CSF-1 induces fos gene transcription and activates the transcription factor Elk-1 in mature osteoclasts.

Mice with targeted deletion of the fos gene fail to develop mature osteoclasts, reflecting an absolute requirement for the c-Fos proto-oncogene in osteoclast precursors. C-Fos is also expressed in mature osteoclasts; however, the regulation of fos in these cells has not been studied. By using cultured murine osteoclast-like cells (OCLs) we found that treatment with colony-stimulation factor 1 (CSF-1) induced a 3.9-fold increase in c-Fos rnRNA at 30 minutes and a 2.6-fold increase at 60 minutes. With use of mature osteoclasts isolated from transgenic mice expressing the bacterial Lac-Z gene under the control of the murine fos promoter, we were able to directly demonstrate transcriptional activation of fos by CSF-1 in these cells. Transcriptional activation was 2.6-fold greater at 5 minutes and 2.8-fold greater at 15 minutes in CSF-1-treated cells than in vehicle-treated cells. CSF-1 induced nuclear protein binding to the fos serum response element that was significantly attenuated by antibodies to the transcription factor Elk-1 but not by Sap-1a. Treatment of mature osteoclasts with CSF-1 for 2 hours resulted in a significant increase in the levels of nuclear c-Fos protein. These data demonstrate that CSF-1 upregulates c-fos expression in mature osteoclasts at least in part via transcriptional activation of fos. CSF-1 induced binding of Elk-1 to the fos gene serum response element appears to be part of the molecular mechanism by which this occurs.

Evidence for a functional association between phosphatidylinositol 3-kinase and c-src in the spreading response of osteoclasts to colony-stimulating factor-1.

Osteoclasts are bone-resorbing cells whose normal function depends in part upon their ability to migrate over the bone surface to initiate new sites of bone resorption. The growth factor/cytokine, colony-stimulating factor-1 (CSF-1), potently stimulates osteoclast motility, in a c-src-dependent fashion. The intracellular signaling molecules that participate with c-src in CSF-1-induced remodeling of the osteoclast cytoskeleton have not been identified. Here we demonstrate, using the inhibitors wortmannin and LY294002, that activation of phosphatidylinositol 3-kinase (PI3-K) is required for CSF-1-induced spreading in osteoclasts. After CSF-1 treatment of osteoclast-like cells, PI3-K activity associated with the CSF-1 receptor c-fms, is increased, and the 85-kDa regulatory subunit of PI3-K and c-src coimmunoprecipitate. CSF-1 induces redistribution of PI3-K to the periphery of the cell. The association between p85 and c-src is due in part to a direct interaction between the proline-rich sequences of p85 and the SH3 domain of c-src. In vitro, the c-src SH3 domain stimulates PI3-K activity. Taken together, the current data suggest that c-src, via its SH3 domain, may participate in CSF-1-induced activation of PI3-K and that PI3-K and c-src are in the signaling pathway that subserves CSF-1-induced cytoskeletal changes in osteoclasts.

Parathyroid hormone increases circulating levels of fibronectin in vivo: modulating effect of ovariectomy.

To explore the effect of PTH on circulating levels of fibronectin (FN), adult female Sprague-Dawley rats were implanted with Alzet minipumps prepared to deliver 7 pmol/h x kg BW of either human PTH (1-34) or human PTH (1-84). Both forms of the hormone led to significant and progressive increases in circulating levels of FN over the 72-h study period (P < 0.001). However, at every time point, circulating levels of FN with human PTH (hPTH) (1-84) infusion were significantly higher than with hPTH (1-34), such that at the end of the infusion, mean levels in the hPTH (1-34) group were 32.2 +/- 1.4 ng/ml, in the hPTH (1-84) group 93.8 +/- 5.4 ng/ml, and in the vehicle infused group 14.6 +/- 0.7 ng/ml. The greater agonist efficacy of hPTH (1-84) was not explained by differences in circulating levels of the hormones, and both forms of the hormone were equipotent at stimulating cAMP production by ROS 17/2.8 cells. However, hPTH (1-84) remained a more effective agonist than hPTH (1-34) at stimulating FN production in these cells (P < 0.001). Nephrectomy did not blunt the ability of PTH to increase circulating FN in vivo, indicating that the kidney was not the source of the FN produced in response to PTH. Pretreament with the potent bisphosphonate APD to block bone resorption also did not blunt the in vivo response to PTH. Parathyroidectomy did not blunt the response. Cultured fetal rat bones showed a significant 2.4-fold increase in FN production when treated with PTH. Consistent with our earlier in vitro studies (Endocrinology, 135: 1639-1644, 1994), estrogen deficiency, induced by ovariectomy, significantly diminished the ability of PTH to increase circulating FN levels in vivo (P < 0.001). We conclude that PTH increases circulating levels of FN in vivo and may be a physiologic regulator for the plasma form of this glycoprotein. The effects of PTH on circulating FN may reflect the anabolic properties of the hormone in bone and the blunted response following estrogen withdrawal could be a manifestation of the diminished bone formation vis-à-vis resorption seen in the estrogen deficient state.

Colony stimulating factor-1 plays a role in osteoclast formation and function in bone resorption induced by parathyroid hormone and parathyroid hormone-related protein.

