Targeting TGF‐β to Attenuate Ischemia–Reperfusion Associated Muscle Fibrosis and Enhance Regeneration: A Preclinical Study

ABSTRACT

Ischemia–reperfusion (IR) injury induces a pro‐inflammatory cascade that disrupts inflammation resolution and exacerbates skeletal muscle fibrosis, leading to impaired regeneration through fibrosis and compromised myofiber regeneration. This study aimed to: (1) Demonstrate IR‐induced dysregulation of muscle regeneration following acute injury; (2) Elucidating how TGF‐β1 signaling within the FAP‐associated inflammatory niche mediates fibrosis and impairs myofiber repair; and (3) Evaluating anti‐TGF‐β neutralization as a mechanism‐based strategy to preserve regenerative capacity. Male C57BL6 mice (8 weeks) received CTX injections into the tibialis anterior (TA) to induce skeletal muscle injury. Transient clamping of femoral artery and vein was induced at 3 days post‐injury to induce IR. Mice were stratified into CTX group and CTX‐IR group depending on the induction of IR injury. TGF‐β neutralizing antibody (TGF‐β NAb) was administered in vivo to evaluate therapeutic potential. HE and Sirius red staining was used to assess cross‐sectional area (CSA) of myofibers and percentage of fibrotic tissue. Western blotting was used to assess the expression of Collagen I, Collagen III and TGF‐β. Flow cytometry was used to assess the number of fibro‐adipogenic progenitors (FAPs). Collagen I and CD90 were assessed using immunostaining of TA cryosections and FAPs isolated from muscle tissues. The result shows that CTX‐IR mice showed smaller myofiber cross‐sectional area and more fibrotic tissue than CTX mice. Western blots revealed higher levels of Collagen I, Collagen III and TGF‐β1 in CTX‐IR samples. TGF‐β1 stimulated the expression of collagen I in FAPs. TGF‐β NAb application ameliorated the fibrosis of skeletal muscle and improved myofiber regeneration after IR. In conclusion, IR impairs muscle regeneration through TGF‐β‐associated FAP activation and collagen deposition. Inhibition of TGF‐β attenuates fibrosis and increases the CSA of myofibers, demonstrating it as a potential target for promoting structural recovery and regenerative architecture in co‐existing muscle injuries during orthopedic trauma.

The murine model of ischemia–reperfusion (IR) superimposed on acute cardiotoxin (CTX) injury disrupts skeletal muscle repair. The IR‐induced inflammatory microenvironment drives macrophages to release high levels of TGF‐β1, which binds to fibro‐adipogenic progenitors (FAPs). Untreated, this leads to FAP persistence, severe fibrosis, and myofiber atrophy. Targeted neutralization of TGF‐β promotes FAP apoptosis, attenuates fibrotic matrix deposition, and successfully restores the trajectory of myofiber regeneration.

Abbreviations

CSA

cross‐sectional area

CTX

cardiotoxin

DMEM

Dulbecco's Modified Eagle Medium

dpi

days post‐injection

ECM

extracellular matrix

FA

femoral artery

FACS

fluorescence‐activated cell sorting

FAP

fibro‐adipogenic progenitor

FBS

fetal bovine serum

FV

femoral vein

H&E

hematoxylin and eosin

IF

immunofluorescence

IgG

immunoglobulin G (isotype control)

IR

ischemia–reperfusion

NAb

neutralizing antibody

NC

negative control

PDGFRα

platelet‐derived growth factor receptor alpha

p‐Smad2/3

phosphorylated Smad2/3

SC

satellite cell

TA

tibialis anterior

TGF‐β

transforming growth factor‐beta

1. Introduction

Skeletal muscle possesses a highly orchestrated, innate capacity for regeneration following acute injury. This repair process relies on a tightly regulated spatiotemporal interplay among resident immune cells, muscle stem cells (satellite cells), and fibro‐adipogenic progenitors (FAPs) to clear necrotic debris and reconstitute functional myofibers [1, 2]. However, when the delicate balance of inflammatory resolution is perturbed, this regenerative program derails. Prolonged or aberrant immune activation creates a microenvironment that favors excessive extracellular matrix (ECM) deposition over myogenesis, ultimately leading to pathological fibrosis and long‐term functional deficits [1, 3].

A critical factor that disrupts the highly coordinated regenerative program is ischemia–reperfusion (IR) injury [4, 5]. In trauma and surgical care, IR is a common consequence of diverse etiologies, including vascular injuries, arterial occlusions, severe contusions leading to compartment syndrome, ischemic shock, and the use of tourniquets in emergency settings and orthopedic procedures [5, 6]. The pathophysiological hallmark of IR injury is a paradoxical exacerbation of tissue damage upon the restoration of blood flow, characterized by an intense burst of oxidative stress and a sterile inflammatory response. Unlike standard mechanical injury, the IR‐induced inflammatory storm disrupts the normal resolution of inflammation and progression of regeneration, thereby sustaining FAP activation and shifting the tissue trajectory toward maladaptive fibrotic remodeling [1, 3, 6] (Figure 1A).

FIGURE 1.

