Macrophage-Derived TGF-β and VEGF Promote the Progression of Trauma-Induced Heterotopic Ossification

Heterotopic ossification (HO) is a pathological bone formation process caused by musculoskeletal trauma. HO is characterized by aberrant endochondral ossification and angiogenesis. Our previous studies have indicated that macrophage inflammation is involved in traumatic HO formation. In this study, we found that macrophage infiltration and TGF-β signaling activation are presented in human HO. Depletion of macrophages effectively suppressed traumatic HO formation in a HO mice model, and macrophage depletion significantly inhibited the activation of TGF-β/Smad2/3 signaling. In addition, the TGF-β blockade created by a neutralizing antibody impeded ectopic bone formation in vivo. Notably, endochondral ossification and angiogenesis are attenuated following macrophage depletion or TGF-β inhibition. Furthermore, our observations on macrophage polarization revealed that M2 macrophages, rather than M1 macrophages, play a critical role in supporting HO development by enhancing the osteogenic and chondrogenic differentiation of mesenchymal stem cells. Our findings on ectopic bone formation in HO patients and the mice model indicate that M2 macrophages are an important contributor for HO development, and that inhibition of M2 polarization or TGF-β activity may be a potential method of therapy for traumatic HO.


INTRODUCTION
Heterotopic ossification (HO) is the abnormal formation of mature bones in extraskeletal tissues. Traumatic HO is a frequent complication caused by trauma, such as fractures, severe burns, and total hip or knee arthroplasty [1]. Clinical treatment is now limited to nonsteroidal anti-inflammatory drugs (NSAIDs), radiation therapy, and surgical resection. However, performing an operation is risky because HO often reoccurs after surgical 203 Macrophages Promote Traumatic Heterotopic Ossification excision. Under certain special circumstances, the HO tissues may be nonresectable due to their sensitive locations [2]. HO lesions are preceded by inflammatory events, including innate immune response and adaptive immunity [3,4]. It is still not clear how inflammatory cells affect HO development. Our recent studies have implicated that macrophages play a critical role in the progression of HO [5,6]. However, in vivo studies on the contribution made by macrophages for HO progression and its underlying mechanisms involved in HO are limited.
Tissue injury leads to an inflammatory response that recruits a variety of cells, such as monocytes and lymphocytes, at different stages [7]. Previous research studies have indicated that the aberrant expression of monocytes/macrophages is essential for wound healing, including that of HO [8,9]. Studies also have demonstrated the participation of macrophages in HO caused by endochondral ossification [10], but have not identified the cell population that responds to HO progression. Macrophages are characterized by high functional plasticity in different environments [11]. They can be polarized into two opposing functional states, known as anti-inflammation (M1) macrophages and pro-inflammation (M2) macrophages [12]. M1 macrophages produce a variety of cytokines, such as IL-12, IFN-g, and TNF-a, which lead to a pro-inflammation pattern. On the contrary, M2 macrophages may promote tissue remodeling, enhance angiogenesis, and regulate immune responses by secreting anti-inflammatory cytokines [13].
M2 macrophages, which express the cell surface marker, CD206, and produce IL-10 and TGF-β, have been described as a reparative and anti-inflammatory factor in trauma [14]. Aberrant activation of TGF-β has been observed in many diseases. In the skeletal system, elevated TGF-β expression is coupled with bone resorption and remolding [15]. It is well known that TGF-β is an important regulator of chondrogenic differentiation, and that endochondral ossification is a common feature of HO [16]. In addition, osteogenesis is a metabolic process that requires angiogenesis. Recent studies have shown that abundant blood vessel formation is a common feature during the development of HO [17].
Given the role of TGF-β and VEGF during normal and ectopic bone development and the infiltration of macrophages throughout the formation of HO, we hypothesized that aberrant macrophage activation is a significant source of TGF-β and VEGF. We found that TGF-β and angiogenesis are enhanced in human and mice HO. The inhibition of macrophages and TGF-β activity effectively mitigates HO formation. In addition, our findings identified that M2 macrophages play a central role in regulating TGF-β and VEGF expression during HO development.

