Osteoclast formation, survival and morphology are highly dependent on exogenous cholesterol/lipoproteins.

Osteoporosis is associated with both atherosclerosis and vascular calcification. No mechanism yet explains the parallel progression of these diseases. Here, we demonstrate that osteoclasts (OCL) depend on lipoproteins to modulate cellular cholesterol levels and that this controls OCL formation and survival. Removal of cholesterol in OCL via high-density lipoprotein or cyclodextrin treatment dose-dependently induced apoptosis, with actin disruption, nuclear condensation and caspase-3 activation. One mechanism linked to the induction of OCL apoptosis was the cell-type-specific failure to induce HMG-CoA reductase mRNA expression, suggesting an absence of feedback regulation of de novo cholesterol biosynthesis. Furthermore, cyclodextrin treatment substantially suppressed essential M-CSF and RANKL-induced survival signaling pathways via Akt, mTOR and S6K. Consistent with these findings, cholesterol delivery via low-density lipoprotein (LDL) significantly increased OCL viability. Interestingly, OCLs from the LDL receptor (LDLR)-/- mouse exhibited reduced size and lifespan in vitro. Remarkably, LDLR+/+ OCL in lipoprotein-deficient medium phenocopied LDLR-/- OCL, while fusion and spreading of LDLR-/- OCL was rescued when cholesterol was chemically delivered during differentiation. With hyperlipidemia being associated with disease of the vascular system and bone, these findings provide novel insights into the selective lipoprotein and cholesterol dependency of the bone resorbing cell. Cell Death and Differentiation (2004) 11, S108-S118. doi:10.1038/sj.cdd.4401399 Published online 12 March 2004

M-CSF, TNFalpha and RANK ligand promote osteoclast survival by signaling through mTOR/S6 kinase.

Multinucleated bone-resorbing osteoclasts (Ocl) are cells of hematopoietic origin that play a major role in osteoporosis pathophysiology. Ocl survival and activity require M-CSF and RANK ligand (RANKL). M-CSF signals to Akt, while RANKL, like TNFalpha, activates NF-kappaB. We show here that although these are separate pathways in the Ocl, signaling of all three cytokines converges on mammalian target of rapamycin (mTOR) as part of their antiapoptotic action. Accordingly, rapamycin blocks M-CSF- and RANKL-dependent Ocl survival inducing apoptosis, and suppresses in vitro bone resorption proportional to the reduction in Ocl number. The cytokine signaling intermediates for mTOR/ribosomal protein S6 kinase (S6K) activation include phosphatidylinositol-3 kinase, Akt, Erks and geranylgeranylated proteins. Inhibitors of these intermediates suppress cytokine activation of S6K and induce Ocl apoptosis. mTOR regulates protein translation acting via S6K, 4E-BP1 and S6. We find that inhibition of translation by other mechanisms also induces Ocl apoptosis, demonstrating that Ocl survival is highly sensitive to continuous de novo protein synthesis. This study thus identifies mTOR/S6K as an essential signaling pathway engaged in the stimulation of cell survival in osteoclasts.

Novel, nonpeptidic cyanamides as potent and reversible inhibitors of human cathepsins K and L.

