Low-density lipoprotein receptor–related protein 5 (LRP5) was identified as a major contributor to bone health when loss-of-function mutations in LRP5 were discovered to be associated with the autosomal recessive osteoporosis-pseudoglioma syndrome, a syndrome characterized by low bone mass1, 2 and when heterozygous missense mutations in LRP5 were observed in individuals with a dominantly inherited high-bone-mass (HBM) phenotype.3-5 Decreased or increased bone formation was shown respectively to be the predominant underlying alteration in bone turnover in these phenotypes. These observations led to studies to elucidate the underlying mechanism of LRP5 effects on bone. Wnt proteins are a family of secreted glycoproteins that are active in many cells and that control multiple developmental processes, including mesoderm induction, cell fate determination, limb patterning, and organogenesis; they also have been implicated in oncogenesis.6 Wnts may induce signals through several intracellular cascades. In the canonical pathway, Wnt binds to a receptor complex that includes both a member of the frizzled (Fzd) family of seven-transmembrane receptors and of the single-pass members of the LRP family of membrane receptors, for example, LRP5 and LRP6 (Fig. 1). Binding of Wnt ligands to the Fzd/LRP5 or LRP6 complex results in phosphorylation of the intracellular cytoplasmic tail of Lrp5 or Lrp6. This disrupts an intracellular protein “destruction” complex that otherwise facilitates the phosphorylation of β-catenin by glycogen synthase kinase 3β (GSK3β) and targets β-catenin for ubiquitin-dependent proteolytic degradation. When β-catenin degradation is reduced, increased levels of nuclear β-catenin result that interact with transcription factors of the Tcf/Lef family to activate specific gene expression programs in target cells. Several in situ, ex vivo, and in vitro studies have reported that LRP5 can function as a coreceptor in the canonical signaling cascade of Wnt in bone,1, 7, 8 ultimately resulting in increased bone formation. In vitro studies also have shown that secreted antagonists for Wnt signaling, such as Dickkopf1 (DKK1), or sclerostin may bind to LRP5 (or LRP6)7, 8 and in so doing antagonize Wnt signaling through LRP5 (or LRP6). LRP5 mutations linked to HBM therefore would cause reduced binding of these antagonists to LRP57, 8 and result in increased Wnt signaling and increased bone formation. HBM shares many common features with sclerosteosis and van Buchem disease, which are also disorders of bone overgrowth and are linked to loss-of-function mutations and downregulation, respectively, of the SOST gene, which encodes sclerostin.9-11 Reduced sclerostin activity resulting in enhanced LRP5 activity and increased canonical Wnt signaling therefore could also underlie the pathophysiology of sclerosteosis and van Buchem disease. Model of the canonical WNT signaling pathway in osseous cells. Wnt signaling is regulated via binding of extracellular WNT ligands to receptors of the low-density lipoprotein receptor–related protein (LRP) family, including LRP4, LRP5, and LRP6, and the Frizzled (Fzd) family. This pathway is tightly controlled in a spatiotemporal manner. (A) Secreted antagonists such a soluble Frizzled-related protein (sFRP) and WNT-inhibitory factor 1(WIF1) physically interact with and inhibit Wnt. Other Wnt anatgonists, such as proteins of the Dickkopf (DKK) family, bind to LRP receptors and also may interact with members of the Kremen (Krm) family of transmembrane proteins, thus inhibiting canonical Wnt signaling. Sclerostin (Scl), produced by osteocytes (Oc), also may bind LRP and antagonize its capacity to transduce Wnt signaling. Mutations in SOST, which encodes Scl, results in decreased binding of Scl to LRP4 and LRP5. In the absence of Wnt signaling, β-catenin (β-Ctnn) is associated with a cytoplasmic complex containing axis inhibitor (AXIN), the adenomatous polyposis coli (APC) protein, casein kinase I (CKI), and glycogen synthase kinase 3β (GSK3β). This promotes phosphorylation of β-Ctnn and its interaction with the ubiquitin ligase β-transducin repeat–containing protein (β-TRCP), leading to the ubiquitination of β-Ctnn and its degradation by the proteosome. Wnt target genes are repressed under these circumstances by the T-cell-specific transcription factor/lymphoid enhancer–binding factor (Tcf/Lef) family of transcription factors in association with other coregulators. (B) In the active state, Wnt binds to the receptor FZD and the coreceptor LRP, which is phosphorylated, leading to the recruitment of Dishevelled (Dsh) proteins and inhibition of the protein complex required for the phosphorylation of β-Ctnn, in part via increased phosphorylation of GSK3β, which inhibits the activity of this enzyme. This