1
COMMENTARY Mechanisms of calcium disposal from osteoclastic resorption hemivacuole H K Datta and B R Horrocks1 School of Clinical and Laboratory Science, The Medical School, University of Newcastle, Framlington Place, Newcastle upon Tyne NE2 4HH, UK 1
School of Natural Sciences, University of Newcastle, Framlington Place, Newcastle upon Tyne NE2 4HH, UK
(Requests for offprints should be addressed to H K Datta; Email:
[email protected])
Abstract One of the most remarkable but neglected aspects of osteoclast function is its unique adaptation that allows the cell to function despite its resorbing surface being exposed to extremely high levels of ambient Ca2+. Recently our studies have provided evidence of continuous transcellular Ca2+ disposal, suggesting that osteoclasts are able to prevent Ca2+ accumulation within the resorptive hemivacuole. It has also been shown that matrix protein degradation products that accumulate within the osteoclast resorptive vacuole are also undergoing transcellular transport by transcytosis. However, both experimental
evidence and theoretical considerations suggest that transcellular transport of Ca2+ and matrix protein is likely to occur via distinct routes. In light of these considerations, we are able to provide convincing explanations for the apparent anomalies of osteoclast intracellular [Ca2+] responses to a variety of endocrine stimuli. The understanding of the mechanisms involved in Ca2+ handling by osteoclasts indicates the lack of a simple link between osteoclast function and changes in overall cytosolic [Ca2+].
Introduction
the resorption hemivacuole by the osteoclast resorptive activity was continually transported out of the resorptive site (Berger et al. 1999, 2001). These in situ studies have also suggested that in a bone-resorbing osteoclast a relatively large amount of calcium enters from the resorption hemivacuole into the cell and is released in a constant steady-state manner at the basolateral surface (Berger et al. 2001). However, the nature of the mechanisms and structures that are involved in the transport of Ca2+ from the hemivacuole to the basolateral surface are not known.
Osteoclasts are multinucleate cells that play a critical role in bone morphogenesis and remodelling. A number of metabolic bone diseases arise due purely to a net increase in osteoclastic activity; the increase in activity may be subtle but insidious, as in osteoporosis, or acute and aggressive, as in hypercalcaemia of malignancy. Despite its central importance in the pathogenesis of a number of metabolic bone diseases, many aspects of osteoclast function remain unclear. During bone resorption, a large amount of Ca2+ is generated within the osteoclast resorptive hemivacuole and [Ca2+] in the resorptive hemivacuole can reach up to 40 mM (Silver et al. 1988). The precise mechanisms involved in the disposal of Ca2+ are not clear. Understanding the mechanisms of Ca2+ disposal is of immense importance as it may lead to the development of novel therapeutic strategies for inhibiting excessive osteoclast resorptive activity. In order to address this problem we used an in vitro model system that employed scanning electrochemical microscopy and transparent resorbable matrices that closely mimic bone-resorbing osteoclast in vivo. This model system demonstrated that the Ca2+ produced in
Journal of Endocrinology (2003) 176, 1–5
Routes of Ca2+ disposal In the light of recent observations of continuous disposal of Ca2+ from the resorptive site (Berger et al. 2001) and the demonstration of bulk trafficking of matrix protein collagen by transcytosis (Nesbitt & Horton 1997, Salo et al. 1997), we have considered the possible routes of Ca2+ disposal from osteoclast hemivacuole. The likely routes are as follows. (1) Bulk transport by transcytosis, i.e. where the osteoclast acts as a conduit allowing the calcium to pass as a part
Journal of Endocrinology (2003) 176, 1–5 0022–0795/03/0176–001 2003 Society for Endocrinology Printed in Great Britain
Online version via http://www.endocrinology.org
2
