THE OSTEOCLAST is the cell that resorbs bone. It derives

Transcription

1 /99/$03.00/0 Vol. 140, No. 4 Endocrinology Printed in U.S.A. Copyright 1999 by The Endocrine Society Prostaglandin E 2 Cooperates with TRANCE in Osteoclast Induction from Hemopoietic Precursors: Synergistic Activation of Differentiation, Cell Spreading, and Fusion MOHAN R. WANI, KAREN FULLER, NACK SUNG KIM, YONGWON CHOI, AND TIM CHAMBERS St. George s Hospital Medical School, London, United Kingdom SW17 ORE; The Rockefeller University (N.S.K., Y.C.), New York, New York 10021; and Howard Hughes Medical Institute (Y.C.), New York, New York ABSTRACT It was recently found that osteoblastic cells express TRANCE (tumor necrosis factor-related activation-induced cytokine), a newly identified member of the tumor necrosis factor superfamily, and that expression was increased by calciotropic hormones. Furthermore, soluble recombinant TRANCE induces osteoclast formation and resorption in stroma-free populations of hemopoietic precursor cells. However, overexpression of the decoy receptor osteoprotegerin in vivo shows that there are substantial differences in the sensitivity of different sites to resorption-inhibition, suggesting that either alternative ligands exist or the sensitivity of osteoclasts to TRANCE can be modified by cofactors. We therefore tested the possibility that cofactors might enhance osteoclast formation by TRANCE. We found that the number of tartrate-resistant acid phosphatase-positive and Received August 27, Address all correspondence and requests for reprints to: Dr. T. J. Chambers, Department of Experimental Pathology, St. George s Hospital Medical School, Cranmer Terrace, London, United Kingdom SW17 ORE. t.chambers@sghms.ac.uk. calcitonin receptor-positive cells was increased by a factor of 10 by the presence of PGE 2 in the absence of stromal cells. Moreover, although the tartrate-resistant acid phosphatase-positive cells that formed in TRANCE alone were typically mononuclear and poorly spread, the addition of PGE 2 induced the formation of large, well spread multinuclear cells. There was an increase in bone resorption that corresponded with the increase in osteoclast number. PGE 2 did not synergize with TRANCE for resorption-stimulation in mature cells. 8-Bromo-cAMP showed a similar syngergistic effect on osteoclastic differentiation. Thus, PGE 2 appears to stimulate bone resorption through a direct effect on hemopoietic precursors, primarily through a synergistic effect on the ability of TRANCE to induce osteoclastic differentiation. (Endocrinology 140: , 1999) THE OSTEOCLAST is the cell that resorbs bone. It derives from a macrophage colony-stimulating factor (M-CSF)- dependent precursor shared with the macrophage, which differentiates into osteoclasts when precursors are incubated in contact with osteoblastic or bone marrow stromal cells. It was recently found that TRANCE (TNF-related activation induced cytokine) (also called RANKL, OPGL, and ODF), originally identified as a member of the tumor necrosis factor (TNF) family that stimulates dendritic cells (1 3), is expressed by osteoblastic cells and substitutes for stromal cells in osteoclast formation (4, 5). Moreover, the ligand is upregulated in osteoblastic cells by calciotropic hormones and stimulates bone resorption by mature isolated osteoclasts (4 6). Osteoprotegerin (OPG) has also been identified as a soluble member of the TNF receptor superfamily that binds to TRANCE (5, 7). OPG inhibits both osteoclast formation and bone resorption in vitro and appears to be a decoy receptor (5, 6). However, although overexpression of OPG in transgenic mice suppresses bone resorption in many sites, bone modeling and tooth eruption, which both require bone resorption, are unaffected (7). This suggests that there may be alternative osteoclast-inductive ligands, or that the sensitivity of resorption to modulation by OPG might be modified in some sites by factors that synergize with TRANCE. PGs have long been considered to play a crucial role in bone physiology. They strongly stimulate bone resorption and osteoclast formation in intact bone tissue and bone marrow cultures (8 12). Many agents that induce bone resorption cause PG production in osteoblastic or bone marrow stromal cells, and suppression of PG synthesis in such systems also suppresses bone resorption (13 16). PGs may be important in inflammatory bone loss in diseases such as rheumatoid arthritis. The target cell for the action of PG on bone resorption has not been identified; PGs have effects on both the osteoblastic/bone marrow stromal cells that stimulate osteoclast formation and resorption and on immature cells of the mononuclear phagocyte series from which osteoclasts derive, but the relationship between these actions and osteoclast formation are unknown. Recently, PGs were found to stimulate expression of messenger RNA for TRANCE in osteoblastic cells (4). It is becoming increasingly clear that the osteoclast is a member of the mononuclear phagocyte family that is specialized for bone resorption. In the inflammatory process, in which the mononuclear phagocyte family plays a crucial part, PGs by themselves have little inflammatory capacity, but in the presence of other mediators they can synergistically amplify the local inflammatory response (17, 18). Moreover, PG cooperates with TNF in the activation of dendritic cells (19). Thus, although PG does not induce osteoclasts 1927

