Muscle and Bone: Combating the Evil Side of the Connection.

Abstract

Muscles and bones form intimate relationships at the developmental, anatomical, mechanical, and biochemical level that are all of critical importance for the appropriate formation and function of the musculoskeletal system.1, 2 Under physiological conditions, pain-free locomotion is facilitated by a proper homeostasis of each tissue and a synergistic and coordinated interaction of the two organ systems. However, disturbances of these tightly regulated processes can cause pain, disability, immobility, and may eventually lead to a lethal fate.2, 3 Thus, providing more insights into the regulation of muscle-bone interactions under physiological and disease conditions is a very powerful approach for the development of treatments to cure debilitating musculoskeletal maladies.4 Earliest evidence for the close relationship between muscle and bone is provided by the mesodermal origin of both tissues that give rise to somites. During development, somites differentiate into the dorsal dermomyotome and the ventral sclerotome that eventually form muscle and bone, respectively.1 While being formed during embryogenesis, muscle and bone develop in a tightly controlled manner under the influence of a plethora of mechanical and biochemical cues. For instance, intrauterine muscle contraction by the embryo determines the specific circumferential shape of each bone and joint development.5 During adolescence and in the adult, muscle contraction imposes the greatest load on bone, with a linear relationship between bone mass and muscle size and force.6, 7 Consequently, loss of bone and muscle mass in the elderly is coupled, leading to osteoporosis and sarcopenia with subsequent falls and fragility fractures.4, 8 These conditions are among the greatest challenges in today's musculoskeletal medicine and the development of effective treatments that provide beneficial effects to both bone and muscle is an important goal of ongoing research in the field.4, 8 Beyond mechanics, muscle and bone are subject to a stringent regulation by soluble ligands and signaling pathways during development and tissue homeostasis in the adult.2 For instance, canonical Wnt signaling,9 insulin-like growth factor 1 (IGF-1),10 transforming growth factor-β (TGF-β),11 and bone morphogenetic proteins (BMPs),11 among other pathways, are well established to contribute to the regulation of the differentiation and function of bone and muscle tissue. Downstream of these pathways, cell type–specific transcription factor networks are engaged to control the differentiation and function of bone and muscle cells.1 Among the signaling pathways that control bone and muscle physiology, the BMP pathway assumes a central role. Historically, in 1965 Marshall R. Urist discovered that the intramuscular implantation of demineralized bone matrix induces the formation of osseous tissue, an observation that led to the discovery of BMPs.12 Furthermore, this experiment clearly demonstrates that bone can be formed within muscle tissue, provided an appropriate stimulus is present. This finding has been confirmed in vitro using C2C12 myoblasts, which transdifferentiate into matrix-producing osteoblasts upon stimulation with BMP-2 or BMP-4.13, 14 Although the discovery of BMPs has paved the way for the field of bone regeneration, formation of osseous structures in soft tissue can occur under benign and malignant conditions and may cause tremendous problems for the affected individuals.15 The most frequent nonmalignant form of heterotopic ossification (HO) is a complication frequently seen after trauma, traumatic brain injury, burns, and major orthopedic surgeries such as hip arthroplasty or acetabular fractures.16 The rate of disability is variable but patient well-being can be compromised by dysfunction, joint contractures, and severe pain.16 It has been reported that an activated immune system plays a critical role in the formation of HO by providing local and systemic inflammatory cues that affect oxygen tension and pH and enhances the osteogenic potential of mesenchymal precursor cells through BMP-mediated SMAD signaling, which ultimately leads to heterotopic bone formation.(16–19) Thus, nonsteroidal anti-inflammatory drugs such as cyclooxygenase (COX)-1 and COX-2 inhibitors and radiation are frequently used in the clinics to prevent HO, whereas a well-planned surgical removal of existing ossifications might alleviate the burden of the patients.16 Fibrodysplasia ossificans progressiva (FOP; MIM# 135100) is the most catastrophic form of extraskeletal ossification, and was first described by the English surgeon John Freke in 1736.20 FOP is a rare autosomal-dominant genetic disorder with a prevalence of about one in two million and it is an intractable and extremely disabling disease with a considerably shortened lifespan.3, 21 Its appearance is characterized by congenital malformation of the great toes and sporadic but progressive postnatal ectopic bone