IBMS/ECTS 2001 - PROGRAM and ABSTRACTS
ABSTRACTS
Short communications will be presented orally and as
posters displayed throughout the conference.
The suffix after the abstract number indicates the poster session in which the poster will
be attended.
(W - Wednesday, T - Thursday, S - Saturday)
Click on the abstract number to view the short communication presentation session.
IDENTIFICATION OF A BISPHOSPHONATE WHICH INHIBITS FPP SYNTHASE AND IPP ISOMERASE
K. Thompson1*, F. P. Coxon1, J. E. Dunford1, F. H. Ebetino2, M. J. Rogers1
1Department of Medicine and Therapeutics, University of Aberdeen, UK
2Procter & Gamble Pharmaceuticals, Cincinnati, USA
Several recent studies have shown that the nitrogen-containing bisphosphonates (N-BPs) are inhibitors of farnesyl diphosphate (FPP) synthase, an enzyme in the mevalonate pathway. Heterocycle-containing N-BPs such as zoledronic acid and risedronate are extremely potent inhibitors of FPP synthase (IC50 3nM and 10nM respectively, with recombinant human enzyme). Inhibition of FPP synthase in osteoclasts causes loss of FPP and geranylgeranyl diphosphate, which are required for the isoprenylation of small GTP-binding proteins. Some N-BPs (ibandronate and incadronate) also inhibit squalene synthase, a distal enzyme in the mevalonate pathway. To date, no other enzymes in the pathway have been shown to be inhibited by N-BPs.
We have screened a large number of N-BPs for the ability to inhibit recombinant human FPP synthase and isopentenyl diphosphate (IPP) isomerase (the proximal enzyme to FPP synthase in the mevalonate pathway). One compound, NE21650 (containing an ortho-amino group attached to an aromatic ring) was found to be a potent inhibitor of FPP synthase (IC50 ~0.03microM) and also a weak inhibitor of IPP isomerase (IC50 ~70microM). By contrast, an isomer of this compound, NE10571 (containing a para-amino group) was found to be a poor inhibitor of FPP synthase (IC50 ~40microM) and, like all other bisphosphonates tested, had no effect on IPP isomerase activity at concentrations up to 300microM. In an MTT cytotoxicity assay with J774 macrophages, NE21650 decreased the number of viable cells (IC50 ~40microM), whilst the isomer NE10571 had no effect at concentrations up to 100microM. NE21650 was also a potent inhibitor of bone resorption by rabbit osteoclasts in vitro, whereas NE10571 had no effect. Furthermore, NE21650 (which inhibits FPP synthase and IPP isomerase) was a more potent inhibitor of bone resorption in vitro that ALN (which only inhibits FPP synthase, with a similar IC50 as NE21650).
These data illustrate that slight modifications to the structure of the nitrogen-containing side-chain of N-BPs determine the potency for inhibition of FPP synthase. In addition, certain side-chain structures (as in NE21650) also allow inhibition of other enzymes in the mevalonate pathway, such as IPP isomerase. Inhibition of more than one enzyme may lead to a slight increase in anti-resorptive potency.
IDENTIFICATION OF A PHOSPHONOCARBOXYLATE ANALOGUE OF RISEDRONATE WHICH INHIBITS GERANYLGERANYL TRANSFERASE II AND SELECTIVELY PREVENTS PRENYLATION OF RAB PROTEINS IN OSTEOCLASTS
F. P. Coxon1*, B. Larijani2, J. E. Dunford1, M. C. Seabra2, F. H. Ebetino3, S. P. Luckman1, M. J. Rogers1
1University of Aberdeen, UK
2Imperial College London, UK
3Procter & Gamble Pharmaceuticals, Cincinnati, USA
Nitrogen-containing bisphosphonates such as risedronate inhibit bone resorption by inhibiting farnesyl diphosphate (FPP) synthase, resulting in loss of isoprenoid lipids required for prenylation of small GTPases. Replacement of a phosphonate group of risedronate with a carboxylate moiety gives rise to a less potent anti-resorptive phosphonocarboxylate compound, NE10790. We examined whether the anti-resorptive property of this analogue could also be attributed to inhibition of FPP synthase.
NE10790 was less potent than risedronate at reducing cell viability of J774 macrophages (IC50 ~1200microM vs ~27microM respectively) and at inhibiting bone resorption by rabbit osteoclasts (IC50 ~900microM vs ~15microM respectively) in vitro. Whereas risedronate inhibited the incorporation of [14C]mevalonate into all prenylated (farnesylated and geranylgeranylated) proteins in J774 cells, NE10790 inhibited only the prenylation of proteins of molecular mass 22-26kDa that were not affected by FTI-277 or GGTI-298, peptidomimetic inhibitors of farnesyl transferase (FTase) and geranylgeranyl transferase (GGTase) I, respectively. Furthermore, when purified osteoclasts or J774 cells were labelled with [3H]geranylgeraniol, NE10790 specifically inhibited geranylgeranylation of 22-26kDa proteins that were not affected by GGTI-298. Together, these observations suggest that NE10790 inhibits GGTase II, the enzyme which geranylgeranylates 22-26kD Rab proteins. This was confirmed directly by immunoprecipitation of Rab6, Ras or Rho A from J774 cells labelled with [14C]mevalonate. NE10790 inhibited incorporation of [14C]mevalonate into geranylgeranylated Rab6, but not into farnesylated Ras or geranylgeranylated Rho A. Finally, NE10790 inhibited the activity of rhGGTase II in vitro (IC50 ~600microM), whereas risedronate had no effect. Neither NE10790 nor risedronate affected the activity of either rhFTase or rhGGTase I. Whilst risedronate was a potent inhibitor of rhFPP synthase (IC50 10nM), NE10790 was far less potent (IC50 ~900microM).
These data demonstrate that NE10790 is a newly-identified inhibitor of GGTase II. Although NE10790 is also a weak inhibitor of FPP synthase, treatment of intact cells with NE10790 results in the selective loss of prenylation of Rab proteins, the only known prenylation inhibitor to show this specificity. The anti-resorptive activity of NE10790 is likely due to disruption of intracellular membrane trafficking in osteoclasts for which Rab function is crucial.
