Portal de Periódicos da CAPES

Calcitonin Physiology, Saved by a Lysophospholipid

2015; Oxford University Press; Volume: 30; Issue: 2 Linguagem: Inglês

10.1002/jbmr.2449

1523-4681

T. John Martin, Natalie A. Sims,

Protein Kinase Regulation and GTPase Signaling

Although the discovery of calcitonin was a major event in endocrinology, its physiological role has been elusive throughout the last 50 years. A recent report in Nature Communications1 provides further impetus to the idea that despite its pharmacologic inhibition of bone resorption, calcitonin also functions physiologically as an inhibitor of bone formation. The advent of calcitonin as a rapidly acting hypocalcemic hormone of thyroid origin2-4 was greeted as the pathway to finer control of calcium homeostasis than could be achieved by parathyroid hormone (PTH) alone. Calcitonin's lowering of blood calcium was achieved exclusively by its action on bone, with inhibition of bone resorption in vitro5 and in vivo6 preventing calcium efflux from bone, rather than promoting its deposition.7, 8 It was some years before it was shown that specific receptors for calcitonin were highly expressed in osteoclasts9 and that the hormone acted directly on osteoclasts to reduce their motility and activity.10 Despite these actions that could readily be demonstrated pharmacologically, evidence for a calcium-regulating role in the mature mammal remained lacking. Nevertheless, the physiological function of calcitonin in maturity has been suggested to be one that regulates the bone-resorptive process, in either a continuous or intermittent manner, a role that might become more important in circumstances in which skeletal integrity is at particular risk (eg, in pregnancy and lactation).11, 12 Then came the surprising discovery that although ablation of the gene that encodes both calcitonin and its gene-related peptide (CGRP) gene resulted in mice (Calca-/-) that were much less able than wild type to overcome hypercalcemia induced by a calcium load,13 these mice also had high bone mass, with histomorphometry indicating more bone formation than controls,13 suggesting that calcitonin might inhibit bone formation during normal bone remodeling. This phenotype was not found in mice rendered deficient for CGRPα alone,14 leaving the deficiency of calcitonin as an explanation for the increased bone formation in Calca-/- mice, something that could never have been predicted. The increased trabecular bone mass was noted particularly in young mice,13 and although still evident at 6 months of age, progressive bone loss occurred with age by increased resorption, resulting in marked cortical porosity, even though bone formation was maintained at a high level.15 These puzzling findings suggested, in addition to lactation, two necessary physiological roles for calcitonin that differed with age, ie, inhibition of bone formation throughout life and inhibition of bone resorption with age. Shortly after this work, young Calcr+/− mice were reported to exhibit a bone phenotype virtually indistinguishable from that of the ligand null mice: high bone mass and increased bone formation by histomorphometry.16 Study of the hemizygotes was necessary because germline ablation of the calcitonin receptor (CTR) resulted in embryo lethality. The same group prepared an incomplete CTR null model that overcame the survival defect, finding some evidence again of increased bone formation.17 Thus, the conclusion was that in mice the removal of calcitonin production or action resulted in a greater amount of bone, implying a physiological role for calcitonin as a tonic inhibitor of bone formation. How the effect on bone formation comes about was puzzling, particularly because there was a lack of evidence for calcitonin receptors and responses in osteoblasts.18 It seemed likely that the physiological roles of calcitonin in bone are brought about through two pathways—a direct effect on osteoclasts to inhibit resorption and an indirect one resulting in the elaboration of a critical, locally active inhibitor of bone formation. One mechanism by which calcitonin could stimulate bone formation arose from a study where calcitonin treatment significantly increased sclerostin expression in bone.19-21 Osteocytes freshly isolated from calvariae and long bones expressed Calcr mRNA,19-21 and immunohistochemistry revealed co-localization of CTR and sclerostin in some osteocytes in calvarial sections. However, because CTR and sclerostin expression are both lost in long-term culture of osteocytes,19, 20 the mechanism by which calcitonin might stimulate sclerostin remains undefined. Entirely new light has been cast on the matter by the findings reported in Keller and colleagues,1 where they prepared a new Calcr-/- mouse that was viable, an outcome that they ascribed to having excised the neomycin cassette contained in the original knockout. High bone mass and bone formation were evident from 3 to 18 months of age, and a similar phenotype resulted when the deletion was restricted to the osteoclast lineage. In contrast to the ligand-deficient mice, aging of the new global Calcr-/- mouse was not accompanied by bone loss. The authors recognized that the ligand-deficient mice would lack not only CT and CGRP but also the CT precursor, procalcitonin, which interacts with receptors other than CTR.22 They found that procalcitonin inhibited osteoclast formation in vitro from hemopoietic precursors of both wild-type and Calcr-/- mice. In sepsis, procalcitonin is expressed widely, confers toxicity at inflammatory sites, and circulates at high levels, giving rise to its use as a biomarker.23, 24 Although procalcitonin circulates in normal conditions at levels similar to CT, its role in bone biology has received very little attention. Keller and colleagues observed that lipopolysaccharide-induced bone resorption occurred in CT/CGRP mice but not in Calcr-/- mice, lending support to a role for procalcitonin or some other product of the calcitonin gene as an inhibitor of bone resorption through a non-CTR–mediated pathway. This warrants further study. Keller and colleagues identified a novel mediator of the inhibitory effect of CT on bone formation by a microarray analysis of wild-type and Calcr-/- osteoclasts treated for 6 hours with murine calcitonin. Among the 29 genes whose regulation by CT was changed by loss of CTR, the authors chose to pursue Spns2, which encodes spinster 2. This transmembrane exporter protein is