It has recently been suggested that interleukin (IL)-11 plays a role in the pathogenesis of glucocorticoid (GC)-induced osteoporosis. IL-11 belongs to the gp130 cytokine family, which includes also IL-6. We have previously investigated GC-IL-6 interplay, showing that GC inhibits IL-6 release and IL-6 up-regulates GC receptor (GR) numbers in the human osteoblast-like cell lines Saos-2 and MG-63, which constitutively have an opposite pattern of expression for GR, IL-11, IL-6, alkaline phosphatase and osteoprotegerin (OPG). The aim of this study was to investigate GC-IL-11 interplay in the same two cell lines. First, cells were incubated with cortisol (0.01-1 microM) for 20 h in the presence and in the absence of a known IL-11 secretagogue (IL-1beta); cell media were assayed for IL-11 by ELISA. Secondly, cells were incubated with IL-11 (0.1-100 ng/ml) or specific anti-IL-11 monoclonal antibody for 20 h, and then assayed for GR by a radioligand binding assay. Similar to IL-6, both constitutive and IL-1beta-inducible IL-11 release were dose-dependently inhibited by cortisol (P<0.01); at variance with IL-6, exogenous IL-11 dose-dependently decreased GR numbers in MG-63 cells (P<0.05), while anti-IL-11 antibody significantly increased GR numbers in both cell lines (P<0.05). IL-11-induced reduction of GR in MG-63 cells was confirmed by Western blot analysis. While exerting opposite effects on GR numbers, neither IL-6 nor IL-11 significantly modified GC-dependent inhibition of OPG release. Our data indicate that even physiological concentrations of cortisol negatively modulate IL-11 secretion and demonstrate, for the first time, an inhibitory effect of the cytokine on GR. Thus, the concept of autocrine-paracrine loops that modulate GC action and involve gp130 cytokines is corroborated. These loops could have clinical relevance for the dynamics of bone loss in patients given GC and having high concentrations of these cytokines in the bone microenvironment.
A Dovio, ML Sartori, RG Masera, B Ceoloni, G Reimondo, P Prolo, S Racca and A Angeli
A Siddiqi, JM Burrin, DF Wood and JP Monson
Hyperthyroidism is associated with increased bone resorption but the mechanisms by which thyroid hormone (T3) affects bone cell metabolism remain unclear. Recently it has been suggested that T3 stimulates osteoclastic resorption indirectly through the release of soluble mediators from osteoblasts. The aim of the present study was to investigate whether the T3-induced increase in bone resorption could be due to the regulation of cytokine production by human osteoblasts (hOb). The effects of T3 (1, 10, 100 nM) and IL-1 beta (100 U/ml) as the positive control were examined on cytokine protein release and mRNA levels in cultured hOb cell lines (MG63, SaOs-2), primary hOb and human bone marrow stromal (hBMS) cells. T3 increased IL-6 and IL-8 mRNA levels as well as IL-6 and IL-8 protein release into the culture media from MG63 and hBMS cells in a time- and dose-dependent manner. The maximal effect on protein release in hBMS cells occurred at 24 h with a dose of T3 10 nM (IL-6 5.5 +/- 1.1-fold above controls; IL-8 3.7 +/- 0.5-fold above controls, P < 0.05). At the same time, mRNA levels in hBMS cells were increased 6.2 +/- 0.8-fold for IL-6 (P < 0.05) and 5.7 +/- 0.8-fold for IL-8 (P < 0.05). Similar results were obtained in MG63 cells but no response was seen in SaOs-2 or hOb cells despite measurable basal production. Nor was there detectable regulation of IL-1 beta, IL-3, IL-11, IL-4 or granulocyte macrophage-colony stimulating factor by T3 in any cell type. In conclusion, T3 increases IL-6 and IL-8 production by MG63 and hBMS cells, suggesting that IL-6 and IL-8 may be T3-regulated genes in osteoblasts.
