SEARCH

SEARCH BY CITATION

Keywords:

  • MUSCLE RESIDENT STROMAL CELLS;
  • SCA1;
  • MUSCLE DAMAGE;
  • µCT SCAN;
  • ALK1

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Heterotopic ossification (HO) is defined as the formation of bone inside soft tissue. Symptoms include joint stiffness, swelling, and pain. Apart from the inherited form, the common traumatic form generally occurs at sites of injury in damaged muscles and is often associated with brain injury. We investigated bone morphogenetic protein 9 (BMP-9), which possesses a strong osteoinductive capacity, for its involvement in muscle HO physiopathology. We found that BMP-9 had an osteoinductive influence on mouse muscle resident stromal cells by increasing their alkaline phosphatase activity and bone-specific marker expression. Interestingly, BMP-9 induced HO only in damaged muscle, whereas BMP-2 promoted HO in skeletal muscle regardless of its state. The addition of the soluble form of the ALK1 protein (the BMP-9 receptor) significantly inhibited the osteoinductive potential of BMP-9 in cells and HO in damaged muscles. BMP-9 thus should be considered a candidate for involvement in HO physiopathology, with its activity depending on the skeletal muscle microenvironment. © 2011 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Heterotopic ossification (HO) is defined as the formation of bone inside soft tissue structures where bone normally does not exist. HO may present with clinical manifestations including increased joint stiffness, limited range of motion, swelling, and pain and even can result in severe functional limitations.1 Several conditions may lead to soft tissue ossification, and these include fibrodysplasia ossificans progressiva and traumatic myositis ossificans. The first is a rare congenital disorder with a frequency of 1 per 2 million2 that is caused by a recently discovered recurrent missense mutation in the BMP type 1 receptor activin receptor 1A/activin-like kinase-2 (ACVR1/ALK2).3 Sporadic mutations and an autosomal dominant pattern of inheritance are responsible for the disease, in which ossification can occur with or without injury and typically grows in a predictable pattern.3

The second and by far most common type of traumatic HO occurs in injured tissues, notably at sites of damaged muscle.4 Many traumatic conditions have been associated with traumatic HO, including fractures, burns, immobilization, total hip replacement surgeries, and spinal cord and brain injuries.5–7 The diversity of these conditions suggests that factors released by distant tissues such as the brain may have a direct osteoinductive influence on soft tissues or may stimulate local cells to produce osteogenic growth factors.

In the early 1970 s, Urist and colleagues reported that the pathogenesis of HO was related to the transformation of primitive mesenchymal cells of soft tissues into osteogenic cells.8, 9 A few years later, Chalmers and colleagues proposed that three essential conditions were required to induce HO: (1) osteogenic precursor cells, (2) inducing agents, and (3) a permissive environment.10 Among the growth factors with osteoinductive potential, the most important are bone morphogenetic proteins (BMPs), which belong to the transforming growth factor β (TGF-β) superfamily. BMPs play an important role in stem cell biology by regulating cell proliferation and differentiation during development.11 BMPs are also involved in the regulation of osteoblast differentiation and bone formation. In addition, genetic disruption of BMPs can result in skeletal and extraskeletal abnormalities.12

More than 20 members of the BMP family have been identified to date, including BMP-2, -4, -6, and -7, that have osteogenic properties.13 A new BMP (BMP-9) has been identified recently and has been shown to be the most potent BMP.13 BMPs signal through serine/threonine kinase receptors composed of type 1 and 2 subtypes. When activated, the receptors phophorylate R-Smad-1, -5, and -8, which are able to recruit co-Smad-4. The R-Smad/co-Smad complex then is able to enter the nucleus and regulate target-gene transcription.14 BMP-9 is the physiologic ligand of the endothelial type I ALK1 receptor in association with the type II BMP receptor (BMPRII).15 ALK1 is one of the seven type I receptors for TGF-β family members.16 The activation of ALK1 induces the phosphorylation of receptor-regulated Smad-1 and Smad-5 and their accumulation in the nucleus.17

Two classes of progenitor cells have been identified in skeletal muscle. The first source, myogenic stem cells, or satellite cells, resides beneath the basal lamina of myofibers and has the potential to repair skeletal muscle throughout life.18–20 The second source, muscle-resident mesenchymal or stromal cells (mrSCs), is located around blood vessels in the connective tissue compartment and is closely associated with satellite cells.21–23 In culture, mrSCs are multipotent and can differentiate into many mesodermal cell lineages.21, 23–26 One important characteristic of mrSCs is their high immunoreactivity with anti-Sca1 antibody. This peculiarity also has been used extensively, alone or in combination with other cell surface antigens, to identify and/or enrich mrSCs.21, 23, 26 In addition, the mrSC population expands during muscle degeneration/regeneration and contributes to muscle regeneration.21, 23, 25, 26 To engage the regeneration process, mrSCs are guided by the changing microenvironment. As hypothesized by Walker and colleagues, the adult stem cell microenvironment is maintained by physical contact and a defined set of diffusible factors.27 When this microenvironment changes, as is the case with damaged muscle, mrSCs can proliferate, differentiate, and actively participate in the regeneration process. As such, mrSCs have to be considered as potent cells involved in HO pathophysiology.