2 1928 PGE 2 COOPERATES WITH TRANCE Endo 1999 Vol 140 No 4 from precursors in the absence of stromal cells, it is possible that it might synergize with TRANCE. We therefore investigated the effects of PGE 2 on osteoclast induction by TRANCE. Materials and Methods Media and reagents MEM (Imperial Laboratories, Southampton, UK) with Earle s salts was used for all incubations supplemented with 10% heat-inactivated FBS, 2 mm l-glutamine, 100 IU/ml penicillin, and 100 g/ml streptomycin (all from Imperial Laboratories). Recombinant human M-CSF (rhm-csf) was provided by Dr. J. Wozney (Genetics Institute, Cambridge, MA). Human CD8-TRANCE was prepared as previously described (1, 2). PGE 2 (Sigma Chemical Co., Poole, UK) was dissolved in ethanol and stored as a m stock solution at 20 C in a dark glass tube. Indomethacin (Sigma Chemical Co.) was dissolved in ethanol and stored as a m stock solution at 20 C. 8-Bromoadenosine camp (8-Br-cAMP; Sigma Chemical Co.) was dissolved in medium 199 (Imperial Laboratories) containing 1 mg/ml BSA (Sigma Chemical Co.) and stored as a m stock solution at 20 C. All incubations were performed at 37 C in a humidified atmosphere of 5% CO 2 in air. Slices of devitalized bovine cortical bone, used as substrates for osteoclastic resorption, were prepared as previously described (20). Bone slices ( mm) were prepared from bovine femora using a low speed diamond edge saw (Buehler, Evanston, IL), cleaned by ultrasonication in sterile water, washed, sterilized by immersion in ethanol, and stored dry at room temperature. Isolation and culture of spleen cells Spleen cells were isolated from 2- to 5-day-old MF 1 mice as previously described (21). In brief, spleen cell suspensions were prepared by mechanically disaggregating spleens with a sterile scalpel blade followed by repeated passage through a 21-gauge needle. The suspension was washed twice and resuspended (10 6 /ml) in MEM and 10% heat-inactivated FBS. This suspension was added (100 l/well) to the wells of 96-well plates (Life Technologies, Uxbridge, UK), each well of which contained either a 6-mm Thermanox coverslip (Life Technologies) or a slice of bovine cortical bone. To each of these wells an additional 100 l medium containing M-CSF, TRANCE, and/or PGE 2 were added. Cultures were fed every 3 days by replacing 100 l culture medium with an equal quantity of fresh medium and reagents. After incubation for 7 days, bone slices were prepared for measurement of bone resorption, and coverslips were prepared for tartrate-resistant acid phosphatase (TRAP) staining and 125 I-labeled salmon calcitonin ([ 125 I]CT) autoradiography as described below. Isolation and culture of bone marrow precursors Bone marrow cells were isolated from 5- to 8-week-old MF1 mice as previously described (22). Mice were killed by cervical dislocation. Femora and tibiae were aseptically removed and dissected free of adherent soft tissue. The bone ends were cut, and the marrow cavity was flushed out into a petri dish by slowly injecting medium 199 at one end of the bone using a sterile 21-gauge needle. The bone marrow suspension was carefully agitated with a plastic Pasteur pipette to obtain a single cell suspension. The bone marrow cells were washed twice, resuspended in MEM containing 10% FBS, and incubated for 24 h in M-CSF (5 ng/ml) at a density of cells/ml in a 75-cm 2 flask. After 24 h, nonadherent cells were harvested, washed, and resuspended (10 6 /ml) in MEM-FBS. This suspension was added (100 l/well) to the wells of 96-well plates containing coverslips and bone slices. To each of these wells an additional 100 l medium containing M-CSF, TRANCE, and/or PGE 2 were added. Cells were incubated for 2 21 days. Cultures were fed every 3 days and assessed after incubation as described for spleen cell cultures. Isolation and culture of blood-derived monocytes MF 1 mice (5 8 weeks old) were anesthetized by injecting Avertin (BDH, Poole, UK) ip. The heart was exposed by opening the chest cavity aseptically, and blood was aspirated into a 2-ml syringe attached to a 19-gauge needle. The blood sample was transferred immediately into a sterile tube containing heparin (5 U/ml). The blood was diluted with Dulbecco s PBS, overlaid onto 3 ml Histopaque-1077 (Sigma Chemical Co.) in a 15-ml centrifuge tube, and centrifuged at 1200 g at 15 C for 10 min. Mononuclear cells were collected from the buffy coat, resuspended in cold PBS, and centrifuged at 200 g at 4 C for 10 min. Cells were resuspended (10 6 /ml) in MEM-FBS and cultured as described for spleen cells. Isolation of mature osteoclasts from neonatal rat bones Osteoclasts were isolated from neonatal rat long bones and sedimented onto bone slices as previously described (23). The number and area of excavations were assessed in the scanning electron microscope after 24-h incubation as previously described (23). Assessment of bone resorption After removing bone slices from cultures, cells on the surface of the bone slice were removed by immersion in 10% (vol/vol) sodium hypochlorite (BDH) for 10 min, followed by washing in water and dehydration in 70% ethanol. After drying, the bone slices were mounted on glass slides and sputter-coated with gold (SC500 sputter coater, Emscope Laboratories, Ashford, UK). The glass slides were examined by reflected