A Novel Mouse Model of Trauma Induced Heterotopic Ossification Xuhui Liu,1,2 Heejae Kang,1 Mohammad Shahnazari,1,3 Hubert Kim,1,2 Liping Wang,1,3 Olla Larm,4 Lars Adolfsson,4 Robert Nissenson,1,3 Bernard Halloran1,3 1 Department of Veterans Affairs, San Francisco Veterans Affairs Medical Center, 4150 Clement Street, San Francisco, CA 94404, 2Department of Orthopedic Surgery, University of California at San Francisco, 500 Parnassus Avenue, San Francisco, CA 94143, 3Department of Medicine, University of California at San Francisco, 500 Parnassus Avenue, San Francisco, CA 94143, 4ExThera AB, Karolinska Institute, Berzelius va¨g 37, SE 171 77, Stockholm, Sweden
Received 15 January 2013; accepted 17 September 2013 Published online 17 October 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22500
ABSTRACT: Severe soft tissue trauma is associated with heterotopic ossification (HO), the abnormal deposition of bone at extraskeletal sites. The pathophysiology of the development of trauma-induced HO remains largely unknown due in part to the lack of appropriate animal models. In this study, we sought to develop a new trauma-induced HO mouse model using muscle impact injury combined with low dose BMP-2. BMP-2 at doses ranging from 0 to 2 mg was injected into quadriceps muscles of adult male C57/BL6 mice. Animals then received a one-time quadriceps impaction injury to mimic the trauma associated with severe injuries. HO was monitored using in vivo microCT scanning at 1, 2, 4, and 8 weeks after treatment. After trauma, the expression of BMP-2, -4, BMP receptor 1, SOX9 and RUNX2 were increased in muscle. Although little or no HO was observed in mice receiving 1 mg BMP-2, combining this dose with muscle trauma produced an abundance of HO. At higher doses of BMP-2, trauma did not augment mineral deposition. These results suggest that BMP-2 signaling can sensitize muscle to trauma-induced HO. They also provide the basis for a new model to study the pathogenesis of trauma-induced HO. ß 2013 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:183–188, 2014. Keywords: heterotopic ossification; trauma; bone morphogenetic protein; mouse model; MicroCT
Heterotopic ossification (HO) is defined as the abnormal deposition of bone in soft tissues.1–15 Severe muscle damage, spinal cord injury, head trauma, neurological trauma and joint arthroplasty are often accompanied by HO. Heterotopic ossification occurs in 24% of hip replacement patients. In injured military personnel returning from the Middle East, the incidence has been reported to be as high as 64%.7,8,10,16 Severe restrictions in joint movement, pain and soft tissue infections are common complications of HO. Heterotopic ossification is hard to prevent and once developed is extremely difficult to treat. Present treatment options are limited to gross surgical excision, radiation therapy and treatment with non-steroidal anti inflammatory drugs (NSAID). Surgical excision is the most popular but recurrence rates requiring further surgery are reported to be as high as 25%.9,17,18 A critical barrier to the development of more effective preventive measures and treatments for HO is the lack of clinically relevant animal models. Current models rely mostly on hydro-gel implants containing bone morphogenetic protein (BMP)-2, -4, and -9.4 These are good for studying BMP-induced ectopic bone formation but commonly do not include a trauma component. A mouse model of fibrodysplasia ossificans progressive (FOP), however, has proven to be a particularly valuable tool to study traumatic HO because of its trauma inducible nature.19–22 In this model an activating mutation in ACVR1/ALK2, a typeGrant sponsor: Department of Defense; Grant number: W81XWH-11-2-0189; Grant sponsor: Veterans Affairs Merit review program; Grant sponsor: Northern California Institute for Research and Education. Correspondence to: Bernard Halloran (T: þ1-415-750-6928; F: þ1-415-750-6929; E-mail: [email protected]) # 2013 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
I BMP receptor sensitizes the muscle to injury-induced HO. Other trauma-associated models include the use of BMPs in combination with cardiotoxin (CTX) or physical damage induced by crushing the muscle with surgical tweezers.23 It has been difficult to induce HO with trauma alone, although a recently developed model using blast amputation holds promise.24 In the studies reported here we set out to develop a new murine model of traumatic HO that more closely mimics the formation of heterotopic bone associated with blunt physical muscle trauma such as occurs in accidents or injuries in a military setting. The model involves a one-time impaction injury of the quadriceps muscle combined with a sub-symptomatic dose of BMP-2.
MATERIALS AND METHODS Animals Male C57B/6 mice, 3 months of age were obtained from Jackson Laboratories (Sacramento, CA). The animals were housed in air-filtered, humidity- and temperature-controlled rooms with equal 12 h light-12 h dark cycles and fed a standard mouse diet. The animal protocol for the study was in accordance with the NIH Guide for the Care and Use of Laboratory Animals and approved by the Animal Care and Use Committee at the Veterans Affairs Medical Center, San Francisco. Induction of HO With BMP-2 In vivo controlled release of BMP-2 was achieved by mixing rhBMP-2 (Medtronic Sofamor Danek USA, Inc., Memphis, TN) with a heparin-chitosan ionic complex, in the form of a hydro gel (ExThera, Inc., Stockholm, Sweden), which has been proven to successfully induce HO in rats.25 In brief, 20 ml of heparin-chitosan hydro gel with 0, 0.25, 0.5, 0.75, 1, 1.5 and 2 mg BMP-2 was injected into the quadriceps muscle of both hindlimbs with a microsyringe. The mice were allowed to move freely about their cages after injection. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
183
184
LIU ET AL.