leads to accumulation of unphosphorylated, stabilized β-Ctnn in the cytoplasm, which then translocates to the nucleus. Hypophosphorylated β-Ctnn in the nucleus interacts with Tcf/Lef and other coregulators to activate Wnt-responsive genes generally involved in cell fate determination, proliferation, differentiation, adhesion, and survival. (C) LRP4 and LRP5 appear to transduce responses that predominantly increase osteoblastic activity and bone formation, whereas LRP6 appears to transduce responses that predominantly inhibit osteoclastic activity and bone resorption, but the precise mechanisms of these effects remain uncertain. In vivo evidence for a mechanistic role for LRP5 in bone formation was provided by a global knockout of Lrp5 (to produce Lrp5−/− mice) in which a decrease in osteoblasts, osteopenia, and persistent embryonic eye vascularisation were observed and which therefore recapitulated the human osteoporosis-pseudoglioma syndrome.12 A direct role for LRP5 in bone has been proposed based on studies in transgenic mice that express the Lrp5 HBM mutation in osteoblast-lineage cells.13 These mice were shown to have a similar phenotype to patients with HBM, including increased bone formation, an increase in osteoblasts with reduced osteoblast and osteocyte apoptosis, but no increase in osteoclastic bone resorption. This report was recently extended by Cui and colleagues,14 who found that inducing the Lrp5 HBM allele in late osteoblasts and osteocytes produced an increase in bone mass, whereas inactivation of Lrp5 in osteocytes caused a decrease in bone mass compared with wild-type mice. Furthermore, induction of an Lrp5 HBM mutant in cells that form the appendicular skeleton, but not in cells that form the axial skeleton, resulted in increased bone mass in the appendicular skeleton but not in the axial skeleton, indicating that signaling of Lrp5 mutants functions locally. In addition, the skeletal anabolic response to mechanical loading has been reported to be completely blocked in Lrp5−/− mice.15 An osteoporosis phenotype also has been described in humans with a mutation in the LRP6 gene,16 and mice carrying a heterozygous null mutation in Lrp6 have low bone mass.17 Mice harboring a spontaneous Lrp6 hypomorphic mutation, ringelschwanz, also exhibit reduced trabecular bone mineral density (BMD) and low bone volume, but with evidence of increased bone resorption.18 These findings potentially could point to an important differential control mechanism in which LRP5 would preferentially enhance osteoblastic activity and bone formation through the canonical Wnt pathway, whereas LRP6 also acting in osteoblastic cells via canonical Wnt signaling but at a different stage of development or in a different cell context would inhibit primarily osteoclastic bone resorption. Conditions for uncoupling resorption and formation therefore would be possible, and bone modeling therefore would be optimized, an especially important consideration during growth. The observation that antibodies to sclerostin both inhibit resorption and stimulate formation19 then might be explained by inhibition of sclerostin interaction with LRP6 as well as with LRP5 in adult bone. The antibody effects then potentially would differ from those of SOST mutations, which appear to produce sclerostin forms that preferentially reduce binding to osteoblast-stimulating LRPs. A critically important observation, however, was that targeted deletion of β-catenin from early20 or late21 osteoblasts or from osteocytes22 in vivo produced increased bone loss predominantly owing to increased bone resorption rather than decreased bone formation. Furthermore, osteoblast-specific deletion of Apc, another component in the canonical Wnt signaling pathway that normally prevents degradation of β-catenin (Fig. 1), produced increases in osteoid deposition, reduction in osteoclasts and an osteopetrosis phenotype, and mice carrying osteoblast-specific deletions of both the Apc and the β-catenin genes developed a bone-resorbing osteoporosis phenotype similar to that in those lacking only the β-catenin gene.21 Thus the β-catenin knockout mice appear to more closely resemble mice carrying hypomorphic Lrp6 mutations rather than Lrp5−/− mice by producing a bone-loss syndrome characterized by high bone resorption rather than reduced bone formation. Recently, an unbiased proteomics approach identified LRP4, another member of the LRP family, as a candidate sclerostin interaction partner.23 Furthermore, mutations in LRP4 were identified in patients affected by sclerosteosis but without SOST mutations. In vitro studies revealed that LRP4 mutants impair the interaction of wild-type LRP4 with sclerostin, indicating that the mutants act as dominant negatives and that LRP4 specifically facilitates