light microscopy. Bone resorption was quantified using an eyepiece graticule. CT receptor (CTR) autoradiography Differentiation of CTR-positive cells was assessed using [ 125 I]CT, as previously described (24). Salmon CT (Novartis, Basel, Switzerland) was iodinated by a modification of the chloramine-t method (25). Coverslips were washed in medium 199-BSA, and then incubated with 0.2 nm [ 125 I]CT in medium 199-BSA for 1hatroom temperature. Controls consisted of coverslips incubated with excess (300 nm) unlabeled sct. After incubation, the coverslips were washed in PBS, fixed in formalin, and washed three times in distilled water. Coverslips were coated with K5 nuclear emulsion (Ilford, Ilford, UK), developed after 3 5 days at 4 C, and counterstained with Mayer s hemalum. CTR-positive cells were scored as those that demonstrated sufficient grain density to clearly outline the cells. For each coverslip, the number of CTR-positive cells present in 10 random fields were counted at 250 magnification. TRAP histochemistry Osteoclast formation was also evaluated by quantification of TRAPpositive cell number using a modification of the method of Burstone (26). After incubation, cells on coverslips were washed in PBS, fixed in 10% formalin for 10 min, and stained for acid phosphatase in the presence of 0.05 m sodium tartrate (Sigma Chemical Co.). The substrate used was napthol AS-BI phosphate (Sigma Chemical Co.). Only those cells that were strongly TRAP positive, showing a bright pink-colored cytoplasm, were counted by light microscopy. PGE 2 assay PGE 2 was assayed in the supernatants from some cultures using a commercially available enzyme immunoassay system (Amersham, Aylesbury, UK) according to the manufacturer s specifications. Culture supernatant (100 l) was taken for assay after incubation in cultures for 7 days. Statistical analysis of data Differences between groups were analyzed using unpaired Student s t test. P 0.05 was considered significant. Results We first tested the ability of TRANCE to induce osteoclastic cells from cells obtained from the spleens of 2- to 5-day-old mice. At this age bone marrow space is still insufficient for hemopoiesis, and the spleen cells are hemopoietic but lack

3 PGE 2 COOPERATES WITH TRANCE 1929 the (osteoblastic) stromal cells required for bone resorption. No TRAP-positive cells or bone resorption was seen in cultures incubated with M-CSF, with or without PGE 2. A small proportion of cells incubated in M-CSF with TRANCE became strongly TRAP positive, and bone slices showed resorption (Table 1). Both TRAP cell production and bone resorption were substantially increased by PGE 2. Multinuclear TRAP-positive cells were also observed in the presence of PGE 2 (Fig. 1). Consistent with previously observed inhibitory effects of PGE 2 on macrophages and M-CSF-dependent precursors (27 31), total cell numbers were reduced in the presence of PGE 2. To determine whether the augmentation of the effect of TRANCE by PGE 2 was attributable to an effect on stroma, the experiment was repeated using bone marrow cells incubated at low cell density in M-CSF for 24 h. This depletes the cell preparations of stroma, as judged by the absence of any cell growth in control cultures incubated subsequently without M-CSF for 7 days. Compared with spleen, a greater proportion of cells became TRAP positive in the presence of TRANCE, and PGE 2 showed a synergistic effect on TRAPpositive cell production at least as great as that seen in the stroma-containing spleen cell cultures (Table 2). In nonadherent bone marrow cell cultures, multinuclear TRAP-positive cells were present without PGE 2, but the majority of TRAP-positive cells remained mononuclear and poorly spread (Fig. 2); sometimes such cells appeared as clusters, which might represent prefusion cells. In the presence of PGE 2, a substantial proportion of the cell population became TRAP positive, and PGE 2 increased the number and degree of multinuclearity of multinuclear cells. TRAP-positive mononuclear cells were frequently observed clustered around multinuclear TRAP-positive cells (Fig. 2). In the presence of PGE 2 very large, well spread polykaryons were frequently seen. Areas occupied by TRAP-positive cell debris within which pyknotic nuclei could be discerned were observed, suggestive of apoptosis among the large polykaryons. We noted in [ 125 I]CT autoradiographs that the multinuclear cells were strongly CTR positive. As previously noted in bone marrow cultures containing stroma, many mononuclear cells were also CTR positive, although these generally showed lower autoradiograph grain density than the multinuclear cells. Such CTR-positive mononuclear cells were always TRAP positive (data not shown). Additionally, many cells that were only weakly TRAP positive were present. CTR expression in these cells was below the level detected in our autoradiographs, but weakly TRAP-positive cells seen in cultures with TRANCE (see Fig. 2B) were nevertheless more strongly TRAP positive than the cells observed in cultures incubated without TRANCE (see Fig. 2A). Neither TRAP-positive nor CTR-positive cells were seen in cultures from which TRANCE was omitted. In view of the synergism between PGE 2 and TRANCE, it seemed possible that the frequent occurrence of TRAP-positive cells in cultures to which PGE 2 had not been added might be due to