Patients and Specimens
This study was approved by the Ethics Committee of Shanghai Jiao Tong University Affiliated Sixth People's Hospital, and written informed consent was obtained from all patients or their legal guardians. Traumatic HO was identified using X-ray and CT radiography conducted on 24 patients (14 males and 10 females, previously healthy, age ranging from 24 to 67 years) who had previously suffered an elbow fracture and were treated using internal fixation, or revision after hip/knee arthroplasty (8 patients per group). The muscle tissues were used as baseline controls. Blood samples (5 ml per person) were collected 1 day before clinical surgery and blood samples obtained from 8 healthy individuals were used as baseline controls. All the samples were processed immediately to collect serum, which was then stored in a −80 °C freezer. The serum specimens were processed for ELISA.

Mice
Male 6-to 8-week-old BALB/c mice were housed under specific pathogen-free conditions at the Animal Experimental Center of Shanghai Sixth People's Hospital, and all experiments conducted were approved by the Animal Research Committee of Shanghai Jiao Tong University Affiliated Sixth People's Hospital. Traumainduced HO was created using a murine tenotomy model, as we previously described [18]. After anesthesia was induced using an intraperitoneal injection of 1% pentobarbital sodium, a skin incision that was 1 cm in length was made on the lateral aspect of the Achilles tendon to expose its full length. Then, the Achilles tendon was transected precisely at its midpoint using a surgical knife. For the sham operation, the incision was made through the skin without touching the Achilles tendon. The incised skin was closed using absorbable sutures. To assess the effect of clodronate liposomes on HO, the animals were randomly divided into 2 groups: control group (tenotomy surgery with PBS liposomes) and clodronate group (tenotomy with clodronate). The mice were intraperitoneally injected at doses of 1.4 mg/20 g body weight twice a week from the day of surgery. For the TGF-β blockage experiments, the mice were intraperitoneally injected with TGF-β antibodies (R&D Systems, Minneapolis, MN, 5 mg/kg body weight) daily twice a week before being sacrificed to be used for the micro CT analysis.

Histological Analysis
Mice were sacrificed using carbon dioxide (CO 2 ) inhalation for an indicated period to be used for histological observations. The ankles with Achilles tendons were dissected and fixed in 4% paraformaldehyde for 24 h. The HO tissues from patients were collected during surgical operation. Then, the tissues were decalcificated using 10% ethylenediaminetetracetic acid (EDTA) solution for 1 month. The decalcified tissues were processed using graded dehydration and were then embedded in paraffin. Five-micrometer thick histological sections were obtained using a microtome and were subsequently processed for staining.
For SOFG staining, the sections were stained with 0.1% Safranin-O and 0.02% Fast Green (Sigma-Aldrich, Oakville, ON, Canada) by following the manufacturer's instructions.

Micro-CT
The tenotomy mice were sacrificed at the durations indicated. The hind limbs of the mice in each group were fixed overnight in 4% paraformaldehyde and were analyzed using a high-resolution Micro-CT scanner Skyscan 1176 (Bruker, Kontich, Belgium). The parameter was set at a resolution of 18 μm and 70 kV voltage. The region of interest (ROI) was set as the entire tibia to ensure that all heterotopic bones were included within the ROI. Threedimensional images were reconstructed and obtained using NRecon software, and HO bone volumes were analyzed using CTAn software, as previously described [19].

Enzyme-Linked Immunosorbent Assay
The concentrations of TGF-β and VEGF in the serum were determined using the Quantikine ELISA Kit (R&D Systems, Minneapolis, MN), by following the manufacturer's instructions.

Cell Culture
Bone marrow-derived macrophages (BMDMs) were isolated from the bone marrow of mice at day 0, week 2, 4, and 8 post-surgery, as previously described [5]. In brief, both the femur and tibia of the mice were excised and the soft tissue was completely removed. Then, the bone marrow cells were flushed from the marrow cavity using a 26G needle. The harvested cells were used in the experiments described below.