H K DATTA
and B R HORROCKS
·
Calcium mobilisation by osteoclasts
of the bulk flow in manner akin to the one described as ‘trafficking’ of collagen matrix (Nesbitt & Horton 1997, Salo et al. 1997). (2) Selective uptake involving specialised Ca2+ channels in the cell plasma membrane at the site of the resorptive vacuole and involving intracellular channels traversing the cell and opening at the ventral surface. The release of the calcium could be facilitated by the presence of Ca2+ pumps at the basolateral surface, thereby accelerating calcium disposal at the surface. (3) Ca2+ may in fact ‘leak’ around the sealing zone of the resorptive hemivacuole, though this seems inconsistent with the requirement for the osteoclast to form a tight seal to create a local microenvironment conducive to initiating resorption. It is also possible that all the above-mentioned mechanisms (1–3) may operate, making variable contributions to the Ca2+ disposal process. Here we consider experimental evidence and theoretical arguments to demonstrate that the ‘selective’ rather than the ‘bulk transcytosis’ or ‘leakage’ is likely to be the main mode of Ca2+ disposal. We believe that possibility (3) is the least likely since the generation of a low pH in the hemivacuole to initiate dissolution of the inorganic component of the matrix would be frustrated by the rapid diffusion of protons out of the hemivacuole. It is worth noting that the diffusion coefficient of protons in aqueous media (9·310 5 cm2/s) is much larger than that of any other small ion, e.g. Ca2+ (7·910 6 cm2/s). Experimental evidence for selective transport Interestingly, fluorescence microscopic experiments have indicated transport of collagen through the osteoclast, a process termed transcytosis, during resorption (Nesbitt & Horton 1997, Salo et al. 1997). These studies also provided evidence for the transcytosis of the inorganic phosphate, thereby suggesting one possible mechanism for the disposal of Ca2+ within the osteoclast hemivacuole. However, a close analysis of the data on the kinetics of Ca2+ disposal at the surface of the bone-resorbing osteoclast and its comparison with fluorescence microscopic experiments reveals some critical differences between the mechanism of Ca2+ disposal (Berger et al. 1999, 2001) and transcytosis (Nesbitt & Horton 1997, Salo et al. 1997). Perhaps the most important and obvious difference is that, unlike collagen trafficking that occurs after hours, Ca2+ flux at the basolateral surface is seen within minutes following the ‘seeding’ of osteoclast on bone and occurs at an approximately constant rate. In cases (1) and (3) stated above, the flow would be expected to take place continuously during the course of the resorptive process or may occur in an intermittent fashion where a quantum of calcium is transferred at certain intervals. In the case where specific Ca2+ channels Journal of Endocrinology (2003) 176, 1–5
and pumps are involved, the most likely mode of functioning is continuous rather than intermittent disposal. Indeed, our recent studies have suggested a continuous Ca2+ disposal via a transcellular route occurring at the surface of bone-resorbing osteoclast. The data also show Ca2+ disposal from basolateral surface that can be detected within minutes of resorption, favouring an explanation based on the ‘selective’ route rather than ‘bulk trafficking’ or ‘leakage’. However, our data do not rule out a quantal disposal where the size of the quanta may be small and we see a net flux averaged over many quanta and also broadened in time by the diffusional blurring as the calcium transits between the basolateral surface and the microelectrode detector. Finally, ‘bulk’ flow of Ca2+ disposal by transcytosis would involve cytosolic vacuoles traversing the osteoclast laden with high concentration of the cation. Such exceptionally high vacuolar [Ca2+] have not been reported, and are likely to be extremely hazardous to the cell. Lessons from other cells There are various documented anomalies and peculiarities of Ca2+ handling by osteoclasts, such