autocrine/paracrine actions of PGE 2 produced in response to TRANCE. To exclude the possibility that PGE 2 is a TRANCE-induced cofactor for TRANCE, we compared TRAP-positive cell production and bone resorption in the presence vs. the absence of indomethacin (10 6 m); TABLE 1. Effect of PGE 2 (10 6 M) on cells from hemopoietic mouse spleen TRAP-positive cells with 3 % of bone surface or more nuclei/cm 2 resorbed TRAP-positive nuclear density (nuclei/cm 2 ) Culture conditions Total cell no./cm 2 TRAP-positive cells/cm 2 TRAP-positive cells as % of total cell no. M-CSF TRANCE 275,000 12,700 3, , M-CSF TRANCE PGE 125,000 6,050 a 15,000 1,450 a a 20,350 1,300 a 1, a a Spleen cells from 2 5-day-old mice were incubated with M-CSF (30 ng/ml) and TRANCE (100 ng/ml) with or without PGE 2 (10 6 M)at10 5 cells/well in 96-well plates containing plastic coverslips or bone slices. TRAP-positive cell production and bone resorption were assessed after 7 days of incubation. Results are expressed as the mean SEM (n 6 cultures/variable). a P 0.05 vs. cultures with no added PGE2.

4 1930 PGE 2 COOPERATES WITH TRANCE Endo 1999 Vol 140 No 4 FIG. 1. TRAP-positive cells in cultures of cells from hemopoietic spleen after incubation in A) M-CSF (30 ng/ml) and TRANCE (100 ng/ml); and B) M-CSF, TRANCE, and PGE 2 (10 6 M) for 7 days. Magnification, 360. FIG. 2. Murine bone marrow cells were incubated ( cells/ml) in M-CSF (5 ng/ml) for 24 h before incubation on coverslips for 7 days in A) M-CSF (30 ng/ml); B) M-CSF plus TRANCE (100 ng/ml); C F) M-CSF, TRANCE, and PGE 2 (10 6 M). All cultures were stained for TRAP, except D, which was assessed for CTR expression by [ 125 I]CT autoradiography. The lower half of F shows part of a large apoptotic cell. The margin of the cell shows strongly TRAP-positive areas within which pyknotic, apoptotic nuclei can be discerned. The remainder of the cell is more weakly TRAP positive, resembling the center of cells such as that seen in E and also contains nuclear debris. Magnification, 240.

5 PGE 2 COOPERATES WITH TRANCE 1931 TABLE 2. Effect of PGE 2 (10 6 M) on production of TRAP-positive cells in cultures of nonadherent M-CSF-dependent bone marrow cells No. of TRAP-positive cells with 3 or more nuclei/cm 2 TRAP-positive nuclear density (no. of nuclei/cm 2 ) Culture conditions Total cell no./cm 2 TRAP-positive cells/cm 2 TRAP-positive cells as % of total cell no. M-CSF TRANCE 400,000 12,350 32,500 2, , , M-CSF TRANCE PGE 175,000 5,850 a 137,500 4,650 a a 189,150 3,900 a 6, a M-CSF-dependent nonadherent bone marrow cells were obtained by incubating bone marrow from adult mice for 24 h at cells/ml in 5 ng/ml M-CSF. Cells were then incubated (7 days) in 96-well plates (10 5 cells/well) with M-CSF (30 ng/ml) and TRANCE (100 ng/ml), with or without PGE 2 (10 6 M) before staining for TRAP. Results are expressed as the mean SEM (n 6 cultures/variable. a P 0.05 vs. cultures with no added PGE2. we found no significant effect (data not shown). Moreover, no ( 20 pg/ml) PGE 2 production in response to TRANCE could be detected in these cultures, even without indomethacin (data not shown), at the single time point measured. PGE 2 synergized with TRANCE for TRAP-positive cell production and bone resorption at concentrations of 10 7 m and above (Fig. 3). Control cultures incubated with M-CSF and PGE 2 alone showed neither TRAP-positive cell production nor bone resorption (data not shown). PGE 2 also reduced the concentration of TRANCE required for TRAP-positive cell production by a factor of 10 (Fig. 4). Similar synergism was seen for CTR-positive cell production and bone resorption (Fig. 4). The number of CTR-positive cells in these cultures correlated with the TRAP-positive cell number; in [ 125 I]CT autoradiographs also stained for TRAP, CTR-positive cells were universally TRAP positive; 50 80% of the mononuclear cells counted as TRAP positive were also CTR positive. Bone resorption, expressed per CTR-positive cell, was not increased by PGE 2 (Fig. 4). Although there appeared to be a trend for bone resorption to be increased to a slightly greater extent than osteoclast formation, as judged by TRAP-positive cell production, this effect was not always seen. We have previously found that PGE 2 does not increase bone resorption by mature osteoclasts disaggregated from neonatal rat bone. We found that PGE 2 did not significantly affect TRANCE-mediated stimulation of bone resorption by osteoclasts disaggregated from neonatal mouse long bones during 24 h of incubation of bone slices (data not shown). Moreover, with or without PGE 2, bone resorption in cultures of nonadherent bone marrow cells, expressed per CTR cell, was at least an order of magnitude lower than that previously noted in cultures of bone marrow cells that had not been depleted of bone marrow stromal cells (11, 24, 32). PGE 2 is known to have effects on macrophages and M- CSF-dependent bone marrow precursors. To determine whether this effect is in some way related to changes in the capacity of M-CSF-dependent cells to form osteoclasts, we compared osteoclast formation by nonadherent bone mar- FIG. 3. Dose dependency of synergism of PGE 2 with TRANCE. Nonadherent bone marrow cells were incubated in M-CSF (30 ng/ml) and TRANCE (100 ng/ml) before assessment of the number of TRAPpositive cells and bone resorption. n 6 cultures/variable. *, P 0.05 vs. no PGE 2.