the inhibitory action of sclerostin on Wnt1/β-catenin signaling. Therefore, questions regarding the bone-forming mechanism of action of LRP5 may well now extend to questions regarding the bone-forming mechanism of action of LRP4. It still remains at least formally possible, however, that targeted deletion of β-catenin in bone reflects the integration of a variety of anabolic and resorptive inputs on canonical Wnt signaling that include signaling via LRP4/LRP5 and LRP6 and via yet-to-be discovered agonists. Nevertheless, in view of this apparent contradiction between the observed effects on bone of loss of LRP5 action and of loss of β-catenin action in mouse models in vivo, as well as other considerations, Yadav and colleagues24 sought alternative mechanisms for the action of LRP5 in enhancing osteoblast activity and increasing bone mass. They found that despite low in vivo numbers of osteoblasts in Lrp5−/− mice, in vitro proliferation was not affected, suggesting that the alteration in osteoblasts was not cell-autonomous. In analysis of a microarray prepared from Lrp5−/− osteoblasts, the rate-limiting enzyme for serotonin synthesis, tryptophan hydroxylase (Tph1), was found to be a target downstream of Lrp5. Although Tph1 therefore is expressed in Lrp5−/− osteoblasts, expression of this enzyme is many times higher in gut than in osteoblasts, including the gut of Lrp5−/− mice. The authors additionally found that targeted deletion of Lrp5 in the duodenal enterochromaffin cells of mice resulted in high circulating levels of serum serotonin caused by upregulation of Tph1. Dietary restriction of tryptophan, which resulted in reduced serum levels of serotonin in Lrp5−/− mice, normalized the suppressed bone-formation parameters in these animals, suggesting that the increase in serum serotonin produced by the gut but active in bone may be responsible for this effect. Consistent with this idea, duodenum-specific expression of an Lrp5 HBM mutant also resulted in a high-bone-mass phenotype in the mice, and an increase in bone mass was reported when Tph1 was conditionally inactivated only in the intestine. In vitro, serotonin appeared to act on osteoblasts through the serotonin receptor Htr1b and inhibited osteoblast proliferation, although the dose used was rather high. The authors also have reported significant increases in serum serotonin in patients with osteoporosis-pseudoglioma syndrome and reductions in plasma serotonin in HBM.25 Furthermore, significant but weak inverse associations have been reported26 between serum serotonin levels and indices of bone mass in pre- and postmenopausal women. A model therefore was developed in which Lrp5 has no direct role in bone but instead regulates duodenal expression of Tph1 and that duodenum-derived circulating serotonin inhibits bone mass accrual. Serotonin or 5-hydroxytryptamine (5-HT) is a monoamine that, as well as being produced by Tph1 in duodenal cells, is synthesized in neurons of the brain stem by the distinct enzyme Tph2. Serotonin in the brain does not cross the brain-blood barrier but influences a broad range of behavioral, cognitive, and physiologic functions. Furthermore, Yadav and colleagues demonstrated that Tph2 null mice develop severe osteoporosis and concluded that brain stem–derived serotonin promotes rather than inhibits bone mass accrual.27 The vast majority of serotonin, however, is found outside the brain, and gut-derived serotonin functions as a paracrine factor to stimulate peristalsis and mucus secretion. The actions of serotonin are mediated by seven families of membrane-bound receptors (5-Htrs). Spatial and temporal serotonin signaling is tightly regulated by a plasma membrane serotonin transporter (5-HTT) that actively transports serotonin into cells using transmembrane ion gradients28, 29 and reduces local extracellular serotonin concentrations available for receptor activation. Pharmacologic agents that antagonize 5-HTT, such as selective serotonin reuptake inhibitors (SSRIs), therefore potentiate local serotonin activity and are used to relieve depressive symptoms. Interestingly, a null mutation in the 5Htt gene produced a skeletal phenotype in vivo characterized by decreased bone formation, reduced mass, altered architecture, and inferior mechanical properties30 consistent with a negative role of serotonin in bone accrual. Furthermore, young and adult rodents treated for 4 weeks with daily doses of an SSRI also exhibit reduced bone mass, altered skeletal architecture, and reduced bone mechanical properties.31 In clinical studies, most32-34 but not all35, 36 cross-sectional and longitudinal studies have shown associations between SSRI use and reductions in BMD. In the Canadian Multicenter Osteoporosis Study (CaMOS), older men and women taking