formation starting during the first decade of life.3, 21 Classical features of this disease are periodic flare-ups, which are painful inflammatory processes with soft tissue swellings that often occur in response to minor soft tissue injuries, viral infections, and intramuscular injections. Flare-ups then convert skeletal muscle, ligaments, tendons, fascia, and aponeuroses into bone, thereby causing ankylosis of the joints of the axial and appendicular skeleton, followed by severe disability and ultimately immobility by the end of the second decade of life.3, 21 FOP is usually lethal around 40 years of age due to a thoracic insufficiency syndrome and subsequent cardiorespiratory failure.3, 21 Although FOP was described centuries ago, the underlying disease mechanism has been elusive for a long time until evidence for the involvement of BMPs was provided.22 The role of BMP signaling in the development of FOP-like lesions was then confirmed by a mouse model in which BMP-4 was overexpressed.23 More precise information about the pathophysiology of FOP was reported in 2006 by Shore and colleagues,24 who identified a missense gain-of-function mutation in the gene encoding the Activin-A receptor type 1 (ACVR1), a BMP type 1 receptor, which causes an arginine to histidine change in amino acid 206 (R206H) and keeps the mutant ACVR1 (mACVR1) in a constitutive active state due to a loss of autoinhibition. This finding provided a rationale for the ectopic bone formation. The discovery of the genetic basis of FOP has led the field to intensify the investigations on trauma-induced HO and FOP with the hope to better understand the underlying pathophysiology and to identify therapeutic approaches because there is so far no treatment for FOP.18, 21 In the context of FOP, bone formation occurs via endochondral bone formation.3 Chondrocytes are the precursors of bone-forming osteoblasts; however, the origin of the cells preceding the chondrocytes in the disease context has not yet been fully clarified. Because ossifications in FOP frequently occur in ligaments, tendons, and muscle, muscle stem cells such as satellite cells or mesenchymal progenitor cells located in the muscle interstitial space have been postulated to cause ectopic bone formation,25 whereas bone marrow transplantation experiments ruled out a role of hematopoietic cells and bone marrow progenitors.26 Based on cell lineage tracing and intramuscular transplantation assays, mesenchymal progenitor cells that are platelet-derived growth factor receptor α (PDGFRα)-positive and responsive to BMP-2 stimulation were reported as the cellular source for HO.25, 27 However, investigations of bone-forming cells from FOP patients and corresponding mouse models revealed the expression of endothelial markers by about 50% of the chondrocytes and osteoblasts located in the affected lesions, suggesting an endothelial origin of these cells.28 Using a cell lineage tracing approach, this hypothesis was confirmed and further supported by the observation that expression of constitutive active ACVR1 in endothelial cells caused endothelial-to-mesenchymal transition (EMT), leading to a stem cell–like phenotype. These stem-like cells were then shown to differentiate into chondrocytes and osteoblasts.28 Thus, mesenchymal stem-like cells or mesenchymal progenitor cells that might also be generated by EMT from vascular endothelial cells rather than muscle stem cells may eventually serve as the main cellular source for chondrocytes and finally for matrix-producing osteoblasts causing HO in FOP.1, 27, 29 Nevertheless, whether these connective tissue progenitor cells are the sole source for HO in muscle and if these precursors give also rise to ectopic bone formation in tendons and ligaments has not yet been fully clarified. Furthermore, it remains to be elucidated if a combination of different precursors is at play in causing HO and if that might be variable under certain conditions. Formation of ectopic bone in individuals with FOP does not occur continuously but rather sporadically following trivial tissue trauma and episodic flare-ups, demonstrating that an aberrant BMP signaling in connective tissue progenitor cells due to a gain-of-function mutation in ACVR1 is necessary but not sufficient to convert one tissue into another. This strongly implicates the participation of altered microenvironmental cues such as local early inflammatory conditions, a circumstance that is well established in the pathophysiology of posttraumatic HO and for which strong evidence also exists in FOP.18, 30 For instance, BMP-4 overexpression in lymphocytes was suggested to have a role in heterotopic bone formation in FOP.22 Furthermore, immunosuppression in the context of bone marrow transplantation was effective to ameliorate HO in a patient with FOP with recurrence of the ossifications once immunosuppression was discontinued.26 These and other observations underscore the prominent role of the innate immune system although the exact mechanisms by which its