6 1932 PGE 2 COOPERATES WITH TRANCE Endo 1999 Vol 140 No 4 FIG. 4. Effect of PGE 2 (10 6 M) on TRAP- and CTR-positive cell production and bone resorption in cultures of nonadherent bone marrow cells incubated for 7 days in varying concentrations of TRANCE. n 6 cultures/variable. FIG. 5. Effect of preincubation of nonadherent bone marrow cells for 5 days in PGE 2 (10 6 M) before addition of TRANCE (100 ng/ml). In A, nonadherent bone marrow cells were incubated, for the times shown, in M-CSF (30 ng/ml) and TRANCE (100 ng/ml) with or without PGE 2 before assessment of total cell number, TRAP-positive cell number, and bone resorption. In B, cultures were similarly treated, but the addition of TRANCE was delayed until day 5. n 6 cultures/variable. row cells to which TRANCE was added either immediately or after a 5-day preincubation period with PGE 2. We found (Fig. 5) that preincubation in PGE 2 did not enhance the proportion of cells that subsequently developed osteoclastic characteristics. We also noted that when TRANCE addition was delayed until the nonadherent precursors had been in culture for 5 days, bone resorption and TRAP-positive cell production were substantially reduced.

7 PGE 2 COOPERATES WITH TRANCE 1933 TABLE 3. Effect of PGE 2 (10 6 M) on production of TRAP-positive cells and bone resorption by murine peripheral blood mononuclear cells Culture conditions Total cell no./cm 2 TRAP-positive cells/cm 2 TRAP-positive cells as % of total cell no. % of bone surface resorbed M-CSF TRANCE 225,000 9, M-CSF TRANCE PGE 2 150,000 9,050 a 2, a a Murine peripheral blood mononuclear cells were incubated with M-CSF (30 ng/ml) and TRANCE (100 ng/ml) with or without PGE 2 (10 6 M) for 7 days before assessment of TRAP-positive cell production and bone resorption. Values are the mean SEM (n 8 cultures/variable). Three repeat experiments showed similar results. a P 0.05 vs. M-CSF TRANCE. TABLE 4. Effect of 8-Br-cAMP on production of TRAP-positive cells and bone resorption by nonadherent bone marrow cells Culture conditions Total cells/cm 2 TRAP-positive cells/cm 2 % of bone surface resorbed M-CSF TRANCE 345,000 2,900 8, M-CSF TRANCE 8-Br-cAMP 190,800 5,050 a 84,150 3,000 a a Nonadherent, M-CSF-dependent bone marrow cells were incubated for 7 days in M-CSF (30 ng/ml) and TRANCE (100 ng/ml) with or without 8-Br-cAMP (10 4 M). Values are the mean SEM of four cultures per variable. This experiment was repeated with similar results. a P 0.05 vs. cultures with no added 8-Br-cAMP. PGE2 2 also increased the number of cells that became TRAP positive in (stroma-free) cultures of peripheral blood mononuclear cells (Table 3). Perhaps consistent with the greater commitment of this population to lineages other than the osteoclast, a relatively small proportion of cells developed TRAP positivity, and correspondingly low levels of bone resorption were observed. 8-Br-cAMP (10 4 m) showed a similar effect on M-CSFdependent bone marrow cells to that observed using PGE 2 (Table 4), although bone resorption per TRAP-positive cell appeared lower than with PGE 2. Discussion PGs have long been known to play a crucial role in the regulation of bone resorption (33). However, until recently, because osteoclast formation and activation required a contact-dependent interaction with osteoblastic cells, it has not been possible to identify the mechanism(s) or even the target cell(s) through which PG exerts its actions. Because osteoblastic cells mediate much of the hormonal responsiveness of osteoclast behavior (34, 35), it has been considered likely that PGs act by inducing osteoblastic cells to stimulate osteoclasts. Consistent with this, PGE 2 has recently been shown to induce TRANCE in osteoblastic cells (4). Our data suggest, however, that PGE 2 additionally acts directly on precursors synergistically with TRANCE in the induction of osteoclastic differentiation and maturation. TRANCE is clearly necessary for osteoclast formation. As previously reported (4, 5), in the presence of M-CSF alone, stroma-depleted M-CSF-dependent osteoclast-macrophage precursors formed only macrophages. In the presence of TRANCE, up to 10% of cells expressed TRAP and CTR, which are characteristic of osteoclasts (24, 26, 36 38). Although PGE 2 was unable to induce osteoclast formation in the absence of TRANCE in these stroma-depleted bone marrow cultures, it synergized with TRANCE to increase the proportion of cells that were strongly TRAP positive