Because BMP-9 has a strong osteogenic potential, we hypothesized that it may play a role in HO. We show for the first time that BMP-9 can efficiently induce the osteogenic program of mrSCs. Interestingly, injecting BMP-9 directly into muscle does not induce ossification unless the muscle is damaged, indicating that cells in regenerating tissue are more permissive to environmental cues than cells in resting tissue. This study establishes for the first time that there is a potential link between BMP-9 and HO and that BMP-9 can target the appropriate cell population and signaling pathway that lead to muscle HO.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Animals

Twelve-week-old male C57Bl/6 mice (Charles River, Saint-Constant, Quebec, Canada) were used. All procedures involving animals met the standards set out in Guide for the Care and Use of Laboratory Animals from the Institute for Laboratory Animal Research (ILAR, 1996) and were approved by the Institutional Animal Care and Use Committee of Université de Sherbrooke (Protocol No. 141-07).

Once the animals were anesthetized, the tibialis anterior (TA) muscles were immediately collected, weighed, and processed for histologic examination or were flash frozen for molecular analysis. Flash-frozen muscles were mechanically crushed in liquid nitrogen using a mortar and pestle and stored at −80°C until used to extract RNA and proteins.

Isolation and culture of primary muscle resident stromal cell

mrSCs were isolated and cultured as described previously.18, 21, 24 Briefly, gastrocnemius muscle was minced and digested with collagenase I (Sigma-Aldrich, Mississauga, Ontario, Canada) for 45 minutes at 37°C. The tissue slurry was washed with DMEM containing 10% fetal bovine serum (FBS; Hyclone, Nepean, Ontario, Canada) and poured through a 100-µm and then a 50-µm cell strainer (BD Falcon, Mississauga, Ontario, Canada). Cells then were plated in uncoated tissue culture dishes (BD Falcon). For cell expansion, cells were cultured in DMEM supplemented with 10% FBS and antibiotics at 37°C with 5% CO2. mrSC cells were passaged at 80% confluence and were used prior to passage two. To induce the osteogenic program, the mrSCs were transferred to differentiation medium (DMEM containing 5% horse serum) with or without BMP factors (BMP-2 3.9 nM or BMP-9 3.9 nM) and incubated for 6 days. The differentiation medium was replaced every 2 days.

To inhibit BMP-9, differentiation medium containing BMP-9 was supplemented with 19.5 or 78 nM of soluble ALK1-Fc receptor, which contains the extracellular domain of ALK1 fused to the Fc portion of a mouse IgG immunoglobulin. Recombinant proteins were purchased from R&D Systems (Minneapolis, MN, USA). Cells differentiated for 6 days in the presence of BMP-2, BMP-9, or a combination of BMP-9 and ALK1-Fc and were further processed for alkaline phosphatase (AlkP) activity staining or quantitative polymerase chain reaction (qPCR) analysis.

Induction of HO in skeletal muscle

HO was assessed in damaged and undamaged TA muscles. To induce damage, 40 µL of 10 µM cardiotoxin (CTX; Latoxan, Valence, France) was injected into the TA muscles, as described previously.18, 21, 28 Contralateral TA muscles injected with 40 µL of 0.9% saline solution were used as controls. Alternatively, the TA muscles were damaged by crushing with surgical tweezers. Three days after the injury, 50 µL of 2.9 µM BMP-2, 2.9 µM BMP-9, or a combination of 2.9 µM BMP-9 and 43.5 µM soluble ALK1-Fc receptor was injected into the TA muscles. The muscles were harvested 7 and 14 days after the BMP injections and were flash frozen for molecular analysis or were fixed in 95% EtOH for in situ alcian blue/alizarin red S staining or micro–computed tomographic (µCT) analysis.

Crushed muscle extract

Crushed muscle extracts were prepared according to Tatsumi and colleagues.29 Undamaged or CTX-damaged muscle were harvested after 3 days, macerated for 2 hours on ice in Tris-buffered saline (TBS; 50 mM Tris, 0.9% NaCl, pH 7.6) at 1 g of muscle per milliliter. Resulting extracts then were centrifuged at 500g for 10 minutes at 4°C to remove cell and tissue debris. Crushed muscle extracts were used to stimulate cells at a dilution of 1:10 in 10% in DMEM (v/v).

Intramuscular transplantation of mrSCs expressing nuclear β-galactosidase (nLacZ)

Lentivirus was produced by transfecting 293T cells seeded 24 hours before at 8 × 104 cells/cm2 with Lipofectamine 2000 (Invitrogen, Burlington, Ontario, Canada) according to the manufacturer's instructions. Briefly, 15 µg of transfer vector pSMPUW-MNDnLacZ (Cell Biolabs, San Diego, CA, USA) or pLenti CMV/TO GFP Puro (w159-1) (gift of Dr É Campeau, University of Massachusetts Medical School, Worcester, MA, USA),30 15 µg of pLP1, 6 µg of pLP2, and 3 µg of pLP/VSVG (Invitrogen) were mixed with Lipofectamine 2000. 293T cells were incubated overnight with the DNA-liposome complex in a final volume of 10 mL of Opti-MEM (Invitrogen). The day after transfection, the medium was changed for 6 mL of fresh Opti-MEM. Viral supernatants were collected at 48 and 72 hours after transfection and filtered through a 0.2-µM syringe filter and stored at −80°C.

Subconfluent mrSCs in culture were infected with a multiplicity of infection of 0.6. Forty-eight hours after infection, cells were trypsinized and injected (1 × 105 cells, final volume of 50 µL containing 2.9 µM of BMP-9) in a TA muscle that was injured with CTX 3 days prior. Fourteen days after injections (17 days after CTX treatment), muscles were harvested and fixed in 4% paraformaldehyde (PFA) for 2 hours at 4°C.

