Macrophage Derived TGF‐β Recruits BMSCs for Initiation of Bone Formation During Fracture Healing

ABSTRACT

Impaired fracture healing remains a major clinical challenge, with approximately 10% of fractures progressing to nonunion or delayed union. Although numerous studies have focused on strategies to promote fracture healing, the mechanisms governing the initiation phase of repair remain poorly understood. This study aimed to investigate the contribution of macrophage‐derived TGF‐β1 to the initiation of early fracture repair. A murine femoral osteotomy model was used to study fracture healing. We examined the temporal relationship among bone healing, callus maturation, pSmad2/3 positive cells, and osteoprogenitors. TGF‐β1 signaling was systemically inhibited using a neutralizing antibody. Macrophages were selectively depleted in LysM‐iDTR mice by diphtheria toxin administration. The source and expression levels of TGF‐β1 at the fracture site were determined by pSmad2/3 staining and RT‐qPCR. Fracture healing was assessed using micro‐CT, histomorphometry, and immunohistochemistry. TGF‐β1 signaling peaked during the inflammatory phase and preceded osteoprogenitor recruitment. Systemic inhibition of TGF‐β1 or selective macrophage depletion before fracture establishment severely disrupted mesenchymal stem cell (MSC) recruitment and bone formation. In contrast, late‐stage macrophage depletion did not substantially affect bone healing. CD86+ macrophages were identified as a key early source of TGF‐β1. Importantly, local delivery of TGF‐β1 via hydrogel microspheres during the early stage rescued osteogenic repair in macrophage‐deficient mice. CD86+ macrophage‐derived TGF‐β1 is essential for initiating fracture repair by recruiting MSCs and promoting osteogenesis. Targeting this pathway may represent a novel therapeutic strategy to enhance bone regeneration in cases of impaired healing.

At the early stage of fracture, the macrophages migrate to the fracture site for anti‐inflammation and tissue reconstruction. Upon arrival, the TGF‐β derived from macrophages induces recruitment of bone marrow mesenchymal stromal cells (BMSCs) from bone marrow to the fracture site. Subsequently, the BMSCs differentiate into osteoprogenitors and osteoblasts and undergo bone modeling and remodeling after new bone formation.

Abbreviations

ANOVA

analysis of variance

BMSCs

bone marrow stromal cells

BV

bone volume

CD206

cluster of differentiation 206

CD86

cluster of differentiation 86

DT

diphtheria toxin

IHC

immunohistochemistry

LAP

latency‐associated peptide

LepR

leptin receptor

M1

classically activated macrophages

M2

alternatively activated macrophages

MMPs

matrix metalloproteinases

MSCs

mesenchymal stem cells

OSX

osterix

POD

post‐operative day

pSmad2/3

phosphorylated Smad2/3

ROI

region of interest

RT‐qPCR

real‐time quantitative polymerase chain reaction

SD

standard deviation

Tb.N

trabecular number

TGF‐β

transforming growth factor‐beta

TRAP

tartrate‐resistant acid phosphatase

μCT

micro‐computed tomography

1. Introduction

Bone fractures are estimated to occur in over 16 million people in the United States each year, imposing a substantial economic and clinical burden [1]. Most fractures heal completely through a regenerative process that restores fully functional and integrated new tissue [2]. However, approximately 10% of fractures result in delayed union or nonunion [3]. Understanding the physiological processes underlying fracture repair is therefore critical for addressing this public health concern. Fracture healing occurs primarily through endochondral ossification, during which mesenchymal stem cells (MSCs) condense, undergo chondrogenesis, and are subsequently replaced by mineralized bone [4]. After fracture, multiple cell types—including immune cells, macrophages, MSCs, endothelial cells, and osteoprogenitors—are sequentially recruited sequentially from the inflammatory phase to the repair phase, ultimately supporting MSC differentiation into callus bone [5].