and CTR positive by 3- to 10-fold. PGE 2 also increased bone resorption and caused increased fusion and increased spreading of TRAP-positive cells. Although the proportion of strongly TRAP-positive cells was increased by PGE 2, TRAP was detectable at lower levels in virtually all cells. Although these weakly positive cells were CTR negative, they were not seen in M-CSF alone, even with PGE 2. Thus, it may be that the entire population of nonadherent M-CSF-dependent bone marrow cells has the potential for osteoclast formation in the presence of a sufficiently potent osteoclast-inductive stimulus. The corollary to this is that expression of even low levels of TRAP in cells that are CTR negative might represent specifically osteoclastic differentiation, but to a level that does not reach full maturation. Thus, our previous observation of TRAP-positive cells in stroma-containing bone marrow cultures that did not develop CTR-positive cells or bone resorption (24, 39) might be interpreted as due to a weak osteoclastogenic stimulus through spontaneous TRANCE expression in nonhormonally stimulated stromal cells. The proportion of precursors that form osteoclasts may thus depend on the intensity of the osteoclast-inductive stimulus. It may also be reduced by prior commitment of precursors to alternative lineages; although many nonadherent M-CSF-dependent precursors became strongly TRAP positive, the proportion was 5-fold less if cells were allowed to differentiate for 5 days in M-CSF alone, and splenocytes and monocytes, which contain progressively smaller proportions of immature vs. mature mononuclear phagocytes, also developed progressively fewer TRAP-positive cells. PGE 2 also increased multinuclearity and cell spreading in TRAP-positive cells. However, bone resorption did not increase substantially when expressed per osteoclast nucleus. This is consistent with the failure of PGE 2 to synergize with TRANCE in stimulation of bone resorption by mature osteoclasts isolated from neonatal rat bone. In fact, the area of bone resorbed per CTR-positive cell was 10-fold less than we have previously observed in bone marrow cultures in which osteoclast-inductive bone marrow stromal cells are present. As TRANCE can substitute for osteoblastic cells in the stimulation of bone resorption by mature cells (6), it may be that stromal cells supply signals, in addition to TRANCE, that facilitate osteoclastic maturation.

8 1934 PGE 2 COOPERATES WITH TRANCE Endo 1999 Vol 140 No 4 The mechanism by which PGs synergize with TRANCE in osteoclast induction remains unknown. Preincubation of precursors in PGE 2 before TRANCE addition did not sensitize the precursors to TRANCE action. Our data suggest the involvement of a camp signaling pathway in the maturation of osteoclasts; addition of 8-Br-cAMP, which increases intracellular levels of camp, mimicked the effects of PGE 2 on osteoclast formation. Similarly, camp synergizes with TNF- in dendritic cell maturation (19), and in monocytes, camp synergizes with TNF- to up-regulate interleukin-1 (40). It has been considered likely for some time that PGE 2 acts through camp in mature osteoclasts; camp induces morphological and functional changes similar to those induced by PGE 2 and CT, both of which induce camp (20, 37, 41, 42). In the present experiments, 8-Br-cAMP stimulated osteoclast production by a factor of 20, suggesting that camp synergizes with TRANCE in osteoclast maturation. We noted that TRAP-positive mononuclear cells formed into foci and aggregates, which may have been the precursors of polykaryons. In other situations, PGs can coordinate the decisions of groups of identical cells by autocrine signaling (43, 44). However, we found no evidence either that TRAP-positive cell production was dependent upon PG synthesis, or that TRANCE induced PG synthesis in precursors. These foci may be the result of local derivation from a precursor with a propensity for TRAP differentiation or may reflect interactions preceding cell-specific fusion. Osteoclasts in these cultures showed a transient existence similar to that previously observed in bone marrow cultures. Macrophages persisted for prolonged periods. Previously, both M-CSF and TRANCE have been shown to be survival factors for osteoclasts (6, 23, 45). We noted morphological evidence of apoptosis in TRAP-positive multinuclear cells despite the presence of M-CSF/TRANCE. It may be that although their immediate survival is