Cell Sorting and Flow Cytometry Analysis
After filtration using RBC lysis and washing with 0.1% BSA in PBS, we counted the cells and incubated equal numbers of cells for 45 min at 4 °C with the primary antibody. For macrophage identification, we used F4/80, CD115, and CD11b antibodies. For M1 macrophage identification, we used F4/80 and CD86 antibodies. For M2 macrophage identification, F4/80 and CD206 antibodies (BioLegend, San Diego, CA, USA) were used. The stained cells were processed on a BD FACSCalibur flow cytometer and were analyzed using FlowJo software.

Cell Culture
The sorted cells were maintained in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37 °C in a humidified atmosphere containing 5% CO 2 . To prepare the conditioned medium (CM), the cells were grown to 80% confluence in 5-cm dishes containing DMEM/10% FBS. The medium was discarded, and the cells were further cultured in serum-free DMEM for 24 h. Then, the medium was collected, centrifuged at 1000 × g for 10 min, and filtered through 0.22-μm filters (Millipore, Billerica, MA). Macrophages Promote Traumatic Heterotopic Ossification

RNA Isolation and Real-Time PCR
MSC cells (purchased from the Chinese Academy of Sciences, Shanghai, China) were exposed to a macrophage conditioned medium for 48 h. Total RNA was isolated using TRIzol Reagent (Thermo Fisher Scientific), and cDNA was synthesized using the Mix-X miRNA First-Strand Synthesis Kit (TaKaRa Bio). Subsequently, realtime PCR was performed using SYBR Green Premix Ex Taq (Takara) to quantify the target gene mRNAs. All procedures were performed by following the manufacturer's instructions. The following primer sequences were used: GAPDH, 5′-ATG GGG AAG GTG AAG GTC G-3′ (forward) and 5′-GGG GTC ATT GAT GGC AAC AATA-3′ (reverse);

Statistical Analysis
GraphPad Prism 8 was used for all statistical analyses of the data obtained. The data are presented as mean ± standard deviation (SD). Comparisons between groups were performed using Student's t-test, while oneway ANOVA was used for comparisons between multiple groups. All experiments were performed in triplicate, and representative experiments are shown. Statistical significance was set at p < 0.05.

Macrophage Infiltration and Activation of TGF-β/Smad2/3 Increased in Human HO Tissues
The HO patients who had undergone elbow injury or knee/hip arthroplasty, and were identified using X-ray or CT imaging, were included in this study. The HO specimens were collected during elbow internal fixation, knee, or hip replacement revision surgeries. The HE staining showed thick cartilage layers adjacent to the bone tissues in the HO samples (Fig. 1A). To examine the effects of macrophages on the progression of HO, CD68 expression was detected using immunohistochemistry (IHC). We observed a significant infiltration of CD68 + macrophages into HO tissues during elbow, knee, and hip surgeries (Fig. 1B, C). Immunostaining showed that TGF-β expression was significantly higher in the HO samples than in the muscle (Fig. 1D, E). Additionally, a significant increase in the expression of p-Smad2/3 was observed in HO tissues (Fig. 1F, G). Finally, the concentration of TGF-β in HO patient serum was higher than that of the healthy control (Fig. 1H). Together, our data revealed that enhanced macrophage accumulation and TGF-β/ Smad2/3 signaling activation are involved in traumatic HO development.