At the molecular level, key signaling molecules, including transforming growth factor‐β1 (TGF‐β1) and bone morphogenetic proteins (BMPs), play pivotal roles in mediating bone formation during fracture repair [6, 7]. TGF‐β1 has been reported to stimulate MSC proliferation as a pleiotropic growth factor [6]. Precise spatiotemporal activation of TGF‐β is essential for successful tissue regeneration and remodeling; conversely, dysregulation of its ligand expression, bioavailability, or activation disrupts physiological processes and impairs healing [8]. Increased concentrations of TGF‐β in serum and fracture hematoma have also been reported during physiological healing in patients, particularly during the early healing period, indicating the importance of TGF‐β in fracture repair [9]. TGF‐β1 has also been shown to induce osteoprogenitor condensation through chemotaxis rather than by directly enhancing osteogenesis [10]. However, whether TGF‐β1 plays an important role in fracture healing by recruiting MSCs, how it fits within the sequence of cellular events during repair and the origin of TGF‐β1 in the healing process remain unclear.

The TGF‐β subfamily comprises three isoforms (TGF‐β1, TGF‐β2, and TGF‐β3), which are secreted as latent complexes bound to latency‐associated protein (LAP). Binding to LAP maintains TGF‐β latency by sequestering it within the extracellular matrix of various tissues [11, 12]. Activation of latent TGF‐β1 is mediated through diverse mechanisms, including proteolytic cleavage by matrix metalloproteinases (MMPs) and plasmin, interactions with integrins and thrombospondin‐1, and exposure to reactive oxygen species or acidic pH conditions [13, 14, 15]. TGF‐β1 has been shown to coordinate bone remodeling, in which osteoclast‐mediated activation of TGF‐β1 stimulates bone formation by recruiting bone marrow‐derived mesenchymal stromal cells [10]. In addition, aberrant activation of TGF‐β1 plays a critical role in several pathological ossification, including enthesopathy, heterotopic ossification and aberrant bone formation [16, 17, 18].

Macrophages are pivotal immune cells endowed with remarkable plasticity, allowing them to adopt diverse functional phenotypes—primarily the pro‐inflammatory M1 state and the anti‐inflammatory, pro‐reparative M2 state [19]. In the context of bone injury, macrophages act as indispensable orchestrators of the healing process [20]. They facilitate the initial inflammatory response, clear debris, and engage in crosstalk with MSCs, thereby influencing MSC recruitment, proliferation, and osteogenic differentiation. This interaction is fundamental to fracture repair because it bridges innate immune responses with skeletal regeneration [20, 21, 22, 23]. A key molecular mediator in this reparative dialogue is TGF‐β. Both M1 and M2 macrophages can serve as sources of TGF‐β. Macrophage‐derived TGF‐β activity may have a critical role in initiating fracture healing, given the high level of TGF‐β release and its activation during bone remodeling. However, how macrophage‐derived TGF‐β directly initiates and coordinates the early osteogenic programs during fracture healing remains inadequately defined. Elucidating this specific axis is crucial for developing targeted therapeutic strategies to enhance bone regeneration.

Here, using a mouse model of fracture healing, we elucidated the specific contribution of TGF‐β1 secreted by macrophages. We identified TGF‐β1 as a critical mediator that orchestrates the homing of MSC to the injury site, thereby promoting bone formation. Notably, systemic inhibition of TGF‐β1 activity with a neutralizing antibody markedly impaired bone regeneration, underscoring the non‐redundant function of this pathway in repair. Furthermore, macrophage depletion during the early post‐fracture stage resulted in significantly compromised healing, directly indicating that macrophages are critical for the initiation of fracture repair. More importantly, this healing impairment was effectively rescued by the local application of TGF‐β1‐loaded microspheres, which restored the osteogenic program and fracture union. Collectively, our findings establish macrophage‐derived TGF‐β1 as a key initiator of the bone formation phase during fracture repair.