dependent on M-CSF and TRANCE, osteoclastic differentiation entrains an apoptotic mechanism resistant to blockade by TRANCE/M-CSF. This might be a component of the regulation of osteoclastic bone resorption through imposition of an upper limit on the activity of individual cells. Although TRANCE is sufficient, with M-CSF, for osteoclast differentiation and activation, PGE 2 strongly synergizes with TRANCE in osteoclast formation. There are probably other agents that similarly synergize with TRANCE in the induction and regulation of osteoclasts, and such agents might provide the potential for diversity and complexity in the cellular and hormonal regulation of bone resorption, such that regulatory inputs from morphogenetic, mechanical, calcium-regulating, and inflammatory stimuli can be made. These putative cofactors might account for the differential sensitivity of bones to suppression of resorption by OPG, and PGE 2 might contribute to the increased bone loss adjacent to inflammatory tissues, as is observed in rheumatoid arthritis and other diseases. References 1. Wong BR, Josien R, Lee SY, Sauter B, Li H-L, Steinman RM, Choi Y 1997 TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor. J Exp Med 186: Wong BR, Rho J, Arron J, Robinson E, Orlinick J, Chao M, Kalachikov S, Cayani E, Bartlett III FS, Frankel WN, Lee SY, Choi Y 1997 TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-jun N- terminal kinase in T cells. J Biol Chem 272: Anderson DM, Maraskovsky E, Billingsley WL, Dougall WC, Tometsko ME, Roux ER, Teepe MC, DuBose RF, Cosman D, Galibert L 1997 A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390: Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki S-I, Tomoyasu A, Yano K, Goto M, Murakami A, Tsuda E, Morinaga T, Higashio K, Udagawa N, Takahashi N, Suda T 1998 Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and identical to TRANCE-RANKL. Proc Natl Acad Sci USA 95: Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess TER, Colombero A, Elliott G, Scully S, Hsu H, Sullivan J, Hawkins N, Davy E, Capparelli C, Eli A, Qian YX, Kaufman S, Sarosi I, Shalhoub V, Senaldi G, Guo J, Delaney J, Boyle WJ 1998 Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93: Fuller K, Wong B, Fox S, Choi Y, Chambers TJ 1998 TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J Exp Med 188: Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang M-S, Lüthy R, Nguyen HQ, Wooden S, Bennett L, Boone T, Shimamoto G, DeRose M, Elliott R, Columbero A, Tan H-L, Trail G, Sullivan J, Davy E, Bucay N, Renshaw-Gegg L, Hughes TM, Hill D, Pattison W, Campbell P, Sander S, Van G, Tarpley J, Derby P, Lee R, Program AE, Boyle WJ 1997 Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89: Klein DC, Raisz LG 1970 Prostaglandins: stimulation of bone resorption in tissue culture. Endocrinology 86: Raisz LG 1982 Prostaglandins and skeletal metabolism. In: Lee JB (ed) Prostaglandins. Elsevier, New York, pp Shinar DM, Rodan GA 1990 Biphasic effects of transforming growth factor on the production of osteoclast-like cells in mouse bone marrow cultures: the role of prostaglandins in the generation of these cells. Endocrinology 126: Collins DA, Chambers TJ 1991 Effect of prostaglandins E 1,E 2 and F 2 on osteoclast formation in mouse bone marrow cultures. J Bone Miner Res 6: Akatsu T, Takahashi N, Debari K, Morita I, Murota S, Nagata N, Takatani O, Suda T 1989 Prostaglandins promote osteoclast like cell formation by a mechanism involving cyclic adenosine 3,5 -monophosphate in mouse bone marrow cell cultures. J Bone Miner Res 4: Tashjian Jr AH, Hohmann EL, Antoniades HN, Levine L 1982 Platelet-derived growth factor stimulates bone resorption via a prostaglandin-mediated mechanism. Endocrinology 111: Tashjian Jr AH, Voelkel EF, Lazzaro M, Singer FR, Roberts AB, Derynck R, Winkler ME, Levine L 1985 and human transforming growth factors stimulate prostaglandin production and bone resorption in cultured mouse calvaria. Proc Natl Acad Sci USA 82: Tashjian Jr AH, Voelkel EF, Lazzaro M, Goad D, Bosma T, Levine L 1987 Tumor necrosis factor- (cachectin) stimulates bone resorption in mouse calvaria via a prostaglandin mediated mechanism. Endocrinology 120: Tashjian Jr AH, Levine L 1978 Epidermal growth factor stimulates prostaglandin production and bone resorption in cultured mouse calvariae. Biochem Biophys Res Commun 85: Williams TJ, Morley J 1973 Prostaglandins as potentiators of increased vascular permeability in