Muscle Impaction Injury While under isoflurane anesthesia, a one-time muscle impaction injury was achieved by dropping a stainless steel ball weighing 16.3 g (15.6 mm in diameter) from a height of 100 cm onto the right quadriceps muscle immediately after BMP-2 injection as described previously.26 The left quadriceps served as a non-injury control. To prevent pain, subcutaneous injection of 0.5 mg/kg Buprenorphine was given twice a day for the first 2 post-injury days. MicroCT Analysis The intact thighs, including the femur and quadriceps, were scanned in vivo at 1, 2, 4, and 8 weeks after treatment using a Scanco Viva CT 40 (Scanco Medical, Basserdorf, Switzerland) microCT. The isotropic voxel size was nominally 10.5 mm and the X-ray energy was 55 kV. A global threshold, set at 245 in the per mille unit or 376 mg of hydroxyapatite/ cm3, was applied to distinguish mineralized from soft tissue. In order to quantify the ossification, microCT images at 2 weeks after treatment was assessed by quantifying the total amount of mineralized tissue and the degree of bone mineralization (segmented density). Histology Separate mice were sacrificed at 8 weeks after treatment for histological analysis. The hindlimbs were harvested and fixed with 10% PBS buffered formalin. The quadriceps were removed from the limb and decalcified in 10% ethylenediaminetetraacetic acid (EDTA) (pH 8.0) for 3 weeks with changes every 3 days. Decalcified samples were embedded in paraffin, sectioned at 7 mm of thickness and stained with H&E. Quantification of Gene Expression in Muscle To examine gene expression in the muscle following impact injury, the right quadriceps was impacted using our model and the entire muscle was collected 24 h, 1 week, and 2 weeks post-injury (n ¼ 6/time point). The contralateral muscle was collected to serve as control (no impact). Total RNA was extracted from the muscle following homogenization in the RNA-Stat 60 reagent (AMSBIO, Lake Forest, CA) and purified using RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s recommendations. Two micrograms of total RNA was reverse transcribed in 100 ml of reaction mixture containing 1 PCR Buffer, 1 mM deoxynucleoside triphosphate (DNTP) mix, 5 mM random primers, 7.5 mM MgCl2, 0.4 U/ml RNAase inhibitor, and 2.5 U/ml MultiScribe Reverse Transcriptase (Reverse Transcription Reagents, Applied Biosystems, Foster City, CA) at a sequence of 25˚C for 10 min, 48˚C for 40 min, and 95˚C for 5 min. The real time PCR quantification of gene expression was performed in triplicate reactions using inventoried Taqman assays for mice including l19, gapdh, tnfa, bmp2, bmp4, bmp9, bmpr1a, bmpr2, sox9, runx2, col2, and colX in a 7300 RT-PCR System (Applied Biosystems). The reactions were performed under the following conditions: 95˚C for 10 min, 40 cycles of 95˚ for 15 s, and 60˚C for 1 min. Analysis was carried
out using the software supplied by the manufacturer and the number of threshold cycles (Ct) required for the FAM fluorophore intensities to exceed a threshold just above background were calculated. Ct values for triplicate reactions were averaged for each target gene and expressed as a ratio to that of the GAPDH Ct value. Statistical Analysis Data are reported as mean SEM with n ¼ 6 mice per experimental group. Analysis of Variance (ANOVA) was used to compare the BV in each leg between animals receiving different doses of BMP-2. Student’s t-test was used to compare the BV between the groups with and without injury at each BMP-2 dose. Statistical significance was considered when p < 0.05.
RESULTS BMP-2 Induced HO HO was first observed 2 weeks after BMP-2 injection (1.5 mg; Fig. 1). Injection of BMP-2 without the gel carrier did not induce mineral formation. At 4 and 8 weeks after injection, there was no further expansion of the mineral deposit but there was condensation and reordering of the mineral into a more organized structure (Fig. 1). No mineral could be detected after 1 week. Dosing of BMP-2 without the heparin-chitosan ionic complex did not induce HO. Increasing amounts of BMP-2 in the gel carrier dose dependently increased the amount of mineral deposited in the muscle (Fig. 2). At BMP-2 doses of 1 mg or less little or no mineral was apparent. Between 1 and 1.5 mg of BMP-2 there was a dramatic increase in mineral deposited from less that 0.5 mm3 to more than 2 mm3. Doses above 1.5 mg did not increase the amount of mineral. In most of the ossified tissues, the mineral was confined to the muscle and independent of the bone (Fig. 1). In some cases (approximately 20%), however, the mineral was closely associated with or on the bone surface (Fig. 3). This typically resulted in a dramatic increase in the amount of mineral formed. Samples with bone involvement were excluded from our analyses. Effect of Trauma on BMP-2 Induced HO Trauma alone or trauma combined with the BMP-2 carrier (0 mg) did not induce HO (Figs. 4 and 5). In mice receiving less than 1 mg of BMP-2, trauma had little or no effect on mineral deposition. In mice receiving 1 mg BMP-2, trauma dramatically increased HO formation (Fig. 4). Quantitative BV analysis showed that injury increased heterotopic bone by sixfold. At higher, saturating doses of BMP-2 muscle injury did not affect the amount of mineral deposited (data not shown).
Figure 1. Representative microCT images of HO at 1, 2, and 4 weeks after 1.5 mg BMP-2 injection without trauma. HO was first observed 2 weeks after BMP-2 injection. At 4 weeks after injection, there was no further expansion of the mineral deposit but there was condensation and reordering of the mineral into a more organized structure. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
A MOUSE HETEROTOPIC OSSIFICATION MODEL
185
Figure 4. Representative microCT images of HO 2 weeks after 0 and 1 mg BMP-2 injection with and without trauma. Trauma significantly augments HO development in the presence of BMP2. Figure 2. Volume of heterotopic bone (BV) formed 2 weeks after BMP-2 injection without trauma. (Mean SEM, N ¼ 6, p < 0.05 compared to other groups.) At BMP-2 doses of 1 mg or less little or no mineral was apparent. Between 1 and 1.5 mg of BMP-2 there was a dramatic increase in mineral deposited from less that 0.5 mm3 to more than 2 mm3.
using GAPDH and L19 for normalization were similar. All reported data were normalized to GAPDH.
