Bone formation takes place only at specific times and sites during the growth and remodeling of the skeleton, implying the existence of exquisite control mechanisms. It is thought that at least one of these mechanisms involves the production of antagonists of the powerful factors that stimulate the differentiation of osteoblasts from their multipotential mesenchymal progenitors. Thus, the activity of insulin-like growth factors (IGFs) are blocked by IGF-binding proteins; the actions of bone morphogenetic protein are antagonized by members of the cysteine knot family of proteins such as noggin and gremlin; and Wnts are inhibited by a variety of factors, including another cysteine knot family member, sclerostin.
Secreted Wnt proteins bind to co-receptors comprised of low-density lipoprotein receptor-related protein 5 (LRP5) or LRP6 associated with a member of the Frizzled family of receptors (of which there are 10 members). This leads to stimulation of a so-called canonical intracellular signaling cascade that culminates in the stabilization of β-catenin, which partners with lymphoid enhancer binding factor/T cell factor (LEF/TCF) transcription factors to activate the expression of specific target genes.(1) Manipulation of the components of this pathway in mice has shown that canonical Wnt signaling stimulates bone formation by promoting commitment of multipotential mesenchymal stem cells to the osteoblast lineage, stimulating the replication and further differentiation of preosteoblasts, and activating prosurvival pathways in osteoblasts and perhaps osteoblast progenitors.(1,2) Antagonists of Wnt signaling include secreted Frizzled-related proteins that act as soluble decoy receptors for Wnt proteins; Dickkopf proteins that downregulate the LRP5/6 co-receptor; and sclerostin (the product of the Sost gene), which binds to LRP5/LRP6 and thereby prevents activation of Wnt signaling.(3) However, Wnts and sclerostin do not compete for the same binding site.
Sclerostin represents a classic negative feedback inhibitor of osteoblast production because it is constitutively expressed exclusively by osteocytes, a final differentiated cell type of the osteoblast lineage. The physiological relevance of sclerostin for the regulation of bone formation was first appreciated when the excessive bone mass seen in patients with Van Buchem's disease or sclerosteosis was associated with absence of SOST gene expression.(4,5) This was confirmed by the high bone mass phenotype of sclerostin-null mice.(6) Interestingly, mutations in the LRP5 domain involved in sclerostin binding result in high bone mass in both mice and humans, and one of them (LRP5G171V) abolished sclerostin binding.(7)
The paper by Li et al.(8) published this month in the Journal of Bone and Mineral Research adds considerable weight to the notion that antagonists of pro-osteogenic growth factors normally suppress bone formation in the adult and shows the feasibility of reversing osteoporosis by inhibiting such antagonists. Specifically, Li et al. report that administration of a neutralizing antibody against sclerostin to 19-mo-old rats, 13 mo after ovariectomy, restored bone mass in the vertebrae and hindlimbs to sham levels within 3 wk. After 5 wk of administration, bone mass was well above sham values. μCT and histomorphometric measurements at the distal femur showed a dramatic increase in trabecular bone in antibody-treated animals. Antibody administration was also accompanied by increased femoral and vertebral bone strength. Histomorphometric measurements of cancellous bone at the proximal tibia and lumbar vertebrae showed that the increased bone mass was associated with increased osteoblast surface and bone formation rate. Mineralizing surface was increased by 4- to 5-fold, reaching 50–60%. Interestingly, osteoclast number was decreased by the anti-sclerostin antibody, suggesting that the increased bone formation was de novo and did not involve activation of remodeling. Cortical thickness at the femoral diaphysis was increased because of increased bone formation on both endosteal and periosteal bone surfaces.