Macrophage Depletion Prevents HO Progression in Mice
Clodronate liposomes have been used as an effective agent to deplete macrophages in various in vivo models [20]. To study the effect of liposomal clodronate on HO, we used a HO mice model with dissection at the midpoint in the Achilles. We treated the mice with intraperitoneal injections of clodronate liposome twice a week until they were euthanized. At week 4 and 8, the mice were sacrificed; tissue and blood samples were subsequently obtained and analyzed. We first examined HO formation, and the micro-CT results showed that treatment with clodronate liposomes almost entirely inhibited HO formation at week 4 and 8 ( Fig. 2A, B). The efficacy of the clodronate liposomes on the depletion of macrophages was confirmed using CD68 staining. Treatment with clodronate liposomes for 8 weeks eliminated CD68 + cells from the HO tissue (Fig. 2C, G). The Safranin O and fast green (SOFG) staining revealed that hardly any cartilage or bone could be observed following macrophage depletion (Fig. 2D). Additionally, immunostaining demonstrated that the expression of the osteogenic differentiation marker, osteocalcin (Ocn, Fig. 2E, H), had decreased significantly after clodronate liposome treatment. A similar result was observed with the chondrogenic differentiation marker, Sox9 (Fig. 2F, I). A monoclonal IgG was served as isotype control ( Supplementary  Fig. 1). Altogether, these data indicate that macrophage depletion prevents HO formation by inhibiting endochondral ossification.

Macrophage Depletion Reduces Angiogenesis in HO
It is well known that angiogenesis is essential for endochondral ossification [21,22]. We investigated the impact of clodronate liposomes on angiogenesis in HO. X-ray photography confirmed that clodronate liposomes significantly inhibited HO formation at week 2, 4, and 8 (Fig. 3A). HE staining demonstrated that blood vessels were present in the control HO group (treated with PBS liposomes, week 8), whereas fewer blood vessels were observed in the clodronate liposome treated groups (Fig. 3B). These results were confirmed using CD31 immunohistochemistry staining. Compared with the control HO group, microvessel density was significantly lower following clodronate liposomal treatment (Fig. 3C, E). Macrophages are a major source of secreted vascular endothelial growth factor (VEGF) that can mediate angiogenesis [23]. Therefore, we determined VEGF expression in HO tissues and confirmed that the reduction in blood vessel density was associated with a significant decrease in VEGF levels upon treatment with clodronate liposomes ( (Fig. 3D, F). Furthermore, a marked reduction in circulating VEGF levels was observed after treatment with clodronate liposomes (Fig. 3G). To examine whether clodronate liposome directly affect VEGF expression in macrophages, bone marrow-derived macrophages (BMDMs) were isolated from the bone marrow and were treated with clodronate liposomes for  2 h. The VEGF expression levels decreased following clodronate liposome treatment (Fig. 3H). A monoclonal IgG was served as isotype control ( Supplementary  Fig. 2). Collectively, these results indicate that depletion of macrophages attenuates angiogenesis and HO formation.

Macrophage Depletion Suppresses TGF-β Signaling
To investigate the role played by TGF-β in the progression of HO, we examined TGF-β expression using the traumatic HO mice model. Ectopic bone was formed  Macrophage depletion suppresses TGF-β signaling. A The mice were subjected to a tenotomy to generate a traumatic HO model. Representative X-ray images of HO in the mice at 0 day, 2 weeks, 4 weeks, and 8 weeks. B Immunohistochemical staining of TGF-β. C Immunohistochemical staining of p-Smad2/3. D Quantification of TGF-β expression at the times indicated. E Serum TGF-β levels in the mice were detected by ELISA. F Quantification of the p-Smad2/3 + cells at the times indicated. G Immunohistochemical staining of TGF-β in HO mice treated with PBS liposomes or clodronate liposomes at week 8. H Immunohistochemical staining for p-Smad2/3 in the two groups at week 8. I Quantification of TGF-β expression in the two groups. J Serum TGF-β levels in the two groups at week 8. K Quantification of p-Smad2/3 expression. n = 5 per group. Data are presented as mean ± SD. *, p < 0.01. Scale bar, 50 μm.
in the Achilles tendon of the mouse model 4 weeks post-puncture (Fig. 4A). The immunohistochemistry staining showed that the expression of TGF-β increased 2 weeks post-puncture and was maintained at a high level at 8 weeks post-puncture (Fig. 4B, D). Meanwhile, the serum TGF-β concentration in the mice increased 2 weeks after trauma (Fig. 4E). In addition, a similar result was observed with p-Smad2/3 staining of the HO tissues (Fig. 4C, F). The HO mice were treated by administrating clodronate liposome injection. Our data demonstrated that TGF-β levels in the tissue (Fig. 4G, I) and serum (Fig. 4J) were significantly inhibited upon treatment with clodronate liposomes at week 8. Finally, the number of pSmad2/3 + cells decreased following macrophage depletion (Fig. 4H, K). Taken together, the results demonstrate that macrophage depletion effectively suppressed the activation of TGF-β signaling, which plays an important role in triggering HO formation.