D Swolin-Eide and C Ohlsson
High levels of glucocorticoids are believed to alter bone remodeling by decreasing bone formation and increasing bone resorption. It has been suggested that different cytokines, like interleukin-6 (IL-6) and interleukin-1 (IL-1), are involved in bone resorption by activating immature osteoclasts, and some studies indicate that IL-6 promotes bone formation by a mitogenic effect on osteoblasts. The aim of the present investigation was to study whether cortisol regulates the expression of IL-6 and IL-1 beta in human osteoblast-like cells. A high dose of cortisol (10(-7)M) decreased, as expected, the C-terminal propeptide of type I collagen released into the culture medium. The IL-6 mRNA levels and IL-6 protein released into the culture medium were also decreased by cortisol in a dose-dependent manner. The maximum effect was seen at 1 microM cortisol (mRNA 23.1 +/- 7.9% of control culture; protein 28.2 +/- 8.3% of control culture). The decrease in IL-6 mRNA levels was apparent 4 h after the addition of cortisol and was still present 20 h later. The decrease in IL-6 protein released into the culture medium was seen 20 h later than the decrease in IL-6 mRNA levels. The production of IL-1 beta protein released into the culture medium was decreased in a dose-dependent manner after the addition of cortisol with a maximum effect at 1 microM. The effect of cortisol on IL-1 beta protein released into the culture medium was seen 16 h after the addition of cortisol. To summarize, cortisol decreases the expression of IL-6 as well as IL-1 beta in human osteoblast-like cells.
S Celic, Y Katayama, PJ Chilco, TJ Martin and DM Findlay
We have previously shown that an exogenous type I collagen matrix can regulate expression of mRNA for parathyroid hormone (PTH)-related protein (PTHrP) and its receptor, the PTH/PTHrP receptor, in the UMR106-06 osteogenic sarcoma cell line, which is considered to be representative of a relatively mature osteoblast phenotype. Consistent with those data, we show here that growth of UMR106-06 cells on type I collagen increased PTH/PTHrP receptor-binding capacity. Analysis of the binding data showed that the number of PTH/PTHrP receptors expressed by cells cultured on collagen was at least 2-fold greater than that of cells cultured on plastic. Expression of mRNA encoding alkaline phosphatase (ALP) and osteopontin (OP) was also upregulated in cells cultured on collagen, suggesting that interaction with collagen promotes the osteoblast phenotype in this cell line. Retinoic acid (RA), which has also been shown to promote osteoblastic differentiation, synergized with type I collagen to cause super-induction of OP mRNA. In contrast, RA abolished the collagen-induced increase in ALP mRNA and PTH/PTHrP receptor mRNA. The collagen-mediated increase in the expression of OP and PTH/PTHrP receptor mRNA, but not that of ALP, was perturbed by prior covalent modification of the collagen by non-enzymatic glycation. The collagen effects did not occur via interaction with RGD amino acid domains in type I collagen, but evidence was obtained for involvement of the DGEA amino acid cell-binding domain. The mechanism by which plating of UMR106-06 cells on a type I collagen substrate affects PTH/PTHrP receptor mRNA levels was investigated. Inhibition of cytoskeletal organization using cytochalasin D, and inhibitors of protein phosphatases, protein kinase C, phospholipase C and cyclooxygenase, did not abrogate the collagen-mediated effects. In contrast, treatment of cells with the protein tyrosine kinase inhibitor genistein, but not herbimycin A, dose-dependently abolished the collagen effects on the expression of PTH/PTHrP receptor, ALP and OP mRNA. These results show that a type I collagen substrate influences the expression of osteoblast-associated genes in a cell model of mature osteoblasts and suggests that this involves, at least in part, changes in intracellular tyrosine phosphorylation.