Establishment and validation of the mouse cardiotoxin ischemia–reperfusion (CTX‐IR) model. (A) Schematic of the experimental timeline for the CTX and CTX‐IR groups. Acute injury was induced in the tibialis anterior (TA) muscles of 8–10‐week‐old C57BL/6 male mice via intramuscular injection of 10 μg/mL cardiotoxin (CTX). At 3 days post‐injection (dpi), the CTX‐IR group underwent 2 h of ischemia–reperfusion (IR) via femoral artery and vein occlusion. TA muscles were harvested at 17 dpi for analysis. The clinical relevance of our model is illustrated in this figure. (B) Illustration of the surgical procedure for IR induction. An 8 mm transverse incision was made 3 mm distal to the inguinal fold to expose the femoral artery (red dot), vein (blue dot), and nerve (yellow dot). A microvascular clamp was used to occlude the femoral artery and vein. (C) Laser Doppler imaging confirmed the effectiveness of the IR procedure. Blood flow in the TA muscle (outlined in white) was significantly reduced during occlusion (middle) and dramatically increased upon reperfusion (right) compared to the pre‐surgery control (left).

Transforming growth factor‐β (TGF‐β) is a central profibrotic cytokine that mediates this immune‐stromal crosstalk. Under pathological conditions, excessive TGF‐β signaling inhibits myoblast differentiation, prevents FAP apoptosis, and aggressively drives collagen production [7]. The therapeutic potential of neutralizing TGF‐β has been demonstrated in models of chronic muscular dystrophy, renal fibrosis, and pulmonary disease [7, 8, 9, 10, 11, 12]. However, the specific contribution of the TGF‐β/FAP axis to the unique regenerative failure caused by acute IR injury in skeletal muscle remains poorly defined.

TGF‐β1 (hereafter referred to as TGF‐β when the context is unambiguous) acts as a primary driver of this pathological shift. Unlike synchronous polytrauma models where ischemia and myonecrosis occur simultaneously, clinical orthopedic scenarios—such as closed limb fractures requiring delayed definitive surgery—often feature a delayed ischemic hit that interrupts an already active regenerative program. Targeted neutralization of TGF‐β during this critical transitional window represents a mechanism‐based strategy to rescue the microenvironment from ischemic stress. Therefore, the objectives of this study were to: (1) establish a clinically relevant murine model of IR injury superimposed on the reparative phase of acute muscle damage; (2) elucidate the role of TGF‐β in driving the associated fibrosis and impairment of myofiber repair; and (3) evaluate the therapeutic efficacy of targeted TGF‐β neutralization as a strategy to attenuate fibrosis and promote muscle regeneration after IR injury.

2. Materials and Methods

Eight weeks old wild‐type C57BL/6 male mice (GemPharmatech Co. Ltd., Nanjing, China) were grouped into CTX group and CTX‐IR group, Tibialis anterior (TA) samples were collected at 5dpi and 17dpi in and CTX‐IR group for histological and flow‐cytometry analysis (Figure  1A ). The research was approved by our institutional ethics committee. TA samples harvested from uninjured muscle were also collected for comparison.

2.1. Mice Skeletal Muscle Injury Model

Wild‐type C57BL6J male mice (8 weeks old) were anesthetized with isoflurane, and the hair of TA and inguinal region was removed using shaver and hair removal cream. A 0.05 mL injection of CTX (Shanghai Boyao Biological Technology Co. Ltd., Shangha, China) solution (10 μg/mL) was administered into the TA. Mice were monitored for anesthesia recovery on a warming plate. Three days after CTX injection, mice were anesthetized with isoflurane again, and an 8 mm transverse incision was made 3 mm caudal to the inguinal fold to expose the femoral artery (FA) and vein (FV) (Figure 1A,B). The femoral nerve was carefully separated from FA and FV with forceps, then microvascular clamps were used to occlude the FA and FV for 2 h (Figure 1B). The effectivity of occlusion was confirmed using Lazer Doppler imaging system (Figure 1C). After the designated time, the microvascular clamps were removed, FA and FV were gently massaged with forceps when necessary until confirmation of blood flow restoration. The incision was then closed using 5–0 sutures. Mice were closely monitored until full recovery from anesthesia. Despite CTX and CTX‐IR group, a group of IR model was established at 3 dpi (without prior CTX injury), and euthanized at 17 dpi (14 days after IR induction) for comparison. CTX‐IR group demonstrated markedly reduced TA/body mass ratio, CSA of regenerated myofiber, and increased collagen deposition marked by Sirius red staining. To demonstrate the effect of anti‐TGF‐β therapy, mice received approximately 0.5 mL (range, 0.4–0.6 mL of a 0.1 mg/mL PBS solution of TGF‐β NAb (Invitrogen, MA5‐23795) (yielding ~2.5 mg/kg per injection)) or IgG via intraperitoneal injection strictly at 6 and 12 h after IR induction in the CTX‐IR model, resulting in a cumulative dose of approximately 5 mg/kg. For variable control, an additional group of mice underwent IR alone (2‐h femoral artery/vein occlusion without prior CTX injection) and were harvested at 17 dpi. Histological data from this group are presented in Figure S1.

2.2. Histological Assessment

Cryosections were made according to published protocol [13]. In brief, mice were euthanized in a CO₂ filled euthanasia chamber, and cervical dislocation was performed when necessary. Body mass and the dry mass of TAs (n = 6 at each time point) were documented. Harvested TA samples were embedded in cutting temperature compound, snap‐frozen in liquid nitrogen, and stored at −80°C. Cryosections with thickness of 8 μm were conducted and mounted on glass slides, stored at −80°C, and subjected to histological analysis including hematoxylin–eosin (HE) staining, Sirius red staining, and Immunofluorescence (IF) staining. CSA of myofibers and collagen area was determined in Sirius red staining images.