DISCUSSION Histology Representative structures of the HO deposits for mice receiving 1 mg BMP-2 with or without muscle trauma are shown in Figure 6. Without injury, only a small amount of bone is formed and it appears relatively unstructured. With muscle impaction injury, there is a dramatic increase in the amount of bone formed, it appears more structured and in some cases includes marrow-like components. The bone is clearly lamellar and contains numerous osteocytes. Gene Expression The effects of muscle injury on gene expression are shown in Figures 7 and 8. Injury stimulated BMP-2 and -4 expression within 24 h and expression levels remained elevated for at least 7 days. By 14 days, levels returned to normal (BMP-4) or were below normal (BMP-2). BMP-9 was undetectable at all times. Expression of BMPR1a also increased by 24 h and returned to normal by 14 days. No change in BMPR2 expression was observed. To determine whether muscle injury could induce expression of genes associated with chondrogenesis and osteoblastogenesis we measured expression of SOX9, Col II, Col X, and RUNX2. Expression of SOX9 and RUNX2 was increased 24 h and 7 days after injury. By 14 days, levels had returned to normal. Col II and Col X were undetectable at all times. Expression of TNFa followed a pattern similar to SOX9 and RUNX2. The results
We have successfully developed a new murine model of traumatic HO that mimics the formation of heterotopic bone associated with blunt physical muscle trauma. Previously developed trauma-inclusive models in normal rodents have used BMPs in combination with CTX or physical damage induced by crushing the muscle with surgical tweezers.23 These have proven to be valuable models for the study of HO but do not effectively mimic the formation of heterotopic bone in response to severe blunt physical injury. The impact force induced in our model is in the range of many injuries in military personnel but is less than many others. Higher impact forces in our model frequently induce fracture and this can lead to dramatically different responses in HO formation. In order to study the mechanisms involved in HO we set out to develop a model that forms a consistent amount of HO and is not complicated by varying degrees of bone injury and healing.26 The development of models of HO that involve severe bone injury or loss of limbs is important but was not a part of goal of this study. Clinical HO develops through a process of endochondral bone formation.27 In our model, mineralization is not apparent 1 week after injury but by 2 weeks is clearly evident in mice receiving BMP-2 alone or BMP-2 with trauma. This mimics the lag time in the appearance of mineral observed in patients with trauma induced HO. The mineral deposit in our model appears to be confined to the site of BMP-2 gel injection suggesting that there is a need for the BMP-2
Figure 3. Representative microCT images of HO 2 weeks after 1.5 mg BMP-2 injection with and without bone involvement (no injury). Bone involvement augments the development of HO. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
186
LIU ET AL.
Figure 5. Volume of heterotopic bone (BV) formed 2 weeks after 1 mg BMP-2 injection with and without trauma. (Mean SEM, N ¼ 6, p < 0.05 compared to non-injury control.) Injury increased heterotopic bone volume by about sixfold.
to remain in the lesion for some time to initiate bone formation. What sustains the heterotopic bone is not clear but once formed there is no indication that it can be resorbed away for at least 8 weeks after treatment in our model. In about 20% of our animals, we found the heterotopic bone attached to the femur. In these instances the volume of bone formed appeared much greater than when the mineral deposit was not in contact with the femur. We think that bone involvement may be due to the needle touching the periosteum during injection. In patients with HO, the mineral deposits are frequently associated with the bone surface. It appears that irritation of the periosteum may augment mineral accumulation during the development of HO. Further studies are needed to fully establish the role of the periosteum in HO development.
Figure 6. Typical H&E staining for heterotopic bone after 1 mg BMP-2 injection with and without muscle injury at 8 weeks after injury with low (5, upper panel) and high (20, lower panel) magnifications. HO was developed at the BMP-2 injection site with and without injury. This is supported by the presence of remaining undigested hydrogel ( ) at the site of HO. Without injury, only a small unstructured bone is formed. With muscle trauma, the HO is significantly larger and better structured. The bone is clearly lamellar and contains numerous osteocytes (arrows) and marrow-like components. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
Figure 7. RT-PCR results of BMP signaling in muscle after impaction (no BMP-2 injection). BMP-2, -4, BMPR1a gene expression significantly increased in muscle at 1 and 7 days after injury. The expression of BMPR2 remains unchanged at all time points. Open bars stand for the values of control group, striped bars stand for the values of the injured group. (Mean SEM, normalized to GAPDH, N ¼ 6, p < 0.05, p < 0.01.)
The amount of heterotopic bone formed in our model is extremely sensitive to the amount of BMP-2 deposited in the muscle. There appears to be a triggering of mineral deposition between 1 and 1.5 mg of BMP-2. If the BMPs are involved in clinical HO, and there is evidence that they are, small differences in the local concentrations may determine whether HO forms following trauma. Our RT-PCR analysis has shown that trauma significantly induces the gene expression multiple BMP family members (including BMP-2; Fig. 7). However, the local concentration of BMPs at the injury site was not measured in this study. Thus, we could not directly compare the physiological concentration of BMP-2 induced by muscle trauma to the doses of BMP-2 we used in this study. The effects of trauma on bone formation were dramatic. Combining a sub-threshold dose of BMP-2 (1 mg) with trauma resulted in a striking increase in mineral deposition (Figs. 4 and 5). The presence of exogenously provided BMP-2 appears to sensitize the muscle to trauma-induced HO. This is much like the circumstances in patients with FOP where constitutive BMP signaling sensitizes muscle to mild trauma.21,22 The cellular source of HO remains undefined. Previous studies have suggested that endothelial progenitors may be the responsible for the development of HO and BMP signaling plays a critical role in promoting endothelial–mesenchymal transition.28,29 In our study, we have observed significantly increased expression of BMP-2 and BMP-4 in muscle after trauma. Other studies suggests that muscle resident stem cells are involved in HO development and inflammation cytokines, like TNFa plays a critical role in this process.30
A MOUSE HETEROTOPIC OSSIFICATION MODEL
187
Figure 8. RT-PCR results of SOX9, RUNX2, and TNFa gene expression in muscle after impaction (no BMP-2 injection). All three genes significantly increased in muscle at 1 and 7 days after injury. Open bars stand for the values of control group, striped bars stand for the values of the injured group. (Mean SEM, normalized to GAPDH, N ¼ 6, p < 0.05, p < 0.01.)
Our data also showed significantly increased TNFa in traumatic muscle. Future studies are needed to identify the cellular source and key signaling pathways responsible for the development of HO and our model can serve as a powerful tool for this purpose. There are some limitations in this study. First, only a single strain (C57/BL6) young male mouse was tested in this study. It has been known that the strain, gender and age have significant effects on mouse bone metabolism. Thus, the conclusion from this study may not fit other mice with different strain, gender, or age. Second, the injection site on the quadriceps muscle was not strictly controlled. We have noticed that in some mice, HO developed at the proximal side of quadriceps, while in other mice, HO developed at the distal side of quadriceps muscle. We believe that this was caused by variation in the site of BMP-2 injection. Third, the histological analysis of HO was only conducted at a late time point at 8 weeks when HO has fully developed. Thus, the early stages of HO development, such as the chondrogenic phase, was not observed. In this study, we have successfully established a new trauma-induced HO mouse model that closely mimics HO in patients with severe blunt soft tissue injury. In our model a onetime muscle impact can increase the volume of HO by sixfold with a subsymptomatic dose of 1 mg BMP-2 in mouse quadriceps muscles. This novel mouse model can serve as a powerful tool in future studies to examine the molecular mechanisms involved in trauma-induced HO. It can also serve as a powerful screening tool in seeking better prevention and treatments for trauma-induced HO in the future.