ANDROGENS AND ESTROGENS SUPPRESS RANKL-INDUCED OSTEOCLAST FORMATION VIA A DIRECT ACTION ON OSTEOCLAST PRECURSORS
J. W. Pike*, A. C. Bendixen, D. M. Huber, K. M. Dienger, S. Srivastava, P. Pathrose, N. K. Shevde
University of Cincinnati, Cincinnati, Ohio, USA
Steroid deficiency in both men and women leads to a substantial increase in bone turnover and a critical imbalance between bone formation and bone resorption. This imbalance results in a progressive loss of trabecular bone mass and eventually osteoporosis, due in part to an increase in osteoclast formation from monocytic precursors in the bone marrow. The discovery of receptor activator of NF-kappa B ligand (RANKL), a TNF-like factor that together with M-CSF stimulates osteoclast formation independent of stromal cells, has enabled us to examine the direct effects of sex steroids on osteoclast differentiation. We have used murine bone marrow monocytes and murine RAW264.7 cells as models of osteoclast differentiation. Gonadectomy leads to a doubling of osteoclast precursors and an increased sensitization of the cells to both estrogens and androgens. Osteoclast formation in turn is dose-dependently suppressed by both 17 beta-estradiol, tamoxifen, raloxifene and the androgens. RANKL-induced osteoclast differentiation in RAW264.7 cells is also suppressed by estrogen, tamoxifen and raloxifene, and by the androgens testosterone and 5-dihydrotestosterone. These effects are mediated by the sex steroids' cognate receptors estrogen receptor alpha or beta and the androgen receptor. Estrogens and androgens both block osteoclast formation through a common mechanism that involves the transcription factors c-Jun, ATF1, ATF2 and CREB. Both sex hormones as well as tamoxifen and raloxifene suppress the activity of the SAPK c-Jun N-terminal kinase and reduce transcription factor activation. c-Jun expression is also dramatically downregulated. The effects of the sex steroids on this pathway appear to be specific in that neither hormone can downregulate RANKL induced activation of NF-kappa B or SAPK p38, both of which are also essential for osteoclast formation. The result of this action is to decrease the expression of AP-1 responsive genes likely to be involved in the differentiation process as well as genes such as TRAP and cathepsin K that are important for osteoclast function. These findings suggest that in addition to their indirect actions on supportive cells, sex hormones can also act directly to modulate osteoclast differentiation from precursors of the monocyte-macrophage lineage.
HUMAN OSTEOPROTEGERIN STIMULATES METALLOPROTEINASE ACTIVITIES OF RABBIT BONE CELLS BY THE RAS/MAPK PATHWAY
A. V. Rousselle*, E. Grimaud, Y. Fortun, M. Padrines, F. Redini, D. Heymann
Medicine Faculty, EE-99/01, Nantes, France
Osteoprotegerin (OPG) was recently identified as an essential factor inhibiting osteoclast-like cell formation by blocking the action of the osteoclast differentiation factor (ODF) expressed in stromal cells. Human osteoprotegerin (hOPG) has also been reported to inhibit the bone-resorbing activity of isolated mature rabbit osteoclasts and a 140 kDa OPG-binding protein has been detected on the plasma membrane of osteoclasts. The present study investigated the effects of hOPG on cysteine-proteinase (cathepsins) and matrix metalloproteinase (MMP-2 and MMP-9) activities, key proteases involved in bone resorption. Cathepsin activities were detected in supernatants by the action of a fluorogenic substrate and metalloproteinase activities by zymography. After 24 h of incubation with total rabbit bone cells, hOPG (1, 10, 50 and 100 ng/ml) stimulated in a dose dependent manner MMP-9 (latent form) and MMP-2 (active form) activities, but did not modulate cathepsin activities. The effects of hOPG were studied in the presence of specific inhibitors of key enzymes of different intracellular signaling pathways to determine the specific pathway involved in hOPG-stimulated MMP activity in bone cells. Gelatin zymograms indicated that Wortmannin (PI 3 kinase inhibitor), NF-kappaB SN50 (inhibitor of the translocation of NF-kappaB), and PKI 5-24 (protein kinase A inhibitor) did not prevent hOPG-induced MMP activities. Conversely, Genistein (inhibitor of tyrosine kinase), H7 (protein kinase C or cAMP-dependent protein kinase inhibitor), PD98059 (MEK inhibitor) and FPT inhibitor II (Ras inhibitor) completely abolished hOPG-induced MMP-9 and MMP-2 activities, indicating that the hOPG-induced ras/MAPK pathway specifically regulates MMP activities. Our experiments also indicated that although the cysteine-proteinase activities released by purified osteoclasts were not modulated by hOPG (1, 10, 50 and 100 ng/ml), MMP-9 activities released by purified osteoclasts were increased. Similarly, specific inhibitors of different enzymes of ras/MAPK pathway blocked hOPG-induced MMP-9 expression. These original data evidence a new hOPG-induced signal-transduction pathway in osteoclasts. hOPG appears to use the ras/ MAPK pathway whereas ODF acts on NF-kappaB pathway. The existence of two different signalling pathways suggests the presence of an OPG-binding receptor distinct from the ODF-binding protein, also known as RANK, expressed on the surface of osteoclasts.
ANTI-RESORPTIVE AND SURVIVAL ACTIONS OF SELECTIVE INTEGRIN PEPTIDOMIMETICS ON AUTHENTIC HUMAN OSTEOCLASTS
P. Collin-Osdoby1*, F. Anderson1, L. Rothe1, X. Yu1, M. Nelson1, Y. Huang1, W. Maloney1, W. Westlin2, G. A. Nickols2, P. Osdoby1
1Washington Unviversity, USA
2Pharmacia, USA
Mature osteoclasts (OC) undergo functional cycles of migration and resorption, processes dependent upon specific cell surface integrins. The integrin avb3 becomes highly expressed in OC development, mediates OC motility and attachment to RGD-containing extracellular matrix proteins, and is crucial for OC bone resorption and survival in vivo and in vitro. Conversely, RGD ligands and avb3 antagonists inhibit OC function and prevent bone loss in ovariectomized animals. However, the precise mechanisms of action and specificity of integrin antagonists on human OC (HOC) have not been well characterized. Here, we investigated two integrin peptidomimetics designated S787 and S783, with S787 (IC50 = 0.7 nM) being a more potent inhibitor than S783 (IC50 = 14.7 nM) for human avb3 binding, for their effects on HOC function and survival. Authentic HOC were isolated from femoral heads obtained during hip replacement surgery, cultured (10 days) on ivory with or without S783 or S787 (0.01 to 10 uM), and the ivory analyzed for TRAP stained HOC and resorption pits. Both peptidomimetics dose-dependently inhibited HOC pit resorption via decreasing the mean number of HOC remaining attached to the ivory, number of pits formed per HOC, area resorbed per HOC, and size of individual pits. Inhibition was more pronounced with S787 and maximal (up to 95%) at 1 uM of either S787 or S783. Whether decreased HOC numbers reflected increased apoptosis was evaluated using HOC or in vitro monocyte-generated OC (HMN-OC) treated with S787 or S783 (0.01 to 10 uM, 20 h) and stained with annexin V-FITC. Both peptidomimetics elicited HOC and HMN-OC apoptosis (not necrosis), S787 was more potent than S783, and maximal stimulation of HOC apoptosis (2 to 4-fold) occurred with 0.1 uM S787 or 10 uM S783. An RPA multiprobe kit was used to quantify potential changes in bcl family gene expression during peptidomimetic-induced apoptosis of HMN-OC (at 6 and 24 h) and revealed decreases in anti-apoptotic bclx (6 h) and bfl1 (6, 24 h). Thus, integrin peptidomimetics potently and dose-dependently inhibit HOC bone pit resorption, induce HOC apoptosis via a bcl-related pathway, and may exert differential effects as a function of their avb3 potency.