2.3. Fluorescence‐Activated Cell Sorting of FAP

For fluorescence‐activated cell sorting (FACS) of FAPs, single cell suspension was prepared according to published protocol [14, 15]. TA samples were harvested from both lower limbs of three unperturbed 8‐week‐old wild‐type C57BL/6 male mice. Briefly, the TA samples were minced thoroughly until a slurry paste of well‐minced tissue was obtained. This slurry was incubated in a mixture solution of Collagenase II (Gibco, Cat#: 17101015) and dispase (Gibco, Cat#: 17105‐041) for 1.5 h at 37°C with agitation The muscle suspension was then aspirated and ejected through a 20‐gauge needle for 10 cycles to further dissociate cells, followed by filtration through a 40‐μm nylon cell strainer. Centrifuge the single‐cell suspension and resuspend the cell pellet using PBS supplemented with 2% fetal bovine serum (FBS) (Gibco, Cat#: 10099158) and incubate with a cocktail of fluorophore‐conjugated antibodies, including CD31 (BioLegend, Cat#: 102418), CD45 (BioLegend, Cat#: 304016), VCAM‐1 (BioLegend, Cat#: 305810), and Sca‐1 (BioLegend, Cat#: 108128), against cell surface markers for fibro‐adipogenic progenitors (FAPs). FAPs were identified as CD31⁻CD45⁻VCAM‐1⁻Sca‐1⁺ cells and sorted using flow‐cytometer (BD Melody).

2.4. Quantification of FAP in TA Samples

The weight of each TA sample was recorded prior to the preparation of the single‐cell suspension. The raw number of FAPs was determined using flow cytometry following the above protocol. The number of FAPs per gram of muscle tissue was calculated as follows: Number of FAPs per gram = (Number of FAPs counted)/(Weight of muscle tissue in grams).

2.5. Cellular Experiments and IF Staining

FACS sorted FAP cells were seeded into Dulbecco's Modified Eagle Medium (DMEM) (Gibco, Cat#: 11965092) supplemented with 10% (v/v) FBS and 2 ng/mL fibroblast growth factor (MCE, Cat#: HY‐P7065) [14] for an initial 72‐h period. Following this culture, cells were subjected to two experimental conditions: one with the addition of 10 ng/mL mouse TGF‐β1 (MCE, Cat#: HY‐P70648) and one without. Medium with and without TGF‐β1 was refreshed every 24 h and continued for 72 h. After the simulation period, cells were fixed and subjected to IF staining. Specifically, collagen I (Proteintech Cat#: 66761‐1‐Ig), p‐Smad2/3 (Proteintech Cat#: 80427‐2‐RR) and DAPI (Beyotime, Cat#: 1006) nuclear stain were employed. The stained cells were visualized under a confocal microscope. To quantitatively assess protein expression, image analysis was performed on six random vision fields for each experimental group. The percentage of collagen I positive area normalized to cell numbers in fluorescence was analyzed using Image J to compare collagen I expression levels between the group with TGF‐β1 simulation and the group without. Nuclear and cytoplasmic p‐Smad2/3 fluorescence intensities were measured using Image J (NIH). For each cell, the nuclear region was defined by DAPI staining, and the total cell region was delineated by the p‐Smad2/3 signal outline. Cytoplasmic intensity was calculated as (total integrated density − nuclear integrated density)/(total area − nuclear area). The nuclear/cytoplasmic (N/C) ratio was computed by dividing the mean nuclear intensity by the mean cytoplasmic intensity. Ki67 (Proteintech, 27309‐1‐AP) was stained to calculate the percentage of proliferating (Ki67⁺) FAPs. the TUNEL test kit (Beyotime, C1091) was used to mark and calculate the percentage of apoptosis FAP.

2.6. RNA Extraction and qPCR

Total RNA was extracted from TA muscle tissues using the Servicebio RNA Extraction Kit (G3013) according to the manufacturer's protocol. cDNA was synthesized from 1 μg of total RNA using the SweScript All‐in‐One SuperMix for qPCR (G3337, Servicebio) according to the manufacturer's instructions. qPCR was performed on a Bio‐Rad CFX Connect Real‐Time PCR Detection System. Each reaction was set up in a 10 μL volume, containing 7.5 μL of 2× Universal Blue SYBR Green qPCR Master Mix (G3326, Servicebio), 1.5 μL of 2.5 μM gene‐specific primers, 2.0 μL of cDNA template, and 4.0 μL of nuclease‐free water. The thermal cycling program consisted of an initial denaturation at 95°C for 30 s, followed by 40 cycles of 95°C for 15 s and 60°C for 30 s. A melting curve analysis was generated from 65°C to 95°C (0.5°C increments) to confirm primer specificity. Data were acquired using Bio‐Rad CFX Maestro software. The primer sequences were as follows: TGFB1 forward: ACTGGTGTTGTGACGCCATTG, reverse: GGCTGATCCCGTTGATTTCCT; TGFB2 forward: CCCCACATCCTGCTCTAATGTTT, reverse: CTAAAGCAGTAGGCAGCATCCA; TGFB3 forward: CCGGATGAGCACATAGCCAA, reverse: TCTCTCTCCAACAGCCACTG; GAPDH forward: CCTCGTCCCGTAGACAAAATG, reverse: TGAGTCAATGAAGGGGTCGT.

Relative mRNA expression levels were calculated using the 2⁻ΔΔCt method with GAPDH as the endogenous reference. For each gene, the fold change was normalized to the Uninjured group (set to 1). Data are presented as mean ± SD from 3 mice per group (randomly selected from 6 per group).

