Building bone to reverse osteoporosis and repair fractures (original) (raw)
Overview of osteoblast differentiation. Numerous secreted factors of paracrine, autocrine, and endocrine origin influence osteoblast development and maturation. These include some bone morphogenetic proteins (BMPs), PTH, FGF, IGF, endothelin-1, and prostaglandin agonists (13–16). Only a few are used clinically in the US to promote bone formation in response to injury or aging. Recombinant human BMP2 is used clinically to mediate spinal fusions, and BMP7 (also known as OP1) is used for the treatment of nonunion of long-bone fractures that occur secondary to trauma and for which an allograft is unsuitable. Neither is approved for osteoporosis therapy, because they both have a short half-life and cannot be administered systemically. As mentioned earlier, PTH is FDA approved to increase bone formation in patients with osteoporosis; however, it can only be prescribed for two years in the US. Thus, there is a clear need for more anabolic therapies to promote osteoblast activity. Targeting the Wnt pathway is one promising avenue. Available evidence indicates that Wnts and PTH stimulate bone formation via complementary pathways (17). BMP2 also synergizes with canonical Wnts to promote bone formation, and there seems to be considerable cross-talk between the Wnt and BMP signaling pathways (18, 19).
Osteoblasts are mesenchymal cells derived from mesodermal and neural crest progenitors. As cells of the osteoblast lineage differentiate, they produce molecules essential for construction of the mineralized bone matrix and for support of hematopoiesis and angiogenesis. Sequential expression of several molecules facilitates the differentiation of the progenitor cell into a proliferating preosteoblast, then into a bone matrix–producing osteoblast, and eventually into a mechanosensory osteocyte or a bone-lining cell (Figure 2). Runt-related transcription factor 2 (Runx2) is the earliest marker of an osteoblast lineage cell and is necessary, but not sufficient, for a progenitor cell to differentiate along the osteoblast lineage (20, 21). Runx2 regulates cellular proliferation, integrates numerous signaling pathways, and controls the expression of many genes (e.g., those encoding osteocalcin, VEGF, RANKL, sclerostin, and dentin matrix protein 1 [DMP1]) throughout osteoblast maturation (22, 23). Osterix is another transcription factor essential for the formation of preosteoblasts from multipotent progenitors (24). It acts downstream of Runx2, but its expression is downregulated in bone marrow stromal cells by endothelial cells (25); osterix, in turn, impairs osteoblast differentiation (24). Osterix is also a suppressor of VEGF expression following endothelin-1 exposure (26); therefore, it might be a crucial mediator of osteoblast function within the context of the BRC. During the matrix-forming and maturation phases of osteoblast differentiation, the simultaneous coexpression of alkaline phosphatase and type I collagen sets the stage for mineralization (27). Mature osteoblasts also produce osteocalcin, RANKL, and the receptor for PTH (PTHR1), which, among other actions, regulate bone formation and resorption. As osteoblasts become embedded in the mineralized matrix, they transform into osteocytes and begin expressing several molecules, including DMP1 and sclerostin, that control bone formation and phosphate metabolism (28).
The central role of canonical Wnt signaling in regulating osteoblast lineage specification, expansion, and terminal differentiation. Osteoblasts are derived from multipotent mesodermal or neural crest progenitors. Activation of the canonical (ca) Wnt signaling pathway, manifest through β-catenin stabilization, prevents the formation of cartilage (chondrogenesis). Wnt10b prevents adipogenesis (40). The canonical Wnt signaling pathway promotes survival of all cells of the osteoblast lineage and induces the proliferation of preosteoblasts. +, canonical Wnt signaling promotes the process; –, canonical Wnt signaling inhibits the process; AP, alkaline phosphatase; Cola1, collagen α1; DMP1, dentin matrix protein 1; OCN, osteocalcin; Osx, osterix; PTHR, receptor for PTH.