ACKNOWLEDGEMENTS This work was supported by the Department of Defense (grant #W81XWH-11-2-0189), the Veterans Affairs Merit review program and the Northern California Institute for Research and Education. We thank John Carney for his helpful ideas and review of this manuscript, and Christian Santa Maria and Alfred Li for their technical support in microCT scanning and image analysis. Dr. Larm and Dr. Adolfsson are share holders of ExThera AB.
REFERENCES 1. Kaplan FS, Glaser DL, Hebela N, et al. 2004. Heterotopic ossification. J Am Acad Orthop Surg 12:116–125. 2. Kaplan FS, Tabas JA, Gannon FH, et al. 1993. The histopathology of fibrodysplasia ossificans progressiva. An endochondral process. J Bone Joint Surg Am 75:220–230. 3. Garland DE. 1991. A clinical perspective on common forms of acquired heterotopic ossification. Clin Orthop Relat Res 263:13–29. 4. O’Connor JP. 1998. Animal models of heterotopic ossification. Clin Orthop Relat Res 346:71–80. 5. Kaplan FS, Shore EM. Derailing heterotopic ossification and RARing to go. Nat Med 17:420–421. 6. Kirsch T. 2008. Determinants of pathologic mineralization. Crit Rev Eukaryot Gene Expr 18:1–9. 7. Stoltny T, Koczy B, Wawrzynek W, et al. 2007. Heterotopic ossification in patients after total hip replacement. Ortop Traumatol Rehabil 9:264–272. 8. Potter BK, Burns TC, Lacap AP, et al. 2006. Heterotopic ossification in the residual limbs of traumatic and combatrelated amputees. J Am Acad Orthop Surg 14:S191–S197. 9. Potter BK, Burns TC, Lacap AP, et al. 2007. Heterotopic ossification following traumatic and combat-related amputations. Prevalence, risk factors, and preliminary results of excision. J Bone Joint Surg Am 89:476–486. 10. Forsberg JA, Pepek JM, Wagner S, et al. 2009. Heterotopic ossification in high-energy wartime extremity injuries: prevalence and risk factors. J Bone Joint Surg Am 91:1084–1091. 11. Garland DE, Dowling V. 1983. Forearm fractures in the head-injured adult. Clin Orthop Relat Res 190–196. 12. Hendricks HT, Geurts AC, van Ginneken BC, et al. 2007. Brain injury severity and autonomic dysregulation accurately predict heterotopic ossification in patients with traumatic brain injury. Clin Rehabil 21:545–553. 13. Cipriano CA, Pill SG, Keenan MA. 2009. Heterotopic ossification following traumatic brain injury and spinal cord injury. J Am Acad Orthop Surg 17:689–697. 14. Jackson WM, Aragon AB, Bulken-Hoover JD, et al. 2009. Putative heterotopic ossification progenitor cells derived from traumatized muscle. J Orthop Res 27:1645–1651. 15. Ji Y, Christopherson GT, Kluk MW, et al. 2011. Heterotopic ossification following musculoskeletal trauma: modeling stem and progenitor cells in their microenvironment. Adv Exp Med Biol 720:39–50. 16. Brown KV, Dharm-Datta S, Potter BK, et al. Comparison of development of heterotopic ossification in injured US and UK Armed Services personnel with combat-related amputations: preliminary findings and hypotheses regarding causality. J Trauma 69:S116–S122. 17. Genet F, Jourdan C, Lautridou C, et al. The impact of preoperative hip heterotopic ossification extent on recur-
JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
188
18.
19.
20.
21.
22.
23.
LIU ET AL. rence in patients with head and spinal cord injury: a case control study. PLoS ONE 6:e23129. Baldwin K, Hosalkar HS, Donegan DJ, et al. Surgical resection of heterotopic bone about the elbow: an institutional experience with traumatic and neurologic etiologies. J Hand Surg 36:798–803. Pignolo RJ, Shore EM, Kaplan FS. Fibrodysplasia ossificans progressiva: clinical and genetic aspects. Orphanet J Rare Dis 6:80. Pignolo RJ, Xu M, Russell E, et al. Heterozygous inactivation of Gnas in adipose-derived mesenchymal progenitor cells enhances osteoblast differentiation and promotes heterotopic ossification. J Bone Miner Res 26:2647–55. Shore EM, Kaplan FS. 2008. Insights from a rare genetic disorder of extra-skeletal bone formation, fibrodysplasia ossificans progressiva (FOP). Bone 43:427–433. Kan L, Hu M, Gomes WA, et al. 2004. Transgenic mice overexpressing BMP4 develop a fibrodysplasia ossificans progressiva (FOP)-like phenotype. Am J Pathol 165:1107– 1115. Leblanc E, Trensz F, Haroun S, et al. BMP9-Induced muscle heterotopic ossification requires changes to the skeletal muscle microenvironment. J Bone Miner Res.
JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
24. Tannous O, Griffith C, O’Toole RV, et al. Heterotopic ossification after extremity blast amputation in a Sprague– Dawley rat animal model. J Orthop Trauma 25:506–510. 25. Engstrand T, Veltheim R, Arnander C, et al. 2008. A novel biodegradable delivery system for bone morphogenetic protein-2. Plast Reconstr Surg 121:1920–1928. 26. Nozaki M, Li Y, Zhu J, et al. 2008. Improved muscle healing after contusion injury by the inhibitory effect of suramin on myostatin, a negative regulator of muscle growth. Am J Sports Med 36:2354–2362. 27. Forsberg JA, Potter BK. 2010. Heterotopic ossification in wartime wounds. J Surg Orthop Adv 19:54–61. 28. Medici D, Shore EM, Lounev VY, et al. 2010. Conversion of vascular endothelial cells into multipotent stem-like cells. Nat Med 16:1400–1406. 29. Matsubara H, Hogan DE, Morgan EF, et al. 2012. Vascular tissues are a primary source of BMP2 expression during bone formation induced by distraction osteogenesis. Bone 51:168–180. 30. Glass GE, Chan JK, Freidin A, et al. 2011. TNF-alpha promotes fracture repair by augmenting the recruitment and differentiation of muscle-derived stromal cells. Proc Natl Acad Sci USA 108:1585–1590.