as erratic and inconsistent hormone-induced intracellular [Ca2+] ([Ca2+]i) mobilisation, heterogeneity in hormone-induced Ca2+ signals, high resting [Ca2+]i and rather active Ca2+ surface efflux (Zaidi et al. 1990, MacIntyre et al. 1991, Bizzari et al. 1994, Rathod et al. 1995, Berger et al. 2001). We here propose a hypothesis that is based on recent findings in other cell types and readily explains these anomalies and peculiarities (Berridge 1993, Parekh & Penner 1997, Zhang et al. 1999, Mogami et al. 2000, Park et al. 2000). One such interesting observation is the role of endoplasmic reticulum (ER) as a functional pool for Ca2+ in pancreatic acinar cells, which shows that ER is one functionally continuous unit, providing a homogeneous environment for the lumenal Ca2+concentrations (Mogami et al. 2000). The transcellular Ca2+ transport in acinar cells is thought to involve uptake through storeoperated Ca2+ channels (SOC) (Parekh & Penner 1997) in the basal membrane and is pumped into the basal ER by Ca2+ ATPases (Mogami et al. 2000). Ca2+ then diffuses from the base to the apical region of the ER lumen, where it is believed to be released into cytosol via specific Ca2+ channels (Mogami et al. 2000, Park et al. 2000). Plasma membrane Ca2+ ATPases, concentrated in the apical membrane, then pump Ca2+ into the apical lumen. Interestingly, such regional-specific specialisation of Ca2+stores has been observed in parotid acinar cells, where ryanodine- and cyclic ADP-ribose (cADPR)-sensitive stores are localised in the basal region whilst inositol triphosphate (IP3)-sensitive stores are present mainly in the apical pole (Moller et al. 1996). Of particular relevance to the Ca2+ handling by osteoclasts is the observation that Ca2+ released from the apical www.endocrinology.org
Calcium mobilisation by osteoclasts ·
H K DATTA
and B R HORROCKS
Figure 1 The schematic diagram shows a possible disposal route of calcium generated within the resorptive hemivacuole by osteoclastic activity. The real-time electrochemical measurements of Ca2+ at the surface of bone-resorbing osteoclasts revealed a steady-state disposal that commences within minutes. The disposal from the resorptive site is likely to involve three main mechanisms. (i) Uptake into the cell via specific Ca2+ channels, such as SOC, followed by uptake into ER by sarco-endoplasmic reticulum Ca2+ (SERCA)-activated ATPase, and efflux from surface by Ca2+ pumps. (ii) Non-specific bulk transport by transcytosis. (iii) ‘Leakage’ through the sealing zone. Some of these mechanisms will be expected to be active in osteoclasts cultured on non-resorbable surface, such as glass and plastic. (1) SOC, (2) SERCA and (3) [Ca2+]i. IP3R, inositol triphosphate receptor. [Ca++]e/40 mM, Ca2+ levels in the hemivacuole.
ER terminal is quickly replenished from the ER at the base of the acinar cells. This observation leads us to propose the existence of a qualitatively similar, but much more active and possibly larger ER network in osteoclasts that allows rapid transcellular transport of Ca2+ from resorptive hemivacuoles to the basolateral surface (Fig. 1). The fact that acinar cells and osteoclasts have a number of common structural and functional features provides further support for our hypothesis. Just like the pancreatic acinar cell, the osteoclast is structurally and functionally polarised (Salo et al. 1997). Both cells have an extensive rough ER dominating the basolateral region, and the whole of the ER in the acinar cell is functionally connected (Glowacki et al. 1986, Park et al. 2000). The resorptive surface in resorbing osteoclasts, just as the apical region in the acinar cells, has no ER as the space is densely packed with secretory granules. However, subtle differences between osteoclasts and the acinar cells exist, such as in the nature and distribution of Ca2+ channels involved in the uptake of Ca2+ from the hemivacuole. www.endocrinology.org