2.7. Western Blot Analysis

At day 5 post‐injury, TA muscle samples of both lower limbs of mice from both CTX (n = 6) and CTX‐IR (n = 6) groups were collected. Proteins were extracted using a buffer containing protease and phosphatase inhibitors. After determining protein concentrations, equal amounts were loaded onto SDS‐PAGE gels for electrophoresis under denaturing conditions. Subsequently, proteins were transferred to membranes, blocked to prevent non‐specific binding, and incubated with primary antibodies against COL1A1 (Invitrogen, Cat#: PA5‐89281), collagen III (Abcam, Cat#: ab184993), TGF‐β (Abcam, Cat#: ab215715), and GAPDH (Abcam, Cat#: ab8245). Following incubation with secondary antibodies and thorough washing, proteins were visualized using an enhanced chemiluminescence substrate, and images were captured. Quantitative analysis of protein expression levels was performed normalized to GAPDH using Image J (NIH, United States).

2.8. Statistical Analysis

Prior to statistical analysis, the normality and homogeneity of variances assumptions were evaluated for all data sets using Shapiro–Wilk and Levene's tests, respectively. For time‐course experiments involving two groups across multiple time points (e.g., TA/body mass, FAP number per gram tissue), two‐way analysis of variance (ANOVA) was performed with group and time as fixed factors, followed by Šídák's multiple comparisons test for predefined pairwise comparisons at each time point. For comparisons among three or more groups at a single time point (e.g., uninjured, CTX, CTX‐IR, and treatment arms), one‐way ANOVA was used, followed by Tukey's honestly significant difference (HSD) test for all pairwise comparisons, or Dunnett's test when comparing against a single control group. Unpaired two‐tailed Student's t‐tests were employed for direct comparisons between two groups at a single endpoint. A p‐value threshold of 0.05 was chosen for determining statistical significance. Data are presented as the mean ± SD, unless otherwise noted. Statistical analyses and generation of charts were performed using GraphPad Prism (Version 9.4.1, San Diego, United States).

3. Results

3.1. IR Impaired Myofiber Regeneration and Exacerbated Fibrosis in CTX‐Induced Skeletal Muscle

TA/body mass was compared between CTX and CTX‐IR across five time points (Uninjured, 3, 5, 7, 17 dpi). At 5 dpi, TA/body mass was significantly higher in CTX‐IR than CTX (p = 0.0007), indicating an acute edematous or inflammatory response to the second hit (Figure 2A). Conversely, at 17 dpi, TA/body mass was significantly lower in CTX‐IR compared to CTX (p = 0.0001), reflecting impaired regenerative recovery (Figure 2A). No significant differences were observed at uninjured, 3 dpi, or 7 dpi (all adjusted p > 0.05). All time‐course data met assumptions of normality and equal variance (Shapiro–Wilk and Levene's tests, p > 0.05), and individual data points are shown alongside group means to illustrate inter‐animal variability. The CSA of myofiber and percentage of collagen area was illustrated by Sirius red staining (Figure 2C), with CTX group demonstrating higher CSA of regenerated myofiber (Figure 2D) and smaller collagen area than CTX‐IR group (CTX group: 5.23% ± 0.80%, CTX‐IR group: 9.31% ± 2.07%, t = 4.500, p < 0.001) (Figure 2E). The illustrated diffuse Western Blot bands in the ~130–180 kDa range further demonstrated the higher protein expression level of COL1A1 (p < 0.001) and collagen III (p = 0.0207), the principal component of fibrosis tissue of skeletal muscle, were significantly higher in CTX‐IR group at 5 dpi, indicating a pro‐fibrotic response of regenerating muscle tissue to IR (Figure 2F–H). Although some variability in COL1A1 and collagen III expression was observed across samples, the overall difference between groups remained statistically significant, indicating a consistent pro‐fibrotic effect of IR.

FIGURE 2.

IR impairs skeletal muscle regeneration and induces fibrosis. (A) Quantification of TA muscle dry weight, normalized to body weight, in uninjured, CTX, and CTX‐IR mice at 3, 5, 7, and 17 dpi. (B) Representative gross morphology of TA muscles from CTX (left) and CTX‐IR (right) groups at 17 dpi. (C) Representative images of Hematoxylin–Eosin (H&E) and Sirius Red staining of TA muscle cross‐sections from CTX and CTX‐IR groups at 17 dpi. (D) Quantification of myofiber cross‐sectional area (CSA) in both groups at 17 dpi. (E) Quantification of the fibrotic area (collagen percentage) from Sirius Red‐stained sections. (F) Representative Western blot of collagen I and collagen III protein levels in TA muscles from uninjured, CTX, and CTX‐IR groups at 5 dpi (n = 3 in each group). (G, H) Quantification of collagen I and III protein levels, normalized to GAPDH. Individual data points are shown in figures; sample sizes were n = 6 animals per group for all panels unless otherwise stated. Time‐course data in (A) were analyzed by two‐way ANOVA with Šídák's multiple comparisons test; Comparisons among three groups were analyzed by one‐way ANOVA followed by Tukey's honestly significant difference (HSD) test. Two‐group comparisons at single endpoints were performed using unpaired two‐tailed Student's t‐test Adjusted or nominal p values: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant. Scale bar: 100 μm.

3.2. IR Caused an Increased Number of FAP in Regenerating Skeletal Muscle Following CTX Injury

Increased expression of Collagen I and collagen III has led us to investigate the population of FAP that contributes to the secretion and deposition of fibrotic tissue. The number of FAPs normalized to muscle mass in CTX and CTX‐IR group at 5 dpi was evaluated. The CTX‐IR group demonstrated increased number of FAPs than CTX group normalized to muscle mass (p = 0.0162) (Figure 3A–C). To validate the association of FAP number and fibrosis, we further stained CD90, the marker of fibrotic FAP, together with Collagen I and collagen III in CTX and CTX‐IR groups at 17dpi, demonstrating increased pro‐fibrotic FAP and increased expression of Collagen I (p = 0.0057), Collagen III (p = 0.0105) and CD90 (p = 0.0003).