2. Materials and Methods

2.1. Mice

Wild‐type C57BL/6J male mice (12 weeks old) were obtained from Shanghai Model Organism Center. For antibody intervention studies, 3‐month‐old male mice received intraperitoneal injections of either the control antibody 13C4 or the TGF‐β1 neutralizing antibody 1D11 (R&D Systems, Minneapolis, MN) at a dose of 5 mg/kg body weight, administered three times per week starting at 1, 2, or 3 weeks post‐fracture. Sham‐operated mice without any treatment served as baseline controls. To generate LysM‐iDTR mice for monocyte depletion, LysM‐Cre mice (GemPharmatech, T003822) were crossed with ROSA26::iDTR mice (JAX Lab, 008040). Only homozygous LysM‐iDTR mice were used in this study. Phenotypic validation was performed via intraperitoneal injection of diphtheria toxin (DT, 25 μg/kg; Calbiochem), and genotyping was confirmed by polymerase chain reaction (PCR). For the early deletion group, the DT was injected 1 week before establishment of the facture model. For the late deletion group, the DT was injected 1 week after establishment of the facture model.

For the fracture model, male mice were anesthetized with ketamine/xylazine. An open femoral fracture model was established by surgically exposing the distal two‐thirds of the femur. A complete osteotomy was performed at the mid‐diaphysis using an oscillating saw to create a standardized 1 mm gap fracture.

All animals were housed in the Animal Facility of Tongji University School of Medicine. All experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Tongji University, Shanghai, China.

2.2. Histochemistry, Immunohistochemistry, and Histomorphometry

Paraffin‐embedded tissue blocks were sectioned at a thickness of 4 μm using a microtome. Sections were stained with Safranin O (Sigma‐Aldrich, S2255) and Fast Green (Sigma‐Aldrich, F7252) staining to evaluate proteoglycan content and tissue morphology. Safranin O stains proteoglycans (cartilage matrix) red, whereas Fast Green stains bone matrix and other proteins green. Tartrate‐resistant acid phosphatase (TRAP) staining was performed to identify mature osteoclasts according to the manufacturer's protocol (Sigma‐Aldrich, 387A‐1KT), followed by counterstaining with Methyl Green (Sigma‐Aldrich, M8884). Immunohistochemical and immunofluorescence analyses were carried out as previously described [24]. Following deparaffinization, sections were subjected to antigen retrieval by heating at 99°C for 10 min in Target Retrieval Solution (Dako, S1699), followed by rehydration. After three washes with PBS, antigen‐retrieved paraffin sections and frozen sections were incubated overnight at 4°C in a humidified chamber with the following primary antibodies: human/mouse pSmad2/3 (Santa Cruz Biotechnology, sc‐11 769; 1:50), mouse CD86 (CST, #91882; 1:100), mouse CD206 (CST, #24595; 1:100), leptin receptor (Invitrogen, MA532685; 1:100), and osterix (Abcam, ab22552; 1:100). After additional PBS washes, immunohistochemical staining was performed by incubating the sections with appropriate secondary antibodies in blocking solution for 1 h at room temperature. Immunoreactivity was visualized using chromogenic substrates (Dako, K3468), and sections were counterstained with hematoxylin (Sigma‐Aldrich, H9627). For quantification, positively stained cells were counted in five randomly selected fields across five sequential sections per mouse within each experimental group. Results were normalized to the number of cells per square millimeter within the fracture area.

2.3. Flow Cytometry Analysis

Flow cytometry was performed using a FACSCalibur instrument (Becton Dickinson) equipped with CellQuest software for data acquisition and analysis. The primary antibodies used for staining included APC‐conjugated anti‐Sca1 (BioLegend, 122 511; dilution 1:100), APC‐conjugated anti‐CD86 (BioLegend, 102 409; dilution 1:100), and APC‐conjugated anti‐CD206 (BioLegend, 120 413; dilution 1:100).

2.4. MicroCT Analysis

Murine femora were fixed in 10% formalin for 48 h and scanned using a high‐resolution μCT system (Skyscan 1172). Scans were acquired at 60 kV with an isotropic voxel size of 12 μm. The region of interest (ROI) was manually delineated to encompass the entire fracture callus region. A series of eight contiguous slices was selected for three‐dimensional reconstruction. The quantified structural parameters included total callus volume, bone volume (BV), and trabecular number (Tb.N). All μCT analyses were performed by an investigator blinded to group allocation.

