Prostaglandin E2, Interleukin 1α, and Tumor Necrosis Factor-α Increase Human Osteoclast Formation and Bone Resorption in Vitro.

Prostaglandin E2 (PGE2) and the cytokines interleukin (IL) 1a and tumor necrosis factor (TNF)a increase bone resorption in vivo, but the effect of these agents on osteoclastic bone resorption has never been studied in an in vitro human system. Our recently described human bone marrow culture system, in which osteoclasts are generated (vitronectin and calcitonin receptor-positive cells which resorb bone), was used to study the effects of these agents. Addition of indomethacin to macrophage colony-stimulating factor (M-CSF)-treated cultures nearly abolished osteoclast parameters, indicating that prostaglandins are virtually essential for human osteoclast formation. Additionally, PGE2 dose responsively increased osteoclast numbers and bone resorption. The effects of M-CSF and PGE2 are independent, as demonstrated by unaltered PGE2 concentrations in culture supernatants in spite of the dose-responsive increase in osteoclast parameters in response to M-CSF. The generation of osteoclasts in the presence of PGE2 occurred in favor of CD 14-positive macrophage formation. IL 1a and TNFa increased osteoclast parameters in a dose-responsive manner. Maximum stimulation yielded culture supernatant levels of PGE2 approximately the same as those concentrations of exogenous PGE2 that dramatically induced osteoclast formation. This osteoclast-inducing effect was inhibited both by indomethacin and by the specific inhibitor of inducible prostaglandin G/H synthase, NS398, and this was reversed by addition of exogenous PGE2. These results demonstrate unequivocally that IL 1a and TNFa enhance human osteoclast formation and suggest that they mediate their effects through PGE2. (Endocrinology 139: 3157–3164, 1998) P (PGs) are ubiquitous local hormones that have been previously reported to exert important effects on the skeleton (for review, see Ref. 1). In particular, in vivo studies in humans and mice provide evidence that PGs enhance bone resorption (for review, see Ref. 2). It has been shown that they increase bone resorption (3) and osteoclast formation (4, 5) in organ culture. It was subsequently found that PGs exert an anticatabolic effect by inhibiting bone resorption by mature isolated neonatal rat osteoclasts (6–8), implying that the PG-induced increase in bone resorption in organ culture was mediated through cells other than the mature osteoclast. The more recent finding that PGs stimulate bone resorption and calcitonin (CT) receptor-positive cell formation (9–11), as well as tartrateresistant acid phosphatase-positive cell formation (12) in murine bone marrow (BM) cultures, suggests that the increase in bone resorption in organ culture may be the result of osteoclast formation. It has been suggested that prostaglandin E2 (PGE2) exerts the opposite effect on human osteoclast formation, having been found to suppress the formation of osteoclast-like 23c6positive cells in human BM cultures (13). 23c6 is a monoclonal antibody that identifies the vitronectin receptor and preferentially identifies osteoclasts (14, 15). However, bone resorption, which is unequivocal evidence that osteoclasts have developed, was not assessed in this study (13). We have previously reported, however, that PGE2 enhances osteoclastic bone resorption in human BM cultures, suggesting that PGE2 induces osteoclast formation in both human and murine species (16). This study was performed, however, before culture conditions had been optimized to enable reproducible formation of substantial numbers of human osteoclasts. Only a very small area of bone was resorbed in these cultures, and for this reason, osteoclast numbers were not assessed. The effect of PGE2 on human osteoclast formation, therefore, remains contentious. It is now possible to generate reproducibly large numbers of human osteoclasts, which resorb bone (17). This is achieved by adding macrophage colony-stimulating factor (M-CSF) to human BM cultures (17, 18). With the availability of a sound and reliable assay, we elected to test the effect of PGE2 on human osteoclast formation, because it is obviously important to resolve the issue of whether this local hormone exerts similar or opposing effects on the generation of human and murine osteoclasts. The proinflammatory cytokines, including interleukin (IL) 1a, tumor necrosis factor (TNF)a, and IL 6, are known to enhance bone resorption in murine cultures and in humans and mice in vivo (19). Evidence also exists, for the murine species, that many of the hormones and local factors that increase bone resorption also increase PG formation in bone cells (9, 20–25). There is evidence that this is regulated by the stimulation of an inducible form of the enzyme PG G/H synthase (PGHS-2) together with control of arachidonic acid release (25). However, we have been unable to demonstrate increased bone resorption in the presence of several of these Received January 2, 1998. Address all correspondence and requests for reprints to: Adrienne M. Flanagan, Department of Histopathology, Imperial College School of Science, Technology and Medicine at St. Mary’s, Norfolk Place, London W2 1PG, United Kingdom. E-mail: [email protected]. * This work was supported by The Wellcome Trust, United Kingdom. 0013-7227/98/$03.00/0 Vol. 139, No. 7 Endocrinology Printed in U.S.A. Copyright © 1998 by The Endocrine Society