DISCOVERY AND ANALYSIS OF OSTEOBLAST RELATED GENES USING CDNA MICROARRAY
A. Raouf1,2, A. Seth1,2,3*
1Department of Laboratory Medicine & Pathobiology, University of Toronto, Canada
2CIHR group in Periodontal Physiology, University of Toronto, Canada
3Laboratory of Molecular Pathology, Sunnybrook and Women's College Health Sciences Centre
The osteoblast maturation is a complex process and involves distinct phenotypic changes that are accompanied by specific alterations in the expression of many genes. MC3T3-E1, a clonal osteogenic cell line, is an excellent model to study osteoblast development. We have shown that Ets1 is expressed at higher levels during the proliferation phase whereas Ets2 is at higher levels during the differentiation and mineralization phases of MC3T3-E1 development. To identify new bone related genes and potential ets targets in osteoblasts, we utilized high-density mouse GEM1 cDNA microarray gene chip from Incytegenomics. We compared expression profiles of 8700 genes during days 3 and 34 of MC3T3-E1 development where we observe substantial difference in Ets1 and Ets2 genes expression. 530 cDNAs were found to be differentially expressed between day 3 and 34. Majority of day 3 genes were consisted of cell cycle regulators, and transcription factors. Analysis of the day 34 genes revealed a number of signal transduction related and ECM associated proteins. The expression profiles of cyclin B1, Pr22, Mad-2, and TACC3, were confirmed by RT-PCR and Northern blots. Since the expression pattern of these genes resemble that of Ets1, it is possible that some of these are Ets1 targets. We also observed the expression of many growth suppressor genes, however, none of those have been previously described in osteoblasts. The expression of lumican, cystatin-C, and signal transduction related Hax-1 gene, and growth arrest gene TIS21/BTG2 were examined by RT-PCR and Northern blots. While lumican and cystatin-C were expressed at higher levels on day 34 compared to day 3, lumican could not be detected on day 3. The expression levels of Hax1 and TIS21/BTG2 were found to peak during differentiation phase of the MC3T3-E1, suggesting that they may have role(s) in osteoblast differentiation. It is possible that some of these genes are potential Ets2 targets since they co-express with Ets2. Our determination of genes active in the process of osteoblast proliferation and differentiation should prove to be of enormous value not only to the understanding of normal bone development but also to the understanding of molecular basis of many bone associated diseases.
IN VIVO EVIDENCE FOR ESTRADIOL-CONTROLLED INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-5 EXPRESSION IN BONE CELLS
E. M. Colin1, D. J. Lindenbergh-Kortleve2, G. J. C. M. van den Bemd1, C. J. Buurman1, J. W. van Neck2, S. L. S. Drop2, H. A. P. Pols1, J. P. T. M. van Leeuwen1*
1Department of Internal Medicine, Erasmus Medical Center Rotterdam, Rotterdam, The Netherlands
2Department of Pediatrics, Erasmus Medical Center Rotterdam, The Netherlands
Bone remodeling is regulated via a complex interplay between hormones and locally produced factors. In vitro studies have shown that estradiol up-regulates IGF-I mRNA in osteoblasts, which could account for the bone preservative effects of estrogens. In contrast to this, in vivo studies have shown a negative relation between serum estradiol and IGF-I levels. The aim of the present study was to assess whether this discrepancy is due to differences between local (bone) and systemic (liver) regulation of IGF-I by estradiol.
Three month old female BN rats (12 per group) were sham operated (SHAM), ovariectomized (OVX) and ovariectomized + estradiol suppletion (E2). After 4 weeks, serum IGF-I was determined by radioimmunoassay and expression of IGF-I and IGF-BP mRNAs was studied by non-radioactive in situ hybridization. Bone sections of each experimental condition were mounted on the same slide and hybridized together in one session. Three independent blinded observers scored the expression (number of cells and intensity) twice. In addition, BMD and serum 1,25-(OH)2D3, PTH, and osteocalcin were measured.
BMD and serum osteocalcin were modified according to that to be expected after OVX and E2 suppletion. IGF-I mRNA was detected in osteoblasts and osteocytes. IGFBP-5 mRNA was only expressed in osteoblasts. IGFBP-2 was weakly expressed while IGFBP-3, -4, and -6 mRNAs were undetectable. OVX resulted in a slight increase in serum IGF-I level compared to SHAM-operated rats. In contrast, IGF-I mRNA expression in bone was depressed while IGFBP-5 mRNA expression was not changed. E2 supplementation caused a significant reduction in serum IGF-I level. However, in bone IGF-I mRNA expression was restored to above SHAM-level and IGFBP-5 mRNA expression was even significantly increased compared to the SHAM group. The effects observed couldn't be explained by changes in 1,25-(OH)2D3 or PTH.
In conclusion, the presence of E2 is important for bone IGF-I and IGFBP-5 mRNA expression and might in this way influence bone remodelling. Also, it was found that serum IGF-I concentrations are not correlated with skeletal IGF-I mRNA expression. The current study thereby shows target tissue specific control of IGF-I and IGFBP-5 expression by estradiol.