2.5. Specimen Collection

Mice were deeply anesthetized and transcardially perfused with 10% buffered formalin for 5 min via the left ventricle. For histological and immunohistochemical evaluation, all harvested specimens were fixed in 10% buffered formalin for 24 h and subsequently decalcified in 10% ethylenediaminetetraacetic acid (EDTA; VWR, 0105; pH 7.4) for 14 days. After decalcification, the tissues were embedded in paraffin.

2.6. Statistical Analysis

All statistical analyses were conducted using SPSS 19.0 software. Data were presented as the mean ± SD. Student's t‐tests or ANOVA were conducted depending on the comparison groups to determine statistical significance. p < 0.05 was considered statistically significant.

3. Results

3.1. Elevated TGF‐β1 Signaling Is Associated With Bone Formation During Fracture Healing

To delineate the temporal dynamics of fracture repair and their association with TGF‐β1 signaling, we first characterized the healing process in a murine fracture model. Micro‐computed tomography (μCT) analysis revealed a progressive restoration of bone architecture over time. Initial callus formation was observed at the fracture site on day 3 and culminated in extensive bony callus formation at 7, 14, and 28 days (Figure 1A). Consistently, μCT analysis showed that whole callus volume, bone volume (BV) and trabecular number (Tb. N) increased significantly from day 3 to 14 and then gradually changed as healing progressed (Figure 1B–D). At day 3, the fracture site was characterized by hematoma formation and the initial appearance of soft callus, as shown by Safranin O/Fast Green (SO/FG) staining. By day 7, a prominent soft callus with evident cartilage formation was observed. This was followed by the transition to a hard callus through endochondral ossification at day 14, culminating in extensive bony callus formation and the initiation of remodeling by day 28 (Figure 1E).

FIGURE 1.

TGF‐β1 signaling is dynamically activated and precedes osteoprogenitor expansion during fracture healing. (A) Representative three‐dimensional micro‐CT reconstruction images of the fracture callus at 3‐, 7‐, 14‐, and 28‐days post‐operative (POD). Scale bar = 1 mm. (B‐D) Quantitative micro‐CT analysis of the fracture callus over time, including whole callus volume (B), bone volume (BV) (C), and trabecular number (Tb.N) (D). (E) Representative histological sections of the fracture callus stained with Safranin O/Fast Green (SO/FG) at the indicated time points. Scale bar = 500 μm. (F, G) Immunohistochemical (IHC) staining (F) and quantitative analysis (G) for phosphorylated Smad2/3 (pSmad2/3) in the fracture callus at POD 3, 7, 14, and 28. Scale bar = 500 μm. (H, I) IHC staining (H) and quantitative analysis (I) for leptin receptor (LepR) in the fracture callus at the indicated time points. Scale bar = 500 μm. (J, K) IHC staining (J) and quantitative analysis (K) for osterix (OSX) in the fracture callus at the indicated time points. Scale bar = 500 μm. Data are presented as the mean ± SD. **p < 0.05, ***p < 0.01, ****p < 0.001.

We next assessed the activation of the TGF‐β1 pathway and its correlation with osteogenic commitment by immunohistochemistry. Staining for phosphorylated Smad2 (pSmad2), a direct indicator of canonical TGF‐β1 signaling activation, showed that pathway activity was significantly elevated at the fracture site. The signal intensity and number of pSmad2‐positive cells increased after injury, peaked at day 7, and subsequently declined to near baseline levels by day 28 as the callus matured (Figure 1F,G). In striking contrast, the expression patterns of key osteogenic markers followed a distinct, subsequent trajectory. The populations of leptin receptor‐positive (LepR+) mesenchymal stromal cells (Figure 1H,I) and osterix‐positive (OSX+) osteoprogenitors (Figure 1J,K) within the callus increased steadily from day 3, reached their peak at POD 14, coinciding with the peak of osteogenic activity, and remained elevated through POD 28.

Collectively, these data demonstrate a precise spatiotemporal activation of TGF‐β1 signaling during fracture healing, which precedes and is temporally associated with the expansion of osteoprogenitor cells and subsequent bone formation.