TGF-β Antibody Inhibits the Formation of Trauma-Induced HO
We investigated whether TGF-β is directly involved in traumatic HO progression. A TGF-β neutralizing antibody or IgG (used as control) was injected into the mice twice a week. To examine the change in HO signaling, the mice were sacrificed at week 2, 4, and 8. μ-CT analysis showed that HO was not formed following TGF-β antibody treatment at week 2 and 4. However, HO was formed at week 8. We observed a significantly reduced HO volume after administration of the TGF-β neutralizing antibody injection, compared with the controls (Fig. 5A, G). SOFG staining demonstrated that endochondral ossification formation was reduced following TGF-β antibody treatment (Fig. 5B). Immunostaining revealed that the expression of VEGF was downregulated (Fig. 5C, H). And a decreased number of pSmad2/3 positive cells was observed after the HO mice were injected with a TGF-β antibody (Fig. 5D, I). The immunostaining of CD31 showed that the number of blood vessels had decreased significantly in TGF-β antibody treatment mice (Fig. 5E, J). In addition, we observed a similar change in Ocn expression in the mice (Fig. 5F, K). A monoclonal IgG was served as isotype control (Supplementary Fig. 3). Collectively, our data demonstrates that TGF-β signaling inhibition attenuates the progression of HO.

M2 Macrophages Promote Osteogenic and Chondrogenic Differentiation
To determine the type of macrophage involved in HO progression, an Achilles tendon mice model was used and bone marrow-derived macrophages were collected at week 0, 2, 4, and 8. M1 and M2 macrophages from the BMDMs were examined using flow cytometry analysis. In the gating strategy used to identify macrophages, a nucleated cell gate was set based on FSC-H and SSC-H to exclude debris. Next, the pan-macrophage marker F4/80 was used and live macrophages were identified and used in the following experiments ( Supplementary Fig. 4A-C). The M1 and M2 macrophages were identified based on CD86 and CD206, respectively. The results showed that the percentage of M2 macrophages increased significantly after trauma, whereas the percentage of M1 macrophages was maintained at a stable level during the whole process (Fig. 6A, B, and C). The immunohistochemistry staining showed that the numbers of CD206 positive cells in the tissues were increased after 2 week and peaked at week 4, then slightly declined at week 8 (Fig. 6D, E). These results indicate that M2 macrophages may be responsible for HO development. Then, M1 and M2 macrophages were sorted and isolated using FASC. The 4-and 8-week M1 and M2 cells were cultured in DMEM medium for 24 h and the conditioned medium was collected. Then, mesenchymal stem cells (MSCs) were incubated in the M1 or M2 conditioned medium for another 48 h; the expression levels of osteogenic and chondrogenic differentiation makers were examined using real-time PCR. The 4-week and 8-week M2-conditioned medium enhanced the expression of level of the osteogenic markers Runx2, Sp7, and OCN, and the chondrogenic marker Sox9. However, the M1-conditioned medium did not affect the expression levels of these genes (Fig. 6F). Similar results were observed in the expressions of TGF-β receptor 1 (TGFBR1) and VEGF receptor 1 (VEGFR1) (Fig. 6G). Taken together, these results suggest that M2 macrophages promote endochondral ossification during HO progression.