In non-excitable and excitable cells, an increase in [Ca2+]i results from the inflow of the cation through the plasma membrane. The influx of Ca2+ into the cells takes place through three types of Ca2+ channels, voltageoperated Ca2+ channels (VOCCs), ligand-gated nonspecific Ca2+ channels and receptor-linked Ca2+ channels (RLCCs) (Barritt 1999). In non-excitable cells, an additional source for the increase in [Ca2+]i is the release of Ca2+ from the ER, whereas in excitable cells it comes from the sacroplasmic reticulum though ryanodine receptor (RyR) Ca2+ channels. The osteoclast plasma membrane expresses VOCC (Teti et al. 1989, Bizzari et al. 1994), RLCC (Bennett et al. 2001), Na+–Ca2+ exchanger (Moonga at al. 2001) and a stretch-activated Ca2+ entry pathway (Wiltink et al. 1995). Interestingly, a number of past observations have suggested osteoclast plasma membrane expressions of RyR, normally found in the sarcoplasmic reticulum in excitable cells, such as skeletal muscle. Indeed, recent studies have demonstrated that osteoclast plasma membranes express type 2 Journal of Endocrinology (2003) 176, 1–5
3
4
H K DATTA
and B R HORROCKS
·
Calcium mobilisation by osteoclasts
RyR, which is involved in gating Ca2+ influx and functions as a ‘sensor’ for extracellular cations (Moonga et al. 2002). However, osteoclast influx channels have not been extensively investigated, and therefore their relative importance in terms of contribution to the overall calcium disposal from the resorptive site is not known. Similarly, although osteoclast plasma membrane Ca2+ ATPase has been characterised, its comparative activity is not known (Becker & Gay 1990). Anomaly of cytosolic Ca2+ transients A number of diverse agents, such as calcitonin (CT), interleukin-4 (IL-4), protein tyrosine kinase inhibitors and elevated extracellular [Ca2+], which acutely increase [Ca2+]i, inhibit the osteoclast (Bizzari et al. 1994, Rathod et al. 1995, Kajiya et al. 2000, Berger et al. 2001). Indeed, an increase in [Ca2+]i has been generally accepted as a potential mechanism by which various agents inhibit osteoclastic activity. However, an elevation in [Ca2+]i is not always associated with the inhibition of osteoclasts, instead it may lead to activation, as seen following exposure to parathyroid hormone (PTH), ionomycin and pertussis (Rathod et al. 1995, Berger et al. 2001). Furthermore, physiological stimulators of osteoclasts may produce variable changes in [Ca2+]i, e.g. PTH has been shown to increase, decrease or have no effect on cytosolic calcium (Rathod et al. 1995, Miyauchi et al. 1990, Berger et al. 2001). A closer scrutiny of the published literature has revealed an unusual association between [Ca2+]i and osteoclastic activity, namely that an elevation of [Ca2+]i in osteoclasts can be stimulatory, inhibitory or without effect (Miyauchi et al. 1990, Zaidi et al. 1990, MacIntyre et al. 1991, Bizzari et al. 1994, Rathod et al. 1994, 1995, Parkinson et al. 1998, Kajiya et al. 2000, Berger et al. 2001). This implies the lack of a simple general association between acute changes in osteoclast functional modalities and elevation in cytosolic calcium. In an active osteoclast there is likely to be a continuous influx of Ca2+ at the resorptive site, balanced by the process of efflux of the cation at the basolateral surface. The dynamic balance between influx and efflux is likely to be disturbed both by the inhibitors and stimulators, such as CT and PTH respectively (Rathod et al. 1995, Berger et al. 2001). Therefore, an imbalance between influx and efflux brought about by both inhibitors and stimulators can result in a net increase in cytosolic [Ca2+] unless both efflux and influx inhibition occurs equally and simultaneously. It should be remembered that the osteoclasts cultured on non-resorbable matrices continue to be active, as exemplified by the presence of an active proton pump and free radical generation, albeit at lower levels. These considerations suggest that dynamic efflux and influx from the osteoclast will occur but at a lower rate than seen on resorbable matrices. The spatial and temporal aspects of the Journal of Endocrinology (2003) 176, 1–5