X-Gal staining

Whole muscles were stained in 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) solution as described previously.21, 31 X-Gal-stained muscles were decalcified and then embedded in optimal cutting temperature compound (OCT; Sakura, Torrance, CA, USA). Cryosections were stained with hematoxylin and eosine, but the alkaline water step was omitted to prevent blue staining of the nuclei by hematoxylin, which can be mistaken for nLacZ+ cells.

Alkaline phosphatase staining

Cells were fixed in a citrate buffer containing 60% acetone, immersed in a fast violet B diazonium salt and naphthol AS-MX phosphate alkaline buffer solution for 30 minutes, as recommended by the manufacturer (Leukocyte Alkaline Phosphatase Kit, Sigma-Aldrich), and then washed gently with PBS. To quantify alkaline phosphatase activity, the pixel densities of the images were measured using a monochrome conversion of the images and a homemade script in ImageJ (Version 1.42q; National Institutes of Health, Bethesda, MD, USA).32

Flow cytometric analysis and cell sorting

mrSCs are adherent cells with a fibroblastic morphology that express the cell surface marker stem cell antigen (Sca1+) but not endothelial and hematopoietic cell surface markers [CD31 and lineage (Lin)]21, 24, 26 (Supplemental Fig. S1). Thus mrSCs also can be analyzed or highly enriched directly from digested skeletal muscles using flow-activated cell sorting (FACS). Cells from digested muscle were incubated on ice for 20 minutes with 1 µg of antibody per 1 × 106 cells: APC-Sca1 (eBiosciences, San Diego, CA, USA), fluorescein isothiocyanate (FITC)–CD31, and phycoerythrin (PE)–Lin. Analysis and/or sorting were performed using a MoFlo cytometer (DakoCytomation, Glostrup, Denmark). After sorting, the Sca1+CD31Lin (Sca1+CL) cells were flash frozen for future RNA extraction or used in cell culture experiments.

Quantitative PCR

Total RNA was extracted from flash-frozen crushed muscles or cultured cells using TRIzol (Invitrogen). The RNA (1 µg) was reverse transcribed using Reverse Transcriptase Superscript II (Invitrogen). qPCR was performed using 50 ng of cDNA under the following conditions: a 5-minute denaturation step at 95°C followed by 40 cycles of 40 seconds at 95°C, 40 seconds at 56°C, and 40 seconds at 72°C. qPCR assays were performed on a Rotor-Gene 6000 (Corbett Robotics, Eight Mile Plains, Australia) using iQSYBR Green Supermix (BioRad Laboratories, Mississauga, Ontario, Canada). The results were analyzed using the 2−ΔΔCT relative quantification method normalized to GAPDH. The primer sets are listed in Table 1.

Table 1. Primer Sets Used for qPCR
 ForwardReverse
Alk1CTCCCAAGGCTCACTTTCTGCAGAAAGTGAGCCTTGGGAG
Alk2CCTGGAAGTTGGCCTTATCAGTGGTGATGAGCCCTTCAAT
Alk3CTTCTCCAGCTGCTTTTGCTATAGCGGCCTTTACCAACCT
Alk5GGCGAAGGCATTACAGTGTTTGCACATACAAATGGCCTGT
Alk6sAGCTGGTTCCGAGAGACTGACAGCATGGACTTTGCGTCTA
AlkPACACCTTGACTGTGGTTACGTCCTTGTAGCCAGGCCCGTTA
BMP2CTCCTCCCCCTGCTCGCTGTTGGGGCTCGGAGATGGCGAA
BMP4TGATACCTGAGACCGGGAAGAGCCGGTAAAGATCCCTCAT
BMP6GAGGATGGAGATGGGACTCATTCTTCAAGGTGAGCGAGGT
BMP7GAAAACAGCAGCAGTGACCAGGTGGCGTTCATGTAGGAGT
BMP9AACCTGGAGGGACACTGATGGGCTGAGCTCCGACTCTATG
GapdhGTGTCCGTCGTGGATCTGACCACCACCCTGTTGCTGTAG
OsterixCCCCTGGCCATGCTGACTGCAGGGAGCTGGGTAGGCGTCC
OsteocalcinGAGGGCAATAAGGTAGTGAACAGAAAGCCATACTGGTCTGATAGCTCG
Runx2CCCAGCCACCTTTACCTACATATGGAGTGCTGCTGGTCTG

Western blot analysis

Cells or tissues were lysed in RIPA buffer (0.5% NP-40, 0.1% SDS, 150 mM NaCl, 50 mM Tris-HCl, pH 7.5). Proteins were separated by SDS-PAGE, transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA), and probed with anti-ALK1 (1:800, H150; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) and anti-GAPDH (1:1000, FL-335; Santa-Cruz) antibodies. Blots were incubated for 1 hour at room temperature with a peroxidase-conjugated anti-rabbit secondary antibody (1:5000; Amersham Health, Oakville, Ontario, Canada), and bands were revealed with ECL according to the manufacturer's instructions on a BioMax ML film (Kodak, Rochester, NY, USA). The autoradiograms were digitized and the bands were quantified by densitometric measurements using ImageJ.32

µCT analysis

TA image analysis was performed using a SkyScan 1072 µCT scanner (energy 20 to 100 kV, current 0 to 250 µA, resolution 4.5 to 20 µm; detector 1024 × 1024 pixels, 12-bit, digital cooled CCD camera; Skyscan, Aartselaar, Belgium) at the Centre for Bone and Periodontal Research (Montreal, Quebec, Canada).