Compounds containing a 1-cyanopyrrolidinyl ring were identified as potent and reversible inhibitors of cathepsins K and L. The original lead compound 1 inhibits cathepsins K and L with IC(50) values of 0. 37 and 0.45 M, respectively. Modification of compound 1 by replacement of the quinoline moiety led to the synthesis of N-(1-cyano-3-pyrrolidinyl)benzenesulfonamide (2). Compound 2 was found to be a potent inhibitor of cathepsins K and L with a K(i) value of 50 nM for cathepsin K. Replacement of the 1-cyanopyrrolidine of compound 2 by a 1-cyanoazetidine increased the potency of the inhibitor by 10-fold. This increase in potency is probably due to an enhanced chemical reactivity of the compound toward the thiolate of the active site of the enzyme. This is demonstrated when the assay is performed in the presence of glutathione at pH 7.0 which favors the formation of a GSH thiolate anion. Under these assay conditions, there is a loss of potency in the 1-cyanoazetidine series due to the formation of an inactive complex between the GSH thiolate and the 1-cyanoazetidine inhibitors. 1-Cyanopyrrolidinyl inhibitors exhibited time-dependent inhibition which allowed us to determine the association and dissociation rate constants with human cathepsin K. The kinetic data obtained showed that the increase of potency observed between different 1-cyanopyrrolidinyl inhibitors is due to an increase of k(on) values and that the association of the compound with the enzyme fits an apparent one-step mechanism. (13)C NMR experiments performed with the enzyme papain showed that compound 2 forms a covalent isothiourea ester adduct with the enzyme. As predicted by the kinetic analysis, the addition of the irreversible inhibitor E64 to the enzyme-cyanopyrrolidinyl complex totally abolished the signal of the isothiourea bond as observed by (13)C NMR, thereby demonstrating that the formation of the covalent bond with the active site cysteine residue is reversible. Finally, compound 2 inhibits bone resorption in an in vitro assay involving rabbit osteoclasts and bovine bone with an IC(50) value of 0.7 M. 1-Cyanopyrrolidine represents a new class of nonpeptidic compounds that inhibit cathepsin K and L activity and proteolysis of bone collagen.

Nonpeptide alpha(v)beta(3) antagonists. 1. Transformation of a potent, integrin-selective alpha(IIb)beta(3) antagonist into a potent alpha(v)beta(3) antagonist.

Modification of the potent fibrinogen receptor (alpha(IIb)beta(3)) antagonist 1 generated compounds with high affinity for the vitronectin receptor alpha(v)beta(3). Sequential modification of the basic N-terminus of 1 led to the identification of the 5,6,7, 8-tetrahydro[1,8]naphthyridine moiety (THN) as a lipophilic, moderately basic N-terminus that provides molecules with excellent potency and selectivity for the integrin receptor alpha(v)beta(3). The THN-containing analogue 5 is a potent inhibitor of bone resorption in vitro and in vivo. In addition, the identification of a novel, nonpeptide radioligand with high affinity to alpha(v)beta(3) is also reported.

Retinoic acid effects on an SV-40 large T antigen immortalized adult rat bone cell line.

Clonal cell lines were established from adult rat tibia cells immortalized with SV-40 large T antigen. One clone (TRAB-11), in which retinoic acid (RA) induced alkaline phosphatase (AP) activity, was selected for further study. The TRAB-11 cells express high levels of type I collagen mRNA, type IV collagen, fibronectin, practically no type III collagen, little osteopontin, and no osteocalcin. RA stimulates proliferation of TRAB-11 cells (starting at 10 pM) and survival (starting at 100 pM). TRAB-11 cells synthesize fibroblast growth factor-2 (FGF-2), which has potent autocrine mitogenic effects on these cells and acts synergistically with RA. TRAB-11 cells attach better to type IV collagen than to fibronectin or laminin. Cell attachment to type IV collagen is increased by RA and decreased (65%) by an antibody directed against alpha1beta1 integrin. RA up-regulates steady-state levels of alpha1, mRNA without affecting beta1 mRNA expression. In conclusion, we report the establishment of a clonal cell line from the outgrowth of adult rat tibiae which is highly sensitive to RA in its growth and survival in culture, apparently as a result of integrin-mediated cell interaction with extracellular matrix proteins.

Alendronate mechanism of action: geranylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro.

