1 COMMENTARY

Mechanisms of 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 1School 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 evidence and theoretical considerations suggest that trans- function is its unique adaptation that allows the cellular transport of Ca2+ and matrix protein is likely to to function despite its resorbing surface being exposed occur via distinct routes. In light of these considerations, to extremely high levels of ambient Ca2+. Recently our we are able to provide convincing explanations for studies have provided evidence of continuous transcellular the apparent anomalies of osteoclast intracellular [Ca2+] Ca2+ disposal, suggesting that are able to responses to a variety of endocrine stimuli. The under- prevent Ca2+ accumulation within the resorptive hemi- standing of the mechanisms involved in Ca2+ handling by vacuole. It has also been shown that matrix protein osteoclasts indicates the lack of a simple link between degradation products that accumulate within the osteo- osteoclast function and changes in overall cytosolic [Ca2+]. clast resorptive vacuole are also undergoing transcellular Journal of Endocrinology (2003) 176, 1–5 transport by transcytosis. However, both experimental

Introduction the resorption hemivacuole by the osteoclast resorptive activity was continually transported out of the resorptive Osteoclasts are multinucleate cells that play a critical role site (Berger et al. 1999, 2001). These in situ studies have in morphogenesis and remodelling. A number of also suggested that in a bone-resorbing osteoclast a rela- metabolic bone diseases arise due purely to a net increase tively large amount of calcium enters from the resorption in osteoclastic activity; the increase in activity may be hemivacuole into the cell and is released in a constant subtle but insidious, as in , or acute and steady-state manner at the basolateral surface (Berger et al. aggressive, as in hypercalcaemia of malignancy. Despite its 2001). However, the nature of the mechanisms and central importance in the pathogenesis of a number of structures that are involved in the transport of Ca2+ from metabolic bone diseases, many aspects of osteoclast func- the hemivacuole to the basolateral surface are not tion remain unclear. During bone resorption, a large known. amount of Ca2+ is generated within the osteoclast resorp- tive hemivacuole and [Ca2+] in the resorptive hemi- vacuole can reach up to 40 mM (Silver et al. 1988). The Routes of Ca2+ disposal precise mechanisms involved in the disposal of Ca2+ are not clear. Understanding the mechanisms of Ca2+ dis- In the light of recent observations of continuous disposal of posal is of immense importance as it may lead to the Ca2+ from the resorptive site (Berger et al. 2001) and the development of novel therapeutic strategies for inhibiting demonstration of bulk trafficking of matrix protein colla- excessive osteoclast resorptive activity. gen by transcytosis (Nesbitt & Horton 1997, Salo et al. In order to address this problem we used an in vitro 1997), we have considered the possible routes of Ca2+ model system that employed scanning electrochemical disposal from osteoclast hemivacuole. The likely routes are microscopy and transparent resorbable matrices that as follows. closely mimic bone-resorbing osteoclast in vivo. This (1) Bulk transport by transcytosis, i.e. where the osteo- model system demonstrated that the Ca2+ produced in clast acts as a conduit allowing the calcium to pass as a part

Journal of Endocrinology (2003) 176, 1–5 Online version via http://www.endocrinology.org 0022–0795/03/0176–001  2003 Society for Endocrinology Printed in Great Britain

Downloaded from Bioscientifica.com at 10/02/2021 06:04:15PM via free access 2 H K DATTA and B R HORROCKS · Calcium mobilisation by osteoclasts