In situ staining

HO in TA muscles was visualized in situ using the alizarin red–alcian blue staining protocol described by Depew.33 Briefly, TA muscles were fixed in 95% EtOH (>20× tissue volume) for at least 5 days, immersed in acetone (>20× tissue volume) for 2 days, and then transferred into a staining solution containing 1 volume of 0.3% alcian blue 8GS in 70% EtOH, 1 volume of 0.1% alizarin red S in 95% EtOH, 1 volume of glacial acetic acid, and 17 volumes of 70% EtOH. After 3 days, the TA muscles were washed in distilled water and clarified in 1% KOH containing graded concentrations of glycerol (20% to 100%) over 3 weeks.

Histology and immunofluorescence

TA muscles were excised, fixed in formalin, and embedded in paraffin. Sections (5 µm) were stained with Masson trichrome (MT). Ossification was evaluated using a TE-2000-S microscope (Nikon, Mississauga, Ontario, Canada). For immunofluorescence analysis, frozen sections (7 µm) were fixed in 100% methanol (10 minutes at −20°C), blocked in PBS containing 10% goat serum, 1% bovine serum albumin (BSA), and 0.2% Triton X-100 and then incubated with rabbit polyclonal anti-ALK1 antibody (1:500, H-150; Santa Cruz), rat monoclonal anti-laminin-2 (1:600, 4H8-2; Sigma), and FITC-conjugated rat monoclonal anti-Sca1 antibody (1:400, D7; BD Bioscience, Mississauga, Ontario, Canada) primary antibodies. After several rinses in PBS-Tween, the sections were incubated with Alexa Fluor 594–conjugated goat anti-rabbit IgG secondary antibody (1:1000; Invitrogen) and Alexa Fluor 647–conjugated goat anti-rat IgG secondary antibody (1:1000; Invitrogen). Primary antibodies were omitted as a control. Cell nuclei were labeled with 4,6-diamidino-2-phenylindole (DAPI) reagent (Sigma-Aldrich). Indirect immunofluorescence was examined without counterstaining using an Axioskop 2 phase-contrast/epifluorescence microscope (Carl Zeiss, Inc., Thornwood, NY, USA). Photomicrographs were processed using Image Pro software (Media Cybernetics, Silver Springs, MD, USA).

Statistical analysis

qPCR expression values from the primary mrSC and CTX injury experiments were stated as means ± SEM. Statistical significance was determined using an ANOVA test, with the correction of Bonferroni for multiple comparisons. The power analysis was made with nQuery Advisor software (Version 4.0; Statistical Solutions, Saugus, ME, USA) and was 99% with a significance level of 0.05. GraphPad Prism 5.0 software (GraphPad Software Inc., LaJolla, CA, USA) was used for all statistical analyses.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

BMP-9 induced the expression of osteogenic differentiation markers in muscle resident stromal cells

We investigated whether the skeletal mrSC population was involved in HO in association with BMP-9. mrSCs were isolated, cultured, and then stimulated or not with BMP-9 to determine whether they were capable of activating the osteogenic program. As a positive control, we used BMP-2, which is known to induce osteogenic differentiation in vitro.34 We used 3.9 nM BMP-9 and 3.9 nM BMP-2 to compare their osteogenic activities. After a 6-day treatment, we used qPCR to measure the expression of Runx2 and osterix mRNA. Runx2 and osterix are early and late osteoblast differentiation transcription factors, respectively35, 36 (Fig. 1A). BMP-2- and BMP-9-stimulated cells expressed significantly higher levels of these transcription factors than untreated cells (Fig. 1A). Moreover, Runx2 mRNA expression was significantly higher in BMP-9-treated cells than in BMP-2-treated cells (Fig. 1A). To make sure that this increased expression was related to the activation of a bone differentiation program, we measured AlkP mRNA in the same cells (Fig. 1B). As predicted, AlkP transcripts also were higher in BMP-2- and BMP-9-treated cells, with significantly higher levels in the BMP-9-treated cells (Fig. 1B). We also measured AlkP activity by exposing the cells to a naphthol substrate. The BMP-2- and BMP-9-treated cells displayed more intense staining (fast violet) than untreated cells (Fig. 1C). As was the case with Runx2, BMP-9 was significantly more effective in inducing AlkP activity than BMP-2 (Fig. 1C).

thumbnail image

Figure 1. Muscle-resident stromal cells treated with BMP-9 upregulated the expression of specific bone markers. mrSCs were isolated and cultured in vitro and then treated for 6 days with 3.9 nM BMP-2 or 3.9 nM BMP-9 (n = 6). (A) Graph showing the expression of Runx2 and osterix mRNA in mrSCs as measured by qPCR. Note that both BMP-2 and BMP-9 significantly increased the expression of the two genes. *p < .05; **p < .005; ***p < .001. (B) Graph showing the expression levels of AlkP mRNA in mrSCs as measured by qPCR (n = 6). *p < .05; **p < .005; ***p < .001. (C) Representative image showing AlkP activity staining for 6 days of control, BMP-2-, and BMP-9-treated cells (n = 3). There was a robust increase in AlkP activity in BMP-9-treated cells, as shown by the intense staining.

