Platelet-derived chemokines promote skeletal muscle regeneration by guiding neutrophil recruitment to injured muscles

Abstract

Skeletal muscle regeneration involves coordinated interactions between different cell types. Injection of platelet-rich plasma is circumstantially considered an aid to muscle repair but whether platelets promote regeneration beyond their role in hemostasis remains unexplored. Here, we find that signaling via platelet-released chemokines is an early event necessary for muscle repair in mice. Platelet depletion reduces the levels of the platelet-secreted neutrophil chemoattractants CXCL5 and CXCL7/PPBP. Consequently, early-phase neutrophil infiltration to injured muscles is impaired whereas later inflammation is exacerbated. Consistent with this model, neutrophil infiltration to injured muscles is compromised in male mice with Cxcl7-knockout platelets. Moreover, neo-angiogenesis and the re-establishment of myofiber size and muscle strength occurs optimally in control mice post-injury but not in Cxcl7ko mice and in neutrophil-depleted mice. Altogether, these findings indicate that platelet-secreted CXCL7 promotes regeneration by recruiting neutrophils to injured muscles, and that this signaling axis could be utilized therapeutically to boost muscle regeneration.

Fig. 1. Platelet thrombi are found in injured muscles and are prevented via antibody-based platelet depletion.

a Experimental strategy to assess the role of platelets in skeletal muscle regeneration. The i.v. injection of a platelet-depleting antibody is done 2 h before muscle injury and is repeated at day 4 after injury. Control mice are injected with an IgG control antibody. Muscle injury is induced via the injection of cardiotoxin (CTX) into a tibialis anterior (TA) muscle whereas the contralateral TA muscle is mock-injected with PBS. TA muscles are retrieved at day 1, 7, and 14 for further analyses. b–d Immunostaining of TA muscles from control and platelet-depleted mice, either injured via the injection of cardiotoxin (CTX) or uninjured (mock-injected with PBS). WGA (red) provides an outline of myofibers whereas platelet aggregates (green) of different sizes are detected with anti-GP1bβ antibodies. Platelet thrombi are found in injured muscles at day 1 and day 7 from CTX-induced injury but their presence is minimal at day 14 from injury and they are not detected in uninjured TA muscles. Antibody-based platelet depletion results in the lack of platelets aggregates in injured muscles, indicating that this strategy is effective for testing the role of platelets in muscle regeneration. The graph displays the mean ±SD with n = 5 (from 5 control independent mice) and n = 4 (from 4 platelet-depleted independent mice) biologically independent uninjured muscles; n = 7 biologically independent CTX-injured muscles obtained from n = 7 independent control mice at day 7 after injury; and n = 10 biologically independent CTX-injured muscles obtained from n = 10 independent mice for each of the other timepoints and conditions analyzed; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Tukey post hoc test); ^&&&P < 0.001 (two-way ANOVA with Sidak post hoc test) refers to the comparison of muscles from control versus platelet-depleted mice at a given timepoint of regeneration. Source data are provided in the Source data file.

Fig. 2. Neutrophil infiltration in injured muscles is impeded by platelet depletion whereas late-phase macrophage infiltration is increased.

a, b H&E staining of TA muscles from control and platelet-depleted mice at day 1, 7, and 14 from cardiotoxin (CTX)-induced injury. In agreement with previous studies, immune infiltration is found at day 1 after CTX in control TA muscles but it is largely reduced in the TA muscles from platelet-depleted mice. At later stages, the process of muscle regeneration is impaired, as indicated by the overall lower size of myofibers and by the ultrastructural defects of TA muscles that are found at day 14. There are no noticeable changes in uninjured muscles from platelet-depleted versus control mice. c Immunostaining of TA muscles from control and platelet-depleted mice for neutrophil markers, i.e., MMP9 (red) and Ly6G (white). Myofiber boundaries are identified with immunostaining for anti-Laminin antibodies (green) whereas nuclei are identified by DAPI (blue). d–e Neutrophil infiltration in injured muscles occurs predominantly at day 1 from injury, and it is significantly reduced by platelet depletion. Similar results are found with the quantitation of both Ly6G and MMP9 immunostaining. f Quantitation of macrophages infiltrating the muscle, as defined with anti-F4/80 antibodies. Macrophage infiltration is predominant at day 7 from CTX-mediated injury and it is exacerbated by platelet depletion. In d-f, the graphs display the mean ±SD with n = 5 (from 5 control independent mice) and n = 4 (from 4 platelet-depleted independent mice) biologically independent uninjured muscles; n = 9 biologically independent CTX-injured muscles obtained from n = 9 independent control mice at day 7 after injury; and n = 10 biologically independent CTX-injured muscles obtained from n = 10 independent mice for each of the other timepoints and conditions analyzed; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Tukey post hoc test); ^&&&&P < 0.0001 (two-way ANOVA with Sidak post hoc test) refers to the comparison of muscles from control versus platelet-depleted mice at a given timepoint of regeneration. Source data are provided in the Source data file.