Nitrogen-containing bisphosphonates were shown to cause macrophage apoptosis by inhibiting enzymes in the biosynthetic pathway leading from mevalonate to cholesterol. This study suggests that, in osteoclasts, geranylgeranyl diphosphate, the substrate for prenylation of most GTP binding proteins, is likely to be the crucial intermediate affected by these bisphosphonates. We report that murine osteoclast formation in culture is inhibited by both lovastatin, an inhibitor of hydroxymethylglutaryl CoA reductase, and alendronate. Lovastatin effects are blocked fully by mevalonate and less effectively by geranylgeraniol whereas alendronate effects are blocked partially by mevalonate and more effectively by geranylgeraniol. Alendronate inhibition of bone resorption in mouse calvaria also is blocked by mevalonate whereas clodronate inhibition is not. Furthermore, rabbit osteoclast formation and activity also are inhibited by lovastatin and alendronate. The lovastatin effects are prevented by mevalonate or geranylgeraniol, and alendronate effects are prevented by geranylgeraniol. Farnesol and squalene are without effect. Signaling studies show that lovastatin and alendronate activate in purified osteoclasts a 34-kDa kinase. Lovastatin-mediated activation is blocked by mevalonate and geranylgeraniol whereas alendronate activation is blocked by geranylgeraniol. Together, these findings support the hypothesis that alendronate, acting directly on osteoclasts, inhibits a rate-limiting step in the cholesterol biosynthesis pathway, essential for osteoclast function. This inhibition is prevented by exogenous geranylgeraniol, probably required for prenylation of GTP binding proteins that control cytoskeletal reorganization, vesicular fusion, and apoptosis, processes involved in osteoclast activation and survival.

Phosphatidylinositol 3-kinase association with the osteoclast cytoskeleton, and its involvement in osteoclast attachment and spreading.

Osteoclast activation involves attachment to the mineralized bone matrix and reorganization of the cytoskeleton, leading to polarization of the cell. Signaling molecules, PI3-kinase, rho A, and pp60c-src, were shown to be essential for osteoclastic bone resorption. In this study we have focused on the involvement of these signaling molecules in the early event of osteoclast activation: attachment, spreading, and organization of the cytoskeleton. Highly purified osteoclasts were fractionated into Triton X-100-soluble or cytosolic and Triton X-100-insoluble or cytoskeletal fractions, and the distribution of above-mentioned signaling molecules between the two fractions was examined. PI3-kinase, rho A, and pp60c-src all showed translocation to the cytoskeletal fraction upon osteoclast attachment to plastic. However, PI3-kinase and rho A, but not pp60c-src, showed further translocation of 2.4- and 3.2-fold, respectively, upon attachment of osteoclasts to bone. PI3-kinase translocation to the cytoskeleton was inhibited by either cytochalasin B or colchicine. Furthermore, treatment of osteoclasts with the PI3-kinase inhibitor wortmannin decreased its translocation, suggesting that PI3-kinase activity was needed for its translocation. Moreover, wortmannin inhibited osteoclast attachment to both bone and plastic and caused drastic changes in osteoclast morphology resulting in rounding of the cells, disappearance of F-actin structures or podosomes, and appearance of punctate or vesicular structures inside the cells. Osteoblastic MB1.8 cells and IC-21 macrophages did not show additional translocation of PI3-kinase or rho A upon attachment to bone or changes in attachment or morphology in response to wortmannin. Finally, PI3-kinase coimmunoprecipitated with alpha v beta 3 integrin from osteoclasts.

Protein-tyrosine phosphatase activity regulates osteoclast formation and function: inhibition by alendronate.

Alendronate (ALN), an aminobisphosphonate used in the treatment of osteoporosis, is a potent inhibitor of bone resorption. Its molecular target is still unknown. This study examines the effects of ALN on the activity of osteoclast protein-tyrosine phosphatase (PTP; protein-tyrosine-phosphate phosphohydrolase, EC 3.1.3.48), called PTPepsilon. Using osteoclast-like cells generated by coculturing mouse bone marrow cells with mouse calvaria osteoblasts, we found by molecular cloning and RNA blot hybridization that PTPepsilon is highly expressed in osteoclastic cells. A purified fusion protein of PTPepsilon expressed in bacteria was inhibited by ALN with an IC50 of 2 microM. Other PTP inhibitors--orthovanadate and phenylarsine oxide (PAO)-inhibited PTPepsilon with IC50 values of 0.3 microM and 18 microM, respectively. ALN and another bisphosphonate, etidronate, also inhibited the activities of other bacterially expressed PTPs such as PTPsigma and CD45 (also called leukocyte common antigen). The PTP inhibitors ALN, orthovanadate, and PAO suppressed in vitro formation of multinucleated osteoclasts from osteoclast precursors and in vitro bone resorption by isolated rat osteoclasts (pit formation) with estimated IC50 values of 10 microM, 3 microM, and 0.05 microM, respectively. These findings suggest that tyrosine phosphatase activity plays an important role in osteoclast formation and function and is a putative molecular target of bisphosphonate action.