DISCUSSION
It has been found that the immune system and macrophages are closely related to the HO formation. However, we still have limited knowledge about the mechanisms that link them. The recruitment of circulatory  Previously, we demonstrated that macrophages are involved in HO progression. In this study, we identified M2 macrophage and TGF-β inhibition as a promising therapeutic method for traumatic HO prevention in a preclinical mice model.
Previous studies have provided evidence that macrophages are an important regulator of HO formation [9,24]. Depletion of macrophages is shown to significantly impair the development of genetic and traumainduced HO [8,25], while it has been shown to exert a promotional effect on HO in other studies [26]. To better understand the role of macrophages during the initiation and development of HO, we analyzed HO tissues from different injury sites of clinical patients. We found that an increase in macrophage numbers was accompanied traumatic HO development, as observed in patients. Our observations provide clinical insights into the contribution made by activated macrophages for the pathophysiological regulation of ectopic bone formation. In addition, we found that the ablation of macrophages caused by clodronate liposomes significantly suppressed the development of HO in vivo. Meanwhile, our data showed that the depletion of macrophages inhibited the endochondral ossification and the expression of the Ocn and Sox9 genes.
Monocyte-macrophage lineage cells are characterized by the diverse functions they exert, which play a crucial role in regulating normal tissue homeostasis, including bone development and repair, wound recovery, and the regulation of angiogenesis [27][28][29]. Clodronate liposomes have been used often for macrophage depletion in several disease models, including cancer, autoimmune disorders, and diseases of the skeletal system [30,31]. Although it is well established that clodronate liposomes can effectively inhibit macrophages in vitro and in vivo, its potential effects on other cell types cannot be ruled out. In traumatic HO, the phenotypic heterogeneity of macrophages remains poorly understood. Our findings showed that the number of M2 macrophages increased during HO development, and provided evidence that M2 macrophages could produce a large amount of VEGF and TGF-β. These data demonstrate that the aberrant activation of M2 macrophages may be an important contributor to traumatic HO.
HO development is an energy-consuming metabolic process that requires blood vessels to deliver nutrients for chondrogenesis and osteogenesis [32]. Angiogenesis and osteogenic processes are tightly coupled during skeletal development [33,34]. Recent research studies have shown that human HO exhibits a consistent pattern of vascularization [35], which involves coordination between osteogenesis and angiogenesis [36]. Our data demonstrated that neovascularization is inhibited following macrophage depletion, indicating that the inhibition of neovascularization may attenuate traumatic HO. It is well known that VEGF plays a critical role in angiogenesis and chondrogenesis [37,38], a process that is essential for HO formation as it occurs primarily through endochondral ossification [39,40]. In this study, we found that VEGF levels in HO and the serum decreased after clodronate liposome treatment. These results suggest that macrophages promote neovascularization and facilitate HO formation.
TGF-β is considered to be a specific marker of regenerative macrophages and can regulate the biological functions of monocytes and macrophages [41]. It is known that TGF-β also plays a critical role in chondrogenic differentiation [42,43]. Several recent studies have implicated TGF-β signaling in genetic and neurogenic HO [44,45]. Our findings demonstrated that TGF-β present at the injury sites appears to play a critical role in aberrant bone formation. Blockage of active TGF-β by a specific antibody significantly inhibited HO development in vivo. Depletion of macrophages effectively reduced the expression of TGF-β. Our data indicates that TGF-β producing macrophages form a critical link between inflammation and aberrant ectopic bone generation.
Our previous study showed that quercetin, a natural flavonoid class of polyphenols, attenuated traumainduced HO by restraining monocyte-to-macrophage transition, indicating that macrophages infiltration plays a pivotal role in HO [5]. In this study, our findings revealed the critical role played by M2 macrophages in the regulation of aberrant ectopic bone formation by secreting excessive amounts of TGF-β and VEGF. Using clinical sample analysis, we found a remarkable level of macrophage infiltration and TGF-β activation in human HO tissues. Furthermore, we showed that TGF-β and VEGF derived from macrophages play an important role in driving traumatic HO formation through endochondral ossification. In addition, we identified M2 macrophages as an important source of TGF-β and VEGF following trauma. Our study proposes a paradigm that can be used to understand the functional impact of macrophage polarization on HO progression.

DATA AVAILABILITY
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

DECLARATIONS
Ethics Approval All experiments were approved by the Ethics Committee of Shanghai Jiao Tong University Affiliated Sixth People's Hospital.