Fig. 3. Platelet depletion impedes the growth of myofibers during regeneration.

a Immunostaining of TA muscles from control and platelet-depleted mice at day 1, 7, and 14 from cardiotoxin (CTX)-induced injury with phalloidin (to detect F-actin; red) and with anti-Laminin antibodies (green) to detect the myofiber boundaries. b, c Quantitation of myofiber sizes (as estimated with the Feret’s minimal diameter) based on anti-Laminin and phalloidin staining. Myofibers detected at day 1 largely consist of necrotic myofibers whereas myofibers found at day 7–14 are new myofibers resulting from de novo myogenesis. There are no significant changes in the size of myofibers found at day 7 in the muscles from platelet-depleted versus control mice, suggesting that platelet depletion does not impair myogenesis (see also Supplementary Fig. 4). However, myofiber size is significantly reduced at day 14 from CTX-induced injury in the muscles from platelet-depleted versus control mice. The graphs display the mean ±SD. In b, n = 4 (from 4 control independent mice) and n = 6 (from 6 platelet-depleted independent mice) biologically independent CTX-injured muscles at day 7 after injury; and n = 10 biologically independent CTX-injured muscles obtained from n = 10 independent mice for each of the other timepoints and conditions analyzed. In c, n = 8 (from 8 control independent mice) biologically independent CTX-injured muscles at day 7 after injury; and n = 10 biologically independent CTX-injured muscles obtained from n = 10 independent mice for each of the other timepoints and conditions analyzed. **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Tukey post hoc test); ^&P < 0.05 (two-way ANOVA with Sidak post hoc test) refers to the comparison of muscles from control versus platelet-depleted mice at day 14 of regeneration. d Gaussian plots that show the size range of all myofibers sourced from all TA muscles here analyzed, stained for F-actin. There are minimal changes in myofiber size (Feret’s minimal diameter) at day 7 whereas platelet depletion leads to a significant reduction in myofiber size at day 14 from CTX injection. Source data are provided in the Source data file.

Fig. 4. Platelet depletion reduces the intramuscular levels of neutrophil chemoattractants in the early phase of muscle regeneration.

a Principal Component Analysis (PCA) of 640 cytokines profiled with Quantibody (Quantitative Multiplex ELISA) arrays from TA muscle homogenates obtained from mice with or without platelet depletion at 1, 7, and 14 days from CTX-induced injury. Uninjured muscles after 7 days from platelet depletion were also analyzed. b Heatmap of 522 cytokines that are prominently regulated (based on the average z-scores) during muscle regeneration and/or in response to platelet depletion. c Cytokine categories that are collectively regulated at different timepoints of muscle regeneration in a platelet-dependent manner include cytokines that promote neutrophil chemotaxis (day 1 from injury) and the inflammatory response (day 7–14 from injury). See also Supplementary Fig. 5. d Chemokines that promote neutrophil chemotaxis peak at day 1 after injury but their levels are reduced by platelet depletion. The graphs display the mean ±SD with n = 4 biologically independent muscles from 4 independent mice for each timepoint and condition; *P < 0.05, **P < 0.01, ***P < 0.001 (unpaired two-tailed t test) refer to the comparison of muscles from control versus platelet-depleted mice at a given timepoint. e In vitro neutrophil chemotaxis assays with recombinant versions of platelet-secreted chemokines. CXCL5 and CXCL7 have the strongest chemoattractant activity. The graphs display the mean ±SD with n = 17 (buffer), n = 18 (rCXCL7), n = 18 (rCXCL5), n = 6 (rCXCL4) biologically independent samples; ***P < 0.001 (one-way ANOVA with Tukey post hoc test), ns = not significant. f Consistent with the cytokine array data in d, ELISA assays indicate that the total levels of CXCL7 increase in skeletal muscle upon injury. The graph displays the mean ±SD with n = 11 biologically independent samples; ***P < 0.0001 (unpaired two-tailed t test). g Additional ELISA assays with antibodies specific for inactive CXCL7 (i.e., which has not been proteolytically processed) indicate a decrease in inactive CXCL7 in injured muscles. The graph displays the mean ±SD with n = 11 biologically independent samples; ***P = 0.0004 (unpaired two-tailed t test). Together, these data indicate that the surge in total CXCL7 observed in injured muscles largely consists of proteolytically-cleaved (and hence active) CXCL7. Source data are provided in the Source data file.