inhibition and stimulation of efflux and influx of Ca2+ are equally applicable to osteoclasts cultured on resorbable and non-resorbable matrices. Since those inhibitors that act first on the basolateral surface, such as CT and IL-4, are likely to inhibit efflux before influx is abolished, they will lead to transient elevations in transit calcium in the osteoclast (Fig. 1). Indeed, experimental data demonstrate that the exposure of osteoclasts cultured on glass matrix to CT and IL-4 is found to produce consistent elevation in cytosolic Ca2+ (Bizzari et al. 1994, Berger et al. 2001). However, inhibitors that can have an instantaneous effect on the influx and efflux due to their rapid diffusion, e.g. nitric oxide, will be expected to have no effect on the in-transit Ca2+ (Park et al. 2000) (Fig. 1). Similar considerations apply to the stimulators, e.g. PTH, where our experimental data show a rise, fall or no effect on the osteoclast cytosolic Ca2+, the variety of responses produced being determined by the extent and kinetics of stimulation of efflux and influx by the particular agent (Zaidi et al. 1990, MacIntyre et al. 1991, Bizzari et al. 1994, Rathod et al. 1994, 1995, Parkinson et al. 1998, Kajiya et al. 2000, Berger et al. 2001). The above arguments are valid and applicable both for bulk transcytosis as well as to selective uptake involving specialised Ca2+ channels. The proposed model is also consistent with experimental data demonstrating increases in osteoclast cytosolic [Ca2+] by diverse non-endocrine agents and stimuli (Fig. 1). In conclusion, recent evidence suggests that relatively large amounts of ionised calcium within the osteoclast hemivacuole, released from the bone matrix by osteoclastic resorptive activity, is disposed of continuously in a steadystate manner. We argue that there are likely to be three routes of cation disposal, namely ‘leak’, bulk transcytsosis and ‘selective’ disposal involving Ca2+ channels and pumps. The experimental evidence and theoretical considerations suggest that the ‘selective’ route is likely to be the most important or even the only route of disposal. In views of recent observations regarding acinar cell ER in Ca2+ handling and the fact that osteoclasts and acinar cells share a number of common structural and functional features, we propose a hypothetical model for the ‘selective’ route of Ca2+ disposal from osteoclast hemivacuole. The proposed hypothetical model is able to reconcile the apparent anomalies of osteoclast [Ca2+]i responses to a variety of endocrine stimuli. Acknowledgement Action Research supported this work. References Barritt GJ 1999 Receptor-activated Ca2+ inflow in animal cells: a variety of pathways tailored to meet different intracellular Ca2+ signalling requirements. Biochemical Journal 337 153–169. www.endocrinology.org
Calcium mobilisation by osteoclasts · Becker PJ & Gay CV 1990 Characterization of a Ca2+-ATPase in osteoclast plasma membrane. Journal of Bone and Mineral Research 5 557–567. Bennett BD, Alvarez U & Hruska KA 2001 Receptor-operated osteoclast calcium sensing. Endocrinology 142 1968–1974. Berger CEM, Horrocks BR & Datta HK 1999 Application of ion-selective microelectrodes to the detection of calcium release during bone resorption. Electrochimica Acta 44 2677–2683. Berger CEM, Rathod H, Gillespie JI, Horrocks BR & Datta HK 2001 Scanning electrochemical spectroscopy at the surface of bone-resorbing osteoclast: evidence for steady state disposal and intracellular functional compartmentalization of calcium. Journal of Bone and Mineral Research 16 2092–2102. Berridge ML 1993 Inositol triphosphate and calcium signalling. Nature 361 315–325. Bizzari C, Shioi A, Teitelbaum SL, Ohara J, Harwalkar VA, Erdmann JM, Lacey DL & Civitelli R 1994 Interleukin-4 inhibits bone resorption and acutely increases cytosolic Ca2+ in murine osteoclasts. Journal of Biological Chemistry 269 13817–13824. Glowacki J, Jasty M & Goldring S 1986 Comparison of multinucleate cells elicited in rats by particulate bone, polyethylene, or polymethylmethacrylate. Journal