Received 15 January 2013; accepted 17 September 2013 Published online 17 October 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22500
ABSTRACT: Severe soft tissue trauma is associated with heterotopic ossification (HO), the abnormal deposition of bone at extraskeletal sites. The pathophysiology of the development of trauma-induced HO remains largely unknown due in part to the lack of appropriate animal models. In this study, we sought to develop a new trauma-induced HO mouse model using muscle impact injury combined with low dose BMP-2. BMP-2 at doses ranging from 0 to 2 mg was injected into quadriceps muscles of adult male C57/BL6 mice. Animals then received a one-time quadriceps impaction injury to mimic the trauma associated with severe injuries. HO was monitored using in vivo microCT scanning at 1, 2, 4, and 8 weeks after treatment. After trauma, the expression of BMP-2, -4, BMP receptor 1, SOX9 and RUNX2 were increased in muscle. Although little or no HO was observed in mice receiving 1 mg BMP-2, combining this dose with muscle trauma produced an abundance of HO. At higher doses of BMP-2, trauma did not augment mineral deposition. These results suggest that BMP-2 signaling can sensitize muscle to trauma-induced HO. They also provide the basis for a new model to study the pathogenesis of trauma-induced HO. ß 2013 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 32:183–188, 2014. Keywords: heterotopic ossification; trauma; bone morphogenetic protein; mouse model; MicroCT
Heterotopic ossification (HO) is defined as the abnormal deposition of bone in soft tissues.1–15 Severe muscle damage, spinal cord injury, head trauma, neurological trauma and joint arthroplasty are often accompanied by HO. Heterotopic ossification occurs in 24% of hip replacement patients. In injured military personnel returning from the Middle East, the incidence has been reported to be as high as 64%.7,8,10,16 Severe restrictions in joint movement, pain and soft tissue infections are common complications of HO. Heterotopic ossification is hard to prevent and once developed is extremely difficult to treat. Present treatment options are limited to gross surgical excision, radiation therapy and treatment with non-steroidal anti inflammatory drugs (NSAID). Surgical excision is the most popular but recurrence rates requiring further surgery are reported to be as high as 25%.9,17,18 A critical barrier to the development of more effective preventive measures and treatments for HO is the lack of clinically relevant animal models. Current models rely mostly on hydro-gel implants containing bone morphogenetic protein (BMP)-2, -4, and -9.4 These are good for studying BMP-induced ectopic bone formation but commonly do not include a trauma component. A mouse model of fibrodysplasia ossificans progressive (FOP), however, has proven to be a particularly valuable tool to study traumatic HO because of its trauma inducible nature.19–22 In this model an activating mutation in ACVR1/ALK2, a typeGrant sponsor: Department of Defense; Grant number: W81XWH-11-2-0189; Grant sponsor: Veterans Affairs Merit review program; Grant sponsor: Northern California Institute for Research and Education. Correspondence to: Bernard Halloran (T: þ1-415-750-6928; F: þ1-415-750-6929; E-mail: [email protected]) # 2013 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
I BMP receptor sensitizes the muscle to injury-induced HO. Other trauma-associated models include the use of BMPs in combination with cardiotoxin (CTX) or physical damage induced by crushing the muscle with surgical tweezers.23 It has been difficult to induce HO with trauma alone, although a recently developed model using blast amputation holds promise.24 In the studies reported here we set out to develop a new murine model of traumatic HO that more closely mimics the formation of heterotopic bone associated with blunt physical muscle trauma such as occurs in accidents or injuries in a military setting. The model involves a one-time impaction injury of the quadriceps muscle combined with a sub-symptomatic dose of BMP-2.
MATERIALS AND METHODS Animals Male C57B/6 mice, 3 months of age were obtained from Jackson Laboratories (Sacramento, CA). The animals were housed in air-filtered, humidity- and temperature-controlled rooms with equal 12 h light-12 h dark cycles and fed a standard mouse diet. The animal protocol for the study was in accordance with the NIH Guide for the Care and Use of Laboratory Animals and approved by the Animal Care and Use Committee at the Veterans Affairs Medical Center, San Francisco. Induction of HO With BMP-2 In vivo controlled release of BMP-2 was achieved by mixing rhBMP-2 (Medtronic Sofamor Danek USA, Inc., Memphis, TN) with a heparin-chitosan ionic complex, in the form of a hydro gel (ExThera, Inc., Stockholm, Sweden), which has been proven to successfully induce HO in rats.25 In brief, 20 ml of heparin-chitosan hydro gel with 0, 0.25, 0.5, 0.75, 1, 1.5 and 2 mg BMP-2 was injected into the quadriceps muscle of both hindlimbs with a microsyringe. The mice were allowed to move freely about their cages after injection. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
183
184
LIU ET AL.