Fig. 5. Neutrophil infiltration in injured muscles is impeded by Cxcl7ko platelets.

a Plasma levels of the neutrophil chemoattractant CXCL7 are reduced in Cxcl7ko (Cxcl7^-/-) mice. CXCL5 and CXCL4 (encoded by adjacent genes) are also reduced whereas CXCL1 is not affected. The graphs display the mean ± SD with n = 4 (CXCL7) and n = 5 (CXCL1, CXCL4, CXCL5) biologically independent samples from n = 4 and n = 5 independent mice, respectively; *P < 0.05 and **P < 0.01 (unpaired two-tailed t test). b, c H&E staining of TA muscles from control and Cxcl7ko mice at day 1, 7, and 14 from cardiotoxin (CTX) injection, and uninjured. Immune infiltration at day 1 after CTX-mediated injury is reduced in the muscles from Cxcl7ko mice. Ultrastructural defects can be seen at day 14 post-injury in muscles from Cxcl7ko versus control mice. d Immunostaining of muscles from control and Cxcl7ko mice for neutrophil markers, i.e., Ly6G (purple). Myofiber boundaries are identified with immunostaining for anti-Laminin antibodies (green) whereas nuclei are stained with DAPI (blue). e Intramuscular neutrophil infiltration at day 1 from injury is significantly reduced in Cxcl7ko mice. Similar results are found with both Ly6G and MMP9 immunostaining. The graphs display the mean ±SD with n = 5 biologically independent samples for each group and condition from n = 5 independent mice; **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Tukey post hoc test); ^&P < 0.05 and ^&&&P < 0.001 (two-way ANOVA with Sidak post hoc test) refer to the comparison of CTX-injured WT and Cxcl7ko at a given timepoint of regeneration. f Normally, platelets (green) are found in association with neutrophils (red) in injured muscles. Although lack of CXCL7 decreases neutrophil recruitment, the area of platelet thrombi that is found in injured muscles of WT and Cxcl7ko mice is similar, indicating that defective neutrophil recruitment does not result from lower platelet number or aggregation in Cxcl7ko mice, consistent with previous studies that have found that platelet numbers and hemostatic functions are not impaired in these mice. The graph displays the mean ±SD with n = 5 biologically independent samples for each group and condition from n = 5 independent mice; ***P < 0.001 (two-way ANOVA with Tukey post hoc test). Source data are provided in the Source Data file.

Fig. 6. Myofiber size and force production are reduced in post-injury muscles from Cxcl7ko mice.

Analysis of TA muscles after 14 days from glycerol-induced injury. a TA muscle weight normalized by the tibia bone length indicates that there is post-injury hypertrophy, although this is not significant for Cxcl7ko mice. b There is no significant difference in the twitch force of uninjured and post-injury muscles from WT mice, indicating that regeneration has recovered muscle function. However, post-injury muscles from Cxcl7ko mice are significantly weaker compared to uninjured muscles. c Similar deficits in muscle force are found in pre-fatigue muscles from Cxcl7ko mice, whereas there are no differences post-fatigue. In a–c, the graphs display the mean ±SD with n = 12 (WT) and n = 11 (Cxcl7ko) biologically independent samples from n = 12 and n = 11 independent mice, respectively; *P < 0.05 (two-way ANOVA with Tukey post hoc test), ns = not significant. d Immunostaining of uninjured and post-injury muscles from control and Cxcl7ko mice with antibodies for myosin heavy chain isoforms to detect type 2a myofibers (green), type 2x myofibers (black), and type 2b myofibers (red). Defects in regeneration (such as space in-between myofibers) are found in post-injury muscles from Cxcl7ko mice compared to post-injury controls. e–g Gaussian plots indicate an overall reduced size (Feret’s minimal diameter) of type 2a (e) and type 2x (f) myofibers in post-injury muscles from Cxcl7ko mice, compared to post-injury muscles from WT mice and uninjured controls. h Quantitation of myofiber sizes based on the average values obtained from the individual muscles in a group. There is a significant decline in the size of myofibers in post-injury muscles from Cxcl7ko mice whereas myofiber size differences between uninjured and post-injury WT muscles are not significant. i There is an overall similar myofiber type composition of uninjured and post-injury TA muscles. j Myofiber number similarly increases in post-injury versus uninjured muscles from WT and Cxcl7ko. In h–j, the graphs display the mean ±SD with n = 12 (WT) and n = 11 (Cxcl7ko) biologically independent samples from n = 12 and n = 11 independent mice, respectively; *P < 0.05, **P < 0.01, ***P < 0.001, ns = not significant (two-way ANOVA with Tukey post hoc test). Source data are provided in the Source data file.