FIGURE 3.

IR expands the fibro‐adipogenic progenitor (FAP) population and exacerbates fibrosis. (A) Flow cytometry gating strategy and representative plots showing the percentage of FAPs in uninjured, CTX, and CTX‐IR groups at 5 dpi. (B) Representative flow cytometry plots comparing FAP and satellite cell (SC) populations in TA muscles from all groups at 5 dpi. (C) Quantification of FAP number per gram of muscle tissue (n = 3 in each group). (D, E) Representative immunofluorescence images of regenerating TA muscle from CTX and CTX‐IR groups at 17 dpi. Pro‐fibrotic FAPs are labeled with CD90 (red), and fibrotic regions are labeled with collagen I (D) and III (E) (green). (F) Quantification of Collagen I, Collagen III and CD90. Individual data points are shown in (C, F). Sample sizes: n = 6 animals per group for (C); n = 3 animals per group for (F). Flow‐cytometry data in (C) were analyzed by two‐way ANOVA with Šídák's multiple comparisons test. Comparisons among three groups (uninjured, CTX, CTX‐IR) in (F) were analyzed by one‐way ANOVA followed by Tukey's honestly significant difference (HSD) test. Adjusted p values: *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant. Scale bar: 100 μm.

3.3. TGF‐β1 Is the Predominant Isoform Upregulated in CTX‐IR Muscle

To identify which TGF‐β isoform is correlated with the fibrotic response in the CTX‐IR setting, we measured the mRNA expression of TGFB1, TGFB2, and TGFB3 in TA muscles at 5 dpi by qPCR. Compared with CTX alone, CTX‐IR muscle showed a significant 2.73‐fold upregulation of Tgfb1 (adjusted p < 0.0001; Figure 4C). In contrast, TGFB2 and TGFB3 levels were not significantly different between the two groups (adjusted p = 0.66 and p = 0.31, respectively). These data indicate that TGF‐β1 is the dominant isoform induced by IR during active muscle regeneration, supporting its selection as the primary therapeutic target.

FIGURE 4.

TGF‐β1 is upregulated in CTX‐IR muscle and activates Smad signaling in FAPs. (A) Western blot analysis of TGF‐β1 (44 kDa) and GAPDH (36 kDa, loading control) in tibialis anterior (TA) muscles from uninjured, CTX 5 dpi, and CTX‐IR 5 dpi groups. The bar graph represents the relative densitometric quantification of TGF‐β1 normalized to GAPDH (n = 3 mice per group); (B) Immunofluorescence staining for TGF‐β1 (green) in TA muscle cryosections at 5 dpi; nuclei counterstained with DAPI (blue). The bar graph quantifies the TGF‐β1‐positive area fraction. (C) qPCR analysis of Tgfb1, Tgfb2, and Tgfb3 mRNA expression in TA muscles at 5 dpi. Values are fold change relative to the uninjured group (set to 1, n = 3 mice per group). (D) Representative immunofluorescence images of primary FAPs treated with recombinant TGF‐β1 (10 ng/mL) or vehicle for 72 h, stained for Ki67 (proliferation) and TUNEL (apoptosis). (E, F) Quantification of Ki67‐positive cells (E) and TUNEL‐positive cells (F) as a percentage of the total DAPI‐stained nuclei; (G) Immunofluorescence staining for p‐Smad2/3 and the FAP marker PDGFRα in TA muscle sections from CTX and CTX‐IR groups (5 dpi). (H) Representative images of FAPs treated with TGF‐β1 (10 ng/mL) or control stained for Collagen I and p‐Smad2/3; (I) Quantification of Collagen I positive area normalized to the number of DAPI‐stained nuclei; (J) Quantification of the nuclear‐to‐cytoplasmic fluorescence intensity ratio of p‐Smad2/3. Individual data points are shown in all bar graphs. Sample sizes: n = 6 mice per group unless otherwise stated; n = 6 microscopic fields from 3 independent experiments (2 fields per experiment) for (E, F, I); n = 30 cells from 3 independent experiments (10 cells per experiment) for (J). Comparisons among three groups in (A) were analyzed by one‐way ANOVA followed by Tukey's honestly significant difference (HSD) test. Two‐group comparisons in (B, E, F, I, J) were performed using unpaired two‐tailed Student's t‐test. Isoform‐specific comparisons in (C) (CTX vs. CTX‐IR for each Tgfb isoform) were analyzed by two‐way ANOVA with Šídák's multiple comparisons test. Adjusted or nominal p values: **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant. Scale bars: 100 μm.

3.4. Increased TGF‐β1 After IR Is Associated With the Shift Toward a Pro‐Fibrotic Phenotype in FAP

Aside from the increased number of FAP following IR, the fibrotic activity of FAPs was investigated in our study. The protein level of TGF‐β1, a major pro‐fibrotic factor, was compared between CTX and CTX‐IR group at 5dpi. Western blot analysis demonstrated statistically higher expression level of TGF‐β1 (migrated at approximately 40–55 kDa) in CTX‐IR group compared to CTX group (p = 0.0006), but no significant difference was observed between uninjured and CTX group (p = 0.3752) (Figure 4A,B). Immunofluorescence quantification further confirmed increased positive area of TGF‐β1 in CTX‐IR model at 5dpi (p = 0.0048) (Figure 4C). To evaluate the downstream fibrotic output, FAPs were stimulated with TGF‐β1. Significantly increased percentage of Ki67⁺ FAP (p < 0.0001) and decreased percentage of TUNEL⁺ FAP (p = 0.0008) were observed in TGF‐β1 treated samples (Figure 4D–F). Immunofluorescence quantification showed a significant increase in Collagen I positive area normalized to cell number (number of nuclei marked by DAPI) in TGF‐β1‐treated cells compared to vehicle (Figure 4H), indicating that TGF‐β1 is sufficient to promote collagen production in FAPs. Co‐staining of p‐Smad2/3 with cytoplasmic Collagen I demonstrated increased positive area of Collagen I (p = 0.00027) and increased nucleus translocation of p‐Smad2/3 (p < 0.0001) in TGF‐β1 treated cells (Figure 4H–J), supporting a direct transcriptional link between Smad activation and collagen synthesis.