Muscle Impaction Injury While under isoflurane anesthesia, a one-time muscle impaction injury was achieved by dropping a stainless steel ball weighing 16.3 g (15.6 mm in diameter) from a height of 100 cm onto the right quadriceps muscle immediately after BMP-2 injection as described previously.26 The left quadriceps served as a non-injury control. To prevent pain, subcutaneous injection of 0.5 mg/kg Buprenorphine was given twice a day for the first 2 post-injury days. MicroCT Analysis The intact thighs, including the femur and quadriceps, were scanned in vivo at 1, 2, 4, and 8 weeks after treatment using a Scanco Viva CT 40 (Scanco Medical, Basserdorf, Switzerland) microCT. The isotropic voxel size was nominally 10.5 mm and the X-ray energy was 55 kV. A global threshold, set at 245 in the per mille unit or 376 mg of hydroxyapatite/ cm3, was applied to distinguish mineralized from soft tissue. In order to quantify the ossification, microCT images at 2 weeks after treatment was assessed by quantifying the total amount of mineralized tissue and the degree of bone mineralization (segmented density). Histology Separate mice were sacrificed at 8 weeks after treatment for histological analysis. The hindlimbs were harvested and fixed with 10% PBS buffered formalin. The quadriceps were removed from the limb and decalcified in 10% ethylenediaminetetraacetic acid (EDTA) (pH 8.0) for 3 weeks with changes every 3 days. Decalcified samples were embedded in paraffin, sectioned at 7 mm of thickness and stained with H&E. Quantification of Gene Expression in Muscle To examine gene expression in the muscle following impact injury, the right quadriceps was impacted using our model and the entire muscle was collected 24 h, 1 week, and 2 weeks post-injury (n ¼ 6/time point). The contralateral muscle was collected to serve as control (no impact). Total RNA was extracted from the muscle following homogenization in the RNA-Stat 60 reagent (AMSBIO, Lake Forest, CA) and purified using RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s recommendations. Two micrograms of total RNA was reverse transcribed in 100 ml of reaction mixture containing 1 PCR Buffer, 1 mM deoxynucleoside triphosphate (DNTP) mix, 5 mM random primers, 7.5 mM MgCl2, 0.4 U/ml RNAase inhibitor, and 2.5 U/ml MultiScribe Reverse Transcriptase (Reverse Transcription Reagents, Applied Biosystems, Foster City, CA) at a sequence of 25˚C for 10 min, 48˚C for 40 min, and 95˚C for 5 min. The real time PCR quantification of gene expression was performed in triplicate reactions using inventoried Taqman assays for mice including l19, gapdh, tnfa, bmp2, bmp4, bmp9, bmpr1a, bmpr2, sox9, runx2, col2, and colX in a 7300 RT-PCR System (Applied Biosystems). The reactions were performed under the following conditions: 95˚C for 10 min, 40 cycles of 95˚ for 15 s, and 60˚C for 1 min. Analysis was carried
out using the software supplied by the manufacturer and the number of threshold cycles (Ct) required for the FAM fluorophore intensities to exceed a threshold just above background were calculated. Ct values for triplicate reactions were averaged for each target gene and expressed as a ratio to that of the GAPDH Ct value. Statistical Analysis Data are reported as mean SEM with n ¼ 6 mice per experimental group. Analysis of Variance (ANOVA) was used to compare the BV in each leg between animals receiving different doses of BMP-2. Student’s t-test was used to compare the BV between the groups with and without injury at each BMP-2 dose. Statistical significance was considered when p < 0.05.
RESULTS BMP-2 Induced HO HO was first observed 2 weeks after BMP-2 injection (1.5 mg; Fig. 1). Injection of BMP-2 without the gel carrier did not induce mineral formation. At 4 and 8 weeks after injection, there was no further expansion of the mineral deposit but there was condensation and reordering of the mineral into a more organized structure (Fig. 1). No mineral could be detected after 1 week. Dosing of BMP-2 without the heparin-chitosan ionic complex did not induce HO. Increasing amounts of BMP-2 in the gel carrier dose dependently increased the amount of mineral deposited in the muscle (Fig. 2). At BMP-2 doses of 1 mg or less little or no mineral was apparent. Between 1 and 1.5 mg of BMP-2 there was a dramatic increase in mineral deposited from less that 0.5 mm3 to more than 2 mm3. Doses above 1.5 mg did not increase the amount of mineral. In most of the ossified tissues, the mineral was confined to the muscle and independent of the bone (Fig. 1). In some cases (approximately 20%), however, the mineral was closely associated with or on the bone surface (Fig. 3). This typically resulted in a dramatic increase in the amount of mineral formed. Samples with bone involvement were excluded from our analyses. Effect of Trauma on BMP-2 Induced HO Trauma alone or trauma combined with the BMP-2 carrier (0 mg) did not induce HO (Figs. 4 and 5). In mice receiving less than 1 mg of BMP-2, trauma had little or no effect on mineral deposition. In mice receiving 1 mg BMP-2, trauma dramatically increased HO formation (Fig. 4). Quantitative BV analysis showed that injury increased heterotopic bone by sixfold. At higher, saturating doses of BMP-2 muscle injury did not affect the amount of mineral deposited (data not shown).
Figure 1. Representative microCT images of HO at 1, 2, and 4 weeks after 1.5 mg BMP-2 injection without trauma. HO was first observed 2 weeks after BMP-2 injection. At 4 weeks after injection, there was no further expansion of the mineral deposit but there was condensation and reordering of the mineral into a more organized structure. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
A MOUSE HETEROTOPIC OSSIFICATION MODEL
185
Figure 4. Representative microCT images of HO 2 weeks after 0 and 1 mg BMP-2 injection with and without trauma. Trauma significantly augments HO development in the presence of BMP2. Figure 2. Volume of heterotopic bone (BV) formed 2 weeks after BMP-2 injection without trauma. (Mean SEM, N ¼ 6, p < 0.05 compared to other groups.) At BMP-2 doses of 1 mg or less little or no mineral was apparent. Between 1 and 1.5 mg of BMP-2 there was a dramatic increase in mineral deposited from less that 0.5 mm3 to more than 2 mm3.
using GAPDH and L19 for normalization were similar. All reported data were normalized to GAPDH.
DISCUSSION Histology Representative structures of the HO deposits for mice receiving 1 mg BMP-2 with or without muscle trauma are shown in Figure 6. Without injury, only a small amount of bone is formed and it appears relatively unstructured. With muscle impaction injury, there is a dramatic increase in the amount of bone formed, it appears more structured and in some cases includes marrow-like components. The bone is clearly lamellar and contains numerous osteocytes. Gene Expression The effects of muscle injury on gene expression are shown in Figures 7 and 8. Injury stimulated BMP-2 and -4 expression within 24 h and expression levels remained elevated for at least 7 days. By 14 days, levels returned to normal (BMP-4) or were below normal (BMP-2). BMP-9 was undetectable at all times. Expression of BMPR1a also increased by 24 h and returned to normal by 14 days. No change in BMPR2 expression was observed. To determine whether muscle injury could induce expression of genes associated with chondrogenesis and osteoblastogenesis we measured expression of SOX9, Col II, Col X, and RUNX2. Expression of SOX9 and RUNX2 was increased 24 h and 7 days after injury. By 14 days, levels had returned to normal. Col II and Col X were undetectable at all times. Expression of TNFa followed a pattern similar to SOX9 and RUNX2. The results
We have successfully developed a new murine model of traumatic HO that mimics the formation of heterotopic bone associated with blunt physical muscle trauma. Previously developed trauma-inclusive models in normal rodents have used BMPs in combination with CTX or physical damage induced by crushing the muscle with surgical tweezers.23 These have proven to be valuable models for the study of HO but do not effectively mimic the formation of heterotopic bone in response to severe blunt physical injury. The impact force induced in our model is in the range of many injuries in military personnel but is less than many others. Higher impact forces in our model frequently induce fracture and this can lead to dramatically different responses in HO formation. In order to study the mechanisms involved in HO we set out to develop a model that forms a consistent amount of HO and is not complicated by varying degrees of bone injury and healing.26 The development of models of HO that involve severe bone injury or loss of limbs is important but was not a part of goal of this study. Clinical HO develops through a process of endochondral bone formation.27 In our model, mineralization is not apparent 1 week after injury but by 2 weeks is clearly evident in mice receiving BMP-2 alone or BMP-2 with trauma. This mimics the lag time in the appearance of mineral observed in patients with trauma induced HO. The mineral deposit in our model appears to be confined to the site of BMP-2 gel injection suggesting that there is a need for the BMP-2
Figure 3. Representative microCT images of HO 2 weeks after 1.5 mg BMP-2 injection with and without bone involvement (no injury). Bone involvement augments the development of HO. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
186
LIU ET AL.