Acid extrusion is induced by osteoclast attachment to bone. Inhibition by alendronate and calcitonin.

Acid extrusion is essential for osteoclast (OC) activity. We examined Na+ and HCO3(-)-independent H+ extrusion in rat- and mouse OCs by measuring intracellular pH (pHi) changes, with the pHi indicator BCECF (biscarboxyethyl-5-(6) carboxyfluorescein) after H+ loading with an ammonium pulse. 90% of OCs attached to glass do not possess HCO3- and Na(+)-independent H(+)-extrusion (rate of pHi recovery = 0.043 +/- 0.007 (SEM) pH U/min, n = 26). In contrast, in OCs attached to bone, the pHi recovery rate is 0.228 +/- 0.011 pHi U/min, n = 25. OCs on bone also possess a NH(4+)-permeable pathway not seen on glass. The bone-induced H+ extrusion was inhibited by salmon calcitonin (10(-8) M, for 2 h), and was not present after pretreating the bone slices with the aminobisphosphonate alendronate (ALN). At ALN levels of 0.22 nmol/mm2 bone, H+ extrusion was virtually absent 12 h after cell seeding (0.004 +/- 0.002 pH U/min) and approximately 50% inhibition was observed at 0.022 pmol ALN/mm2 bone. The Na(+)-independent H+ extrusion was not inhibited by bafilomycin A1 (up to 10(-7) M), although a bafilomycin A1 (10(-8) M)-sensitive H+ pump was present in membrane vesicles isolated from these osteoclasts. These findings indicate that Na(+)-independent acid extrusion is stimulated by osteoclast attachment to bone and is virtually absent when bone is preincubated with ALN, or when osteoclasts are treated with salmon calcitonin.

Isolation and characterization of highly enriched, prefusion mouse osteoclastic cells.

This study describes the isolation and characterization of highly enriched mammalian osteoclast precursors, released by the "disintegrin" echistatin, from an osteoclast formation culture. Incubation of a 6-day coculture of mouse bone marrow cells and mouse osteoblastic cells (MB1.8) with echistatin (30 nM), an RGD-containing snake venom, for 20 min yielded an 88-95% pure population of tartrate-resistant acid phosphatase-positive cells, 1.5 x 10(5) cells per 150 cm2 culture dish. These cells were mostly mononucleated and based on the following characteristics are considered to be prefusion osteoclasts (pOC cells): (i) presence of calcitonin (CT) receptors documented by 125I-sCT autoradiography and cAMP generation in response to salmon calcitonin; (ii) expression of mRNAs for alpha v beta 3 integrin, osteopontin, 92-kDa type IV collagenase (matrix metalloproteinase 9), carbonic anhydrase II, OC-2 (an "osteoclastic" cysteine proteinase), and protein tyrosine phosphatase epsilon; and (iii) high level expression of pp60c-src protein. The pOC cells resorb bone (form "pits" on bone slices) but only in the presence of osteoblastic MB1.8 cells and 1,25(OH)2D3. Resorption was inhibited by CT. In conclusion, we describe a rapid, reproducible procedure to isolate virtually pure mammalian prefusion osteoclasts, which should help in the study of osteoclast formation, composition, and function.

Soluble alpha v beta 3-integrin ligands raise [Ca2+]i in rat osteoclasts and mouse-derived osteoclast-like cells.