various cellular players affect heterotopic bone formation is not completely understood. Nevertheless, investigating the precise role of the immune system, its components and interactions with other cells of the microenvironment as well as the effect on signaling pathways is part of ongoing research and very promising to unravel novel approaches to treat FOP.30 Inflammation is inextricably linked to oxygen tension and tissue hypoxia. Hypoxia is a very strong stimulus for which cells have developed sensing and responding mechanisms.31 Important mediators of this regulatory system are prolyl-hydroxylases (PHDs) that hydroxylate hypoxia-inducible factors (HIFs), which causes subsequent polyubiquitination by an E3 ubiquitin ligase containing the von Hippel-Lindau tumor suppressor protein, followed by proteasomal degradation.32 HIFs (HIF1 to HIF3) are heterodimeric transcription factors consisting of an oxygen-sensitive HIF-1α subunit and a constitutively expressed HIF-1β subunit and represent the central components in response to hypoxia.33 Under hypoxic conditions, PHD-mediated hydroxylation ceases, thereby stabilizing HIF-1α that translocates into the nucleus where it dimerizes with HIF-1β and forms a complex with transcriptional cofactors to regulate the expression of hypoxia-sensitive genes.34 Inflammation and consecutive tissue hypoxia on the background of a constitutively active ACVR1 signaling seem to be the main triggers for episodic flare-ups and subsequent HO in FOP. However, the underlying molecular mechanisms remain elusive but would provide a tremendous potential for therapeutic concepts to treat individuals suffering from FOP. In this issue of the Journal of Bone and Mineral Research, two reports provide compelling breakthroughs for the better understanding and treatment of FOP.35, 36 In an elegant study using patient material, in vitro assays and in vivo models, Wang and colleagues35 report profound novel insights into the altered molecular mechanism by which cellular hypoxia promotes HO in FOP. The authors made the interesting observation that early inflammatory lesions in FOP are hypoxic, which increases the nuclear localization of HIF-1α in connective tissue progenitor cells. By transcriptional repression, HIF-1α reduces the expression of RABEP1, which encodes Rabaptin-5. In the cytoplasm, Rabaptin-5 interacts with Rab5 and modulates early endosomal dynamics, a mechanism that is evolutionary highly conserved.37 Regulation of signaling activity by receptor endocytosis is a widespread mechanism that has been described for several pathways including TGF-β signaling.38 Due to the disturbed Rab5-mediated early endosome trafficking, constitutively active ACVR1 bearing the R206H gain-of-function mutation is retained in early endosomes, leading to a continued ligand-independent enhanced signaling. The ligand-independent nature of the continued mACVR1 signal was established in vitro using SHED cells, a FOP patient-derived chondro-osseous progenitor cell population,39 in which SMAD1/5/8 phosphorylation remained increased in the presence of noggin, a well-known BMP ligand antagonist. However, receptor inhibition using Dorsomorphin discontinued the signaling cascade, demonstrating a ligand-independent continued mACVR1 activity in a model of hypoxic cells of FOP lesions. In addition to these findings, cell surface density of the mACVR1 was increased under hypoxic conditions compared to a normoxic environment. Furthermore, mACVR1 was found to have a reduced capacity to bind FKBP12, a protein that is necessary to maintain the autoinhibition of ACVR1.40 However, in contrast to the concept of a ligand-independent signaling mechanism, Hatsell and colleagues41 recently reported that the R206H mutation in ACVR1 converts the non-signaling Activin-ACVR1 complex into a ligand-sensitive signaling complex, which induces SMAD1/5/8 phosphorylation. Although the exact contribution of these mechanisms to HO are unclear, it is very likely that all these features synergize toward an augmented BMP signaling in FOP SHED cells, which demonstrated an improved ability to differentiate into chondrocytes under hypoxia compared to control SHED cells.35 The implication of Rabaptin-5 in the overall mechanism was confirmed by a rescue experiment in which Rabaptin-5 was exogenously expressed in FOP SHED cells, leading to a normalization of BMP signaling activity to normoxic levels. Collectively, these findings provide very strong evidence that hypoxia paired with an augmented BMP signaling due to a persistent ligand-independent mACVR1 activity may promote the transformation of connective tissue progenitor cells into an obligate cartilage template, which then differentiates into bone-forming osteoblasts that eventually cause heterotopic bone formation (Fig. 1). The observations made by Wang and colleagues35 attribute a large part of the disease mechanism in FOP to a deregulated signaling in response to tissue hypoxia. Given the central role of HIF-1α in mediating the cellular