of the bulk flow in manner akin to the one described as and pumps are involved, the most likely mode of func- ‘trafficking’ of matrix (Nesbitt & Horton 1997, tioning is continuous rather than intermittent disposal. Salo et al. 1997). Indeed, our recent studies have suggested a continuous (2) Selective uptake involving specialised Ca2+ channels Ca2+ disposal via a transcellular route occurring at the in the cell plasma membrane at the site of the resorptive surface of bone-resorbing osteoclast. The data also show vacuole and involving intracellular channels traversing Ca2+ disposal from basolateral surface that can be detected the cell and opening at the ventral surface. The release of within minutes of resorption, favouring an explanation the calcium could be facilitated by the presence of Ca2+ based on the ‘selective’ route rather than ‘bulk trafficking’ pumps at the basolateral surface, thereby accelerating or ‘leakage’. However, our data do not rule out a quantal calcium disposal at the surface. disposal where the size of the quanta may be small and we (3) Ca2+ may in fact ‘leak’ around the sealing zone of see a net flux averaged over many quanta and also the resorptive hemivacuole, though this seems inconsistent broadened in time by the diffusional blurring as the with the requirement for the osteoclast to form a tight seal calcium transits between the basolateral surface and the to create a local microenvironment conducive to initiating microelectrode detector. resorption. Finally, ‘bulk’ flow of Ca2+ disposal by transcytosis It is also possible that all the above-mentioned mech- would involve cytosolic vacuoles traversing the osteoclast anisms (1–3) may operate, making variable contributions laden with high concentration of the cation. Such excep- to the Ca2+ disposal process. Here we consider exper- tionally high vacuolar [Ca2+] have not been reported, and imental evidence and theoretical arguments to demon- are likely to be extremely hazardous to the cell. strate that the ‘selective’ rather than the ‘bulk transcytosis’ or ‘leakage’ is likely to be the main mode of Ca2+ disposal. Lessons from other cells We believe that possibility (3) is the least likely since the generation of a low pH in the hemivacuole to initiate There are various documented anomalies and peculiarities dissolution of the inorganic component of the matrix of Ca2+ handling by osteoclasts, such as erratic and would be frustrated by the rapid diffusion of protons out of inconsistent hormone-induced intracellular [Ca2+] ff 2+ the hemivacuole. It is worth noting that the di usion ([Ca ]i) mobilisation, heterogeneity in hormone-induced ffi  5 2+ 2+ 2+ coe cient of protons in aqueous media (9·3 10 Ca signals, high resting [Ca ]i and rather active Ca cm2/s) is much larger than that of any other small ion, e.g. surface efflux (Zaidi et al. 1990, MacIntyre et al. 1991,  Ca2+ (7·910 6 cm2/s). 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 Experimental evidence for selective transport anomalies and peculiarities (Berridge 1993, Parekh & Penner 1997, Zhang et al. 1999, Mogami et al. 2000, Park Interestingly, fluorescence microscopic experiments have et al. 2000). One such interesting observation is the role of indicated transport of collagen through the osteoclast, a endoplasmic reticulum (ER) as a functional pool for Ca2+ process termed transcytosis, during resorption (Nesbitt & in pancreatic acinar cells, which shows that ER is one Horton 1997, Salo et al. 1997). These studies also provided functionally continuous unit, providing a homogeneous evidence for the transcytosis of the inorganic , environment for the lumenal Ca2+concentrations thereby suggesting one possible mechanism for the dis- (Mogami et al. 2000). The transcellular Ca2+ transport in posal of Ca2+ within the osteoclast hemivacuole. How- acinar cells is thought to involve uptake through store- ever, a close analysis of the data on the kinetics of Ca2+ operated Ca2+ channels (SOC) (Parekh & Penner 1997) in disposal at the surface of the bone-resorbing osteoclast and the basal membrane and is pumped into the basal ER by its comparison with fluorescence microscopic experiments Ca2+ ATPases (Mogami et al. 2000). Ca2+ then diffuses reveals some critical differences between the mechanism from the base to the apical region of the ER lumen, where of Ca2+ disposal (Berger et al. 1999, 2001) and transcytosis it is believed to be released into cytosol via specific Ca2+ (Nesbitt & Horton 1997, Salo et al. 1997). Perhaps the channels (Mogami et al. 2000, Park et al. 2000). Plasma most important and obvious difference is that, unlike membrane Ca2+ ATPases, concentrated in the apical collagen trafficking that occurs after hours, Ca2+ flux membrane, then pump Ca2+ into the apical lumen. at the basolateral surface is seen within minutes following Interestingly, such regional-specific specialisation of the ‘seeding’ of osteoclast on bone and occurs at an Ca2+stores has been observed in parotid acinar cells, where approximately constant rate. ryanodine- and cyclic ADP-ribose (cADPR)-sensitive In cases (1) and (3) stated above, the flow would be stores are localised in the basal region whilst inositol expected to take place continuously during the course of triphosphate (IP3)-sensitive stores are present mainly in the the resorptive process or may occur in an intermittent apical pole (Moller et al. 1996). fashion where a quantum of calcium is transferred at Of particular relevance to the Ca2+ handling by osteo- certain intervals. In the case where specific Ca2+ channels clasts is the observation that Ca2+ released from the apical

Journal of Endocrinology (2003) 176, 1–5 www.endocrinology.org

Downloaded from Bioscientifica.com at 10/02/2021 06:04:15PM via free access Calcium mobilisation by osteoclasts · H K DATTA and B R HORROCKS 3

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 2+ ++ 2+ and (3) [Ca ]i.IP3R, inositol triphosphate receptor. [Ca ]e/40 mM, Ca levels in the hemivacuole.