3.5. Neutralizing TGF‐β Ameliorates IR‐Induced Fibrosis and Impaired Myofiber Regeneration

We used TGF‐β NAb and IgG intraperitoneally at 6 and 12 h after IR to investigate the effect of TGF‐β in fibrosis and regeneration deficits of skeletal muscle and validate the efficiency of anti‐TGF‐β therapy to ameliorate the effect of IR in regenerating muscle (Figure 5A). Figure 5B illustrated reduced nucleus translocation of p‐Smad2/3 in PDGFRα+ cells in TGF‐β NAb treated mice compared to IgG control. The number of FAPs normalized to muscle mass was significantly reduced after administration of TGF‐β NAb (p = 0.0067) (Figure 5C). Higher TA/body mass ratio was observed in TGF‐β NAb treated mice (p = 0.0028) (Figure 5D) HE and Sirus red staining of TAs demonstrated increased CSA and decreased collagen deposition (p = 0.008) (Figure 5E–G), decreased positive area of Collagen 1 (p = 0.0164), Collagen III (p = 0.0196) and CD90 (p = 0.0189) in TGF‐β NAb treated muscle (Figure 5H–J).

FIGURE 5.

TGF‐β neutralization ameliorates IR‐induced fibrosis and improves skeletal muscle regeneration. (A) Schematic of the experimental timeline. Mice received TGF‐β neutralizing antibody (TGF‐β NAb) or isotype control IgG at 6 h and 12 h after IR induction, and were euthanized at 5 dpi and 17 dpi for analysis; (B) Representative immunofluorescence images of TA muscle sections stained for PDGFRα (red) and p‐Smad2/3 (green) at 5 dpi. Nuclei were counterstained with DAPI (blue); (C) Left: Representative flow cytometry plots gated on PDGFRα⁺ cells from TA muscle single‐cell suspensions at 5 dpi. Right: Quantification of FAP numbers per gram of muscle. (n = 3 mice per group); (D) Bar graph showing TA muscle weight normalized to body weight at 17 dpi; (E) Representative H&E (top) and Sirius Red (bottom) staining of TA muscle cross‐sections at 17 dpi; (F) Frequency distribution histogram of muscle fiber cross‐sectional area (CSA) at 17 dpi; (G) Quantification of Sirius Red‐positive area (collagen deposition) as a percentage of total tissue area; (H) Representative immunofluorescence images of type I collagen at 17 dpi. (I) Representative immunofluorescence images showing co‐localization of type III collagen and CD90 at 17 dpi. Nuclei were counterstained with DAPI; (J) Quantification of the percentage area positive for type I collagen, type III collagen, and CD90. Data are presented as mean ± SD. Individual data points are shown in all bar graphs. Sample sizes: n = 6 mice per group for (D); n = 3 mice per group for (C, F, G, J), 2 microscopic vision fields for each mice, yielding 6 fields per group for statistical analysis. Normality was confirmed by Shapiro–Wilk test. Two‐group comparisons (NAb vs. IgG) in (C, D, F, G, J) were performed using unpaired two‐tailed Student's t‐test. Adjusted or nominal p values: *p < 0.05, **p < 0.01; ns, not significant. Scale bars: 100 μm.

4. Discussion

The regeneration of skeletal muscle is a critical determinant of functional recovery in orthopedic trauma. However, the management of concomitant muscle injuries is often neglected in clinical practice, particularly when they are overshadowed by more salient fractures. While the impaired regeneration of muscle tissue was traditionally attributed to factors like disuse due to immobilization or comorbidities like diabetes, fragility and aging, growing evidence emphasizes the critical role of disrupted inflammatory resolution in disturbing the balance between regeneration of myofibers and ECM deposition [1, 2, 16, 17].

The regeneration of skeletal muscle is a critical determinant of functional recovery in orthopedic trauma. While prior models induced ischemia concurrently with myonecrosis to simulate acute synchronous polytrauma, our study introduces ischemia–reperfusion (IR) at 3 days post‐injury to model a distinct clinical challenge: the delayed surgical “second hit”. In clinical practice, definitive internal fixation is routinely delayed 3 to 5 days to allow for soft‐tissue edema resolution, during which the application of a surgical tourniquet inflicts an unavoidable iatrogenic IR injury [18, 19]. Biologically, this 3‐day post‐injury window represents a critical inflection point, marking the transition from a pro‐inflammatory state to a reparative phase characterized by peak fibro‐adipogenic progenitor (FAP) activation [16, 20]. By superimposing IR stress as the local microenvironment coordinates structural repair, our model demonstrates how this surgical interruption derails FAP‐myoblast crosstalk and aggressively shifts the tissue trajectory toward pathological fibrosis.

