Macrophage activation and skeletal muscle healing following traumatic injury

Abstract

Following injury to different tissues, macrophages can contribute to both regenerative and fibrotic healing. These seemingly contradictory roles of macrophages may be related to the markedly different phenotypes that macrophages can assume upon exposure to different stimuli. We hypothesized that fibrotic healing after traumatic muscle injury would be dominated by a pro-fibrotic M2a macrophage phenotype, with M1 activation limited to the very early stages of repair. We found that macrophages accumulated in lacerated mouse muscle for at least 21 days, accompanied by limited myofibre regeneration and persistent collagen deposition. However, muscle macrophages did not exhibit either of the canonical M1 or M2a phenotypes, but instead up-regulated both M1- and M2a-associated genes early after injury, followed by down-regulation of most markers examined. Particularly, IL-10 mRNA and protein were markedly elevated in macrophages from 3-day injured muscle. Additionally, though flow cytometry identified distinct subpopulations of macrophages based on high or low expression of TNFα, these subpopulations did not clearly correspond to M1 or M2a phenotypes. Importantly, cell therapy with exogenous M1 macrophages but not non-activated macrophages reduced fibrosis and enhanced muscle fibre regeneration in lacerated muscles. These data indicate that manipulation of macrophage function has potential to improve healing following traumatic injury.

Keywords: inflammation; injury; macrophage; skeletal muscle; tissue repair.

Figure 1. Lacerated muscle heals by both regeneration and fibrosis

Gastrocnemius muscles were lacerated and collected for histological analysis at the indicated time-points. For each muscle, sections with the largest percent damaged area were selected for analysis. Uninjured muscles served as controls. (A) Percent of total cross-sectional area occupied by centrally nucleated (“Regenerating”) or peripherally nucleated (“Normal”) fibers was quantified in five 40× fields per muscle in hematoxylin and eosin stained sections. “Damaged” area was defined as area not occupied by either type of myofiber. (B) Number of myofibers per mm² in uninjured and injured muscles. (C) Average cross-sectional area of individual myofibers. (D) Collagen accumulation was quantified as percent blue pixels in three to six 20× fields per muscle in Masson's trichrome stained sections. (E) Representative images of trichrome stained sections in uninjured muscle and at indicated time points post-injury. Data are presented as mean +/− SD. * p<0.05 versus uninjured. n=2–5 per time point. In (A), * indicates significance for percent damage area.

Figure 2. Muscle laceration results in prolonged accumulation of macrophages but not neutrophils

Gastrocnemius muscles were lacerated and collected for histological analysis at the indicated time-points. For each muscle, sections with the largest percent damaged area were selected for analysis. Uninjured muscles (0d) served as controls. (A) Macrophage accumulation was quantified as percent F4/80-stained area in three to six 20x fields per muscle. (B) Neutrophil accumulation was quantified as percent Ly6G-stained area in three to six 20x fields per muscle. Data are presented as mean +/− SD. * p<0.05 versus uninjured. n=2–6 per time point. Representative 20x images are shown for F4/80 (A) and Ly6G (B) labeling in uninjured muscle, at peak macrophage or neutrophil accumulation, and after peak accumulation. Scale bar =100μm.

Figure 3. Muscle macrophage phenotype

Macrophages were isolated by magnetic separation from uninjured (0d) or lacerated gastrocnemius muscles at the indicated time points. Muscle macrophages were obtained as the CD11b-positive, Ly6G/CD3/CD19-negative cell fraction. As positive and negative controls, bone marrow derived macrophages (right side of vertical bar in panels A–J) were activated with IFNγ and TNFα (M1), IL-4 (M2a), or IL-10 (M2c). Total RNA was isolated and reverse transcribed, and expression of IL-1β (A), TNFα (B), iNOS (C), IDO1 (D), CXCL10 (E), CD206 (F), CD36 (G), TGFβ (H), Ym1 (I), and IL-10 (J) was analyzed by real-time PCR. Expression of each gene was determined by the 2^−ΔΔCT method using GAPDH as endogenous control. M1-associated genes (A–E) were normalized to in vitro activated M1 macrophages, and M2-associated genes (F–J) were normalized to in vitro activated M2a macrophages. (K) Macrophages were isolated by magnetic separation from uninjured or injured muscles, equal numbers of macrophages were incubated for 20 hours, and IL-10 secretion was measured by ELISA on the conditioned medium. Data did not pass tests of normality and equal variance, and are presented with center line as median, boxes representing the 25^th and 75^th percentiles, whiskers representing the 10^th and 90^th percentiles, and outliers as dots. * p<0.05 versus macrophages from uninjured (0d) muscle. n=5–10 per time-point. In vitro activated macrophages were not included in statistical comparisons.