We evaluated the possible involvement of intracellular Ca2+ concentration ([Ca2+]i) changes in the action of alpha v beta 3-ligands, known to regulate osteoclast function. Rat osteoclasts or mouse osteoclast-like cells, as examined by microfluorimetry and fura 2, showed a transient [Ca2+]i increase when perfused with (all 0.1 microM) vitronectin, osteopontin, polypeptide echistatin, fibronectin, and Arg-Gly-Asp-Asp and Arg-Gly-Asp-Ser peptides (10(-4) M) but not with laminin, collagen I, collagen IV, or [Ala24]echistatin, in which Ala was substituted for Arg in the Arg-Gly-Asp complex. The threshold for echistatin was 10 pM, the 50% effective concentration was 1 nM, and the median [Ca2+]i increase was 420 nM above the resting level (217 +/- 22 nM) at saturating concentration of 0.1 microM. Echistatin did not cause Mn2+ influx, and 10 microM nifedipine, 10 nM omega-conotoxin, 5 mM Ni2+, or Cd2+ did not prevent [Ca2+]i change. However, extracellular Ca2+ was needed for the [Ca2+]i increase, probably enabling ligand-integrin interaction. Polyclonal and monoclonal (LM609) antibody as well as depletion of [Ca2+]i stores with 5 microM thapsigargin and Ca(2+)-free medium abolished the [Ca2+]i increase, after restoring extracellular Ca2+. Furthermore, the LM609 antibody induced a Ca2+ signal in the presence or absence of extracellular Ca2+, suggesting that the alpha v beta 3-ligand interaction is mediated at least partially by Ca2+ mobilized from intracellular stores.

Leukemia inhibitory factor binds with high affinity to preosteoblastic RCT-1 cells and potentiates the retinoic acid induction of alkaline phosphatase.

This study examines the effect of leukemia inhibitory factor (LIF) on preosteoblastic rat calvaria (RCT-1) cells, which acquire osteoblastic properties when treated with retinoic acid (RA). LIF potentiated the increase in alkaline phosphatase (AP) activity produced by RA. The LIF effect was time and dose dependent (EC50, approximately 1 pM). The earliest effects on AP activity were detected at 48 h, and maximal effects were observed after 72 h. RA increased AP mRNA about 2-fold at 3 h and 6-fold at 6 and 12 h. LIF further increased AP mRNA to 18-fold at 12 h. After RA treatment AP mRNA returned to control levels at 24 h, but in the presence of LIF, AP mRNA remained elevated at 24 and 72 h of treatment. When given alone, LIF had no effect on either AP activity or mRNA levels. Tumor necrosis factor-alpha and 1,25-dihydroxyvitamin D3 also potentiated the RA induction of AP, and interleukin-6 had a small effect, whereas granulocyte macrophage colony-stimulating factor had no effect. LIF alone had a small inhibitory effect on type 1 collagen mRNA, but did not oppose the stimulatory effect of RA. Consistent with these biological actions, LIF receptors were demonstrated on these cells. [125I]LIF bound to RCT-1 cells at 0 C with an apparent dissociation constant of 20 pM, and it was found that these cells express an average of 300 receptors/cell. Scatchard analyses showed a single class of high affinity binding site. LIF was internalized with an endocytic rate constant for occupied receptors of 0.03 min-1, and the apparent equilibrium dissociation constant at 37 C was 358 pM. These findings suggest that osteoblast precursor cells are among the target cells of LIF.

IL-1 binds to high affinity receptors on human osteosarcoma cells and potentiates prostaglandin E2 stimulation of cAMP production.