Wnts and osteoblast differentiation. Wnts are secreted glycoproteins crucial for the development and homeostatic renewal of many tissues, including bone. Figure 2 depicts the central role played by the canonical Wnt signaling pathway in regulating osteoblast development. Wnts stimulate several signaling pathways by binding a receptor complex consisting of LDL receptor–related protein 5 (LRP5) or LRP6 and one of ten Frizzled (Fz) molecules (18). The canonical Wnt signaling pathway has been the most extensively studied Wnt signaling pathway in osteoblasts. It involves the stabilization of β-catenin and regulation of multiple transcription factors, namely lymphoid enhancer factor 1 (Lef1) and the T cell factors Tcf7 (often referred to as Tcf1), Tcf7L1 (often referred to as Tcf3), and Tcf7L2 (often referred to as Tcf4) (Figure 3). This pathway is active in all cells of the osteoblast lineage, including preosteoblasts, bone-lining cells, and osteocytes (29). Notably, Wnt/β-catenin signaling is a normal physiological response to mechanical loading (30) and participates in the fracture healing process (31). Wnt signaling has three major functions in osteoblast lineage cells: dictating osteoblast specification from osteo-/chondroprogenitors; stimulating osteoblast proliferation; and enhancing osteoblast and osteocyte survival. Within the BMU, Wnts influence osteoclast maturation by regulating RANKL levels in osteoblasts (32).
A partial view of the canonical Wnt signaling pathway. Wnts bind a receptor complex consisting of LRP5 or LRP6 and one of ten Fz proteins. This prevents phosphorylation of β-catenin by GSK3β and other kinases and its subsequent degradation. Of note, mutating the residues that can be phosphorylated to alanine creates stable, gain-of-function β-catenin proteins. Stabilized β-catenin accumulates and translocates to the nucleus, where it interacts with Tcf7 and related transcription factors (Lef1, Tcf7L1, Tcf7L2) to regulate gene expression. Outside the cell, molecules that sequester either LRP5 (e.g., Dkk1 and sclerostin) or the Wnt ligand (e.g., Sfrp) negatively control the canonical Wnt signaling pathway. Lithium chloride inhibits GSK3β inside the cell. APC, adenomatosis polyposis coli.
The crucial role of the canonical Wnt signaling pathway in bone cells was revealed earlier this decade in seminal studies showing that loss-of-function mutations in the gene encoding LRP5 decrease bone mass, whereas gain-of-function mutations increase bone mass in both humans and mice (33–36). The mouse models also revealed that Lrp5 is mechanistically essential for regulating osteoblast number, through both increasing the proliferation of progenitors and enhancing the survival of committed osteoblasts and osteocytes (35, 36). In accordance with these findings, adding Wnts or deleting secreted Wnt inhibitors enhances osteoblast and osteocyte survival (37–39). Since the LRP5 discoveries, many other components of the Wnt signaling pathway (e.g., Wnt10b, β-catenin, Tcf1, adenomatosis polyposis coli [APC], Axin-2, secreted frizzled-related protein 1 [Sfrp1], Sfrp4, and dickkopf homolog 1 [Dkk1]) have been deleted or overexpressed in mice. The general conclusion derived from all these studies is that activation of the canonical Wnt signaling pathway facilitates osteoblast specification from mesenchymal progenitors at the expense of adipogenesis and enhances bone mass and strength, whereas suppression causes bone loss (38, 40–50). Several of these studies also revealed that the canonical Wnt signaling pathway cooperates with the essential transcription factors Runx2 and osterix to maintain and promote osteoblast maturation (41–43, 45, 50).
Although the final conclusion (that stimulation of the canonical Wnt signaling pathway increases bone mass) was similar in the above-mentioned studies (38, 40–50), the underlying mechanisms sometimes differed. For example, constitutively active β-catenin increases bone mass by inducing the expression of the gene encoding osteoprotegerin (OPG) and thereby inhibits osteoclast maturation (41, 50). In contrast, Lrp5 gain-of-function mutations or administration of Wnt3a enhance the proliferation of preosteoblasts and prevent apoptosis of osteoblasts and osteocytes without affecting osteoclasts (35, 37, 51, 52). Thus, activation of Wnt signaling pathways through Wnt proteins binding to their receptors triggers different signaling pathways than if β-catenin is the starting point (37). β-Catenin participates in other signaling pathways, including ones induced by activated receptor tyrosine kinases and nuclear hormone receptors, and physically links E-cadherin to the actin cytoskeleton (53). Additional studies are needed to unravel the complex roles that individual Wnts have within the BMU and to determine the extent to which stabilized β-catenin can be used as a reliable surrogate for measuring Wnt activity.