ER terminal is quickly replenished from the ER at the In non-excitable and excitable cells, an increase in 2+ base of the acinar cells. This observation leads us to [Ca ]i results from the inflow of the cation through the propose the existence of a qualitatively similar, but much plasma membrane. The influx of Ca2+ into the cells takes more active and possibly larger ER network in osteoclasts place through three types of Ca2+ channels, voltage- that allows rapid transcellular transport of Ca2+ from operated Ca2+ channels (VOCCs), ligand-gated non- resorptive hemivacuoles to the basolateral surface (Fig. 1). specific Ca2+ channels and receptor-linked Ca2+ channels The fact that acinar cells and osteoclasts have a number of (RLCCs) (Barritt 1999). In non-excitable cells, an 2+ common structural and functional features provides further additional source for the increase in [Ca ]i is the release support for our hypothesis. Just like the pancreatic acinar of Ca2+ from the ER, whereas in excitable cells it cell, the osteoclast is structurally and functionally polarised comes from the sacroplasmic reticulum though ryanodine (Salo et al. 1997). Both cells have an extensive rough ER receptor (RyR) Ca2+ channels. dominating the basolateral region, and the whole of the The osteoclast plasma membrane expresses VOCC ER in the acinar cell is functionally connected (Glowacki (Teti et al. 1989, Bizzari et al. 1994), RLCC (Bennett et al. et al. 1986, Park et al. 2000). The resorptive surface in 2001), Na+–Ca2+ exchanger (Moonga at al. 2001) and a resorbing osteoclasts, just as the apical region in the acinar stretch-activated Ca2+ entry pathway (Wiltink et al. 1995). cells, has no ER as the space is densely packed with Interestingly, a number of past observations have suggested secretory granules. However, subtle differences between osteoclast plasma membrane expressions of RyR, normally osteoclasts and the acinar cells exist, such as in the nature found in the sarcoplasmic reticulum in excitable cells, such and distribution of Ca2+ channels involved in the uptake of as skeletal muscle. Indeed, recent studies have demon- Ca2+ from the hemivacuole. strated that osteoclast plasma membranes express type 2 www.endocrinology.org Journal of Endocrinology (2003) 176, 1–5

Downloaded from Bioscientifica.com at 10/02/2021 06:04:15PM via free access 4 H K DATTA and B R HORROCKS · Calcium mobilisation by osteoclasts