PROTEIN KINASE C-DEPENDENT UP-REGULATION OF N-CADHERIN EXPRESSION BY PHORBOL ESTER IN HUMAN OSTEOBLASTS
P. Delannoy, J. Lemonnier, E. Hay, D. Modrowski, P. J. Marie*
INSERM U349, Hopital Lariboisiere, Paris, France
Cell-cell adhesion mediated by cadherins is believed to play an essential role in the control of cell growth and differentiation. However, the mechanisms and signaling molecules that regulate cell-cell adhesion molecules in bone cells are not known. Osteoblasts express several cadherins, including E- and N-cadherins and N-CAM. Our recent studies indicate that N-cadherin is regulated by BMP-2 and FGF-2 in human calvaria cells and is involved in the expression of osteoblast markers and differentiation. We now tested the possibility that N-cadherin expression and function may be regulated by activation of PKC in human osteoblasts. Treatment of immortalized human neonatal calvaria (IHNC) cells with phorbol ester (100 nM) increased PKC activity at 5 and 10 min, as determined by the transfer of 32P-labeled phosphate to a biotinylated peptide substrate. This was followed by down-regulation of PKC activity at 20-120 mn. RT-PCR analysis of mRNA corrected for GAPDH showed that transient treatment of IHNC cells with phorbol ester (10 min) increased N-cadherin mRNA levels at 4-8 h but not at later time-points (24-48 h). Western blot analysis showed that N-cadherin protein levels were increased by phorbol ester (10 min) at 24 h, and this was confirmed by immunocytochemical analysis. In contrast, neither E-cadherin, nor N-CAM mRNA or protein levels were affected by phorbol ester. Moreover, transient treatment of IHNC cells with phorbol ester (10 min) increased cell-cell aggregation by 33%, measured in a cell aggregation assay, indicating that the increased N-cadherin synthesis was functional. In addition, treatment with phorbol ester dose dependently increased alkaline phosphatase (ALP) activity in IHNC cells with a maximal effect at 10 nM. This effect was comparable to the promoting effect of BMP-2 (25-50 ng/ml), a known activator of osteoblast differentiation. The data provide the first evidence that direct activation of PKC by transient treatment with phorbol ester increases N-cadherin expression and function, associated with increased ALP activity in human calvaria osteoblasts. This provides a signaling mechanism by which N-cadherin is selectively regulated in human osteoblasts and suggests a possible role of PKC in the control of N-cadherin and osteoblast differentiation.
ACTIVATION OF THE PROTEIN KINASE B (PKB/AKT) SIGNALLING PATHWAY IN OSTEOBLASTS BY MECHANICAL STRAIN
T. S. Grewal*, R. J. Tolley, T. M. Skerry
Department of Biology, University of York, York, UK
Little is known about the intracellular pathways that regulate the mechanisms behind physiological responses of bone cells to mechanical stimuli, although it is known that loading induces NO release (an essential early event) and proliferation of osteoblasts. In endothelial cells, endothelial Nitric Oxide Synthase (eNOS) is activated by the Protein Kinase B (PKB/Akt) intracellular signalling pathway that is also known to be a key mediator of cell survival. The PKB signalling pathway is initiated by activation of cell membrane-associated phosphatidylinositol-3-kinase (PI-3-K) which in turn phosphorylates adjacent phosphatidylinositol lipids to generate PIP3. This event leads to the recruitment of PIP3-dependent kinases (PDK-1 and PDK-2) and PKB from the cytoplasm to the cell membrane. The assembly of these proteins in the same microenvironment facilitates the phosphorylation of PKB at residues Ser473 and Thr308 and hence its activation. Since our gene array analysis revealed mechanically regulated expression of PKB in osteoblasts, we hypothesised that this pathway could play the same key role in strain related events in bone.
Clonal and primary osteoblasts were subjected to mechanical strain using an FX-3000 Flexercell device and Type I collagen coated wells. Cells were subjected to cyclical strains (peak 0.004) at 1 Hz for 10 minutes and total protein extracted at various times after loading. Lysates were separated and analysed by immunoblotting using a specific phospho-PKB (Ser473) antibody (phosphorylation of Ser473 of PKB occurs during activation of the kinase). The blots were stripped and incubated with a PKBalpha antibody to control for protein loading.
Rapid, transient phosphorylation of PKB at Ser473 was evident within 2 minutes of the end of the straining regimen and peaked at 20-40 minutes. This response to strain was completely blocked by pre-treatment of cells for 1 hr with specific PI-3-K inhibitors (wortmannin, 500 nM and LY294002, 10 microM). These findings suggest that activation of the PKB intracellular signalling pathway following mechanical strain in osteoblasts, is responsible for the regulated NO release and control of cell survival that is stimulated by loading.
C-SRC AND PKC ISOFORMS REGULATE OSTEOBLAST DIFFERENTIATION
S. Migliaccio1*, M. Marzia2, N. Sims3, A. Taranta4, S. Bernardini4, R. Baron3, A. Teti1
1Department of Experimental Medicine, University of L'Aquila, L'Aquila, Italy
2Department of Experimental Medicine, University "La Sapienza", Rome, Italy
3Department of Cell Biology and Orthopaedics, Yale University, New Haven, USA
4Istituto Dermopatico dell'Immacolata, Rome, Italy
We have demonstrated that deletion/reduction of c-Src expression enhances osteoblast differentiation and bone formation. We observed by in vivo bone histomorphometry, that bone formation was increased in c-Src null compared to wild-type mice, and in vitro, alkaline phosphatase activity and nodule mineralization were stimulated in c-Src deficient osteoblasts. c-Src deletion inhibited cell proliferation, without affecting the osteoblast life-span. Up-regulation of a series of osteoblast markers, alkaline phosphatase, Osf2/Cbfa1 transcription factor, PTH/PTHrP receptor, osteocalcin and pro-alpha 2(I) collagen, was observed in c-Src-deficient cells, confirming a negative role of c-Src in osteoblast differentiation. Protein kinase C (PKC) total activity is enhanced in mature osteoblasts, and PKC is reported to phosphorylate c-Src. Therefore, we explored the molecular interactions between c-Src and PKC during osteoblast differentiation. A time-dependent increase in expression and/or activity of the conventional (c)PKCalpha and the novel (n)PKCepsilon, associated with increased inositol phospholipid turnover, and a reduction of the atypical (a)PKCzeta, were indeed apparent with the progression of differentiation. Interestingly, in similar experimental conditions, c-Src expression was transiently increased, reaching a maximum at 24 hrs of culture when cell proliferation was elevated, and progressively declining with differentiation. In addition, c-Src activity, evaluated by enolase phosphorylation tests, was inhibited in mature cells, again consistent with a negative role of this kinase in osteoblast differentiation. c-Src co-immunoprecipitated with PKCalpha with an increased pattern in mature osteoblasts. Kinase assay showed that PKCalpha immunoprecipitated from differentiated cells, induced c-Src phosphorylation in a dose-dependent manner. This phosphorylation is hypothesized to occur on Ser12 of c-Src, a residue conserved in several PKC substrates, and could have a negative role on c-Src activity. In contrast, c-Src was unable to phosphorylate the PKCalpha, indicating that no cross-molecular phosphorylations occurred between the two kinases. However, in mature osteoblasts, PKCalpha tyrosine phosphorylation was increased, suggesting that other tyrosine kinases, but not c-Src, induced this event. Based on these results, we conclude that c-Src has a negative role in osteoblast differentiation and bone formation, and that PKCalpha may regulate c-Src activity contributing to the control of bone growth.