of Bone and Mineral Research 1 327–331. Kajiya H, Okabe K, Okamoto F, Tsuzuki T & Soeda H 2000 Protein tyrosine kinase inhibitors increase cytosolic calcium and inhibit actin organization as resorbing activity in rat osteoclasts. Journal of Cell Physiology 183 83–89. MacIntyre I, Zaidi M, Alam AS, Datta HK, Moonga BS, Lidbury PS, Hecker M & Vane JR 1991 Osteoclastic inhibition: an action of nitric oxide not mediated by cyclic GMP. PNAS 88 2936–2940. Miyauchi A, Hruska KA, Greenfield EM, Duncan R, Alvarez J, Barattolo R, Colucci S, Zambonin-Zallone A, Teitelbaum SL & Teti A 1990 Osteoclast cytosolic calcium, regulated by voltage-gated calcium channels and extracellular calcium, controls podosome assembly and bone resorption. Journal of Cell Biology 111 2543–2552. Mogami H, Nakano K, Tepikin AV & Peterson OH 2000 Ca2+ flow via tunnels in polarised cells: recharging of apical Ca2+ stores by focal Ca2+ entry through basal membrane patch. Cell 88 49–55. Moller JV, Juul B & Maire M 1996 Structural organization, ion transport, and energy transduction of ATPases. Biochimica et Biophysica Acta 1286 1–51. Moonga BS, Davidson R, Sun L, Adebanjo OA, Moser J, Abedin M, Zaidi N, Huang CL & Zaidi M 2001 Identification and characterization of a sodium/calcium exchanger, NCX-1, in osteoclasts and its role in bone resorption. Biochemical and Biophysical Research Communications 283 770–775. Moonga BS, Li S, Iqbal J, Davidson R, Shankar VS, Bevis PJ, Inzerillo A, Abe E, Huang CL & Zaidi M 2002 Ca (2+) influx
www.endocrinology.org
H K DATTA
and B R HORROCKS
through the osteoclastic plasma membrane ryanodine receptor. American Journal of Physiology — Renal Physiology 282 F921–F932. Nesbitt SA & Horton MA 1997 Trafficking of matrix collagens through bone-resorbing osteoclasts. Science 276 266–269. Parekh A & Penner R 1997 Store depletion and calcium influx. Physiological Reviews 77 901–930. Park MK, Peterson OH & Tepikin AV 2000 The endoplasmic reticulum as one continuous Ca2+ pool: visualization of rapid Ca2+ movement and equilibration. EMBO Journal 19 5729–5739. Parkinson N, Bolsover S & Mason W 1998 Nuclear and cytosolic calcium changes in osteoclasts stimulated with ATP and integrin-binding peptide. Cell Calcium 24 213–222. Rathod H, Malcolm AJ, Gillespie JI, Berry V, Pooley J, Piggott NH & Datta HK 1994 Characterization of a subtype of primary osteoclastoma: extracellular calcium but not calcitonin inhibits aggressive HLA-DR-positive osteoclastoma possessing ‘functional’ calcitonin receptors. Journal of Pathology 175 293–299. Rathod H, Gillespie JI & Datta HK 1995 Evidence for parathyroid hormone-induced mobilisation of intracellular calcium in the rat osteoclast. Journal of Physiology 489P 71. Salo J, Lehenkari P, Mulari M, Metsikko K & Vaananen HK 1997 Removal of osteoclast bone resorption products by transcytosis. Science 276 270–273. Silver IA, Murrills RJ & Etherington DJ 1988 Microelectrode studies on acid microenvironment beneath adherent macrophage and osteoclasts. Experimental Cell Research 175 266–276. Teti A, Grano M, Colucci S, Argentino L, Barattolo R, Miyauchi A, Teitelbaum SL, Hruska KA & Zambonin-Zallone A 1989 Voltage dependent calcium channel expression in isolated osteoclasts. Bollettino – Societa Italiana Biologia Sperimentale 65 1115–1118. Wiltink A, Nijweide PJ, Scheenen WJ, Ypey D & Van Duijn B 1995 Cell membrane stretch in osteoclasts triggers a self-reinforcing Ca2+ entry pathway. Pflugers Archiv – European Journal of Physiology 429 663–671. Zaidi M, MacIntyre I & Datta HK 1990 Intracellular calcium in the control of osteoclast function: paradoxical elevation of cytosolic free calcium by verapamil. Biochemical and Biophysical Research Communications 167 807–812. Zhang X, Wen J, Bidasee KR, Besch HR Jr, Wojcikiewicz RJ, Lee B & Rubin RP 1999 Ryanodine and inositol triphosphate receptors are differentially distributed and expressed in rat parotid gland. Biochemical Journal 340 519–527.
Received in final form 30 October 2002 Accepted 30 October 2002
Journal of Endocrinology (2003) 176, 1–5
5