Although colony stimulating factor-1 (CSF-1) plays a key role in osteoclast recruitment, studies examining the effect of CSF-1 on mature osteoclasts indicate that it may directly inhibit bone resorption by isolated rat osteoclasts. To define further CSF-1's role in bone remodeling, we examined the effect of neutralizing antisera to CSF-1 on basal and parathyroid hormone (PTH)-induced bone resorption using two organ culture assays designed to examine the recruitment of osteoclast precursors and the activation of mature osteoclasts, respectively. We first assessed whether PTH increases CSF-1 production from bone in organ culture by examining conditioned medium from 19-day-old fetal rat long bones in a mitogenesis assay employing a CSF-1-responsive cell line, CRX-1. Conditioned medium from untreated bones induced a titratable increase in CRX-1 cell proliferation, and treatment of bones with PTH for 72 h caused a significant increase in mitogenic activity. CSF-1 antiserum caused a significant decrease in mitogenic activity in conditioned medium, indicating that bone in organ culture produces CSF-1 constitutively and in response to PTH. To examine bone-derived CSF-1's role in bone resorption, we examined the effect of neutralizing antisera to CSF-1 on basal and PTH-induced bone resorption in the fetal rat long bone assay, which reflects activation of mature osteoclasts. Anti-CSF-1 caused a significant increase in unstimulated and PTH-induced bone resorption compared with control. By contrast, in the fetal mouse metacarpal assay, which examines proliferation and recruitment of osteoclast progenitors and precursors, anti-CSF-1 caused significant inhibition of PTH related protein (PTHrP)-induced bone resorption after 3 and 6 days of incubation. Consistent with these findings, histological examination of cultured 17-day-old fetal metacarpals demonstrated that anti-CSF-1 inhibits the formation of tartrate-resistant acid phosphatase-positive osteoclasts in PTHrP-treated explants, whereas it has no effect on unstimulated bones. We conclude that bone-derived CSF-1 may have a dual role in PTH/PTHrP-induced bone resorption by enhancing the appearance of osteoclast precursors while restraining the resorptive function of mature osteoclasts.

Estrogen modulates parathyroid hormone-induced fibronectin production in human and rat osteoblast-like cells.

Fibronectin (FN) is an important constituent of the extracellular matrix in bone. Its hormonal regulation in this tissue has not been extensively studied. We, therefore, examined the effects of PTH, estrogen, and transforming growth factor beta on the production of FN by human and rat osteoblast-like cells. Confluent cells were stabilized for 48 h under estrogen-replete (10(-9) M 17 beta-estradiol) conditions and then continued under these same conditions or withdrawn from estrogen for varying periods of time. PTH over the range 10(-11)-(10(-8) M caused a dose-dependent increase in FN production [P < 0.001 by analysis of variance (ANOVA)] such that at the highest dose, FN production was increased 11-fold. Estrogen withdrawal for 96 h caused a significant diminution in PTH-induced FN production (P < 0.005 by two-way ANOVA). Estrogen withdrawal over the of period 48-144 h caused a progressive diminution in PTH-induced FN production such that differences in mean values for estrogen-replete vs. deficient conditions were greater at 144 than 48 h (P < 0.05). The estrogen effect was titratable over the range 10(-11)-10(-9) M, and the inactive congener 17 alpha-estradiol failed to prevent the inhibitory effect of estrogen withdrawal on PTH-induced FN production. Interestingly, estrogen withdrawal had absolutely no effect on transforming growth factor-beta-induced FN production. Northern analysis demonstrated no effect of PTH on steady-state FN messenger RNA levels in Saos-2 cells under either estrogen-replete or estrogen-deficient conditions, suggesting that PTH effects an increase in FN production via a posttranscriptional mechanism in these cells. We conclude that PTH stimulates FN production in human and rat osteoblast-like cells, and under estrogen-deficient conditions this effect is significantly diminished. The modulatory effect of estrogen is not a universal phenomenon because transforming growth factor-beta-induced FN production is unaffected by estrogen withdrawal.

Parathyroid hormone-related protein stimulates prostaglandin E2 release from human osteoblast-like cells: modulating effect of peptide length.

Parathyroid hormone-related protein (PTHrP) is a potent bone-resorbing protein that frequently mediates the humoral hypercalcemia of malignancy syndrome. Since prostaglandins may mediate the bone-resorptive action of certain hormones, we examined the effect of PTHrP on prostaglandin E2 (PGE2) secretion by human osteoblast-like cells. There was low-level basal secretion of PGE2 by Saos-2 cells (8.1 +/- 0.6 pg/ml). Using four different preparations of PTHrP, it was observed that with increasing peptide length, from 36 to 141 amino acids, a significant increase in efficacy for PGE2 release was seen in these cells. All forms of PTHrP were agonists for PGE2 release, with effects seen at concentrations as low as 10(-12) M in 48 h conditioned media. The amino terminus of the molecule appeared critical for this effect since the truncated derivative PTHrP-(7-34) did not induce significant PGE2 secretion. However, the influence of peptide length could not be explained by differential activation of adenylate cyclase since [Tyr36]PTHrP-(1-36)amide was equipotent to the longest peptide preparation, PTHrP-(1-141), in stimulating cyclic AMP accumulation in the Saos-2 cells. In contrast, PTHrP-(1-141) was significantly more effective than [Tyr35]PTHrP-(1-36)-amide in inducing a rise in cytosolic calcium. Further, this effect was noted at concentrations lower than those that caused significant cyclic AMP accumulation in the Saos-2 cells. PTHrP-(1-141) induced the release of PGE2 from primary human bone cell cultures to levels entirely comparable to those seen in the Saos-2 cells. PTHrP-(1-141) also induced PGE2 release by cultured fetal rat long bones at 72 h. We conclude that the carboxy-terminal region of PTHrP has important effects on cellular signal transduction pathways and on the release of a potent bone-active cytokine, PGE2.