ANALYSIS OF THE MURINE MODEL OF THE SAETHRE-CHOTZEN SYNDROME SUGGESTS A FUNCTIONAL RELATIONSHIP BETWEEN TWIST AND FGFR2
V. El Ghouzzi1*, R. Quillet2, C. Stoetzel2, A. Munnich1, J. Bonaventure1, F. Perrin-Schmitt2
1INSERM U393, Hôpital Necker, Paris, France
2LGME-CNRS/INSERM U184, Faculté de Médecine, Strasbourg, France
Early ossification of the skull vault usually results in premature fusion of the sutures thus leading to skull deformations of variable severity named craniosynostosis. Most of the genetically determined craniosynostosis syndromes are accounted for by gain-of-function mutations in three members of the Fibroblast Growth Factor Receptor family (Fgfr1-3). However, the Saethre-Chotzen syndrome (SCS) is caused by loss-of-function mutations in the developmental Twist gene which encodes a basic-helix-loop-helix (b-HLH) transcription factor. Phenotypic overlaps between Twist- and Fgfrs-associated craniosynostoses suggest a possible involvement of these factors in a common signalling pathway. Since D-twist is required for proper expression of fgfr homologues in Drosophila, we reasoned that Twist could regulate fgfrs expression in vertebrates. To test this hypothesis we used twi-null heterozygous mice that mimic the human SCS phenotype. RT-PCR analysis and whole mount in situ hybridisation of 10.5 dpc twi-null heterozygous embryos showed that haploinsufficiency at M-twist resulted in up-regulation of fgfr2 in the CD1 genetic background.
Although no direct Twist target genes have been yet identified in human, the basic region of the protein is thought to serve as a DNA binding domain through the recognition of CANNTG motives (E-boxes). Analysis of 3.7kb of the human fgfr2 promoter disclosed a single palindromic E-box (CAGCTG). Using EMSAs, we found that the human Twist protein failed to interact with this E-box but specifically binds to the CATATG NdeI E-box. Hence, Twist appears to up-regulate fgfr2 expression although indirectly. Whether Twist modulates fgfr1 and fgfr3 expression is now under investigation.
EFFECT OF TRAUMA ASSESSMENT ON NONVERTEBRAL FRACTURE RISK REDUCTION IN THE RECOMBINANT HUMAN PARATHYROID (1-34) FRACTURE PREVENTION STUDY
E. F. Eriksen1*, S. L. Myers2, G. A. Gaich2, O. Wang2, B. H. Mitlak2
1Aarhus Amtssygehus, Aarhus, Denmark
2Eli Lilly and Company, Indianapolis, IN, USA
We evaluated the effect of assessing trauma on overall nonvertebral fracture risk reduction in a double-blind, randomized, placebo-controlled study of recombinant human parathyroid hormone [rhPTH(1-34)] in 1637 women with postmenopausal osteoporosis. Women with one or more prevalent vertebral fractures were randomly assigned to 1 of 3 equal study arms: placebo, 20 (PTH20) or 40 microgram (PTH40) of rhPTH(1-34) which was self-injected once daily. All patients received supplemental calcium (1000 mg) and vitamin D (400-1200 IU) daily. The median follow-up was 21 months. Nonvertebral fragility fracture was a protocol-specified endpoint. Nonvertebral fractures were documented at the study sites by review of radiographs or reports, and were to be considered fragility fractures in the investigator’s opinion, if the injury would not have caused a fracture in an otherwise healthy person.
One or more incident nonvertebral fractures occurred in 119 women (7.3%). Relative to women receiving placebo, those receiving PTH20 or PTH40 had relative risks for overall nonvertebral fractures of 0.65 (95% CI 0.43 to 0.98) and 0.60 (0.39 to 0.91), respectively. Of the women with new nonvertebral fractures, 58 were judged at the time of reporting to have had fragility fractures. Compared with placebo, the relative risks for fragility fractures were 0.47 (0.25 to 0.88) and 0.46 (0.25 to 0.86), for the PTH20 and PTH40 groups, respectively. The reduction in fracture risk was not different between PTH doses.
By an analysis of time to first event, the rhPTH(1-34) treatment effect was evident earlier for fragility fractures than for overall nonvertebral fractures.
In conclusion, treatment with both PTH20 and PTH40 reduced the risk of overall nonvertebral fractures (35 and 40%, respectively) and nonvertebral fragility fractures (53 and 54%, respectively). The assessment of trauma enhanced the ability to detect the treatment-related effects of rhPTH(1-34) on nonvertebral fracture risk reduction.
*For the PTH Fracture Prevention Study Investigators
VALIDATION OF CHANGE IN BONE MINERAL DENSITY AS A SURROGATE FOR FRACTURE RISK AFTER STATISTICAL ADJUSTMENT FOR IMPRECISION IN BMD MEASUREMENTS
N. Watts1*, S. Sarkar2, W. Wu2, M. Wong2, K. Harper2
1Emory University, Atlanta, GA, USA
2Eli Lilly and Company, Indianapolis, IN, USA
Although low bone mineral density (BMD) is correlated with an increased risk of osteoporotic fractures, the relationship between increased BMD and decreased fracture risk observed with use of antiresorptive agents is poorly understood. Freedman’s statistic quantifies the extent to which changes in the intermediary marker (BMD) can estimate observed changes in the clinical outcome (fracture), but it does not account for imprecision in BMD measurements, such as that due to temporal variation or types of instruments used. To improve on Freedman’s existing method, the Excess Relative Odds (ERO) statistic, with a simulation-based 95% confidence interval (CI), was developed to determine the extent to which treatment effects on fracture can be explained by changes in BMD. The Corrected Excess Relative Odds (CERO) statistic was developed to additionally correct the ERO statistic for imprecision in BMD measurements. The present analyses used the ERO and CERO statistics to evaluate how well 3-year changes in BMD act as surrogates for estimating vertebral fracture (VF) risk. Data from the MORE (Multiple Outcomes Raloxifene Evaluation) trial, in which 7705 postmenopausal women with osteoporosis were randomized to placebo (PL) or raloxifene 60 or 120 mg/day [JAMA 282(1999):637-45], were used. Freedman's statistic found that increases in femoral neck BMD accounted for 5% of the observed VF risk reduction with raloxifene [J Bone Min Res 14 (1999): S158]. Using the ERO statistic, the change in femoral neck BMD at 3 years explained 7.2% (95% CI 2.1%, 16.5%) of the reduction in VF risk with 3 years of raloxifene therapy. After adjustment for imprecision in BMD measurements using the CERO statistic, change in femoral neck BMD accounted for 9.0% (95% CI 4.3%, 22.2%) of the decrease in VF risk with raloxifene therapy. Therefore, even after adjustment for imprecision in BMD measurements, change in femoral neck BMD explains <10% of the fracture risk reduction observed with 3 years of raloxifene treatment. Based on these data, changes in BMD do not statistically qualify as reliable surrogate markers for estimating fracture risk reduction with raloxifene therapy.