Figure 5. Volume of heterotopic bone (BV) formed 2 weeks after 1 mg BMP-2 injection with and without trauma. (Mean SEM, N ¼ 6, p < 0.05 compared to non-injury control.) Injury increased heterotopic bone volume by about sixfold.
to remain in the lesion for some time to initiate bone formation. What sustains the heterotopic bone is not clear but once formed there is no indication that it can be resorbed away for at least 8 weeks after treatment in our model. In about 20% of our animals, we found the heterotopic bone attached to the femur. In these instances the volume of bone formed appeared much greater than when the mineral deposit was not in contact with the femur. We think that bone involvement may be due to the needle touching the periosteum during injection. In patients with HO, the mineral deposits are frequently associated with the bone surface. It appears that irritation of the periosteum may augment mineral accumulation during the development of HO. Further studies are needed to fully establish the role of the periosteum in HO development.
Figure 6. Typical H&E staining for heterotopic bone after 1 mg BMP-2 injection with and without muscle injury at 8 weeks after injury with low (5, upper panel) and high (20, lower panel) magnifications. HO was developed at the BMP-2 injection site with and without injury. This is supported by the presence of remaining undigested hydrogel ( ) at the site of HO. Without injury, only a small unstructured bone is formed. With muscle trauma, the HO is significantly larger and better structured. The bone is clearly lamellar and contains numerous osteocytes (arrows) and marrow-like components. JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
Figure 7. RT-PCR results of BMP signaling in muscle after impaction (no BMP-2 injection). BMP-2, -4, BMPR1a gene expression significantly increased in muscle at 1 and 7 days after injury. The expression of BMPR2 remains unchanged at all time points. Open bars stand for the values of control group, striped bars stand for the values of the injured group. (Mean SEM, normalized to GAPDH, N ¼ 6, p < 0.05, p < 0.01.)
The amount of heterotopic bone formed in our model is extremely sensitive to the amount of BMP-2 deposited in the muscle. There appears to be a triggering of mineral deposition between 1 and 1.5 mg of BMP-2. If the BMPs are involved in clinical HO, and there is evidence that they are, small differences in the local concentrations may determine whether HO forms following trauma. Our RT-PCR analysis has shown that trauma significantly induces the gene expression multiple BMP family members (including BMP-2; Fig. 7). However, the local concentration of BMPs at the injury site was not measured in this study. Thus, we could not directly compare the physiological concentration of BMP-2 induced by muscle trauma to the doses of BMP-2 we used in this study. The effects of trauma on bone formation were dramatic. Combining a sub-threshold dose of BMP-2 (1 mg) with trauma resulted in a striking increase in mineral deposition (Figs. 4 and 5). The presence of exogenously provided BMP-2 appears to sensitize the muscle to trauma-induced HO. This is much like the circumstances in patients with FOP where constitutive BMP signaling sensitizes muscle to mild trauma.21,22 The cellular source of HO remains undefined. Previous studies have suggested that endothelial progenitors may be the responsible for the development of HO and BMP signaling plays a critical role in promoting endothelial–mesenchymal transition.28,29 In our study, we have observed significantly increased expression of BMP-2 and BMP-4 in muscle after trauma. Other studies suggests that muscle resident stem cells are involved in HO development and inflammation cytokines, like TNFa plays a critical role in this process.30
A MOUSE HETEROTOPIC OSSIFICATION MODEL
187
Figure 8. RT-PCR results of SOX9, RUNX2, and TNFa gene expression in muscle after impaction (no BMP-2 injection). All three genes significantly increased in muscle at 1 and 7 days after injury. Open bars stand for the values of control group, striped bars stand for the values of the injured group. (Mean SEM, normalized to GAPDH, N ¼ 6, p < 0.05, p < 0.01.)
Our data also showed significantly increased TNFa in traumatic muscle. Future studies are needed to identify the cellular source and key signaling pathways responsible for the development of HO and our model can serve as a powerful tool for this purpose. There are some limitations in this study. First, only a single strain (C57/BL6) young male mouse was tested in this study. It has been known that the strain, gender and age have significant effects on mouse bone metabolism. Thus, the conclusion from this study may not fit other mice with different strain, gender, or age. Second, the injection site on the quadriceps muscle was not strictly controlled. We have noticed that in some mice, HO developed at the proximal side of quadriceps, while in other mice, HO developed at the distal side of quadriceps muscle. We believe that this was caused by variation in the site of BMP-2 injection. Third, the histological analysis of HO was only conducted at a late time point at 8 weeks when HO has fully developed. Thus, the early stages of HO development, such as the chondrogenic phase, was not observed. In this study, we have successfully established a new trauma-induced HO mouse model that closely mimics HO in patients with severe blunt soft tissue injury. In our model a onetime muscle impact can increase the volume of HO by sixfold with a subsymptomatic dose of 1 mg BMP-2 in mouse quadriceps muscles. This novel mouse model can serve as a powerful tool in future studies to examine the molecular mechanisms involved in trauma-induced HO. It can also serve as a powerful screening tool in seeking better prevention and treatments for trauma-induced HO in the future.
