Harnessing the Parathyroid Hormone,Wnt,and Bone Morphogenetic Protein Signaling Cascades for Successful Bone Tissue Engineering

Introduction

Tissue engineering holds great promise as a way of enhancing the normal regenerative potential of bone. By deconstructing the skeleton into its components and examining how each component influences the reparative response, it is clear that cells resident in bone, bioactive molecules produced by these cells and those brought into bone via the circulation, together with the unique extracellular matrix (ECM) that makes up the bone itself are involved in a continuous and ever-changing set of reciprocal interactions during regeneration. It is not surprising then that the speed and completeness of bone healing can be improved by the exogenous application of bioactive factors known to participate in bone formation. To date, fibroblast growth factors, Wnts, Indian hedgehog, insulin-like growth factor, transforming growth factor β, platelet-derived growth factor, vitamin D, cyclo-oxygenase 2, bone morphogenetic proteins (BMPs) and parathyroid hormone (PTH) have been reported to enhance the quality of repair, augment the size of the repair tissue, shorten the time to needed remodel the bone regenerate, and affect the strength of the resulting bone when applied in rodents, pigs, sheep, and monkeys.^1–15 Collectively, these studies tell us that a series of interactive signaling cascades function during normal healing and influence discrete steps in the healing process.

What is currently missing from our analysis of bone regeneration is a clear understanding of how the interactions of the signaling pathways activated during repair result in a successful healing response, knowledge that is likely to be required for us to develop optimal tissue engineering solutions aimed at enhancing bone regeneration. Fortunately, advances in our understanding of the biology of tissue regeneration using a variety of organ systems including the skeleton, coupled with data obtained in the clinical evaluation of drug candidates that regulate adult bone mass and affect bone repair are available, suggest that we focus our attention on three signaling pathways with potent effects on new bone production in the adult skeleton; PTH, Wnts, and BMPs. While other bioactive factors will certainly have important effects on bone regeneration, a thorough understanding of the ways in which these pathways cooperate in building new bone is likely to provide us with novel combinatorial therapies that can be used to enhance bone regeneration.

PTH Signaling in Bone Regeneration

Intermittent PTH (iPTH) is the only FDA approved skeletal anabolic agent. Clinical studies have demonstrated its effectiveness for increasing bone mass and reducing fracture risk in osteoporotic individuals. ¹⁶ In addition to success in fracture prevention, iPTH has been shown to affect bone healing in several preclinical settings. In fracture studies, daily injections of iPTH accelerate long bone repair by enhancing fracture callus size in both young and aged animals. ¹⁷ In mandibular fractures, iPTH has been demonstrated to speed up the healing process and increase the size of the fracture callus when given in the first 7 days after fracture. ¹⁸ These local effects of iPTH on the fracture callus were coupled to gains in trabecular bone mass throughout the skeleton as would be expected for an agent that is delivered in the circulation. Additional data from preclinical models show that iPTH enhances spine fusion in a rabbit model, where daily injection augmented fusion mass volume and fusion mass quality. ¹⁹ iPTH also stimulated bone formation in a rat model of stress fracture healing by accelerating intracortical bone remodeling induced by microdamage. ²⁰ An examination of the effects of iPTH on femoral allograft integration into host bone in a mouse bone-grafting model showed that daily subcutaneous injections of iPTH 1 week after surgery increased union rate and graft strength, suggesting it will be a potent adjunct to bone grafting for massive allograft surgeries. ¹⁵ It should be noted that in these preclinical studies, the dose of iPTH that produced positive effects on bone regeneration greatly exceeded the dose approved for fracture prevention. This discrepancy may reflect a difference in target cell response or may be due to the activity of other signaling pathways active at the repair site that influence the efficacy of iPTH that are not active during bone remodeling.

Clinical studies in patients receiving iPTH for enhancement of fracture healing have shown promising but equivocal results. Significant benefit was observed in the treatment of femoral neck fractures, where iPTH therapy was associated with decreased time to walking, reduced reporting of pain upon movement, increased hip bone mineral density, and an overall increase in the quality of life as assessed by the patient. ¹⁶ In contrast, results from the first prospective, randomized, double-blind placebo-controlled clinical trial of iPTH (1-34) treatment for acceleration of distal radius fracture healing produced confusing results; the highest dose of iPTH used failed to reduce the time to healing while a lower dose of iPTH produced significant improvement. ¹⁶ This lack of dose-related effectiveness remains unexplained and may be due to the patient population chosen for study. In contrast to these results, several reports of the successful use of iPTH as adjuvant therapy for delayed healing suggest it may be a reliable treatment option in this very difficult setting.^21,22