PREVENTION OF OSTEOPOROTIC FRACTURES IN DUTCH MEN AND WOMEN: A COST EFFECTIVENESS EVALUATION
C. E. De Laet1*, H. A. P. Pols2
1Institute for Medical Technology Assessment, Rotterdam, the Netherlands
2Dept of Internal Medicine, Erasmus University Medical School, Rotterdam, the Netherlands
Osteoporosis has little impact on overall mortality in the population, but an important impact on health related quality of life (QOL). Therefore cost effectiveness (CE) evaluations have to take this into account. This study analyses costs and effects of interventions to reduce fracture risk in a Dutch setting.
Costs for hip and vertebral were derived from published estimates. QOL estimates were obtained from a Dutch expert panel. Threshold for intervention were either a risk double that of the total population of that age, or a tripled risk. CE evaluated using a model fitted with Dutch data. Cost of intervention was assumed as either 100 euros or 450 euros, the former cost corresponding to interventions such as HRT and Vitamine D, the later corresponding to newer drugs such as bisphosphonates and SERM’s. Risk reduction was assumed to be either 20% or 50% on hip and vertebral fractures only. Intervention during 5 year was evaluated, with an estimated 5-year residual and declining effect. Costs and effects were discounted at 4%. Thresholds for considering an intervention cost-effective were either 20,000 euros per quality adjusted life year (QALY) gained or 30,000 euros per QALY.
In women, assuming the fracture risk of the individual to be double that of the population (an estimated 13% of the population of that age) an intervention with a yearly cost of 100 euros and 20% risk reduction would result in a cost effective intervention starting at the ages of 60 or 65 depending upon the threshold used. In men this would only be at ages 70 or 75 respectively. An intervention costing 450 euros with 50% risk reduction in the same population would become cost effective at ages 65 or 70 in women and to 75 or 80 in men. Limiting the intervention to individuals with a triple risk would greatly enhance the cost-effectiveness and move these thresholds to lower ages, but would only include 5% of the population of that age.
When using a combined strategy of selecting individuals at very high risk at younger ages, and selecting individuals at moderately increased risk at higher ages, cost effective interventions for preventing osteoporotic fractures are possible.
DIGITAL X-RAY RADIOGRAMMETRY OF THE METACARPALS PREDICTS WRIST, HIP AND VERTEBRAL FRACTURE RISK: A PROSPECTIVE ANALYSIS FROM THE STUDY OF OSTEOPOROTIC FRACTURES
M. L. Bouxsein1*, L. Palermo2, C. Yeung2, D. M. Black2
1Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, USA
2Prevention Sciences Group, Dept. of Epidemiology and Biostatistics, UCSF, San Francisco, CA, USA
Digital x-ray radiogrammetry (DXR) is a technique that uses standard hand radiographs to estimate bone mineral density (DXR-BMD). Previous studies have shown that DXR-BMD measurements are strongly correlated to forearm BMD (Black et al, 1999) and have high precision (Jørgensen et al, 2000). To determine whether DXR-BMD measurements predict wrist, hip and vertebral fracture risk we conducted a case-cohort study within a prospective study of 9704 community-dwelling elderly women (the Study of Osteoporotic Fractures). We compared DXR-BMD, and BMD of the radius (proximal and distal), calcaneus, femoral neck, and PA lumbar spine in women who subsequently suffered a wrist (n=200), hip (n=200), or vertebral fracture (n=200) with randomly selected controls from the same cohort (n=392 to 398). DXR-BMD was estimated from hand radiographs acquired at the baseline visit. The radiographs were digitized and the Pronosco X-posure System was used to compute DXR-BMD from the 2nd through 4th metacarpals. All non-vertebral fractures were confirmed by radiographic reports. Vertebral fractures were defined using morphometric analysis of lateral spine radiographs acquired at baseline and an average of 3.7 years later. Age-adjusted odds ratio (vertebral fracture) or relative hazard (wrist and hip fracture) for a 1 SD decrease in BMD are shown in the Table. All BMD measurements were similar for prediction of wrist (OR = 1.5 to 2.0) and vertebral fracture (OR = 1.8 to 2.5). Femoral neck BMD best predicted hip fracture (OR = 3.0), while the odds ratios for all other BMD measurements were similar (OR = 1.5 to 1.9). These prospective data indicate that DXR-BMD performs as well as other peripheral BMD measurements for prediction of wrist, hip, and vertebral fractures.
| Measurement | Rel Hazard for Wrist Fx * |
Rel Hazard for Hip Fx * |
Odds Ratio for Vertebral Fx * |
| DXR-BMD | 1.6 (1.3, 2.0) | 1.8 (1.4, 2.2) | 1.9 (1.5, 2.3) |
| Prox Radius BMD | 1.8 (1.5, 2.1) | 1.7 (1.4, 2.1) | 1.8 (1.5, 2.2) |
| Dist Radius BMD | 2.0 (1.7, 2.5) | 1.8 (1.4, 2.2) | 1.9 (1.6, 2.4) |
| Calcaneus BMD | 1.7 (1.4, 2.0) | 1.9 (1.5, 2.4) | 2.2 (1.8, 2.7) |
| Fem Neck BMD | 1.8 (1.2, 2.3) | 3.0 (2.1, 4.4) | 2.5 (2.0, 3.3) |
| Spine BMD | 1.5 (1.2, 1.9) | 1.5 (1.2, 2.0) | 2.3 (1.9, 2.9) |
| * (95% confidence intervals) | |||
SIT-TO-STAND TEST, A SIMPLE PREDICTOR OF HIP AND OTHER FRACTURES IN OLDER WOMEN
K. Kayan*, E. V. McCloskey, R. U. Ashford, A. Dey, J. A. Kanis
WHO Collaborating Centre for Metabolic Bone Diseases, University of Sheffield, UK
Inability to rise from a chair without using one’s arms has been identified as a potential risk factor for falls and hip fractures. We wished to assess whether a simple sit to stand assessment is associated with hip and other fragility fracture risk in elderly women in England.