The findings of Li et al. imply that Wnt ligands are constitutively produced in both cancellous and periosteal bone and that their proosteogenic actions are continually opposed by sclerostin secreted by osteocytes. This indicates that the Wnt signaling necessary for induction of osteoblast differentiation at the correct time and place must be triggered either by local downregulation of sclerostin or by a local increase in Wnt ligand expression to overcome the sclerostin shield. In support of the former, sclerostin production by osteocytes is downregulated by application of mechanical strain and increased by unloading and is associated with corresponding changes in bone formation.(9)
Sost gene transcription is also sensitive to circulating level of PTH.(10,11) Continuous elevation of PTH by infusion or with a calcium-deficient diet causes rapid sustained suppression of Sost gene expression resulting in dramatic lowering of sclerostin protein levels in osteocytes. This response may contribute to the increased generation of osteoblasts needed to refill resorption cavities created by the increased RANKL-mediated osteoclastogenesis that occurs with continuous PTH elevation. Consistent with this contention, our group has shown that expression of a constitutively active mutant of the PTH receptor specifically in osteocytes decreased sclerostin expression, increased Wnt signaling, and greatly augmented bone formation and bone mass.(12) Local production of other cAMP-inducing factors, for example, prostaglandins, could also be involved in the regulation of sclerostin.(13) Taken together with the control of sclerostin synthesis by mechanical strain, these considerations point to an important role of osteocytes in mediating the effects of mechanical strain, locally produced factors, and systemic hormones on bone formation.(14)
Injection of PTH (which is rapidly cleared from the circulation) also induces a rapid transient suppression of Sost gene expression in mice.(10) Consistent with a role of sclerostin in the anabolic effect of intermittent PTH, transgenic mice overexpressing sclerostin exhibit a reduced anabolic response, in both cortical and trabecular bone, to intermittent PTH compared with controls; the anabolic response in Sost null mice is also impaired.(15) The fact that the anabolic response was not completely abolished in the Sost transgenic mice indicates that that the well-known stimulatory effects of PTH on the expression of pro-osteogenic growth factors, and the direct effects on osteoblast survival, also contribute to PTH anabolism.(16)
The anabolic effect of anti-sclerostin antibody resembles the response of the skeleton to daily injections of PTH but differs in several important respects. For example, the anti-sclerostin-induced increase in bone mass appears considerably greater than that induced by intermittent PTH. This may reflect the fact that the suppression of Sost expression by injected PTH is only transient, whereas anti-sclerostin administration inhibits the antagonist in a sustained fashion. Another difference is that the anabolic effect of intermittent PTH is attenuated by bisphosphonates,(17) perhaps because the stimulatory actions of the hormone on osteoblast number occur mainly at sites of remodeling.(18) Consistent with the view that the predominant response to sclerostin inhibition is de novo bone formation on previously quiescent surfaces, the effect of anti-sclerostin antibody in osteoporotic rats is not affected by alendronate.(19) This finding also implies that blockade of sclerostin will lead to bone anabolism in the many osteoporotic individuals already taking bisphosphonates to preserve skeletal assets.
Despite these advances, much remains to be learned about Wnt signaling and the role of locally produced antagonists in the regulation of bone formation. For example, it is possible that autocrine actions of sclerostin on the production of another osteocyte secretory product explains the inhibitory effect of sclerostin on the development and survival of osteoblasts. Furthermore, bone formation rate declined and osteoclasts increased in mice 8 days after selective ablation of osteocytes (and therefore sclerostin production),(20) indicating that osteocytes produce other factors besides sclerostin that control bone formation and bone resorption. Finally, recent studies showed that the pro-osteogenic effect of LRP5 signaling is caused by suppression of serotonin production in enterochromaffin cells in the gut, leading to increased osteoblast progenitor replication secondary to release from the anti-mitotic effects of serotonin.(21) Interestingly, expression of the sclerostin-insensitive mutant LRP5G171V protein in the gut mimics the high bone mass phenotype of the globally expressed mutant. If the principal impact of this mutation is diminished ability to bind sclerostin,(7) and LRP5 signaling in the gut is a predominant regulator of bone formation, it follows that the bone anabolism elicited by anti-sclerostin therapy could be caused by sequestration of circulating osteocyte-derived sclerostin. Further clarification of exactly how anti-sclerostin therapy stimulates bone formation will require determination of whether sclerostin is actually present in the circulation, a better understanding of the full impact of the G171V mutation on LRP5 function, and elucidation of the pro-osteogenic functions of LRP5 and LRP6 in the gut versus bone.(22)
Reduction of fracture incidence in osteoporotic patients can be achieved by anti-remodeling agents such as bisphosphonates, but the holy grail of osteoporosis therapy remains the restoration of the low bone mass and strength caused by sex steroid deficiency, aging, glucocorticoid therapy, or other chemotherapeutic regimens. The findings of Li et al. show that suppression of sclerostin activity represents a feasible route to bone anabolism. Moreover, they open the possibility that pharmacological approaches to inhibit other pro-osteogenic inhibitors might be similarly effective.