Figure 4. Macrophages at 3 days after muscle laceration are not separable into M1 and M2a subsets

Cells were isolated from gastrocnemius muscles at 3 days post-injury and labeled for flow cytometry. Macrophages were defined as FITC-F4/80+ cells (A) or PEF4/80+ cells (C), with lower threshold set based on background FL1 (B) or FL2 (D) fluorescence in unlabeled cells. Density plots display macrophage expression of TGFβ versus TNFα (E), IL-10 versus TNFα (F), CD36 versus IL-10 (G), CD36 versus IL-1b (H) and CD36 versus TNFα (I). In panel G, cells in the upper left quadrant (arrow) likely represent non-specific labeling, as this population was also seen in IgG control plots (not shown). Data are representative of 5 independent experiments of n=1–2 per experiment.

Figure 5. Dose-dependent reduction in collagen accumulation with M1 macrophage cell therapy

PBS vehicle or 0.5×10⁶, 1.0×10⁶, or 2.0×10⁶ M1-activated bone marrow-derived macrophages were injected into lacerated gastrocnemius muscles at 7 days post-injury. (A) Donor macrophages were labeled with CFSE (green) prior to injection, muscles were collected at the indicated time points, and cryosections were mounted with DAPI to label nuclei (blue). (B) Collagen accumulation at 14 days post-injury (i.e. 7 days post-injection) was quantified as percent blue pixels in three to six 20x fields per muscle in Masson's trichrome stained sections. (C) Damaged area (i.e. area not occupied by myofibers) at 14 days post-injury was quantified in five 40x fields per muscle in hematoxylin and eosin stained sections. Data are presented as mean +/− SD. * p<0.05 versus PBS control. n=4–8 muscles per time point and treatment condition.

Figure 6. Cell therapy with M1 macrophages reduces collagen accumulation and enhances muscle regeneration

(A–E) M1-activated bone marrow-derived macrophages or PBS vehicle were injected into lacerated gastrocnemius muscles at 7 days post-injury and analyzed at 14 or 21 days post-injury. (A) Representative trichrome stained images of lacerated muscles treated with PBS (left) or 2 ×10⁶ M1 macrophages (right) at 14 (top) or 21 (bottom) days post-injury. Scale bar = 100μm. (B) Collagen accumulation was quantified as percent blue pixels in three to six 20x fields per muscle in Masson's trichrome stained sections from muscles injected with PBS vehicle or 2 ×10⁶ M1 macrophages. (C) Damaged area (i.e. area not occupied by myofibers) was quantified in five 40x fields per muscle in hematoxylin and eosin stained sections. (D) Average cross-sectional area of individual myofibers in hematoxylin and eosin stained sections. (E) Number of myofibers per mm² in hematoxylin and eosin stained sections. (F&G) 2 ×10⁶ non-activated bone marrow-derived macrophages or PBS vehicle were injected into lacerated gastrocnemius muscles at 7 days post-injury and analyzed at 14 days post-injury. (F) Collagen accumulation was quantified as percent blue pixels in trichrome stained sections. (G) Damaged area (i.e. area not occupied by myofibers) was quantified in hematoxylin and eosin stained sections. Data are presented as mean +/− SD. * p<0.05 versus PBS control at same time point; # p<0.05 versus 14d time point within same treatment group. n=4–7 per time point and treatment condition.