IL-1 is a potent bone resorbing agent. Its mechanism of action is unknown, but the presence of osteoblasts was shown to be necessary for IL-1 stimulation of bone resorption by isolated osteoclasts. This study examines the presence of IL-1R and IL-1 effects in osteoblastic cells from a clonal human osteosarcoma cell line, Saos-2/B-10. We found that the binding affinity and the number of binding sites increases substantially during the postconfluent stage. Scatchard and curve-fitting analysis revealed one class of high affinity binding sites, with Kd/Ki's of 40 +/- 17 pM (mean +/- SD) for IL-1 alpha (n = 5) and 9 +/- 7 pM for IL-1 beta (n = 5) and 2916 +/- 2438 (n = 6) receptors/cell. Incubation of the cells with 125I-IL-1 alpha (100 pM) at 4 degrees C, followed by incubation at 37 degrees C up to 4 h, revealed internalization of receptor-bound IL-1 alpha. Chemical cross-linking studies showed that the IL-1R in Saos-2/B-10 cells had a molecular mass of approximately 80 kDa. To assess the biologic effect of IL-1 in Saos-2/B-10 cells, we determined PGE2 content and adenylate cyclase activity. Although IL-1 had no effect on PGE2 synthesis, both IL-1 alpha and IL-1 beta enhanced PGE2 stimulation of adenylate cyclase two- to four-fold in a dose-dependent manner. The half-maximal effect for IL-1 alpha was seen at 8 to 10 pM and for IL-1 beta at 0.6 to 1.8 pM. IL-1 did not enhance basal adenylate cyclase or stimulation by parathyroid hormone, isoproterenol, or forskolin. IL-1 enhancement of PGE2-stimulated adenylate cyclase was detected between 1 to 2 h, was maximal at 4 to 5 h, was not prevented by cycloheximide treatment, and was seen in membranes from IL-1 pretreated cells. These data show effects of IL-1 on a human osteoblast-like cell line that are mediated by high affinity receptors. These IL-1 effects could contribute to the biologic action of IL-1 on bone.

Opposing effects of fibroblast growth factor and pertussis toxin on alkaline phosphatase, osteopontin, osteocalcin, and type I collagen mRNA levels in ROS 17/2.8 cells.

In rat osteosarcoma (ROS 17/2.8) cells, which express osteoblastic features in culture, basic fibroblast growth factor (bFGF) reduces the level of alkaline phosphatase, type I collagen, and osteocalcin mRNA and increases osteopontin mRNA, independent of growth stimulation. The fibroblast growth factor (FGF) effects are dose dependent (EC50 about 6 pM) and are detected 24 h after addition of the growth factor. bFGF also reduces parathyroid hormone-stimulatable adenylate cyclase and alkaline phosphatase activity in these cells. Concomitant treatment with pertussis toxin (20 ng/ml) opposes the FGF effects. Although cyclic AMP elevating agents mimic pertussis toxin action on some parameters, they produce opposite effects on others, indicating that antagonism between pertussis toxin and bFGF is not mediated by cyclic AMP. bFGF caused a small reduction in steady state NAD-dependent ADP-ribosylation and had no detectable effects on the steady-state levels of the Gi alpha (alpha subunit of the inhibitory G protein) 1, 2, and 3, visualized with specific antibodies in these cells. Although the site of interaction of pertussis toxin and FGF remains to be determined, the findings presented here suggest separate control of growth and differentiation by bFGF and show that pertussis toxin treatment can modulate differentiation in these cells, presumably via Gi proteins.

Production of parathyroid hormone-like peptide in a human osteosarcoma cell line: stimulation by phorbol esters and epidermal growth factor.

