by macrophage activation.

Data Availability:

Mendeley Data Repository Link to Data and Videos: DOI: 10.17632/4zbzb3rfg3.1