Women aged 75 years and above were enrolled to a prospective, single centre, randomised placebo-controlled study of the effect of the bisphosphonate, clodronate, on hip fracture risk. At entry, all women were assessed for their ability to rise from a chair and classified into those who had difficulty and/or were unable to rise from a chair and those who had no difficulty. They were followed at six monthly intervals for a median of 3.1 years and all incident fractures were verified radiologically from hospital or GP records. The present analysis remains blinded to the bisphosphonate therapy.
Of a total of 4346 women studied, 1381 (31.8%) were classified as having difficulty rising from a chair. These women were significantly older (81.2±4.4 vs. 79.4±3.5 years, p<0.001), shorter (155.2±6.4 vs.156.1±6.0 cms, p<0.001), heavier (67.3±13.3 vs. 64.0±11.3 kg, p<0.001) and had a lower mean total hip BMD (0.74±0.15 vs. 0.76±0.13 g/cm2, <0.001) than those rising without difficulty. In women rising with difficulty, the incidence of hip fracture was increased two-fold compared to those rising without difficulty (3.3% vs.1.6%, p=<0.001; OR 2.18, 95% CI 1.45-3.30, p=0.0002). The risk of any non-hip fracture was also increased in women rising with difficulty (8.0 vs.6.1%, p=0.015; OR 1.36, 95% CI 1.06-1.74, p=0.015). The increased risk persisted after adjustment for age, weight and hip BMD (OR 1.55, 95% CI 0.98-1.23, p=0.058 and OR 1.29, 95% CI 0.99-1.68, p=0.056 for hip and non-hip fractures respectively).
We conclude that a simple test of the ability to rise from a chair to a standing position, as an assessment of neuromuscular function, is a good predictor of hip and other fragility fractures in elderly women.
POSITIVE ASSOCIATION BETWEEN NET ENDOGENOUS NONCARBONIC ACID PRODUCTION (NEAP) AND BONE HEALTH: FURTHER SUPPORT FOR THE IMPORTANCE OF THE SKELETON TO ACID-BASE BALANCE
S. A. New1*, H. M. Macdonald2, D. A. Grubb3, D. M. Reid2
1University of Surrey, England, UK
2University of Aberdeen, Scotland, UK
3Rowett Research Institute, Scotland, UK
The positive influence of alkaline-forming foods (e.g. fruit & vegetables) on bone health has been gaining increasing prominence in the literature from a combination of experimental (at the human, animal & cellular level), clinical and observational studies. Determination of the net endogenous noncarbonic acid production (NEAP) using simple algorithms based on dietary protein and potassium ratios is a useful method for exploring further the role of the bone "reservoir" in acid-base balance (1).
Using a population-based cohort of women, we examined, for the first time, the association between NEAP and indices of bone health. Baseline results of the Aberdeen Perimenopausal Osteoporosis Screening Study (APOSS) (2,3) were used and included measurements of lumbar spine (LS) & femoral neck (FN) bone mineral density (BMD) using DXA in 1056 pre/peri-menopausal women aged 45-54 years and measurement of forearm bone mass using pQCT and the urinary markers of bone resorption, pyridinoline (Pyd) & de-oxypyridinoline (Dpd) in 62 women. Dietary intakes of energy, protein & potassium were assessed using our ‘validated’ food frequency questionnaire (FFQ).
As shown in the table below, women in the highest quartile of NEAP intake were found to have lower LS BMD (P<0.04), forearm total (pQCTTOT) bone mass (P<0.03) and higher bone resorption (P<0.02), differences which remained significant after adjustment for important confounding factors (including age, weight, height, total energy intake, smoking and socio-economic status).
Diets of higher ‘acidity’ were associated with poorer indices of bone health. These novel findings provide: (i) further evidence of a positive link between alkaline-forming foods (e.g. fruit & vegetables) and skeletal health; (ii) further support for the hypothesis that the skeleton plays an important role in acid-base maintenance.
1. Frassetto et al.,(1998). Am J Clin Nutr 68, 576-583.
2. New et al., (1997). Am J Clin Nutr 65, 1831-1837.
3. New et al., (2000). Am J Clin Nutr 71, 142-151.
| LSBMD (g/cm) | pQCTTOT (g/cm) |
Pyd/Cr (nmol/mmol) |
Dpd/Cr (nmol/mmol) |
|
| Lowest Quartile NEAP | 1.074a [0.166] | 414a [44.1] | 42.1a [10.4] | 9.3 a [3.9] |
| Highest Quartile NEAP | 1.045b [0.164] | 380 b [40.4] | 59.2 b [12.5] | 15.5 b [4.8] |
| Values with unlike superscripts (a,b) are significantly difference (P<0.05) | ||||
LOW DOSE SUPPLEMENTATION WITH CHOLECALCIFEROL AND CALCIUM DURING WINTER TIME PREVENTS SEASON-RELATED BONE LOSS
H. W. Woitge1*, K. Witte2, B. Doschko1, B. Auler1, A. Knothe1, R. Ziegler1, B. Lemmer2, M. J. Seibel1
1Department of Medicine I, University of Heidelberg, Heidelberg, Germany
2Department of Pharmacology, University of Heidelberg, Heidelberg, Germany
In countries of the northern hemisphere, 25hydroxy-vitamin D3 (25OHD) varies by season and leads to changes in bone turnover. To test the hypothesis that increased bone turnover in winter is associated with a decrease in bone mineral density (BMD) and that winter time supplementation with cholecalciferol will prevent season-related bone loss, we designed a randomized, non-blinded prospective trial following 60 healthy participants (20 males and 40 postmenopausal females aged 40 – 75 years) over a period of two years. Follow-up included the completion of a questionnaire, a physical examination and the collection of serum and urine specimen every 4 weeks at the same time of the day, and lumbar (LBMD) and total hip (HBMD) BMD measurements every 6 months. Subjects in group 2 (10 males, 20 females) were treated with 500 I.E. cholecalciferol and 500 mg calcium daily from Nov-March during year 2 of the study. Biochemical measurements included serum 25OHD, parathyroid hormone (PTH), total (TAP) and bone-specific (BAP) alkaline phosphatase, and urinary pyridinoline (PYD) and deoxypyridinoline (DPD). Statistics were performed using the Pharmfit method. In year 1, significant circannual rhythms were found for 25OHD (acrophase in August) and for PTH, BAP, PYD and DPD (acrophases in winter). During the winter of year 1 (October-April), subjects in both groups lost approximately 1% in BMD at all sites (p<0.05), while a <1% increase was seen during the following summer (April-October). A similar pattern was seen in group 1 (no treatment) during the second year. In group 2, treatment with Vitamin D and calcium during the winter period of year 2 abolished the seasonal changes of all biochemical markers, and LBMD and HBMD increased (LBMD: p<0.01 vs. year 1; HBMD: p<0.05 vs. year 1).