H Tanaka, A Wakisaka, H Ogasa, S Kawai and CT Liang
In order to establish the cellular basis for using growth factors as possible therapeutic agents for the age-dependent deficit in bone formation activity, we examined the individual and combined effects of IGF-I and/or platelet-derived growth factor (PDGF) on the gene expression of osteoblast-related markers in male rats. The expression of osteoblast markers was examined in the femurs of adult and old rats following marrow ablation, which amplifies gene expression activity. The mRNA levels of collagen(alpha1) (I) (COLI), alkaline phosphatase (AP), osteopontin (OP) and osteocalcin (OC) were significantly lower in the old as compared with the adult rats. To determine whether growth factors can abolish the age-related deficits in mRNA expression in old bone, PDGF and/or IGF-I were infused directly into the right femur for 5 days following marrow ablation. The contralateral femur was infused with vehicle only and used as a control. PDGF stimulated the expression of OP mRNA in both adult and old rats, whereas COLI, AP and OC mRNAs were not affected. IGF-I infusion did not have a significant effect on mRNA expression in adult rats. In contrast, treatment with IGF-I significantly enhanced the mRNA levels of COLI, AP and OP in old rats. To examine whether the combination of both factors could affect the expression of osteoblast markers synergistically, PDGF and IGF-I were infused together. In adult bones, the combined treatment with PDGF and IGF-I caused a slight increase in the level of OP gene expression but no change in AP, OC or COLI genes. Although neither IGF-I nor PDGF alone was effective in stimulating the expression of OC, the combined treatment in old bones enhanced OC expression significantly. The expression of COLI, AP and OP was also stimulated, but the stimulation was no different from that of IGF-I alone. In PDGF plus IGF-I treatment with a high dose, no dose-response effects were observed. Within the limits of the present study, it is suggested that IGF-I and, to a much lesser extent, PDGF may partially restore the deficit in the expression of osteoblast markers in old bones, and that the combination of both factors is slightly better than IGF-I alone in stimulating OC expression.
H Tokuda, K Hirade, X Wang, Y Oiso and O Kozawa
We previously reported that basic fibroblast growth factor (FGF-2) activates p44/p42 mitogen-activated protein (MAP) kinase resulting in the stimulation of vascular endothelial growth factor (VEGF) release in osteoblast-like MC3T3-E1 cells and that FGF-2-activated p38 MAP kinase negatively regulates VEGF release. In the present study, we investigated the involvement of stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK) in FGF-2-induced VEGF release in these cells. FGF-2 markedly induced the phosphorylation of SAPK/JNK. SP600125, an inhibitor of SAPK/JNK, markedly reduced the FGF-2-induced VEGF release. SP600125 suppressed the FGF-2-induced phosphorylation of SAPK/JNK without affecting the phosphorylation of p44/p42 MAP kinase or p38 MAP kinase induced by FGF-2. PD98059, an inhibitor of upstream kinase of p44/p42 MAP kinase, or SB203580, an inhibitor of p38 MAP kinase, failed to affect the FGF-2-induced phosphorylation of SAPK/JNK. A combination of SP600125 and SB203580 suppressed the FGF-2-stimulated VEGF release in an additive manner. These results strongly suggest that FGF-2 activates SAPK/JNK in osteoblasts, and that SAPK/JNK plays a part in FGF-2-induced VEGF release.
M Nasu, T Sugimoto, H Kaji and K Chihara
Although there is clinical evidence showing that combined therapy with parathyroid hormone (PTH) and estrogen is additively effective in increasing the bone mass of patients with osteoporosis, the mechanism of the interaction between these hormones remains unclear. The present study was performed to determine whether estrogen would affect osteoblast proliferation and function modulated by PTH in human osteoblastic SaOS-2 cells. Human PTH-(1-34) significantly inhibited [(3)H]thymidine (TdR) incorporation, which was attenuated by 24 h pretreatment with 10(-10) to 10(-7) M 17 beta-estradiol (17 beta-E(2)) in a concentration-dependent manner. PTH significantly stimulated alkaline phosphatase (ALP) activity, collagen synthesis and type-1 procollagen mRNA expression after pretreatment with 17 beta-E(2 )in these cells. Tamoxifen, an anti-estrogen, antagonized these 17 beta-E(2)-induced effects. Pretreatment with insulin-like growth factor-I (IGF-I) mimicked estrogen action, and coincubation of 3 microg/ml anti-IGF-I antibody antagonized the effects of 17 beta-E(2 )as well as those of IGF-I. In the presence of 17 beta-E(2 )pretreatment, PTH strongly stimulated IGF-binding protein (IGFBP)-5 mRNA expression in these cells, and recombinant IGFBP-5 increased type-1 procollagen mRNA expression and ALP activity. In conclusion, estrogen attenuates PTH-induced inhibition of osteoblast proliferation and PTH stimulates osteoblast function in the presence of estrogen pretreatment. IGF-I and/or IGFBP-5 seemed to be involved in the estrogen-induced modulation of PTH action on osteoblast proliferation and function.