3.2. Systemic Inhibition of TGF‐β1 Signaling Disrupts Callus Maturation and Bone Formation

To directly investigate the functional requirement for TGF‐β1 signaling in fracture repair, we systemically administered a TGF‐β1 neutralizing antibody (1D11) to mice immediately after fracture establishment. μCT analysis at the endpoint of the healing period revealed a profound impairment of bone regeneration in the antibody‐treated group compared with the vehicle‐control group. The control group exhibited robust bony callus formation with evidence of advanced remodeling, successfully bridging the fracture gap. In stark contrast, the TGF‐β1 neutralizing antibody‐treated group displayed a significantly smaller and less mineralized callus, indicating a substantial failure to form adequate new bone (Figure 2A–D). Histological assessment with SO/FG staining provided a clear explanation for this radiological deficit. At the time of bony union in the controls, the fracture callus was predominantly composed of mature bone matrix (stained green), with only minimal remnants of transitional cartilage. Conversely, the 1D11‐treated group showed a persistent and disorganized cartilaginous callus (stained red), indicative of a profound disruption of endochondral ossification. The failure to resolve this cartilage and replace it with bone underscores the critical role of TGF‐β1 in driving callus maturation (Figure 2E). Immunohistochemical analysis confirmed that this phenotype resulted from effective pathway inhibition. The number of pSmad2/3‐positive cells within the callus was significantly lower in the TGF‐β1 neutralizing antibody‐treated group, confirming the successful blockade of canonical TGF‐β1 signaling (Figure 2F,G). Consequently, this ablation of TGF‐β1 signaling severely compromised the osteogenic lineage. OSX staining revealed a significant reduction in the number of osteoprogenitors and osteoblasts within the fracture callus of antibody‐treated mice (Figure 2H,I). This finding indicates that TGF‐β1 is required for sustained osteogenic lineage activity.

FIGURE 2.

Systemic inhibition of TGF‐β1 signaling disrupts callus maturation and osteogenesis. (A) Representative μCT images of the fracture callus in mice treated with vehicle or TGF‐β1 neutralizing antibody. (B–D) Quantitative μCT analysis of whole callus volume (B), BV (C), and Tb.N (D). (E) Representative SO/FG‐stained sections of the fracture callus at POD28. (F, G) IHC staining (F) and quantitative analysis (G) for pSmad2/3 in the fracture callus. Scale bar = 500 μm. (H, I) IHC staining (H) and quantitative analysis (I) for OSX in the fracture callus. Scale bar = 500 μm. Data are presented as the mean ± SD. **p < 0.05, ***p < 0.01, ****p < 0.001.

Taken together, these results demonstrate that systemic neutralization of TGF‐β1 activity severely disrupts the normal progression of fracture healing, and prevents the formation of a mature bony union.

3.3. Dynamic Alterations in Macrophage Polarization Are Coupled With Fracture Healing Progression

We next sought to identify the potential source of TGF‐β1 at the fracture site during bone healing process. Activation of inactive TGF‐β1 embedded in bone tissue through osteoclast‐mediated resorption represents an important source of active TGF‐β1 during bone remodeling. Tartrate‐resistant acid phosphatase (TRAP) staining revealed that osteoclast activity was dynamically regulated and distributed (Figure 3A). A small number of TRAP‐positive multinucleated osteoclasts appeared only at the fracture ends, but not in the hematoma or surrounding tissues. The TRAP+ cells within the callus increased, and reached a peak at day 7, and subsequently declined as the bony callus matured and consolidated by day 28 (Figure 3B).

FIGURE 3.

Dynamic macrophage polarization and osteoclast activity during fracture healing. (A) Representative TRAP staining of the fracture callus at POD 3, 7, 14, and 28. (B) Quantitative analysis of TRAP⁺ multinucleated osteoclasts per mm². (C, D) IHC staining (C) and quantitative analysis (D) for CD86 at POD 3, 7, 14, and 28. Scale bar = 500 μm. (E, F) IHC staining (E) and quantitative analysis (F) for CD206 at POD 3, 7, 14, and 28. Scale bar = 500 μm. (G‐J) RT‐qPCR analysis of TGF‐β1 expression from total cells (G), sorted macrophages at different time points (H), and sorted M1 vs. M2 macrophages at POD3 (I) and POD14 (J). Data are presented as the mean ± SD. **p < 0.05, ***p < 0.01, ****p < 0.001.