ACKNOWLEDGEMENTS This work was supported by the Department of Defense (grant #W81XWH-11-2-0189), the Veterans Affairs Merit review program and the Northern California Institute for Research and Education. We thank John Carney for his helpful ideas and review of this manuscript, and Christian Santa Maria and Alfred Li for their technical support in microCT scanning and image analysis. Dr. Larm and Dr. Adolfsson are share holders of ExThera AB.
REFERENCES 1. Kaplan FS, Glaser DL, Hebela N, et al. 2004. Heterotopic ossification. J Am Acad Orthop Surg 12:116–125. 2. Kaplan FS, Tabas JA, Gannon FH, et al. 1993. The histopathology of fibrodysplasia ossificans progressiva. An endochondral process. J Bone Joint Surg Am 75:220–230. 3. Garland DE. 1991. A clinical perspective on common forms of acquired heterotopic ossification. Clin Orthop Relat Res 263:13–29. 4. O’Connor JP. 1998. Animal models of heterotopic ossification. Clin Orthop Relat Res 346:71–80. 5. Kaplan FS, Shore EM. Derailing heterotopic ossification and RARing to go. Nat Med 17:420–421. 6. Kirsch T. 2008. Determinants of pathologic mineralization. Crit Rev Eukaryot Gene Expr 18:1–9. 7. Stoltny T, Koczy B, Wawrzynek W, et al. 2007. Heterotopic ossification in patients after total hip replacement. Ortop Traumatol Rehabil 9:264–272. 8. Potter BK, Burns TC, Lacap AP, et al. 2006. Heterotopic ossification in the residual limbs of traumatic and combatrelated amputees. J Am Acad Orthop Surg 14:S191–S197. 9. Potter BK, Burns TC, Lacap AP, et al. 2007. Heterotopic ossification following traumatic and combat-related amputations. Prevalence, risk factors, and preliminary results of excision. J Bone Joint Surg Am 89:476–486. 10. Forsberg JA, Pepek JM, Wagner S, et al. 2009. Heterotopic ossification in high-energy wartime extremity injuries: prevalence and risk factors. J Bone Joint Surg Am 91:1084–1091. 11. Garland DE, Dowling V. 1983. Forearm fractures in the head-injured adult. Clin Orthop Relat Res 190–196. 12. Hendricks HT, Geurts AC, van Ginneken BC, et al. 2007. Brain injury severity and autonomic dysregulation accurately predict heterotopic ossification in patients with traumatic brain injury. Clin Rehabil 21:545–553. 13. Cipriano CA, Pill SG, Keenan MA. 2009. Heterotopic ossification following traumatic brain injury and spinal cord injury. J Am Acad Orthop Surg 17:689–697. 14. Jackson WM, Aragon AB, Bulken-Hoover JD, et al. 2009. Putative heterotopic ossification progenitor cells derived from traumatized muscle. J Orthop Res 27:1645–1651. 15. Ji Y, Christopherson GT, Kluk MW, et al. 2011. Heterotopic ossification following musculoskeletal trauma: modeling stem and progenitor cells in their microenvironment. Adv Exp Med Biol 720:39–50. 16. Brown KV, Dharm-Datta S, Potter BK, et al. Comparison of development of heterotopic ossification in injured US and UK Armed Services personnel with combat-related amputations: preliminary findings and hypotheses regarding causality. J Trauma 69:S116–S122. 17. Genet F, Jourdan C, Lautridou C, et al. The impact of preoperative hip heterotopic ossification extent on recur-
JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
188
18.
19.
20.
21.
22.
23.
LIU ET AL. rence in patients with head and spinal cord injury: a case control study. PLoS ONE 6:e23129. Baldwin K, Hosalkar HS, Donegan DJ, et al. Surgical resection of heterotopic bone about the elbow: an institutional experience with traumatic and neurologic etiologies. J Hand Surg 36:798–803. Pignolo RJ, Shore EM, Kaplan FS. Fibrodysplasia ossificans progressiva: clinical and genetic aspects. Orphanet J Rare Dis 6:80. Pignolo RJ, Xu M, Russell E, et al. Heterozygous inactivation of Gnas in adipose-derived mesenchymal progenitor cells enhances osteoblast differentiation and promotes heterotopic ossification. J Bone Miner Res 26:2647–55. Shore EM, Kaplan FS. 2008. Insights from a rare genetic disorder of extra-skeletal bone formation, fibrodysplasia ossificans progressiva (FOP). Bone 43:427–433. Kan L, Hu M, Gomes WA, et al. 2004. Transgenic mice overexpressing BMP4 develop a fibrodysplasia ossificans progressiva (FOP)-like phenotype. Am J Pathol 165:1107– 1115. Leblanc E, Trensz F, Haroun S, et al. BMP9-Induced muscle heterotopic ossification requires changes to the skeletal muscle microenvironment. J Bone Miner Res.
JOURNAL OF ORTHOPAEDIC RESEARCH FEBRUARY 2014
24. Tannous O, Griffith C, O’Toole RV, et al. Heterotopic ossification after extremity blast amputation in a Sprague– Dawley rat animal model. J Orthop Trauma 25:506–510. 25. Engstrand T, Veltheim R, Arnander C, et al. 2008. A novel biodegradable delivery system for bone morphogenetic protein-2. Plast Reconstr Surg 121:1920–1928. 26. Nozaki M, Li Y, Zhu J, et al. 2008. Improved muscle healing after contusion injury by the inhibitory effect of suramin on myostatin, a negative regulator of muscle growth. Am J Sports Med 36:2354–2362. 27. Forsberg JA, Potter BK. 2010. Heterotopic ossification in wartime wounds. J Surg Orthop Adv 19:54–61. 28. Medici D, Shore EM, Lounev VY, et al. 2010. Conversion of vascular endothelial cells into multipotent stem-like cells. Nat Med 16:1400–1406. 29. Matsubara H, Hogan DE, Morgan EF, et al. 2012. Vascular tissues are a primary source of BMP2 expression during bone formation induced by distraction osteogenesis. Bone 51:168–180. 30. Glass GE, Chan JK, Freidin A, et al. 2011. TNF-alpha promotes fracture repair by augmenting the recruitment and differentiation of muscle-derived stromal cells. Proc Natl Acad Sci USA 108:1585–1590.