Download figure to PowerPoint

We also verified whether cell confluence, a stage where differentiation ensues, potentiates the effect of BMP-9 on mrSCs (Supplemental Fig. S2). Our results show variation in basal expression levels of Runx2, AlkP, and osterix between proliferative and highly confluent mrSCs prior to BMP-9 treatment. Highly confluent mrSCs have a higher expression of Runx2 but lower expression of AlkP and osterix. These variations are slight but significant. Treatment of mrSCs with BMP-9 increased mRNA expression of Runx2, AlkP, and osterix in both confluent and proliferating cultures. Significantly higher Runx2 levels were detected in confluent compared with proliferating cultures after BMP-9 treatment. There were no significant differences in the induction of AlkP and osterix in proliferating and confluent cultures. Together, these results suggest that the state of confluency is not a factor that drastically influences differentiation of mrSCs whether they are treated or not with BMP-9 (Supplemental Fig. S2). Because mrSC is a cell population enriched under the tissue culture conditions used here, and because it was shown previously to be related to the adherent fraction of FACS-sorted S+CL cells,21, 24, 26 we verified whether S+CL also could initiate an osteogenic program. Our results show that similar to mrSCs, cultured FACS-sorted S+CL cells display high AlkP activity staining following BMP-9 stimulation compared with unstimulated S+CL cells (Supplemental Fig. S3). Interestingly, AlkP activity was more strongly induced by BMP-9 in S+CL cells isolated from cardiotoxin-injured muscle than in noninjured control muscles. Hence cultured mrSCs are highly related to the FACS-sorted S+CL cell population.

Damaged skeletal muscle has the potential to induce HO

To verify the potential of mrSCs in damaged muscle to commit to and differentiate into osteoblasts, we measured the expression of the bone transcription factor Runx2, a major transcription factor involved in the induction of bone differentiation,36, 37 in damaged muscle using qPCR. Although no HO was observed in CTX-injured skeletal tissue, Runx2 levels were higher 3 and 10 days after injury than in uninjured muscle (Fig. 2A). We also measured the expression of ALK1, the receptor for BMP-9. ALK1 levels were significantly higher in damaged muscles 3 and 10 days after injury than in uninjured contralateral muscles (Fig. 2B). To confirm the increase in ALK1 mRNA transcripts, we performed a Western blot using undamaged and injured muscle extracts. The expression of ALK1 protein was significantly higher in injured muscle than in uninjured muscle (Fig. 2C). The increase in the expression of ALK1 and Runx2 suggested that damage to muscles alters the microenvironment of mrSCs, which, in turn, may make them permissive to HO. We also assessed whether other known osteogenic BMPs and their receptors were influenced by the muscle-damage microenvironment. We screened the mRNA expression of BMP2, -4, -6, -7, and -9, as well as for their receptors ALK2, -3, -5, and -6. We found that only BMP-2 and -6 were significantly increased at 3 days after damage, whereas at 10 days, all studied BMPs were increased except for BMP-7. For ALK receptors, ALK2, -5, and -6 were upregulated 3 days after muscle damage. However, only ALK2 and -5 remained upregulated (Supplemental Fig. S4).

thumbnail image

Figure 2. Damaged skeletal muscle is permissive to HO. Cardiotoxin (CTX)–damaged TA muscles were harvested and analyzed 3 and 10 days after CTX treatment. Saline was injected into the contralateral TA muscles as a control. (A) Graph showing the expression of Runx2 mRNA as measured by qPCR in CTX-damaged and saline-injected control muscles (n = 3). (B) Graph showing the expression of ALK1 mRNA as measured by qPCR (n = 3). There was a rapid increase in Runx2 and ALK1 transcripts in damaged muscles compared with undamaged muscles. (C) Image showing a Western blot of ALK1 protein in CTX-damaged and control TA muscles 10 days after injection. The expression of ALK1 increased significantly in the CTX-damaged muscles compared with the control muscles (n = 4). (D) Graphs showing the expression of Alk-1 and several osteogenic markers from mrSCs cultured in DMEM supplemented with 10% (v/v) of damaged or undamaged muscle extracts (n = 6).

Download figure to PowerPoint

Finally, we also verified whether the soluble factors present in the microenvironment of the damaged muscles can lead to a more pronounced mrSC osteogenic phenotype. This was accomplished by exposing the mrSCs to damaged and undamaged muscle extracts. qPCR results indicated that the soluble factors present in the damaged muscle led to a significant increase in the expression level of several osteogenic markers (Fig. 2D). These results support the idea that damaged skeletal muscle possesses locally produced factors that may contribute to or potentiate HO.

BMP-9 induces HO in damaged but not in undamaged skeletal muscles

Traumatic HO occurs mainly in damaged skeletal muscle.35 To determine whether BMP-9 can induce HO in skeletal muscle, we injected a single dose of BMP-9 in the TA muscles that were either undamaged or had been damaged 3 days prior to the injection. To visualize HO in the muscle, we used an in situ staining technique consisting of alizarin red for bone and alcian blue for cartilage. As denoted, HO was visible 14 days after BMP-9 injection in CTX-injured muscles but not in undamaged muscles. HO appeared as a red mass in these muscles (arrow), with the blue and orange representing nonspecific staining. The saline- and CTX-injected control muscles showed no ossification (Fig. 3A).