Because pSmad2+ cells were mainly distributed in the soft tissue around the fracture ends, we further characterized the inflammatory microenvironment, with a specific focus on macrophage dynamics, during the fracture healing process. To track macrophage infiltration and polarization, we performed immunohistochemistry for phenotype‐specific markers. Analysis of polarization states showed a distinct temporal shift. CD86 staining (an M1 marker) was predominant at day 3, whereas CD206 staining (an M2 marker) peaked at day 14 during the reparative phase (Figure 3C–F). Thus, the macrophage population was predominantly M1‐like (CD86+) at day 3 and transitioned to a predominantly M2‐like (CD206+) state by day 14.

We then investigated the source of TGF‐β1 within this macrophage population. RT‐PCR analysis of TGF‐β1 levels in sorted macrophage populations and total cell populations from the fracture site revealed that macrophage‐derived TGF‐β1 expression was significantly higher than that in control groups (Figure 3G). Furthermore, TGF‐β1 expression in sorted macrophages was much higher at day 3 than at day 14 (Figure 3H). Specifically, at both day 3 and at day 14, TGF‐β1 expression was much higher in M1 macrophages than in M2 macrophages (Figure 3I,J).

Collectively, these findings illustrate a precise temporal program of macrophage polarization that is tightly coupled with the stages of fracture repair. The identification of early M1‐like macrophages as a significant source of TGF‐β1 reveals a key upstream regulator of the TGF‐β1‐dependent healing cascade initiated during the inflammatory phase.

3.4. Macrophage Depletion at Early Stage Impairs the Progression of Fracture Healing

To directly investigate the functional contribution of macrophages to fracture repair, we utilized the LysM‐iDTR mouse model, which allows specific and inducible macrophage ablation upon diphtheria toxin (DT) administration. μCT analysis demonstrated that mice with early macrophage depletion (DT group) exhibited a significantly smaller and less mineralized bony callus compared with controls (Figure 4A–D). SO/FG staining further showed that the fracture callus in DT‐treated mice was characterized by a persistent and unresolved cartilaginous matrix, in contrast to the well‐ossified, mature bony callus observed in control mice (Figure 4E). This failure to progress through the endochondral ossification program was mechanistically linked to a marked reduction in canonical TGF‐β1 signaling, as evidenced by a much lower number of pSmad2/3‐positive cells in the callus of macrophage‐depleted mice (Figure 4F,G). Consequently, the osteogenic lineage was severely impaired, with a significant reduction in OSX‐positive osteoprogenitors observed in the DT group compared to controls (Figure 4H,I). In contrast, the μCT analysis demonstrated that there were comparable bone healing status and similar levels of the whole callus volume, BV and Tb. N between the control group and the late‐stage macrophage depletion group (Figure S1A–D).

FIGURE 4.

Macrophage depletion abrogates TGF‐β1 signaling and impairs fracture repair at early stage. (A) Representative μCT images of the fracture callus in LysM‐iDTR mice treated with DT or vehicle. (B–D) Quantitative μCT analysis of whole callus volume (B), bone volume (C), and Tb.N (D). (E) Representative SO/FG‐stained sections of the fracture callus. (F, G) IHC staining (F) and quantitative analysis (G) for pSmad2. Scale bar = 500 μm. (H, I) IHC staining (H) and quantitative analysis (I) for OSX. Scale bar = 500 μm. Data are presented as the mean ± SD. **p < 0.05, ***p < 0.01, ****p < 0.001.

Taken together, these results demonstrate that macrophages are indispensable for the initiation of fracture healing. The absence of macrophages leads to profound failure of TGF‐β1 signaling activation, ultimately abrogating recruitment of BMSCs and osteoprogenitors.