necessary for the secretion of sphingosine-1-phosphate (S1P), a lysophospholipid (Fig. 1) {FIG1} that has been suggested as an osteoclast-secreted coupling factor that stimulates bone formation.25-27 Although a number of other coupling factors have been proposed from many in vivo and in vitro studies (eg, reviewed in Sims and Martin28 and Martin29), none of these other pathways featured in the array data. There is very little existing data regarding Spns2 in bone, but an earlier genetic screen for bone phenotypes in mice revealed that knockout of Spns2 results in thick cortical bone.30 This phenotype is the reverse of what would be predicted from the results of Keller, but the cellular basis of this phenotype is not yet known. No skeletal examination was undertaken with another global knockout of Spns2.31 The authors investigated whether calcitonin might decrease bone formation by reducing Spns2-mediated S1P release from osteoclasts. Fig. 1 illustrates the key steps in formation, secretion, and degradation of S1P, which has anti-inflammatory, pro-proliferative, and anti-apoptotic effects,32 as well as being necessary for the egress of T and B lymphocytes to the circulation.31, 33 The first suggestion of S1P as a coupling factor came from the finding that osteoclast precursors treated with receptor activator of NF-κB ligand (RANKL) exhibited enhanced expression of sphingosine kinase-1 (Sphk1), an enzyme responsible for the generation of S1P from its precursor, sphingosine.26 In the present work, no effect of CT on expression of Sphk1 or Sphk2 was noted. Osteoclast differentiation was accompanied by increased mRNA levels of Spns2 without any change in either Sphk1 or Sphk2. Calcitonin decreased expression of Spns2 and decreased culture medium levels of S1P in wild-type but not in Calcr-/- osteoclasts. Although the proposed model focuses on the role of Spns2 on the release of S1P from the osteoclast, it should be noted that Spns2 is broadly expressed and would also regulate S1P release from osteoblasts31, 34 and endothelial cells,35 although these are unlikely to be regulated by calcitonin because CTR is lacking in those cells. Of the five receptors for S1P, S1pr1 and S1pr3 mRNA levels increased during osteoblast differentiation.1 Because S1pr1-/- mice are embryo lethal and there were no changes in bone mass detected when S1pr1 was knocked out specifically in osteoblasts, the authors focused on S1pr3. Although S1pr3-/- mice were indistinguishable from controls at age 3 months, by age 8 months they had less bone and less bone formation, without any change in resorption. Although the formation phenotype was predicted from the hypothesis, the fact that it was not evident at age 3 months contrasts with the Calca-/- and Calcr-/- mice, at which time increased bone mass and bone formation were very clearly evident, even as early as age 1 month in the Calca-/- mice.13 Irrespective of this difference in timing, the phenotype of Calcr-/- mice was lacking when they were crossed with S1pr3-/- mice, suggesting a dependency on S1pr3 signaling for that phenotype. Further evidence that the anabolic effect of S1P is mediated by S1PR3 was provided by the use of FTY720 in S1pr3-/- mice. Originally a traditional Asian medicine,36 FTY720 is marketed widely as the first-line treatment for relapsing–remitting multiple sclerosis.37 Phosphorylation of FTY720 by Sphk2 generates a structural analog of S1P that activates all five S1PRs bar S1PR2.38 This agent has been reported to enhance BMP2-induced osteoblast differentiation39 to stimulate osteoblast chemotaxis26 and to regenerate cranial bone defects40 by promoting vasculogenesis41 but was ineffective in a fracture-healing model.42 Other work has suggested that S1P acts through both S1P1R and S1PR2 to facilitate chemotaxis of osteoclast precursors, although the evidence that this favors reduced bone resorption and that FTY720 inhibits resorption is based on studies limited to very few animals.43, 44 Overall, the results suggest that alternative S1P agonists might lead to a bone anabolic, although off-target effects on the immune system may be a concern.38 A notable conclusion from this work is that the physiological function of calcitonin in bone differs entirely from its pharmacological function. This is also the case with PTH, a physiological stimulator of bone resorption that stimulates bone formation in pharmacologic use. In that case, however, the pharmacologic effect of PTH might be reproducing the effect on bone that is the physiological function of locally generated PTHrP.45 Thus a question that might be addressed is whether a paracrine ligand for the CTR is produced in bone. A physiological role of circulating CT in inhibiting bone formation requires new data to explain how it can regulate events in BMUs distributed asynchronously throughout the skeleton and therefore at different stages of the remodeling cycle. Furthermore, circulating levels of CT are very low, <10 pg/mL in adults and slightly higher in infants.46 This has made CT-deficient individuals, and the effects of CT deficiency, extremely difficult to identify in human subjects. This work raises very interesting questions that relate to intercellular communication processes in bone. Does S1P occupy a dominant place in the list of "coupling factors"? There are likely many contributors to this process at different stages of remodeling,29, 47 and establishing dominance might prove difficult. Is S1P effective in the BMU in the resorptive, reversal, or formative phases? If changes in Spns2 production are indeed important, is there an anabolic treatment that could increase its production and provide more S1P? PTH could achieve this through its early activation of osteoclasts in remodeling.48 It has been argued by one of the co-discoverers of calcitonin that the hormone is not involved in calcium homeostasis or in any other important physiologic function, except possibly in protection of the skeleton under conditions of calcium stress.12 Although many outstanding questions remain, the work of Keller and colleagues could give calcitonin its best chance to date of achieving physiological respectability. Both authors state that they have no conflicts of interest. Work from the authors' laboratories is supported by project grants from the National Health and Medical Research Council of Australia and the Victorian Government OIS Program. The authors agreed on the content and writing of the manuscript.