Fig. 7. Neutrophil depletion impairs skeletal muscle regeneration.

Analysis of TA muscles after 10 days from glycerol-induced injury. a Immunostaining for myosin heavy chain isoforms was utilized to detect type 2a (green), type 2x (black), and type 2b myofibers (red): these histological analyses indicate defective muscle regeneration in mice with neutrophil depletion. b–d Gaussian plots indicate that neutrophil depletion impairs the growth of newly formed myofibers in post-injury muscles whereas there is no effect of neutrophil depletion on myofiber size in contralateral uninjured muscles. e Quantitation of myofiber sizes (Feret’s minimal diameter) based on the average values obtained from the individual muscles in a group. There is an overall significant decline in the size of type 2x and 2b myofibers in post-injury muscles from neutrophil-depleted mice compared to post-injury muscles from mock-treated mice. f There is an overall similar myofiber type composition of uninjured and post-injury TA muscles. However, post-injury muscles (both from neutrophil-depleted and mock-treated mice) display higher levels of type 2x myofibers. g The number of type 2x and 2b myofibers increases in post-injury versus uninjured muscles and the number of type 2b myofibers is significantly higher in the muscles from neutrophil-depleted mice. In e–g, the graphs display the mean ±SD with n = 6 (from 6 independent control mice) and n = 4 (from 4 independent neutrophil-depleted mice) biologically independent muscles; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-way ANOVA with Tukey post hoc test), ns = not significant. h The normalized twitch force of post-injury muscles from neutrophil-depleted mice is reduced compared to that of post-injury muscles from mock-treated mice. i Similar deficits in muscle force production are found for the normalized tetanic force of muscles from neutrophil-depleted versus mock-treated mice. Neutrophil depletion does not impact twitch and tetanic force production by uninjured muscles (h, i). In h, i, the graphs display the mean ± SD with n = 6 (from 6 independent control mice) and n = 4 (from 4 independent neutrophil-depleted mice) biologically independent muscles; *P < 0.05, ****P < 0.0001 (two-way ANOVA with Sidak post hoc test). Source data are provided in the Source data file.

Fig. 8. Platelets promote skeletal muscle regeneration by guiding the recruitment of neutrophils to injured muscles via the platelet-released chemokine CXCL7.

In response to injury, platelets localize to and form thrombi in skeletal muscles and promote the recruitment of neutrophils via the release of platelet-specific chemokines (e.g., CXCL7) that are neutrophil chemoattractants. Neutrophil infiltration is known to promote muscle repair via the removal of cellular debris and by setting the stage for the subsequent steps of regeneration, which include the infiltration of monocytes and macrophages and myogenesis, i.e., the de novo formation of myofibers. In the absence of platelets or when platelets lack CXCL7 (CXCL7KO), the recruitment of neutrophils to injured muscles is defective. This in turn leads to unresolved tissue damage, excessive recruitment of macrophages at later phases of regeneration, and to high levels of atrophic ligands that stunt the growth of newly-formed myofibers. Neo-angiogenesis is also reduced. Consequently, post-injury muscles arising from regeneration in the absence of early-stage platelet-initiated chemokine signaling display reduced myofiber size and lower muscle force production. Similar results are found with the experimental depletion of neutrophils. Altogether, these findings indicate a key role for platelet-induced chemokine signaling in ensuring optimal muscle regeneration by guiding the recruitment of neutrophils to injured muscles in the early phase after injury.