3.5. TGF‐β1 Supplementation Rescues Osteoprogenitor Recruitment in Macrophage‐Depleted Mice

We next asked whether the administration of exogenous TGF‐β1 could phenotypically rescue the healing impairment caused by macrophage depletion at early stage. In macrophage‐ablated (LysM‐iDTR + DT) mice, we locally administered TGF‐β1‐loaded hydrogel microspheres (TGF‐β1 group) or blank microspheres (control group) to the fracture site. μCT analysis at the end of the healing period revealed that the systemic absence of macrophages created a significant deficit in bone repair that was largely fully rescued by local TGF‐β1 supplementation. Whole callus volume, BV and Tb. N were not significantly different between the TGF‐β1‐treated and control microsphere groups (Figure 5A–D). This was corroborated by SO/FG staining, which showed a larger proportion of persistent cartilaginous matrix and immature bony tissue in the control group (Figure 5E). A critical and specific rescue was also observed at the cellular level. Immunohistochemical analysis of OSX demonstrated a significant increase in the number of osteoprogenitors within the fracture callus of mice treated with TGF‐β1‐loaded microspheres compared with mice receiving blank control microspheres (Figure 5F,G). This indicates that local TGF‐β1 delivery was sufficient to restore the recruitment of mesenchymal stromal cells even in the absence of macrophages during the early stage. However, TGF‐β1 supplementation did not affect osteoclast maturation or status (Figure 5H,I).

FIGURE 5.

Local TGF‐β1 delivery rescues impaired fracture healing in macrophage depleted mice. (A) Representative μCT images in macrophage‐ablated (LysM‐iDTR + DT) mice treated with TGF‐β1‐loaded or blank vehicle microspheres. (B–D) Quantitative μCT analysis of whole callus volume (B), bone volume (C), and Tb.N (D). (E) Representative SO/FG‐stained sections. (F, G) IHC staining (F) and quantitative analysis (G) for OSX. Scale bar = 500 μm. (H, I) TRAP staining (H) and quantitative analysis (I) of osteoclasts. Scale bar = 500 μm. Data are presented as the mean ± SD. **p < 0.05, ***p < 0.01, ****p < 0.001.

In summary, local application of TGF‐β1 successfully restored the osteoprogenitor pool in macrophage‐deficient mice, confirming the functional primacy of TGF‐β1 as a key macrophage‐derived osteo‐inductive signal.

4. Discussion

The physiology of bone regeneration and effective therapeutic strategies for enhancing this process have been studied for decades. Although TGF‐β1 activity is critical in bone homeostasis and has been extensively investigated [11, 25, 26], knowledge of its role in fracture repair remains limited. Our study provides direct evidence that TGF‐β1 signaling is dynamically activated during fracture repair and that macrophage‐derived TGF‐β1 is indispensable for initiating fracture healing progression. We demonstrate that systemic inhibition of TGF‐β1 activity or depletion of macrophages at early stage severely impairs bone formation, whereas local TGF‐β1 application can specifically rescues osteoprogenitor recruitment in macrophage‐deficient mice.

According to the current concept of fracture repair, coordinated populations of cells participate in fracture healing together with neural and vascular ingrowth after removal of necrotic tissue and the provisional fibrin matrix [23, 27]. Recruited macrophages play a critical role in orchestrating the transition from inflammation to tissue repair and regeneration, a process facilitated by their phenotypic transition from the pro‐inflammatory M1 state to the reparative M2 state [28, 29]. During the reparative process, the cells and soluble mediators provide signals that promote bone formation by recruiting MSCs and driving their differentiation. In the hypoxic microenvironment, neural and vascular invasion is promoted and MSCs differentiate into osteoblast lineage and chondrocyte lineage cells for subsequent soft callus formation [7].