inflammation. Nature 246: Moncada S, Ferreria SH, Vane JR 1973 Prostaglandins, aspirin-like drugs and the oedema of inflammation. Nature 246: Rieser C, Böck G, Klocker H, Bartsch G, Thurnher M 1997 Prostaglandin E 2 and tumor necrosis factor cooperate to activate human dendritic cells: synergistic activation of interleukin 12 production. J Exp Med 186: Chambers TJ, McSheehy PMJ, Thomson BM, Fuller K 1985 The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116: Owens JM, Gallagher AC, Chambers TJ 1996 IL-10 modulates formation of osteoclasts in murine hemopoietic cultures. J Immunol 157: Chambers TJ, Owens JM, Hattersley G, Jat PS, Noble MD 1993 Generation of osteoclast-inductive and osteoclastogenic cell lines from the H-2K b tsa58 transgenic mouse. Proc Natl Acad Sci USA 90: Fuller K, Owens JM, Jagger CJ, Wilson A, Moss R, Chambers TJ 1993 Macrophage colony-stimulating factor stimulates survival and chemotactic behavior in isolated osteoclasts. J Exp Med 178: Hattersley G, Chambers TJ 1989 Calcitonin receptors as markers for osteoclastic differentiation: correlation between generation of bone-resorptive cells and cells that express calcitonin receptors in mouse bone marrow cultures. Endocrinology 125: Hunter WM, Greenwood FC 1962 Preparation of iodine 131 I labelled human growth hormone of high specific activity. Nature 194:495

9 PGE 2 COOPERATES WITH TRANCE Burstone MS 1958 Histochemical demonstration of acid phosphatases with naphthol AS-phosphate. J Natl Cancer Inst 21: Kurland J, Moore MAS 1977 Modulation of hemopoieses by prostaglandins. Exp Hematol 5: Taetle R, Koessler A 1980 Effects of cyclic nucleotides and prostaglandins on normal and abnormal human myeloid progenitor proliferation. Cancer Res 40: Taetle R, Mendelsohn J 1980 Modulation of normal and abnormal myeloid progenitor proliferation by cyclic nucleotides and PGE 1. Blood Cells 6: Williams N 1979 Preferential inhibition of murine macrophage colony formation by prostaglandin E. Blood 53: Oropeza-Rendon RL, Speth V, Hiller G, Weber K, Fischer H 1979 Prostaglandin E 1 reversibly induces morphological changes in macrophages and inhibits phagocytosis. Exp Cell Res 119: Hattersley G, Chambers TJ 1990 Effects of interleukin 3 and of granulocytemacrophage and macrophage colony stimulating factors on osteoclast differentiation from mouse hemopoietic tissue. J Cell Physiol 142: Raisz LG 1996 Interaction of local and systemic factors in the pathogenesis of osteoporosis. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis. Academic Press, San Diego, pp Chambers TJ 1992 Regulation of osteoclast development and function. In: Rifkin BR, Gay CV (eds) Biology and Physiology of the Osteoclast. CRC Press, Boca Raton, pp Suda T, Takahashi N, Martin TJ 1992 Modulation of osteoclast differentiation. Endocr Rev 13: Warshawsky J, Goltzman D, Rouleau MF, Bergeron JJM 1980 Direct in-vivo demonstration by autoradiography of specific binding sites for calcitonin in skeletal and renal tissues of the rat. J Cell Biol 85: Nicholson GC, Moseley JM, Sexton PM, Mendelsohn FAO, Martin TJ 1986 Abundant calcitonin receptors in isolated rat osteoclasts. J Clin Invest 78: Kaye M 1984 When is it an osteoclast? J Clin Pathol 37: Hattersley G, Chambers TJ 1989 Generation of osteoclastic function in mouse bone marrow cultures: multinuclearity and tartrate-resistant acid phosphatase are unreliable markers for osteoclastic differentiation. Endocrinology 124: Lorenz JJ, Furdon PJ, Taylor JD, Verghese MW, Chandra G, Kost TA, Haneline SA, Roner LA, Gray JG 1995 A cyclic adenosine 3,5 -monophosphate signal is required for the induction of IL-1 by TNF- in human monocytes. J Immunol 155: Chambers TJ, Dunn CJ 1983 Pharmacological control of osteoclastic motility. Calcif Tissue Int 35: Suzuki H, Nakamura I, Takahashi N, Ikuhara T, Matsuzaki K, Isogai Y, Hori M, Suda T 1996 Calcitonin-induced changes in the cytoskeleton are mediated by a signal pathway associated with protein kinase A in osteoclasts. Endocrinology 137: Gordon J, Tiller E, Roberts J, Kato K 1993 A community effect in muscle development. Curr Biol 3: Smith WL, Borgeat P 1985 The eicosanoids: prostaglandins, thromboxanes, leukotrienes, and hydroxyeicosaenoic acids. In: Vance DE, Vance JE (ed) Biochemistry of Lipids and Membranes. Benjamin-Cummings, Menlo Park, pp Jimi E, Shuto T, Koga T 1995 Macrophage colony-stimulating factor and interleukin-1 maintain the survival of osteoclast-like cells. Endocrinology 136: