Osteodasts are polarized, usually multinucleate cells that resorb bone by actively extruding large quantities of protons to help dissolve mineral, and by secreting proteolytic agents to break down the collagen matrix. Release occurs at a specialized, convoluted region of osteoclast membrane into a tightly sealed space, akin to a secondary lysosome, adjacent to the bone surface. Resorption is co-ordinated with bone formation to remodel the skeletal frame during growth and bring about a balanced turnover of bone in early and middle adult life. In later life, most notably in women after the menopause, osteoclast activity increases, leading to bone loss (osteoporosis) and fractures. Osteoporosis, much the commonest bone disorder, is a huge and increasing problem in the ageing populations of the developed world.
Not surprisingly, the regulation of osteoclast activity is a subject that now attracts much attention. Bone destruction is a function not only of the recruitment of new osteoclasts from monocytic precursors but also of the activity and lifespan of mature resorptive cells. The list of factors known to influence osteoclast activity is a long one which includes calcium-regulating, growth, adrenal and sex hormones, eicosanoids, cytokines and growth factors, electrolytes and free radicals, as well as mechanical loading. In vitro, mature osteo clasts are remarkably sensitive to pHo and are reversibly activated to form resorption pits when pH is reduced below 7.2 (Arnett & Spowage, 1996). Except for calcitonin and prostaglandins, the actions of endocrine or paracrine agents on mature osteoclasts seem to be indirect and the means by which the resorptive activity of these cells is controlled in vivo remain unclear.
ATP is now recognised as a key extracellular messenger for cell–cell signalling that involves two families of ATP receptors on the cell membrane: the P2Y receptor family couple to G proteins to stimulate phospholipases (PLA2, PLCβ and PLD) and activate a series of intra-cellular signalling pathways, including inositol 1,4,5-trisphosphate-dependent mobilization of intracellular Ca2+; the P2X receptor family gate cation channels permeable to Ca2+, Na2+, K+ and, probably, H+. At least seven main subtypes exist for both P2X and P2Y receptor families (Burnstock & King, 1996). ATP is released into the extracellular space via synaptic vesicles from nerve cells, but also as a result of cell damage and by active excretion via transport proteins -the ATP binding cassette superfamily-such as P-glycoproteins and sulphonylurea receptors.
The paper by Weidema et al. (1997) in this issue of The Journal of Physiology shows that ATP activates both non-selective cation and Ca2+-dependent K+ channels in rat osteoclasts, providing the first electrophysiological evidence for the co-existence of both P2X and P2Y receptors on bone-resorbing cells. This report draws attention to a novel reporter current, a calcium-activated K+ current, associated with P2Y receptor activation. Rat osteoclast K+ channels are shown to exhibit unique biophysical and pharmacological profiles distinct from K+ channels on avian osteoclasts. Ca2+-dependent K+ channels were insensitive to apamin, charybdotoxin and kaliotoxin, which otherwise block known examples of Ca2+-dependent K+ channels, thus broadening the interest of this paper to physiologists and biophysicists outside the field of bone formation and resorption. The potential for cross-talk between P2X and P2Y receptors in osteoclasts remains to be investigated. Conceivably, the calcium released by P2Y receptor activation may have an impact on the rate of inactivation of P2X receptors in osteoclasts.
The work of Weidema et al. (1997) builds on recent findings from other groups. First, Bowler et al. (1995) showed that the P2Y2 receptor (which is stimulated by UTP and ATP) is expressed in osteoclasts purified from human osteoclastoma tissue. Second, using rabbit osteoclasts preloaded with intracellular fluorochromes, Yu & Ferrier (1995) showed that ATP not only produced an intracellular Ca2+ pulse but also caused a transient decrease in pH1, indicating the existence of two distinct intracellular signalling pathways. Yu & Ferrier (1995) found little effect of β,γ-methylene ATP (which was once considered a prototypic agonist for native P2X receptors) on rabbit osteoclasts and cautiously avoided the specific mention of P2X receptors in osteoelasts. Of the known recombinant P2X receptors, however, only two of seven subtypes, P2X1 and P2X3, are activated by β,γ-methylene ATP.
What might be the implications of these findings for the ultimate function of osteoclasts, i.e. bone resorption? The data of Weidema et al. (1997) and Yu & Ferrier (1995) were derived from osteoclasts cultured on glass surfaces rather than bone and dentine surfaces. The observations by Weidema et al. (1997) were made predominantly on cells of a rounded phenotype thought to be characteristic of active osteoclasts, but such cells obviously lack the polarization and active proton pumping associated with resorption. It is unclear whether the pHi decrease reported by Yu & Ferrier (1995) in response to ATP is related to the transient opening of a non-selective (P2X-type) cation channel permeable to H+ and the transient inward current described by Weidema et al. (1997) or to other phenomena. In the absence of data, it is tempting to speculate that an ATP-elicited pHi decrease might favour resorption-pit formation - if a suitable mineralized substrate had been present - by facilitating the extrusion of H+. Following this argument, these phenomena might be expected to be only transient in non-polarized cells that were not actively extruding H+.
Is it possible that P2X receptors could be involved in gating some of the intracellular supply of H+ needed for resorption? There are some tantalizing hints. Recent data show that recombinant P2X2 receptors are sensitized by extracellular acidification, with a pH-activation profile (King et al. 1996) reminiscent of the powerful stimulatory action of H+ on pit formation by osteoclasts. Independent evidence for a H+-stimulated in selective conductance in osteoclasts is provided by the experiments of Nördstrom et al. (1995). Weidema et al. (1997) have provided important new clues to understanding the role for extracellular nucleotides in modulating osteoclast membrane currents and function. Hopefully, some of the technical difficulties associated with patch clamping and ion imaging of osteoclasts (actively resorbing or otherwise) on bone or dentine surfaces (which are opaque, calcium rich and autofluorescent) might soon be overcome.