thumbnail image

Figure 3. BMP-9 induced HO only in damaged muscle. Single intramuscular injections of BMP-9 were made in undamaged TA muscles and TA muscles that had been damaged with cardiotoxin (CTX) for 3 days prior to the injection. (A) Representative images of in situ alcian blue/alizarin red S staining of TA muscles 14 days after a single injection of 2.9 µM BMP-9 in CTX-damaged and undamaged TA muscles. The arrow indicates alizarin red–stained heterotopic bone. Note that no staining was observed in control and CTX-injured muscles that did not received BMP-9. Scale bar = 2.5 mm. (B) Representative anteroposterior (AP) and lateral (Lat) reconstructions of TA muscle µCT scans 2 weeks after a single injection of 2.9 µM BMP-2 or BMP-9 in CTX-damaged or undamaged saline control TA muscles. Note the absence of HO in undamaged BMP-9-treated TA muscles. (C) Micrograph showing Masson trichrome staining of CTX-injured and undamaged saline control muscles that had been injected with 2.9 µM BMP-2 or BMP-9. Heterotopic bone (arrow) and bone marrow (asterisk) were observed in both control and CTX-treated muscles that had been injected with BMP-2, but only in CTX-damaged muscle that had been injected with BMP-9. No HO was observed in control and CTX-injured muscles that did not received BMP-9. Scale bar = 100 µm. (D) Representative AP and Lat reconstructions of TA muscle µCT scans 2 weeks after the crush-injured and saline control TA muscles had been injected with BMP-9. HO was apparent only in the crushed muscle.

Download figure to PowerPoint

We confirmed these results using µCT scans, which are more sensitive for detecting mineralized bone. Fourteen days after the BMP-9 injection, HO was visible only on the CTX-damaged muscle in µCT scans. On the other hand, an equimolar injection of BMP-2 caused HO in both damaged and undamaged TA muscles (Fig. 3B).

Masson trichrome (MT)–stained sections revealed that both damaged and undamaged muscles that had been injected with BMP-2 contained significant quantities of collagen organized as bone matrix (arrow, Fig. 3C). MT staining revealed collagen deposits in BMP-9-treated damaged muscles that were organized as bone containing marrow (asterisk, Fig. 3C). In both saline and damaged controls, no HO was observed. Some fibrosis between myofibers was denoted in CTX muscles that were not treated with BMP-9 (Fig. 3C).

Muscle crush injuries, which more closely mimic physiologic damage, also were used to show that BMP-9-induced HO occurs in damaged muscle19, 38 (Fig. 3D). Our results confirmed that BMP-2 induces HO in both injured and injured muscles, whereas BMP-9 only induced HO in damaged muscles.

BMP-9 induced HO in damaged skeletal muscle through mrSCs

We assessed the potential contribution of mrSCs to the development of HO in damaged muscles in the presence of BMP-9. We first isolated an enriched population of mrSCs from CTX-damaged muscle using FACS. mrSCs in damaged muscles were identified and isolated based on the immunolabeling of Sca1, which is expressed on mrSCs. To avoid cross-contamination with endothelial and hematopoietic cells, anti-CD31 and anti-blood lineage (Lin) antibodies, respectively, also were added (Fig. 4A). Lin+ inflammatory cells and CD31+ endothelial cells were discarded to obtain a Sca1+ CD31 Lin (S+CL) population of mrSCs. There was a significant increase in the fraction of S+CL cells per mg of damaged muscle compared with undamaged muscle (3121 ± 621 versus 374 ± 57, respectively; Fig. 4B). This strongly suggested that mrSCs may be involved in the tissue remodeling that occurs in damaged muscle and may contribute to HO under the appropriate stimulus.

thumbnail image

Figure 4. The muscle-resident stromal cell population increased in damaged muscles and contributed to HO. TA muscles were harvested and analyzed 3 days after they had been damaged by the injection of CTX. (A) Diagram showing the flow cytometric method used to enrich the mrSC population from digested muscle tissues by sorting out Sca1+CD31Lin (S+CL) cells. Briefly, events were collected with a log amplifier and displayed on a four-decade log scale. Samples were forward (FSC) and side-scatter (SSC) gated to exclude debris. Sorting gates were determined for CD31-FITC, Lin-PE, and Sca1-APC after a comparison with control samples. (B) Graphs showing the absolute number of S+CL cells per milligram of damaged and undamaged muscles (n = 3). There was a robust and significant increase in the S+CL population following injury. (C) Photograph showing a cross section of undamaged and damaged skeletal muscle immunostained with anti-Sca1 (green) and anti-ALK1 (red) antibodies. Nuclei were stained with DAPI (blue). Sca1/ALK1-double-stained cells (yellow, empty arrowhead) are localized between muscle fibers, identified by laminin immunostaining (purple). Damaged/regenerating fibers are characterized by the presence of centrally located nuclei (full arrowhead). Scale bar = 20 µm. (D) Graph showing osteogenic differentiation program marker Runx2 and AlkP mRNA expression as measured by qPCR in freshly sorted S+CL mrSC cells from injured TA muscles that had had been treated or not with 2.9 µM BMP-9 (n = 3). There was a significant increase in Runx2 and AlkP mRNA expression in S+CL enriched from BMP-9-injected damaged muscles. (E) Picture showing transplanted mrSCs expressing nLacZ (full arrowhead) in newly formed bone matrix (empty arrowhead) within an injured muscle that was stimulated with BMP-9 (scale bar = 25 µm).

Download figure to PowerPoint

To confirm that the Sca1-derived cell population might be involved in bone differentiation, we immunostained cross sections of undamaged and damaged skeletal muscles using anti-Sca1 and anti-ALK1 antibodies. As shown in Fig. 4C, Sca1 and ALK1 staining colocalized (yellow, empty arrowhead) on the cells between fibers as immunostained for laminin (purple). Regenerating fibers are recognized by their characteristic centrally located nuclei (full arrowhead). These results suggest that mrSCs, which are Sca1+, may be able to transduce BMP-9 signals based on the presence of the ALK1 receptor.