RyR, which is involved in gating Ca2+ influx and inhibition and stimulation of efflux and influx of Ca2+ are functions as a ‘sensor’ for extracellular cations (Moonga equally applicable to osteoclasts cultured on resorbable and et al. 2002). However, osteoclast influx channels have not non-resorbable matrices. Since those inhibitors that act been extensively investigated, and therefore their relative first on the basolateral surface, such as CT and IL-4, are importance in terms of contribution to the overall calcium likely to inhibit efflux before influx is abolished, they will disposal from the resorptive site is not known. Similarly, lead to transient elevations in transit calcium in the although osteoclast plasma membrane Ca2+ ATPase has osteoclast (Fig. 1). Indeed, experimental data demonstrate been characterised, its comparative activity is not known that the exposure of osteoclasts cultured on glass matrix to (Becker & Gay 1990). 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 Anomaly of cytosolic Ca2+ transients on the influx and efflux due to their rapid diffusion, e.g. nitric oxide, will be expected to have no effect on the A number of diverse agents, such as (CT), in-transit Ca2+ (Park et al. 2000) (Fig. 1). Similar con- interleukin-4 (IL-4), protein tyrosine kinase inhibitors and siderations apply to the stimulators, e.g. PTH, where our elevated extracellular [Ca2+], which acutely increase experimental data show a rise, fall or no effect on the 2+ 2+ [Ca ]i, inhibit the osteoclast (Bizzari et al. 1994, Rathod osteoclast cytosolic Ca , the variety of responses pro- et al. 1995, Kajiya et al. 2000, Berger et al. 2001). Indeed, duced being determined by the extent and kinetics of 2+ ffl an increase in [Ca ]i has been generally accepted as a stimulation of e ux and influx by the particular agent potential mechanism by which various agents inhibit (Zaidi et al. 1990, MacIntyre et al. 1991, Bizzari et al. 2+ osteoclastic activity. However, an elevation in [Ca ]i is 1994, Rathod et al. 1994, 1995, Parkinson et al. 1998, not always associated with the inhibition of osteoclasts, Kajiya et al. 2000, Berger et al. 2001). The above instead it may lead to activation, as seen following expo- arguments are valid and applicable both for bulk trans- sure to (PTH), ionomycin and cytosis as well as to selective uptake involving specialised pertussis (Rathod et al. 1995, Berger et al. 2001). Further- Ca2+ channels. The proposed model is also consistent with more, physiological stimulators of osteoclasts may produce experimental data demonstrating increases in osteoclast 2+ 2+ variable changes in [Ca ]i, e.g. PTH has been shown to cytosolic [Ca ] by diverse non-endocrine agents and increase, decrease or have no effect on cytosolic calcium stimuli (Fig. 1). (Rathod et al. 1995, Miyauchi et al. 1990, Berger et al. In conclusion, recent evidence suggests that relatively 2001). A closer scrutiny of the published literature has large amounts of ionised calcium within the osteoclast 2+ revealed an unusual association between [Ca ]i and hemivacuole, released from the bone matrix by osteoclastic 2+ osteoclastic activity, namely that an elevation of [Ca ]i in resorptive activity, is disposed of continuously in a steady- osteoclasts can be stimulatory, inhibitory or without effect state manner. We argue that there are likely to be three (Miyauchi et al. 1990, Zaidi et al. 1990, MacIntyre et al. routes of cation disposal, namely ‘leak’, bulk transcytsosis 1991, Bizzari et al. 1994, Rathod et al. 1994, 1995, and ‘selective’ disposal involving Ca2+ channels and Parkinson et al. 1998, Kajiya et al. 2000, Berger et al. pumps. The experimental evidence and theoretical con- 2001). This implies the lack of a simple general association siderations suggest that the ‘selective’ route is likely to be between acute changes in osteoclast functional modalities the most important or even the only route of disposal. In and elevation in cytosolic calcium. views of recent observations regarding acinar cell ER in In an active osteoclast there is likely to be a continuous Ca2+ handling and the fact that osteoclasts and acinar cells influx of Ca2+ at the resorptive site, balanced by the share a number of common structural and functional process of efflux of the cation at the basolateral surface. features, we propose a hypothetical model for the ‘selec- The dynamic balance between influx and efflux is likely to tive’ route of Ca2+ disposal from osteoclast hemivacuole. be disturbed both by the inhibitors and stimulators, such as The proposed hypothetical model is able to reconcile the 2+ CT and PTH respectively (Rathod et al. 1995, Berger apparent anomalies of osteoclast [Ca ]i responses to a et al. 2001). Therefore, an imbalance between influx and variety of endocrine stimuli. efflux brought about by both inhibitors and stimulators can 2+ result in a net increase in cytosolic [Ca ] unless both Acknowledgement efflux and influx inhibition occurs equally and simul- taneously. It should be remembered that the osteoclasts Action Research supported this work. cultured on non-resorbable matrices continue to be active, as exemplified by the presence of an active proton pump References and free radical generation, albeit at lower levels. These considerations suggest that dynamic efflux and influx from Barritt GJ 1999 Receptor-activated Ca2+ inflow in animal cells: a the osteoclast will occur but at a lower rate than seen on variety of pathways tailored to meet different intracellular Ca2+ resorbable matrices. The spatial and temporal aspects of the signalling requirements. Biochemical Journal 337 153–169.

Journal of Endocrinology (2003) 176, 1–5 www.endocrinology.org

Downloaded from Bioscientifica.com at 10/02/2021 06:04:15PM via free access Calcium mobilisation by osteoclasts · H K DATTA and B R HORROCKS 5