A clonal cell line (Saos-2/B-10) derived from human osteosarcoma Saos-2 cells had the same osteoblastic characteristics as the mother line, but lacked sensitivity to parathyroid hormone (PTH) at early passages. At later passages (greater than 70) the cells became very sensitive to PTH (0.1 nmol/l). The absence of PTH-stimulatable adenylate cyclase correlated with the secretion of an adenylate cyclase-stimulatory activity which had the properties of the recently characterized PTH-like peptide (PTH-LP). This activity was inhibited by the PTH antagonist [8norleucyl,18norleucyl,34tyrosinyl]bovine PTH-(3-34)amide and could be neutralized by an antiserum raised against the synthetic PTH-LP-(1-34). Hybridization with a human PTH-LP cDNA showed that these cells produce two PTH-LP mRNAs of approximately 1.5 and 1.8 kb. The production of PTH-LP was stimulated by 12-O-tetradecanoyl-phorbol-13-acetate (TPA; 150 nmol/l) and epidermal growth factor (EGF; 10 ng/ml). The increased accumulation of PTH-LP in conditioned media in response to TPA was seen after 1 h and levelled off at 6 h. In contrast, EGF stimulation was lower at 3 and 6 h but continued for 24 h. Both agents increased PTH-LP mRNA levels in Saos-2/B-10 cells. A TPA analogue which does not stimulate protein kinase C had no effect on PTH-LP production. Cycloheximide blocked the stimulatory effect of both TPA and EGF and the TPA effect was blocked by actinomycin D, suggesting transcriptional control. The regulation of PTH-LP by these agents may offer clues regarding the association of this protein with malignancy.

Effects of acidic and basic fibroblast growth factors on osteoblastic cells.

Acidic (a) and basic (b) fibroblast growth factors (FGFs) are two related mitogenic and angiogenic factors. They are multifunctional in that they can affect proliferation and induce or delay differentiation. Both aFGF and bFGF were shown to stimulate proliferation of calvaria cells in situ as well as osteoblast-enriched calvaria-derived cells. bFGF was also found to suppress the expression of alkaline phosphatase, parathyroid hormone stimulatable adenylate cyclase, osteocalcin, and type I collagen in the osteoblastic ROS 17/2.8 cells. To explore a possible role for guanine nucleotide binding proteins we assessed the effects of pertussis toxin (PT) on FGF action. PT had opposite effects to those of bFGF on all parameters examined.

Comparison of postreceptor effects of 1-34 human hypercalcemia factor and 1-34 human parathyroid hormone in rat osteosarcoma cells.

A tumor-derived factor believed to cause hypercalcemia by acting on the parathyroid hormone (PTH) receptor was recently purified, cloned, and found to have NH2-terminal sequence homology with PTH. The 1-34 region of this protein was synthesized, evaluated for its postreceptor effects on the ROS 17/2.8 cell line, and its properties were compared to 1-34 PTH. Both 1-34 human humoral hypercalcemia factor (HCF) and 1-34 PTH stimulated adenylate cyclase with an effective concentration (EC)50 of approximately 1 nM. The extent of stimulation by both peptides was equally enhanced by dexamethasone. They both had a pronounced inhibitory effect on growth in the presence of dexamethasone, with an EC50 of approximately 0.1 nM, reduced alkaline phosphatase (AP) activity by approximately 70% in the absence of dexamethasone and by approximately 80% in the presence of dexamethasone with an EC50 of 0.03 nM, and when present at a concentration of 10 nM, reduced AP mRNA levels (estimated by Northern analysis) by approximately 80% in the presence or absence of dexamethasone. Thus, in addition to similar dose-response curves for adenylate cyclase stimulation, both HCF and PTH produced identical postreceptor effects in ROS 17/2.8 cells. These effects of HCF are probably mediated by the interaction of the tumor-derived factor with the PTH receptor.

Growth stimulation of rat calvaria osteoblastic cells by acidic fibroblast growth factor.