We next investigated the potential of the S+CL mrSC population to transduce an osteogenic signal in vivo when stimulated with BMP-9. We first induced CTX injuries in both TA muscles and injected one with BMP-9 and the contralateral TA muscle with saline 24 hours after injury. The muscle cells were harvested after 24 hours and FACS sorted for the S+CL population. qPCR analyses revealed that the S+CL population was able to specifically trigger an osteogenic differentiation program when stimulated with BMP-9 (Fig. 4D). There was a significant increase in Runx2 and AlkP levels in BMP-9-treated S+CL cells relative to those treated with saline (Fig. 4D).

Finally, we verified whether mrSCs were the cells that contributed to HO in damaged muscles following BMP-9 stimulation. This was accomplished by the transplantation of cultured mrSCs overexpressing nuclear β-galactosidase (nLacZ) by lentiviral infection. nLacZ+ cells were injected into TA muscles that were uninjured or CTX-injured in the presence or absence of BMP-9. As a control, we injected cells that were not infected or infected with GFP expressing lentivirus. Our results show that mrSCs participated in HO because nLacZ+ cells were apparent in newly formed bone matrix with BMP-9 stimulation (Fig. 4E).

BMP-9-induced HO is prevented by the soluble ALK1 receptor

To determine whether BMP-9 was responsible for initiation of the osteogenic program and HO, we inhibited its activity in mrSCs using a soluble ALK1-Fc receptor. Figure 5A shows staining for AlkP activity in mrSCs treated with BMP-9 in the presence of 5× and 20× molar equivalents of soluble ALK1-Fc receptor, which can bind and specifically inhibit BMP-9.15 Our results indicated that the level of osteogenesis of mrSCs induced by BMP-9 was inversely proportional to the level of soluble ALK1-Fc receptor added.

thumbnail image

Figure 5. ALK1-Fc receptor is able to inhibit BMP-9-induced differentiation. (A) Representative image of AlkP staining of cultured mrSCs treated for 6 days with 3.9 nM BMP-9 alone or with a combination of 3.9 nM BMP-9 and soluble 19.5 nM (5 × ) or 78 nM (20 × ) ALK1-Fc receptor. A graphic representation of pixel density indicating a robust, significant reduction in AlkP staining in mrSCs treated with BMP-9 and soluble ALK1-Fc receptor compared to mrSCs treated with BMP-9 alone (n = 3). *p < .05; **p < .005; ***p < .001. (B) Representative AP and Lat reconstructions of TA muscle µCT scans 2 weeks after a single injection of 2.9 µM BMP-9 alone or a combination of 2.9 µM BMP-9 and 43.5 µM ALK1-Fc receptor in CTX-injured TA muscles. Only the combination of BMP-9 and Alk1-Fc receptor decreased the formation of HO.

Download figure to PowerPoint

To determine whether the inhibition caused by soluble ALK1-Fc receptor was possible in situ, we performed serial intramuscular injections of BMP-9 in the absence or presence of 15× molar equivalents of soluble ALK1-Fc receptor in TA muscles that had been damaged previously for 3 days. ALK1-Fc receptor consistently inhibited HO in the damaged muscles, as shown by the µCT scans (Fig. 5B), suggesting that BMP-9-induced HO in damaged skeletal muscle may occur via the ALK1 receptor.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Heterotopic ossification (HO) is defined as the formation of bone in soft tissue that often leads to tissue function impairment.1 Clinical indications suggest that HO is observed mostly in fast and mixed muscles of the limbs, such as around the hip, elbow, and shoulder.4 The traumatic form of HO in injured skeletal muscle is often linked to brain and spinal cord injuries. The association of brain injury and HO strongly suggests that HO is caused by an osteogenic factor released into the circulation. Indeed, Gautschi and colleagues recently showed the osteogenic potential of serum from brain-injured patients.39 Postinjury serum samples induced in vitro proliferation of the human fetal osteoblastic cell line hFOB1.19 and of primary human osteoblasts. Furthermore, the samples significantly increased mRNA expression of specific osteoblastic markers.39 Owing to their well-known role in inducing the osteogenic program in cells, bone morphogenetic proteins (BMPs), notably BMP-2, -4, -6, and -7, may be key players in HO. However, there is currently no evidence to suggest that they do indeed play a role in HO. In this study, we showed that BMP-9, a recently identified osteogenic BMP, is highly efficient in triggering the osteogenic program of mrSCs and that it may be involved in the formation of HO in damaged skeletal muscle.

Skeletal muscles display many abnormal nonmyogenic characteristics in pathophysiologic conditions. For example, the muscle fibers of aged and dystrophic skeletal muscles are infiltrated by large quantities of adipogenic tissue and fibrogenic deposits.18, 23, 26, 40 In addition, following extensive trauma, the rate of HO development in skeletal muscles can be as high as 16%.41–43 These examples suggest that progenitor cells are able to transduce osteogenic signals and activate the bone differentiation program once their microenvironment has been altered. There are two well-known classes of progenitor cells in adult skeletal muscles. Satellite cells, which are juxtaposed between the plasma membrane of the muscle fibers and the basement membrane, are known to provide a source of myogenic progenitor cells that contribute to lifetime muscle fiber repair,18–20 whereas stromal progenitor cells (mrSCs), which reside in niches located between the muscle fibers, are closely associated with the vasculature.21–23, 26 Many markers are used to identify and enrich the mrSC population, the most common being Sca1.21, 23, 26