Becker PJ & Gay CV 1990 Characterization of a Ca2+-ATPase in through the osteoclastic plasma membrane ryanodine receptor. osteoclast plasma membrane. Journal of Bone and Mineral Research 5 American Journal of Physiology — Renal Physiology 282 F921–F932. 557–567. Nesbitt SA & Horton MA 1997 Trafficking of matrix Bennett BD, AlvarezU&HruskaKA2001 Receptor-operated through bone-resorbing osteoclasts. Science 276 266–269. osteoclast calcium sensing. Endocrinology 142 1968–1974. ParekhA&PennerR1997Storedepletion and calcium influx. Berger CEM, Horrocks BR & Datta HK 1999 Application of Physiological Reviews 77 901–930. ion-selective microelectrodes to the detection of calcium release Park MK, Peterson OH & Tepikin AV 2000 The endoplasmic during bone resorption. Electrochimica Acta 44 2677–2683. reticulum as one continuous Ca2+ pool: visualization of rapid Ca2+ Berger CEM, Rathod H, Gillespie JI, Horrocks BR & Datta HK movement and equilibration. EMBO Journal 19 5729–5739. 2001 Scanning electrochemical spectroscopy at the surface of ParkinsonN,BolsoverS&MasonW1998Nuclear and cytosolic bone-resorbing osteoclast: evidence for steady state disposal and calcium changes in osteoclasts stimulated with ATP and intracellular functional compartmentalization of calcium. Journal of integrin-binding peptide. Cell Calcium 24 213–222. Bone and Mineral Research 16 2092–2102. Rathod H, Malcolm AJ, Gillespie JI, Berry V, Pooley J, Piggott NH Berridge ML 1993 Inositol triphosphate and calcium signalling. Nature & Datta HK 1994 Characterization of a subtype of primary 361 315–325. osteoclastoma: extracellular calcium but not calcitonin inhibits Bizzari C, Shioi A, Teitelbaum SL, Ohara J, Harwalkar VA, Erdmann aggressive HLA-DR-positive osteoclastoma possessing ‘functional’ JM, Lacey DL & Civitelli R 1994 Interleukin-4 inhibits bone calcitonin receptors. Journal of Pathology 175 293–299. resorption and acutely increases cytosolic Ca2+ in murine osteoclasts. Journal of Biological Chemistry 269 13817–13824. Rathod H, Gillespie JI & Datta HK 1995 Evidence for parathyroid Glowacki J, JastyM&GoldringS1986 Comparison of multinucleate cells hormone-induced mobilisation of intracellular calcium in the rat elicited in rats by particulate bone, polyethylene, or polymethyl- osteoclast. Journal of Physiology 489P 71. methacrylate. Journal of Bone and Mineral Research 1 327–331. Salo J, Lehenkari P, Mulari M, MetsikkoK&Vaananen HK 1997 Kajiya H, Okabe K, Okamoto F, TsuzukiT&SoedaH2000 Protein Removal of osteoclast bone resorption products by transcytosis. tyrosine kinase inhibitors increase cytosolic calcium and inhibit Science 276 270–273. actin organization as resorbing activity in rat osteoclasts. Journal of Silver IA, Murrills RJ & Etherington DJ 1988 Microelectrode studies Cell Physiology 183 83–89. on acid microenvironment beneath adherent macrophage and MacIntyre I, Zaidi M, Alam AS, Datta HK, Moonga BS, Lidbury PS, osteoclasts. Experimental Cell Research 175 266–276. HeckerM&VaneJR1991 Osteoclastic inhibition: an action of Teti A, Grano M, Colucci S, Argentino L, Barattolo R, Miyauchi A, nitric oxide not mediated by cyclic GMP. PNAS 88 2936–2940. Teitelbaum SL, Hruska KA & Zambonin-Zallone A 1989 Voltage Miyauchi A, Hruska KA, Greenfield EM, Duncan R, Alvarez J, dependent calcium channel expression in isolated osteoclasts. Barattolo R, Colucci S, Zambonin-Zallone A, Teitelbaum SL & Teti Bollettino – Societa Italiana Biologia Sperimentale 65 1115–1118. A 1990 Osteoclast cytosolic calcium, regulated by voltage-gated Wiltink A, Nijweide PJ, Scheenen WJ, YpeyD&VanDuijnB1995 calcium channels and extracellular calcium, controls Cell membrane stretch in osteoclasts triggers a self-reinforcing Ca2+ assembly and bone resorption. Journal of Cell Biology 111 2543–2552. entry pathway. Pflugers Archiv – European Journal of Physiology 429 Mogami H, Nakano K, Tepikin AV & Peterson OH 2000 Ca2+ flow 663–671. 2+ via tunnels in polarised cells: recharging of apical Ca stores by Zaidi M, MacIntyreI&Datta HK 1990 Intracellular calcium in the 2+ focal Ca entry through basal membrane patch. Cell 88 49–55. control of osteoclast function: paradoxical elevation of cytosolic free Moller JV, JuulB&Maire M 1996 Structural organization, ion calcium by verapamil. Biochemical and Biophysical Research transport, and energy transduction of ATPases. Biochimica et Communications 167 807–812. Biophysica Acta 1286 1–51. Zhang X, Wen J, Bidasee KR, Besch HR Jr, Wojcikiewicz RJ, Lee Moonga BS, Davidson R, Sun L, Adebanjo OA, Moser J, Abedin M, B & Rubin RP 1999 Ryanodine and inositol triphosphate receptors Zaidi N, Huang CL & Zaidi M 2001 Identification and are differentially distributed and expressed in rat parotid gland. characterization of a sodium/calcium exchanger, NCX-1, in Biochemical Journal 340 519–527. 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, Received in final form 30 October 2002 Inzerillo A, Abe E, Huang CL & Zaidi M 2002 Ca (2+) influx Accepted 30 October 2002

www.endocrinology.org Journal of Endocrinology (2003) 176, 1–5

Downloaded from Bioscientifica.com at 10/02/2021 06:04:15PM via free access