Much remains to be learned about the molecular mechanisms of Wnt signaling, but it is clear that Wnts are potential targets for therapeutics designed to increase bone mass. One possible therapeutic approach would be to deliver Wnt agonists. The limitation of this approach, however, is that Wnts are historically very difficult and expensive to purify; thus, using Wnts as an anabolic agonist is impractical at this time. Efforts to therapeutically target the Wnt signaling pathway have instead focused on two alternative approaches (Figure 3): the first is to inhibit Wnt antagonists (e.g. Dkk1, sclerostin, and Sfrp1) with neutralizing antibodies (discussed below), and the second is to inhibit glycogen synthase kinase 3β (GSK3β), which is a kinase that phosphorylates β-catenin and promotes its degradation. The latter approach has been accomplished in mice, whereby administration of lithium chloride and other small-molecule inhibitors of GSK3β reversed bone loss caused by aging, estrogen deficiency, and Lrp5 mutations (54); increased the sensitivity of osteoblasts and osteocytes to mechanical loading (30); and improved fracture healing (31). Lithium chloride is a commonly prescribed mood-stabilizing drug. Two studies examining fracture risk in humans found that lithium chloride use decreased fracture incidence in a dose-dependent manner (55); however, one study found that fracture incidence did not remain lower and actually increased after lithium chloride use was discontinued (56). Thus, the long-term effects of GSK3β inhibitors and other Wnt pathway agonists on bone formation and quality require careful evaluation before they can be used clinically as bone anabolic agents.
The other therapeutic approach being actively pursued to stimulate Wnt signaling to increase bone mass is the use of humanized monoclonal antibodies to neutralize antagonists of the Wnt signaling pathway. The canonical Wnt signaling pathway can be inhibited extracellularly by sequestering either the ligand or the receptor (Figure 3). Dkk1 and sclerostin inhibit Wnt signaling by dissociating LRP5 from Fz and Wnts. Sfrps, in contrast, bind Wnts and prevent them from associating with the LRP5/Fz complex. Serum Dkk1 levels are inversely proportional to bone mass in mice (48, 49, 57) and are a prognostic biomarker of the osteolysis that is associated with multiple myeloma in humans (58). Similarly, Sfrp1 levels are inversely related to bone formation, but in mouse gene deletion studies, the results were more compelling in female mice than in male mice, indicating that Sfrp1 effects might be tied to hormonal influences (38). Sfrp1-specific neutralizing antibodies and antagonists have not yet been described. In contrast, Dkk1-specific neutralizing antibodies have been tested in several studies (59, 60). Diarra and colleagues reversed bone destruction in a mouse model of rheumatoid arthritis by administering Dkk1-specific antibodies (59). This was accompanied by increased numbers of osteoblasts, rates of bone formation, and expression of OPG and by decreased numbers of osteoclasts. Interestingly, osteophytes (bone spurs) developed in the animals treated with Dkk1-specific antibodies, which effectively converted the phenotype to an osteoarthritic model. In a different study using a mouse model of multiple myeloma, a Dkk1-specific antibody promoted bone formation in tumor-bearing and non–tumor-bearing femurs (60). The Dkk1-specific antibody reduced the number of osteoclasts expressing tartrate-resistant acid phosphatase (TRAP) and increased the number of osteoblasts producing osteocalcin. Testing of Dkk1-specific neutralizing antibodies is ongoing in preclinical animal models and these agents are likely to enter phase I clinical trials.
Although early results indicate that Dkk1-specific antibodies can promote bone formation, there are some concerns about their safety and potency. The development of osteophytes following administration of Dkk1-specific antibodies in the mouse model of rheumatoid arthritis suggests a role for Dkk1 in preventing osteoarthritis and bone spurs. However, optimization of dosage and delivery methods might control this outcome, and more studies are needed to define the appropriate doses and routes of delivery. Tumor formation is also a general concern, because the Wnt signaling pathway is activated by mutations in many cancers. However, it is well established that tumor formation requires multiple events; therefore, even though overstimulation of Wnt signaling pathways might contribute to tumorigenesis, it would be highly unlikely to be the sole initiating event. In anecdotal support of this, neither aged Dkk1 haploinsufficient mice nor humans using lithium chloride or with high bone mass due to gain-of-function mutations in LRP5 have increased cancer incidence (61).