SSRIs had higher rates of fracture over 5 years compared with nonusers,32 and increases in fractures also have been reported in other prospective cohort studies,37, 38 along with dose effects. Increased depression and falls in patients treated with SSRIs, however, could confound studies with SSRIs, but conflicting data have been reported regarding the effects of depression and falls on BMD and fractures.36, 39, 40 Polymorphisms in 5HTT also perhaps could increase or decrease SSRI responses in bone.41 Overall, however, the bulk of the preclinical and clinical data would suggest that SSRIs could act to enhance serotonin levels and inhibit normal bone function. If serotonin acting in the central nervous system promotes bone accrual, this could further complicate the skeletal action of these drugs, and negative skeletal effects of SSRIs presumably would act via peripheral serotonin action. In contrast to the findings of Yadav and colleagues,24 who reported that the actions of Lrp5 were extraskeletal and via augmentation of circulating serotonin, Cui and colleagues14 reported that expression of Lrp5 HBM alleles in intestinal stem cells that give rise to serotonin-synthesizing enterochromaffin cells did not produce HBM and that deletion of Lrp5 from these cells did not produce a bone phenotype. They also found no differences in blood serotonin levels in different Lrp5 genotypes and only small differences in serotonin content in small intestine of Lrp5−/− mice relative to wild-type mice. It is unclear whether some of the discrepant findings between these two groups are due to differences in the promoters used to drive expression of transgenes or to localize cre recombinase for developing conditional (tissue-specific) knockouts. Additionally, in view of the fact that more than 99% of serotonin in blood is sequestered in platelets and even minor perturbation can cause platelet activation or lysis leading to massive serotonin release and elevations in “free” serotonin concentrations, another possible source of discrepancy in these studies could be in the different methods used for analysis of circulating serotonin. Cui and colleagues14 also found that there were no significant differences in bone mass between wild-type mice and mice with global deletion of Tph1 that lack serotonin in blood. If confirmed, these results would have impact not only on the role of LRP5 in bone via serotonin but also on the role of 5-HTT and SSRIs in bone. Both groups then evaluated the effect of pharmacologic inhibition of Tph1 on the skeleton. Yadav and colleagues42 synthesized and used a small-molecule inhibitor (LP533401) orally once daily in a dose-dependent manner for up to 6 weeks in ovariectomized rodents and found reduced serum serotonin and an increase in bone formation with no increase in resorption. More recently, this group reported the minimal effective dose of this Tph1 inhibitor, found that it rescued the low-bone-mass phenotype of Lrp5 deficient mice, demonstrated that long-term treatment continued to result in improved bone mass, and showed that its anabolic effect persisted for several weeks after discontinuation.43 In contrast, Cui and colleagues14 treated sham-operated and ovariectomized mice and rats for 6 weeks with a small-molecule Tph1 inhibitor (LP923941) that is the active enantiomer of LP533401. Reductions in whole blood and intestinal serotonin content were observed in response to the inhibitor, but the inhibitor did not influence the amount of serum procollagen type 1 N-terminal peptide (a marker of bone formation), trabecular bone mass, or cortical bone mass either in the femur or in vertebrae. In contrast, parathyroid hormone (PTH) increased bone formation and bone mass in sham-operated and ovariectomized mice. Contradictory data therefore exist on the fundamental mechanisms of the anabolic action of LRP5, that is, as to whether the effects are osseous or extraosseous and whether they are or are not mediated by gut-derived circulating serotonin. In this regard, additional mouse studies to more clearly elucidate the temporal sequence of Lrp5 activation in bone cells using newer genetic tools may be helpful. The discordance in data presented on mechanisms of Lrp5 action clearly extends, however, to the issue of whether pharmacologic inhibition of gut-derived or even bone-derived serotonin is or is not beneficial for bone in rodents. Studies on the efficacy of TPH1 inhibition on skeletal anabolism in humans therefore should at least help to address the translational impact of serotonin production by TPH1 as an inhibitor of bone formation, whether mediated or not by LRP5. The author states that he has no conflicts of interest.
No takes yet. Share an insight, caveat, or question.
David Goltzman (2011) studied this question.