Purified acidic fibroblast growth factor (aFGF) from bovine brain stimulates the proliferation of calvaria-derived osteoblastic cells. Maximum stimulation, relative to corresponding controls, was seen at 0.2% serum (2- to 3-fold), and no stimulation was seen in the absence of serum or under serum replete conditions. The effect was dose-dependent with an ED50 of around 750 pg/ml (47 pM). aFGF (5 ng/ml) sustained the growth of calvaria cells in culture during multiple passages (72 days) at 0.2% serum. In DNA synthesis assays aFGF produced 2- to 4-fold stimulation; insulin-like growth factor I had a slight effect on DNA synthesis on its own, but enhanced the effect of aFGF 2-fold. In cells fully stimulated by epidermal growth factor (5-fold), aFGF had no further effect. Stimulation of DNA synthesis peaked at 5 ng/ml, while higher concentrations were inhibitory. Recombinant aFGF (bovine sequence) also stimulated cell proliferation (1.5-fold), and its potency was augmented by heparin (50 micrograms/ml), about 2-fold. Using simultaneous histochemical staining for alkaline phosphatase activity and [3H]thymidine nuclear uptake we found that aFGF stimulates DNA synthesis to the same extent in alkaline phosphatase-rich (osteoblastic) and alkaline phosphatase-poor (nonosteoblastic) cells. However, after cell division there is a significant decrease in PTH-responsive adenylate cyclase (2- to 3-fold) and in alkaline phosphatase levels (4- to 8-fold). These findings indicate that aFGF is mitogenic to rat calvaria osteoblastic cells, its action requires additional factors, and its growth stimulation is associated with a reduction in phenotypic expression.

Characterization of a human osteosarcoma cell line (Saos-2) with osteoblastic properties.

This study examines the osteoblastic properties of the established human osteosarcoma cell line Saos-2. Saos-2 cells inoculated into diffusion chambers, which were implanted i.p. into nude mice, produced mineralized matrix in 4 of 6 chambers at 8 weeks. In 5 of 6 chambers there was a strong positive alkaline phosphatase reaction. In culture the alkaline phosphatase levels increased with time and cell density, reaching very high levels at confluence: 4-7 mumol/mg protein/min. The cells show a sensitive adenylate cyclase response to parathyroid hormone, 50% effective dose = 2.8 nM, which increases with cell density and is further raised by dexamethasone treatment. They also exhibit typical binding of 1-25-dihydroxyvitamin D3 to 3.2S receptor protein with an apparent Kd of 0.21 nM; the numbers of sites per cell were 3,300 at 50,000 cells/cm2 and 1,800 at 280,000 cells/cm2. The presence of osteonectin was visualized with a monoclonal antibody which revealed a reticular pattern on the cell surface. Osteonectin was also detected in the medium by Western blots, migrating at around Mr 40,000 in nonreduced gels and Mr 44,000 in reduced gels. The Saos-2 cells thus possess several osteoblastic features and could be useful as a permanent line of human osteoblast-like cells and as a source of bone-related molecules.

Clonal differences in prostaglandin synthesis among osteosarcoma cell lines.

This study compares the metabolism of [14C]-arachidonic acid between PGE2 synthesizing (ROS 17/2.8) and nonsynthesizing (ROS 25/1) osteosarcoma cell lines. In both cell lines: (a) 90% of [14C]-arachidonic acid was taken up at 24 h. (b) More than 90% of the label was associated with phospholipids. (c) [14C]-arachidonic acid was rapidly taken up by phosphatidylcholine which reached the highest specific activity around 5 h while the labeling of other phospholipids was still increasing at 24 h. (d) Twenty-four hours after addition of [14C]-arachidonic acid only 4% of the label was associated with triacylglycerols in ROS 25/1 and 0.3% in ROS 17/2.8 cells. The calcium ionophore A23187 enhanced the release of [14C]-arachidonic acid from phospholipids in the PGE synthesizing osteoblastic cells (ROS 17/2.8 and 2/3) but had no effect in nonosteoblastic cells (ROS 24/1 and 25/1). ROS 17/2.8 and 2/3 cells converted the released arachidonic acid as well as exogeneously added arachidonic acid into PGE2. PGE2 synthesis depended on arachidonic acid concentration. Among bone resorbing agents, parathyroid hormone and 1,25(OH)2D3 had no effect on PGE synthesis, whereas thrombin and rabbit serum stimulated PGE2 production. The effect of rabbit serum was abolished by heat inactivation. The findings of this study indicate that the difference in PGE production between the osteoblastic and nonosteoblastic osteosarcoma cells are due mainly to differences in arachidonic acid conversion to PGE2.