4.1. IR Exacerbates Fibrosis and Impairs Regeneration of Myofibers

IR injury is a common consequence of vascular injuries, arterial occlusions, and the improper use of surgical tourniquets in extremity trauma and orthopedic procedures [4, 5]. Inflammation is the central mechanism through which IR injury inflicts damage on skeletal muscle. The process begins with an oxidative burst that triggers a powerful innate immune response, characterized by neutrophil infiltration, cytokine release, and complement activation [21, 22, 23]. This inflammatory cascade causes direct cellular toxicity, microvascular obstruction, and compartment syndrome, ultimately resulting in muscle fibrosis or even necrosis [21]. Prior work in skeletal muscle has recognized that IR injury could delay regenerative kinetics and impair recovery compared to CTX injury alone. In a recent study by Stepien et al. [22], FA and FV were occluded using Ackland clamp for 3 h to induce IR injury immediately following CTX injection of TAs to induce a combined CTX and IR injury. The combined injury demonstrated increased infiltration of neutrophils, monocytes and mature macrophages, and exacerbated damage and fibrosis were observed compared to muscle samples of CTX‐only injuries.

Successful regeneration of skeletal muscle balances the infiltration of inflammatory cells, clearance of cellular debris, the regulation of proliferation and differentiation of muscle stem cells to regenerate myofibers with reconstitution of the extracellular matrix (ECM) architecture. Perturbation of this process that amplifies or prolongs the initial inflammatory response could result in compromised regeneration of injured skeletal muscle [1, 2, 16, 17]. In our study, the introduction of IR at 3 days after CTX injury disturbed the inflammation‐regeneration translation process. With increased fibrosis and decreased CSA distribution of myofibers, as well as elevated protein levels of collagen I, collagen III, and TGF‐β1 in western blots, IR at 3 dpi of CTX injury demonstrably exacerbates tissue damage and promotes pathological fibrosis. The variability observed in certain histological measures including collagen area, myofiber CSA, etc. likely reflects inherent biological heterogeneity among individual mice, as well as technical variation in tissue section orientation. Despite this, the effect sizes were large enough to achieve statistical significance with the sample sizes used, supporting the robustness of the conclusions.

4.2. TGF‐β Is a Pivotal Regulator in Muscle Fibrosis and Myogenesis

Resident myeloid cells become activated and recruit neutrophils and monocytes following acute skeletal muscle injury, which differentiate into macrophages that clear necrotic fibers and release inflammatory cytokines (e.g., via IL‐6, TNF‐α, TGF‐β). As regeneration proceeds, these inflammatory macrophages gradually transition toward a reparative, anti‐inflammatory phenotype, fostering an environment permissive for myogenesis [11, 22, 24, 25]. Concurrently, FAPs transiently expand and secrete ECM components to scaffold nascent myofibers. In normal regeneration conditions, FAPs undergo apoptosis or are cleared after the regenerative period [16]. However, sustained or excessive FAP activation drives pathological ECM deposition and fibrosis. Dysregulation of FAP homeostasis is increasingly recognized as a central node in fibrotic muscle disease.

TGF‐β, particularly the β1 isoform, is a key mediator orchestrating the immune–stromal crosstalk. To investigate which TGF‐β isoform(s) contribute to the profibrotic response, we measured mRNA expression of Tgfb1, Tgfb2, and Tgfb3 in NC, CTX, and CTX‐IR muscles at 5 dpi (Figure 4C). Only Tgfb1 was significantly upregulated in CTX‐IR compared to CTX (adjusted p < 0.0001, one‐way ANOVA with Tukey's test), whereas Tgfb2 and Tgfb3 showed no significant changes (adjusted p = 0.66 and p = 0.31, respectively). This selective upregulation of TGF‐β1 is consistent with previous reports identifying it as the primary profibrotic isoform in skeletal muscle injury. Accordingly, the pan‐TGF‐β neutralizing antibody (clone 1D11) used in subsequent experiments effectively blocks all three isoforms, with its predominant effect attributable to neutralization of the elevated TGF‐β1 ligand. TGF‐β released by reparative macrophages or activated connective‐tissue cells stimulates ECM gene expression and modulates the survival, proliferation, and differentiation of both FAPs and myogenic progenitors. Notably, TGF‐β signaling is reported to promote proliferation and reduce apoptosis in FAPs, thereby sustaining their population and favoring a fibrogenic fate, which is confirmed in our result (Figure 4D–F) [11, 25].

In our study, imposure of IR at 3 days post‐CTX was associated with a robust rise in TGF‐β1 at 5 dpi. Flow cytometry showed a higher proportion of FAPs in CTX‐IR models and increased Collagen I expression in sorted FAPs after stimulation of TGF‐β1. These data support a model in which IR enhances activation of latent TGF‐β and thereby drives FAP survival, expansion, and ECM production, tipping the regenerative balance toward fibrosis. Additionally, TGF‐β was reported to inhibit myoblast differentiation and impair myotube fusion in vitro. In our study, the impaired CSA in delayed regeneration in IR muscles is consistent with excessive TGF‐β signaling interfering with myofiber maturation.

4.3. TGF‐β as Therapeutic Target in IR‐Associated Regeneration Impairment of Skeletal Muscle

Given the above evidence, therapeutic inhibition of TGF‐β signaling becomes a candidate option to mitigate the deposition of fibrotic tissue and facilitate myofiber regeneration. Mázala et al. [25] reported that inhibiting TGF‐β reduced FAP accumulation and fibrosis of skeletal muscle in D2‐mdx mice after intramuscular injection of neotoxin. Although the regeneration of myofiber was not fully rescued after inhibition of TGF‐β signaling, we suppose that the deficit of myoblast and myotube in D2‐mdx mice could compromise the effect of TGF‐β inhibition, and the intramuscular injection of ITD (TGF‐β receptor antagonist) 2 and 4 days after NTX injury could be an additional injury that impaired the regeneration of myofibers. In our study, therapeutic intraperitoneal administration of TGF‐β NAb at 6 and 12 h after IR successfully attenuated fibrosis, reduced the percentage of collagen area from 14.33% ± 2.33% in IgG group to 10.79% ± 1.15% in TGF‐β NAb treated samples (p = 0.008), with improved CSA of regenerated myofibers (Figure 5C). Therefore, our study demonstrated the therapeutic potential of TGF‐β in facilitating muscle regeneration in context of clinical settings where IR can be unavoidable.