When administered, iPTH is reported to target several of the cell populations within bone. iPTH treatment increases the commitment of mesenchymal stem cells resident in the bone marrow to the osteoblast lineage; iPTH also increases osteoblast maturation, and reduces osteocyte production of sclerostin.^23–26 This effect on sclerostin connects the PTH signaling pathway to that of Wnts, as sclerostin is a potent Wnt signaling antagonist that is made almost exclusively by osteoblasts and osteocytes. By reducing sclerostin production, iPTH positively regulates local Wnt signaling in the bone microenvironment, leading to increased bone formation and a stimulation of bone remodeling. In addition to sclerostin, iPTH also regulates the expression of other secreted Wnt signaling antagonists including Sfrp4, Dkk1 and Wif1, and reducing these inhibitory signals may also be involved in iPTH induced increases in bone formation. As Wnt signaling antagonists other than sclerostin are not exclusive products of bone cells, it is likely that systemic administration of iPTH increases Wnt activity in multiple tissues, although no deleterious effects have been reported due to extraskeletal PTH activity to date. iPTH has also been shown to target periosteal osteoprogenitors in intact bone, enhancing their differentiation into osteoblasts. These periosteal cell effects appear to work through iPTH signaling interactions with both the Wnt and BMP signaling pathways.^23–26

Wnt Pathway Activators

The importance of Wnt signaling in the skeleton and the complexity of the Wnt signaling pathway are well documented. ³ In general, increasing Wnt signaling correlates with enhanced bone formation through increased differentiation of osteoblasts from bone marrow stromal progenitors, activation of osteoblast activity, and inhibition of osteoclast activity. Current clinical data support the strong anabolic effects of Wnts on bone mass. Treatment of healthy men and postmenopausal women with Amgen 785, a humanized anti-sclerostin monoclonal antibody, resulted in statistically significant increases in bone mineral density of up to 5.3% at the lumbar spine and 2.8% at the hip 85 days after initial treatment. The magnitude of the effect of one dose of anti-sclerostin therapy on bone formation was similar or greater than observed for 6 months of daily iPTH injections, suggesting that anti-sclerostin therapy has great potential in combating osteoporosis. ²⁷ These results also suggest that directly activating Wnt signaling in the skeleton may be more robust than modulating it indirectly through iPTH administration. As more information on the molecular pathways activated by anti-sclerostin therapy in humans emerges, it will be easier to determine if changing the iPTH dose and or timing of administration can produce comparable results. ²⁶

A large and ever-growing number of preclinical studies have examined the utility of enhancing Wnt signaling as a means of promoting bone regeneration. A general conclusion from these experiments is that increasing Wnt signaling has a positive effect on bone repair. For example, LiCl treatment, through activation of Wnt signaling, improves fracture repair in mice, as does targeted inhibition of the Wnt pathway antagonists sFRP1, Dkk1 or sclerostin.^2,4,28,29 Most recently, the efficacy of biweekly anti-sclerostin therapy was examined in a rat femoral fracture model and in a monkey fibula osteotomy model. ³⁰ As would be predicted from prior data, anti-sclerostin treatment resulted in a general anabolic effect on the skeleton and there was an observed increase in bone mass and callus volume at the regenerate site in both species. However, anti-sclerostin treatment did not appear to change the defect union rate in either model, consistent with the idea that Wnt signaling has less pronounced effects on the periosteal surface than on more mature skeletal cells. ³⁰ These data suggest that efficacious use of anti-sclerostin therapy will be dependent on the specifics of the bone regeneration scenario.

Moreover, it appears that the optimal amount of Wnt signaling changes at each stage of bone regeneration, and having too much or too little Wnt signaling can interfere with healing. ³¹ For example, local administration of Wnt3a, a protein that directly activates canonical Wnt signaling has both positive and negative effects on calvarial bone regeneration. Quarto et al. ³² reported that a healing benefit of Wnt3a treatment was found in adult mice but not in younger animals where increasing Wnt signaling reduced healing of parietal and frontal bone defects. Equivocal effects on healing were also observed in studies using mice engineered to express increased or decreased levels of β-catenin, a signal mainly activated by Wnts, highlighting the complex nature of Wnt signaling in bone. These results are not unexpected based on the knowledge that Wnt signaling targets many different cell types within the bone microenvironment including skeletal progenitors residing in the periosteum that are capable of becoming chondrocytes and osteoblasts, bone marrow stromal osteoprogenitors, osteoblasts and osteocytes. ³³ While endogenous Wnt signaling occurs in a temporally and spatially specific manner, current treatment methods do not allow for this level of control, making it possible to enhance some aspects of bone regeneration while inadvertently inhibiting others. As bone regeneration occurs over a fairly long time period, it will be important to determine when and for how long Wnt signaling should be increased to create an optimal healing environment at each site of repair. ³⁴ It is also important to recognize that Wnt signaling is not specific for bone but is a pathway active in most tissues. Our ability to enhance it only in the skeleton by using agents with high specificity for bone such as anti-sclerostin therapy will provide the control needed without complications arising from enhanced signaling in non-target tissues.