response to low oxygen tension and in regulating the Rabaptin-5-Rab5–mediated continued mACVR1 signaling, therapeutic approaches aimed at inhibiting HIF-1α were evaluated. Consistent with the mechanistic findings, inhibition of HIF-1α using an adenovirus-Cre–induced deletion of HIF-1α in mouse embryonic fibroblasts and a clinically more relevant chemical inhibition of HIF-1α using apigenin, imatinib, or PX-478 attenuated the ligand-independent BMP signaling activity in hypoxic FOP SHED cells evidenced by a reduced SMAD1/5/8 phosphorylation. More importantly, using a transgenic mouse model of FOP in which a constitutively active mACVR1 is ubiquitously expressed and HO was induced by intramuscular injection of cardiotoxin, treatment with apigenin or imatinib greatly reduced the heterotopic bone formation, thereby maintaining the locomotion of the animals.35 These findings are consistent with a recent report describing the use of PX-478 to prevent HO in mouse models of FOP and trauma-induced HO by inhibition of HIF-1α signaling.42 Although pharmacological inhibition of HIF-1α to prevent HO is very successful in mice, the applicability in humans remains to be investigated. Heterotopic bone formation in FOP occurs exclusively through an obligate cartilage template. Among several pathways, retinoid acid (RA) signaling plays an important negative role for chondrogenesis because RA-receptor (RAR) antagonism accelerates chondroblast differentiation.43 Previous studies have therefore explored the possibility to prevent the formation of the cartilage template in FOP by perpetuating RA signaling. For instance, using a RARγ agonist, cells were rendered insensitive to BMP-2 treatment and injury-induced HO was prevented in mice bearing a constitutively active mACVR1.44 However, in this study a ACVR1Q207D mutation was used to obtain a gain-of-function mutation, a model that does not fully recapitulate the human FOP phenotype,45 which is due to a ACVR1R206H mutation. In this context, Chakkalakal and colleagues36 report in this issue the use of a novel knock-in mouse model that bears the ACVR1R206H mutation globally and in Prrx1-positive mesenchymal cells of the appendicular skeleton and recapitulates the human disease more precisely. As a consequence of an ongoing mACVR1 signaling during growth, the structure of the growth plate was altered, leading to a retarded chondrocyte-to-bone differentiation and subsequent mild reduction of longitudinal growth. Indicating the implication of RAR signaling as one of the underlying mechanisms, palovarotene, a RARγ agonist, restored the skeletal growth retardation. More importantly, palovarotene was very effective in reducing cardiotoxin-induced and spontaneous HO in mutant mice, thereby preserving the mobility of the animals.36 In addition to its function as RARγ agonist, palovarotene might also exert direct or indirect effects on other pathways such as dampening BMP signaling,44 which might elicit synergistic effects to reduce the transformation of activated connective tissue progenitor cells into an obligate cartilage template (Fig. 1). Thus, further characterization of a context-dependent signaling crosstalk in FOP will be a very important line of future investigations to provide more insights into the molecular aberrations of the disease and to unravel more potential therapeutic targets. In summary, FOP is the most devastating form of HO and a debilitating disease that causes early lethality for which no treatment exists so far. Concurrently, two studies report novel insights into the disease-causing molecular mechanisms and provide compelling evidence to interfere with an undesired connection between muscle and bone by preventing a detrimental tissue conversion. Wang and colleagues35 report new knowledge on the thus far unknown role of HIF-1α in the hypoxia-induced continued BMP signaling and subsequent induction of HO in FOP, thereby identifying HIF-1α as a molecular target for a novel treatment. Given that apigenin, imatinib, and PX-478 are in clinical use for the treatment of various malignant diseases,46-48 this opens the possibility that they could perhaps also be used to treat individuals with FOP. Chakkalakal and colleagues36 describe the use of palovarotene as an effective approach to prevent ectopic ossification and mobility in a novel FOP disease model. Of note, palovarotene is currently being tested in clinical trials for the treatment of FOP (www.clinicaltrials.gov). Overall, it will be of tremendous interest to determine if these compounds will indeed be effective to alleviate the burden of patients with FOP, for instance by the time flare-ups arise or as adjuvant therapy following surgical excision of debilitating ectopic ossifications. Furthermore, it would be interesting to explore whether these approaches might also be of use in treating posttraumatic forms of HO. The author states that he has no conflicts of interest. This work was supported by a grant from Deutsche Forschungsgemeinschaft (HE 5208/2-1).