However, the translation of systemic inhibition into a safe clinical reality is challenged by the cytokine's profound pleiotropic roles in health [16, 22]. The TGF‐β family is a central regulator of immune homeostasis, inhibiting both innate and adaptive immune cell function, including T cell proliferation and function. Global inhibition carries the risk of inducing severe immune dysregulation, autoimmunity, and impairing host defense. Due to the above concern, we adapted the short but high‐dose administration of TGF‐β NAb at 6 and 12 h after IR (5dpi of CTX) to highlight the effectiveness of TGF‐β inhibition strategy and minimize the impact of global health.

5. Limitations

Despite establishing the mechanistic link between IR and regeneration impairment, and demonstrating therapeutic efficacy, the current study has intrinsic limitations common to preclinical models. Firstly, while clinically relevant, the CTX‐IR model remains an acute, localized injury model in young, genetically uniform mice which might not fully recapitulate the heterogeneity of human musculoskeletal injury, which often involves compounding issues like nerve damage, fractures, significant soft tissue avulsion, and patient comorbidities. Secondly, while our targeted, a formal dose–response optimization study was not conducted due to resource constraints, the precise optimal therapeutic threshold and delivery window remain to be fully defined. Meanwhile, our low‐dose administration of TGF‐β NAb (cumulative 5 mg/kg over 12 h) successfully attenuated fibrosis while avoiding the long‐term architectural deficits associated with chronic, high‐dose pan‐TGF‐β blockade (10 mg/kg of long‐term administration) [24, 26]. Third, while the Day 3 IR induction is clinically and biologically anchored, we did not perform a systematic temporal comparison, which remains a limitation, and future temporal gradient studies are required to fully define the sensitivity of the regenerative microenvironment to ischemic interruptions. Additionally, functional analysis (ex vivo isometric contractile force measurement) was not performed in our study to analyze the functional recovery injury models and TGF‐β NAb treated mice.

6. Conclusion

We demonstrate that IR superimposed on acute muscle injury provokes maladaptive fibrotic remodeling and suppresses functional myogenesis via TGF‐β associated mechanism. Inhibition of TGF‐β reduces fibrosis and rebalances the response toward regeneration. These results provide mechanistic insight into how derails muscle repair and offer a translationally relevant antifibrotic strategy for musculoskeletal trauma settings.

Author Contributions

Hongru Ma: conceptualization, methodology, investigation, software, data curation, formal analysis, writing – original draft. Zhenjia Zhong: investigation, data curation; Kaiyang Wang: resources, writing – review and editing. Heng Zheng: resources, project administration; Yutong Dai: data curation, resources; Benlong Shi: resources; Xu Sun: resources; Zhen Liu: resources; Yong Qiu: resources, project administration, funding acquisition. Zezhang Zhu: supervision, resources, conceptualization, project administration. Fengfeng Li: conceptualization, writing – review and editing, funding acquisition.

Funding

This research was supported by Jiangsu Provincial Medical Innovation Center of Orthopedic Surgery (CXZX202214, 2022–2025); National Natural Science Foundation of China (81873996); Special Medical Expert Project of Jiangsu Province (SB2199202203).

Ethics Statement

All animal experiments were approved by the Institutional Ethics Committee of the Sixth Affiliated Hospital of Guangzhou Medical University (Qingyuan People's Hospital), Qingyuan, China (approval No. LAEC‐2022‐042), and were conducted in accordance with relevant institutional and national guidelines for the care and use of laboratory animals. All animal procedures, including CTX injection, ischemia‐reperfusion surgery, tissue harvesting, and euthanasia, were conducted at the Sixth Affiliated Hospital of Guangzhou Medical University (Qingyuan People's Hospital). A subset of subsequent tissue section processing and staining was conducted at Nanjing Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School. The Laboratory for Myology, Department of Human Movement Sciences, Faculty of Behavioral and Movement Sciences, Vrije Universiteit Amsterdam, Amsterdam Movement Sciences, provided intellectual input and contributed to study design but did not conduct any animal or laboratory experiments.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Ischemia–reperfusion (IR) alone induces minimal fibrosis and does not recapitulate CTX‐IR‐induced regeneration impairment. (A) Schematic of the experimental timeline for the CTX, CTX‐IR and IR groups; (B) TA/body mass ratio of CTX, CTX‐IR and IR group; (C) Representative H&E (top) and Sirius Red (bottom) staining of TA muscle cross‐sections at 17 dpi, (D) Frequency distribution histogram of muscle fiber cross‐sectional area (CSA) at 17 dpi. (E) Quantification of Sirius Red‐positive area (collagen deposition) as a percentage of total tissue area. Data are presented as mean ± SD. Individual data points are shown in (B, D, E). Sample sizes: n = 6 mice per group for (B); Sample size: n = 3 for (C, D, E), 2 microscopic vision fields for each mice, yielding 6 fields per group for statistical analysis. Comparisons among three groups (CTX, CTX‐IR, IR) were analyzed by one‐way ANOVA followed by Tukey's honestly significant difference (HSD) test. Adjusted p values: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant. Scale bars: 100 μm.

Data Availability Statement

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