The BMP Signaling Pathway

BMPs are the most thoroughly studied of the bone regeneration molecules and have received FDA approval for applications related to bone regeneration in humans. It is surprising then that after years of study, our knowledge of how to use BMPs is remains incomplete, and as such, their clinical utility is limited by problems with optimal delivery and an understanding of the target cells that BMPs act on to affect repair. In normal bone, BMPs are made by osteoblast lineage cells and when secreted may exert their actions locally or be bound up by ECM proteins such as heparan sulfate proteoglycans, fibrillin and matrix Gla-protein, that serve to sequester or enhance their activity in a context and concentration-dependent manner. ³⁵ In addition to BMP:ECM interactions, BMPs are tightly regulated by the presence in the bone microenvironment of several structurally distinct BMP antagonists that alter the ability of BMPs to bind to their receptors, effectively blocking BMP action. To date, the BMP ligand antagonists noggin, gremlin and follistatin and the BMP receptor antagonist BMP3 have been identified as osteoblast products that block BMP activity, and at least one BMP antagonist, BMP3 appears to be a physiological regulator of adult bone mass.^36,37

While it has been established that BMPs are required for the rapid bone formation that occurs after birth, and also for the development and maintenance of the stem cell niches necessary for bone regeneration, ³⁸ we are still learning of the roles that individual BMPs have in adult bone. The availability of mice with conditional inactivation of specific BMP genes has pointed to a specific role for BMP2 in the initiation of fracture healing but not in the maintenance of bone mass, suggesting that the target cell responsive to BMP2 in the adult skeleton is the periosteal osteoprogenitor. Mice lacking periosteal expression of BMP2 exhibit spontaneous fractures and do not form a viable fracture callus, findings directly supportive of the observations that delivery of BMP2 to the fracture site in wildtype mice enhances fracture healing by increasing the chondrogenic potential of periosteal cells.^39,40 This is also the case in studies of bone graft incorporation where addition of BMP2 alone or in combination with stem cells improves graft incorporation into host bone. ¹⁴ The fact that BMP2 administration is most effective at the time of fracture or in the early phases of the healing response is consistent with idea that periosteal cell is the primary target of BMP activity, and in fact lineage tracing studies in mice have shown that local periosteal skeletal progenitors respond to BMP2 by differentiating into chondrocytes.^40,41 It is important to realize that other progenitor cells can respond to BMP2, so local delivery may be key in obtaining desired response and avoiding the possibility of heterotopic ossification. ⁴²

It is likely that the BMP signaling pathway intersects with both Wnt signaling and PTH signaling, as the pathways share skeletal target cells. Within the periosteal cell target population, changes in BMP2 levels within periosteal cells appears to be one of the earliest indicators of initiation of a response after fracture but the signal that encourages periosteal cells to initiate local BMP2 synthesis has not been identified.^43,44 In osteoblasts, it appears that BMP signaling is downstream of Wnt signaling. ⁴⁵ For the most part, how the BMP and Wnt signaling pathways interact at various stages of osteoblast differentiation and function remains largely undefined, and data from several in vitro and in vivo model systems is sometimes contradictory about how these pathway are interconnected. As a whole, these studies suggest a complex and perhaps constantly changing relationship between the requirement for each signal during the repair process, and one that may be greatly influenced by the site and extent of the injury and age of the patient.

The BMP and PTH signaling pathways appear to intersect at the level of the osteoprogenitor cell where iPTH treatment increases BMP activity by directly enhancing mRNA levels of BMP2 in vitro ⁴⁶ and indirectly through reduction of BMP3 expression (Kokabu S and Rosen V, unpublished) and these observations may be the basis for the finding that the treatment of bone defects in rats or rabbits with combinations of BMP2 or BMP7 and iPTH is more efficacious than treatment with either single factor.^47,48 However, like the BMP/Wnt pathway interactions, those of BMPs and PTH are likely to be complex. For example, it has recently been suggested that in human bone PTH inhibits BMP target gene expression, leading to the hypothesis that an inhibition of BMP signaling by PTH may over time limit the availability of mature osteoblasts on bone surfaces and thereby to the observed waning of the anabolic response to PTH. ⁴⁹

Conclusions and Future Directions

Carefully defining target cells and signal pathway interactions at each stage of bone regeneration will be key to successful use of bioactive factors for bone tissue engineering. This knowledge will be required for key therapeutic decisions such as whether sequential delivery or combinatorial delivery of several bioactive factors is required. As regeneration scenarios can be quite different from one another, and differences such as age of the patient, defect size, and underlying medical problems all influence successful repair, the optimal signaling cocktail is likely to be indication specific and based on the dominant target cell population that will be affected. ⁵⁰ A number of important unanswered questions that will influence our decisions remain to be answered. These include: (1) is there one initiating signal or are multiple pathways activated simultaneously; (2) is there a single target cell that responds to the signal or are there many different target cells that respond simultaneously; and, (3) once target cells are activated do they drive the repair by providing signaling molecules in a temporally specific manner? These are difficult questions to address and it is likely that the answers will be somewhat repair specific. One way to develop a data set for iPTH, anti-sclerostin antibody and BMPs, agents currently in hand, would be to utilize these factors in novel combinatorial ways to determine how healing is affected. This could quickly be accomplished in a series of animal models that target important clinical indications. Once completed, these experiments would provide a starting point for improving our understanding of the utility of bioactive molecules in demanding bone regeneration indications.