Inhibition of osteoblast apoptosis by thrombin

Abstract

The multifunctional serine protease thrombin has been shown to be a specific agonist for a variety of functional responses of cells including osteoblasts. The current study was conducted to determine if thrombin was capable of inhibiting apoptosis in osteoblasts, and if so, to examine the mechanism by which this occurred. Thrombin (20–100 nM) significantly inhibited apoptosis in serum-starved cultures of the human osteoblast-like Saos-2 cell line and cultures of primary osteoblasts isolated from mouse calvariae, as well as dexamethasone-treated primary mouse osteoblasts. Inhibition of serum deprivation-induced apoptosis was shown to require thrombin's specific proteolytic activity. Primary mouse osteoblasts were found to express two functional thrombin receptors, PAR-1 and PAR-4. Thrombin inhibited serum deprivation-induced apoptosis in osteoblasts isolated from PAR-1 null mice to the same degree as in osteoblasts isolated from wild-type mice. Treatment of serum-deprived osteoblasts, isolated from either PAR-1 null or wild-type mice, with a PAR-4-activating peptide failed to significantly inhibit apoptosis compared to the relevant control. Medium conditioned by thrombin-treated osteoblasts, in which thrombin had been inactivated, was able to inhibit serum deprivation-induced osteoblast apoptosis almost as well as thrombin itself. Blocking protein synthesis, by cycloheximide pretreatment of the conditioning cells, prevented this action. The ability of known osteoblast survival factors, such as transforming growth factor β1, fibroblast growth factor-2, insulin-like growth factor-II, and interleukin-6, to inhibit serum deprivation-induced osteoblast apoptosis was also tested. None of these factors was able to inhibit serum deprivation-induced osteoblast apoptosis to the same extent as thrombin. The results presented here demonstrate that thrombin treatment of osteoblasts inhibits apoptosis induced either by dexamethasone or by serum deprivation. Furthermore, it does so independently of the known thrombin receptors by bringing about the synthesis and/or secretion of an unknown survival factor or factors, which then act in an autocrine fashion to inhibit apoptosis.

Introduction

The number of osteoblasts present in bone tissue is dependent not only on their proliferation rate, but also on their rate of apoptosis. A number of recent observations demonstrate that regulation of the osteoblast's lifespan is critical for the maintenance of the skeleton. Osteoblast apoptosis is increased in association with inflammation-mediated osteoporosis [1] and glucocorticoid-induced osteoporosis [2], and agents effective in managing the latter condition, such as bisphosphonates and calcitonin, inhibit glucocorticoid-induced osteoblast apoptosis [3]. Oestrogen prevents glucocorticoid-induced osteoblast apoptosis in vivo and in vitro [4], while another agent capable of increasing bone mass in vivo, parathyroid hormone (PTH), appears to exert this effect in part through the inhibition of apoptosis [5].

The coagulation protease, thrombin, exerts multiple effects on osteoblast behaviour. Thrombin stimulates proliferation, prostaglandin release, plasminogen activator inhibitor-1 synthesis, Ca_i²⁺ mobilization, and phosphoinositide metabolism in osteoblast-like cell lines and/or primary cultures of osteoblasts [6], [7], [8], [9], [10]. Many of thrombin's cellular actions in other cells are now known to be mediated by protease-activated receptors 1, 3, and 4 (PARs 1, 3, and 4) [11], [12], [13], [14], [15]. The other identified PAR (PAR-2) is not activated by thrombin, but by other proteases including trypsin [16]. The PARs are closely related members of the G protein-coupled seven-transmembrane domain family of receptors, and share a unique mechanism of activation involving proteolysis of the N-terminal extracellular domain. This cleavage leads to the creation of a new N-terminus described as a “tethered ligand,” which interacts with the second extracellular loop and activates intracellular signalling [11], [13], [14], [15], [16]. In the case of PARs 1, 2, and 4, synthetic peptides corresponding to the tethered ligand sequence are able to activate the relevant receptor in the absence of cleavage [11], [14], [15], [16]. Previously, we have demonstrated that osteoblasts express PAR-1 both in vivo and in vitro and that PAR-1 mediates thrombin-induced calcium mobilization and proliferation in these cells [10], [17], [18].

We have recently observed that thrombin inhibits serum deprivation-induced apoptosis in primary cultures of skeletal myoblasts [19]. The current study was undertaken to determine whether thrombin is also able to exert a similar effect in osteoblasts. Since initial studies indicated that thrombin indeed inhibits apoptosis of serum-deprived osteoblasts, further studies were conducted to shed light on the mechanism of this effect.