This study showed, for the first time, that BMP-9 is able to induce the osteogenic differentiation program in primary mrSCs. Using recombinant BMP-9, we were able to induce a robust increase in the expression of Runx2, a member of the Runt domain family of transcription factors that is known to be the earliest and most powerful molecular determinant of osteoblast differentiation.36 In addition, the Runx2-dependent expression of osterix, a transcription factor whose activity is absolutely required for osteoblast differentiation in mice, was upregulated.36 We also observed massive staining of AlkP activity in treated cells. The activation of the osteogenic program in mrSCs by BMP-9 is significant because it was as effective as BMP-2, which is considered the “gold standard” for fully inducing the bone differentiation program.44

Kang and colleagues, who studied the adenoviral expression of 14 different BMPs, identified BMP-9 as the most osteogenic BMP.13 Its osteogenic potential has been clearly demonstrated in vitro with an increase in AlkP activity in infected C2C12 cells and the development of HO in mouse quadriceps injected with C2C12 cells overexpressing BMP-9.13 Surprisingly, our study showed that a single intramuscular injection of recombinant BMP-9 was unable to induce HO. This was unexpected because BMP-2, which displayed a similar capacity to induce osteogenic program in vitro, was able to induce robust HO in resting skeletal muscle. Since HO usually occurs in damaged skeletal muscle, we also injected BMP-9 into CTX-injured TA muscles. In this case, we observed marked HO in damaged muscles compared with resting undamaged muscles, suggesting that the mrSC microenvironment is altered in damaged muscles. Interestingly, our data corroborate previously published work by Kan and colleagues demonstrating an enrichment of Sca1+ cells in the early phase of HO,45 suggesting that muscle-resident mrSCs give rise to osteogenic progenitors.

The differences between damaged and undamaged muscles suggest that an alteration of the mrSC microenvironment makes mrSCs more permissive to transducing BMP-9 signals through ALK1, which was confirmed by the inhibition of HO by ALK1-Fc receptor. Acute skeletal muscle damage, as is the case with CTX and crush injuries, leads to an early inflammatory response. We used flow cytometry to show that the total number of cells increased significantly in damaged muscle and was mainly due to the presence of inflammatory cells, as indicated by the fact that the population of Lin+ cells increased (Supplemental Fig. S5).

There is a strong link between brain and spinal injuries and HO formation. Cerebrospinal fluid (CSF) has an important osteogenic effect in the hFOB cell line,46 which suggests that a breach of the brain-blood barrier may release an osteogenic factor into the CSF that migrates to damaged muscle tissue. However, Gautschi and colleagues were unable to demonstrate by ELISA that BMP-2, -4, and -7 levels are higher in human CSF from patients suffering from traumatic brain injury and nontraumatic brain pathology than in CSF from healthy patients.46 They concluded that BMPs are not present in high enough concentrations to be responsible for the osteogenic cell response that triggers HO.

Muscle damage–dependent bone formation is a classic clinical presentation of traumatic HO.4 Several factors appear to be working in concert at the site of injury that may alter the milieu of muscle-resident progenitor cells and induce HO. First, the modification of blood flow could induce ischemia and stasis at the site of injury. Such a modification in gas exchange was well described in paraplegic individuals47, 48 and after total hip-replacement procedures,49 two conditions that are often associated with HO. Second, inflammation is known to occur following muscle damage. For example, the inflammatory factor prostaglandin E2 is known for its ability to induce both bone resorption and bone formation depending on the activated pathway.50 Furthermore, nonsteroidal anti-inflammatory drugs, such as cyclooxygenase 2 (COX-2) receptor inhibitors, were shown to reduce HO formation after injury.51–53 Other important contributing factors are hypercalcemia, changes in sympathetic nerve activity, prolonged immobilization, and disequilibrium between parathyroid hormone and calcitonin.4 Importantly, the presence of other soluble growth factors produced locally may potentiate the signaling of BMPs, such as insulin-like growth factor 2 (IGF-2) or retinoic acid, which were shown recently to potentiate the osteogenic effect of BMP-9.54, 55

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Our study indicates that BMP-9, like BMP-2, has a strong osteogenic effect on mrSCs. Interestingly, the in vivo osteogenic effect of BMP-9 was observed only in damaged skeletal muscles. We show that the induction of HO was associated with mrSCs, which increased the expression of ALK1, the BMP-9 receptor. To our knowledge, this is the first report to show that alterations to the muscle microenvironment make mrSCs sensitive to BMP-9. More studies are needed to clearly demonstrate the link between BMP-9 and HO pathophysiology.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

We are grateful to Dr Anthony Scimè for his critical reading of the manuscript. We thank Paul Oleynick for technical assistance with the MoFloCytometer as well as Nancy Charland at the Anatomo-pathology Service. EL received scholarships from the Fondation pour la Recherche et l'Enseignement en Orthopédie de Sherbrooke (FREOS) and the Canadian Institutes of Health Research (CIHR). GG received a New Investigator Award from Fonds de la Recherche en Santé du Québec (FRSQ). This work was supported by grants from the CIHR, the National Science and Engineering Research of Canada (NSERC), the Canada Foundation for Innovation (CFI), and the Cell and Tissue Therapy Network of FRSQ.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosures
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting information may be found in the online version of this article.

FilenameFormatSizeDescription
JBMR_311_sm_SuppFigS1.eps2033KSupplementary Figre S1
JBMR_311_sm_SuppFigS2.eps346KSupplementary Figre S2
JBMR_311_sm_SuppFigS3.eps12120KSupplementary Figre S3
JBMR_311_sm_SuppFigS4.eps407KSupplementary Figre S4
JBMR_311_sm_SuppFigS5.eps3659KSupplementary Figre S5
JBMR_311_sm_SuppData.doc26KSupplementary Data

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.