JW Gunnet, K Granger, E Cryan and KT Demarest
Progestins are believed to exert positive effects on bone density through receptors located in osteoblasts. In the present studies, the binding characteristics and regulation of the progestin receptors in two osteoblast-like cell lines were compared with those in human breast lines. Human TE85 and murine MC3T3-E1 osteoblast-like cells contain a single, high-affinity progestin binding site whose affinity and concentration are lower than in human breast cells. The osteoblastic progestin binding sites showed the expected steroid specificity and associated with the cell nuclei when occupied by ligand. The progestin receptors in osteoblastic cells also had sedimentation coefficients similar to those receptors in breast cells. The regulation of the progestin receptor in the osteoblast-like cells was explored by treating them with estradiol. In contrast to the large, rapid change seen in the breast cells, the progestin receptor levels in the MC3T3-E1 cells showed only a small, delayed up-regulation with estradiol treatment. The progestin receptor number in the TE85 cells was unaffected by estradiol. Down-regulation of the progestin receptors was explored by treating the cells with the progestin, norethindrone (NET). NET administration produced a rapid drop in progestin binding sites in the breast cells and a smaller, more gradual decline in MC3T3-E1 progestin binding. While the maximal decrease in receptor number occurred within 24 h in the breast cells, the receptor number was still continuing to fall after 72 h in the MC3T3-E1 cells. The data presented here demonstrate that both human and murine osteoblast-like cells contain a functional progestin receptor whose binding characteristics and regulation are similar, but not identical, to those receptors in other progestin target tissues such as the breast.
S Celic, P J Chilco, J D Zajac, T J Martin and D M Findlay
We have previously shown that the response of osteoblasts to parathyroid hormone (PTH) can be influenced at the receptor level by growth on the physiological substrate, type I collagen, or by treatment with retinoic acid. We have also shown differential expression of genes when cells of the osteoblast lineage are grown on type I collagen. The aim of this study was therefore to examine the effect of retinoic acid and growth on type I collagen on PTH/PTH-related protein (PTHrP) receptor mRNA expression in the osteosarcoma osteoblast-like cell line UMR106–06.
PTH/PTHrP receptor mRNA levels, as assessed by Northern blot, of cells grown on collagen were increased up to 2-fold compared with cells on plastic and in a concentration-dependent manner with respect to collagen. An increase was seen as early as 6 h and was maintained over a 24 h period. This was not due to increased mRNA stability. Retinoic acid decreased the level of receptor mRNA on both plastic and collagen at each time but did not alter mRNA stability. For all treatments PTH/PTHrP receptor mRNA abundance, relative to glyceraldehyde-3-phosphate dehydrogenase, increased steadily over 24 h after subculture of cells. In contrast, PTHrP mRNA levels were reduced in cells on collagen, compared with plastic. PTH-stimulated cAMP levels of cells grown on collagen were increased compared with plastic at 24 h, but not earlier. Consistent with the mRNA data, retinoic acid decreased the amplitude of cAMP responses in cells on plastic and collagen. There was no evidence for changes in adenylate cyclase per se, since forskolin-induced cAMP levels did not change with either treatment. This study shows that known modulators of osteoblast maturation also affect signal transduction in these cells by regulating gene expression of the PTH/PTHrP receptor as well as the PTHrP ligand.
Journal of Endocrinology (1996) 150, 299–308
M. C. Slootweg, C. M. Hoogerbrugge, T. L. de Poorter, S. A. Duursma and S. C. van Buul-Offers
Specific binding to and proliferative actions of insulinlike growth factors-I and -II (IGF-I and -II) on fetal mouse osteoblasts were tested. Membranes of mouse osteoblasts were shown by binding competition studies to possess specific binding sites for IGF-I and IGF-II. When IGF-I was used as a tracer, half-maximal displacement was obtained with 1·11 μg IGF-I/1 and with 14 μg IGF-II/1. Displacement of 125I-labelled IGF-I was accomplished with 2·33 μg IGF-II/1 and with 55 μg IGF-I/1. Affinity cross-linking showed bands of 130 kDa 125I-labelled IGF-I and 260 kDa 125I-labelled IGF-II under reducing conditions, further indicating the presence of classical type-I and -II receptor sites on mouse osteoblasts. Mitogenic effects of IGFs were weak; a combination with epidermal growth factor or fibroblast growth factor showed strong synergistic action however. The possibility of autocrine/paracrine actions of IGFs is discussed.
Journal of Endocrinology (1990) 125, 271-277