Several growth factors, including TGFs, BMPs, FGFs, PDGFs and IGFs, mediate the reparative process. In the current study, TGF‐β1 signaling was tightly associated with bone formation. TGF‐β activity has been reported to be required for multiple phases of repair, including mesenchymal cell condensation, chondrocyte proliferation and differentiation, and extracellular matrix composition, and can be activated or produced after injury [9, 30]. Previous studies have demonstrated that spatiotemporal activation of latent TGF‐β in the matrix is essential for skeletal tissue homeostasis [10, 31]. During the osteogenic phase, latent TGF‐β is activated through osteoclastic bone resorption, which in turn induces the recruitment of mesenchymal stem cells (MSCs) within the bone remodeling microenvironment [10]. Overactive TGF‐β can induce and promote ectopic bone formation, and initiate pathological changes such as enthesopathy, aberrantly high levels of active TGF‐β can even deteriorate subchondral bone microstructure by disturbing MSC recruitment, indicating that active TGF‐β is critical for MSC recruitment during homeostatic osteogenesis and that its disruption contributes to pathological progression [16, 17, 32]. TGF‐β1 knockout mice exhibit reduced osteoblast numbers and trabecular bone density. However, TGF‐β1 increases bone formation by expanding the osteoprogenitor pool rather than by directly promoting osteogenesis in mesenchymal pluripotent cells [10].

In mammals, TGF‐β is secreted in an inactive form and deposited in the extracellular matrix, with broad expression in bone and cartilage [14]. At the early stage of fracture healing, TGF‐β may be produced by local cells, such as immune cells, and then activated through inflammatory degradation of LAP or the extracellular matrix [33]. Therefore, the high local concentration of TGF‐β within 3 days after fracture observed in the current study is not unexpected. The TGF‐β1 concentrations are elevated at specific time points during fracture healing in humans [9]. Consistently, we found elevated local TGF‐β1 levels and increased TGF‐β1/pSmad2/3 signaling after fracture of long bones, especially a rapid peak immediately after fracture surgery. Several studies have reported that increased levels of TGF‐β and other cytokines during fracture healing, which may be attributed to cytokines released and activated by bone resorption [34, 35, 36].

Macrophages are critical players with the ability to transition from inflammatory to anti‐inflammatory phenotypes after damage or infection. Macrophages are recruited to injured tissue and then undergo phenotypic and functional changes to coordinate injury inflammation and tissue repair after the tissue injury [37]. During fracture healing, macrophages migrate to the fracture site, and their depletion compromises the healing process [38, 39]. A novel finding of our research is the identification of macrophages, particularly early M1‐like macrophages, as a crucial cellular source of TGF‐β1 during the initial phase of fracture healing. This challenges the conventional paradigm that primarily attributes beneficial effects in tissue repair to M2 macrophages [20, 40, 41]. Our temporal analysis revealed that M1‐like macrophages dominate during the early inflammatory phase (day 3) and are the predominant source of TGF‐β1 at this critical time point. M1‐like macrophages not only participate in pro‐inflammatory responses but also synchronously prepare the microenvironment for osteogenesis by clearing debris and secreting related factors. This early TGF‐β1 production appears essential for initiating the downstream signaling cascade that drives MSC recruitment and osteogenic differentiation. Our findings align with emerging concepts in bone biology that emphasize the importance of immune‐bone crosstalk in skeletal repair [42, 43, 44]. The temporal polarization of macrophages from M1 to M2 phenotypes appears to orchestrate a coordinated healing response, with early M1 macrophages initiating repair through TGF‐β secretion [20, 45, 46].

In conclusion, our study elucidates a critical mechanism whereby macrophage‐derived TGF‐β1, particularly from early M1‐like macrophages, initiates the osteogenic program during fracture repair. These findings not only advance our fundamental understanding of bone regeneration but also suggest novel therapeutic strategies targeting the macrophage‐TGF‐β1 axis to enhance fracture healing.

Author Contributions

Conceptualization: Kaihang Song and Qi Sun. Data curation: Kaihang Song, Lingxuan Deng and Qiyun Chen. Formal analysis: Kaihang Song. Funding acquisition: Qi Sun. Investigation: Bo Wang. Methodology: Bo Wang. Project administration: Ming Cai. Software: Lingxuan Deng. Supervision: Ming Cai and Qi Sun. Validation: Kaihang Song, Bo Wang and Jing Chen. Visualization: Qi Sun. Writing – original draft: Kaihang Song. Writing – review and editing: Ming Cai and Qi Sun. All authors have read and agreed to the published version of the manuscript.

Funding

Funding for this work was provided by the National Natural Science Foundation of China (Grant nos. 82102602, 82272176).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Macrophages deletion did affects bone formation of bone fracture at late stage.

Data Availability Statement

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