We conclude that in northern latitudes, circannual changes in bone turnover are associated with bone loss during winter time. Supplementation with vitamin D and calcium during winter time prevents accelerated bone loss and turnover.
TWO YEAR RESULTS OF ONCE-WEEKLY ADMINISTRATION OF ALENDRONATE 70 MG FOR THE TREATMENT OF POSTMENOPAUSAL OSTEOPOROSIS
R. Rizzoli1*, C. Roux2, S. Greenspan3, H. Bone4, T. Schnitzer5, N. Watts6, S. Adami7, J. Foldes8, B. Uebelhart1, C. Peverly9, A. Santora9, A. Kaur9, M. Wu9, J. Orloff9
1Hopital Cantonal, Geneva, Switzerland
2Hopital Cochin, Paris, France
3University of Pittsburgh, Pittsburgh, PA, USA
4Michigan Bone Clinic, Detroit, MI, USA
5Northwestern U., Chicago, IL, USA
6Emory University, Atlanta, GA, USA
7University of Verona, Italy
8Hadassah University Hospital, Jerusalem, Israel
9Merck Research Labs, Rahway, NJ, USA
The therapeutic equivalence of alendronate (ALN) 70 mg once weekly (OW) (provided by Merck and Co., Inc., Whitehouse Station, NJ, USA), ALN 35 mg twice weekly (TW), and ALN 10 mg once daily (OD) in the treatment of postmenopausal osteoporosis for one year has been reported previously (Schnitzer et al, Aging 12:1-12, 2000). We will present preliminary two-year BMD results from a one-year extension. We compared the efficacy of treatment with OW ALN 70 mg (n=519), TW ALN 35 mg (n=369), and OD ALN 10 mg (n=370) over 2 years in a double-blind, multicenter study of postmenopausal women (age 40 to 90) with osteoporosis (BMD of either the lumbar spine or femoral neck greater than or equal 2.5 SDs below peak mean, or prior vertebral or hip fracture). The primary efficacy endpoint was change in lumbar spine BMD. Secondary endpoints included changes in BMD at the hip and total body.
Mean BMD increases at the lumbar spine, total hip, femoral neck, hip trochanter, and total body sites for each treatment regimen after two years were similar, with no clinically meaningful differences between groups.
These findings confirm that once-weekly ALN 70 mg is therapeutically equivalent to once-daily ALN 10 mg in patients with postmenopausal osteoporosis. In addition, the two-year results demonstrate that once-weekly ALN 70 mg is generally as safe and well tolerated as ALN 10 mg once daily.
QUANTITATIVE 3D SYNCHROTRON RADIATION MICRO-TOMOGRAPHY EVALUATION OF OSTEON MINERALIZATION DEGREE DURING ETIDRONATE TREATMENT IN OSTEOPOROTIC PATIENTS
S. Nuzzo1, F. Peyrin1,2*, E. Martín-Badosa1, G. Boivin4, C. Alexandre3, M. H. Lafage-Proust3
1ESRF, Grenoble, France
2CREATIS, UMR CNRS 5515, Villeurbanne, France
3INSERM EM9901, J Monnet University, St-Etienne, France
4INSERM U403, Lyon, France
Bone quality is a factor controlling bone mechanical resistance to fracture. Boivin et al. (Bone, 2000) showed that bisphosphonate induce alteration in bone secondary mineralization, (assessed by micro-radiography on one thin slice of biopsy per patient), that might partly explain the reduction in fracture incidence.
The Synchrotron radiation (SR) microCT system is well suited to image bone architecture with a high spatial resolution (<10microm). The reconstructed images are accurate maps of the 3D-distribution of linear absorption coefficient. Since absorption depends of the bone mineral content, we proposed a calibration method to estimate the 3D-distribution of hydroxyapatite concentrations within the whole sample. Thus, homogeneous water solutions with different K2HPO4 concentrations were used. The theoretical linear relationship between concentrations in the range (0.1-0.8g/cm3 and absorption (1-9cm-1 was verified from experimental measures. The accuracy of this calibration was tested by comparing results from 2D-SR radiography images to those obtained from a reference microradiograph using an aluminum calibration step-wedge
A series of 32 iliac crest biopsy 3D-images from osteoporotic patients before, and after one (n=12) and two years (n=4) of Etidonate was analyzed. 900 projection images were acquired per sample. The energy was set at 20KeV. The transmitted X-ray images were recorded using a bidimensional CCD based camera (1024x1024 pixels). These acquisitions were processed to get the 3D-reconstruction volumes of about 600*500*900 using a 3D filtered backprojection algorithm. Trabecular and cortical ROIs were accurately selected for the analysis of 3D-images of biopsies. Slices of the 3D-reconstructed images showed clearly differences in gray levels, that reflect osteons with different degree of mineralization within bone trabecular or cortical regions Histograms of hydroxyapatite concentration in the trabecular and cortical regions were performed. The bone absorption observed both in cortical and trabecular region is in the range 4-10cm-1. The mineral concentration exhibited an increase throughout therapy of about 3% and 7%, after one year and two years of treatment, respectively, while no significant changes were observed regarding 3D architectural parameters (bone volume, trabecular thickness and number).
Thus, beside microarchitecture evaluation, SRmicroCT, may provide valuable information on the mineralization degree of bone.
THE RECHALLENGE OF PATIENTS WHO PREVIOUSLY DISCONTINUED ALENDRONATE DUE TO UPPER GASTROINTESTINAL SYMPTOMS
P. D. Miller1*, A. A. Licata2, G. Woodson3, M. P. Ettinger4, B. Mako5, M. E. Smith5, L. Wang5, J. Yates5, M. E. Melton5, J.Palmisano5
1Colorado Center for Bone Research, Lakewood, CO, USA
2Cleveland Clinic Foundation, Cleveland, OH, USA
3Atlanta Research Center, Decatur, GA, USA
4Clinical Research Center of South Florida, Stuart, FL USA
5Merck & Co., Inc., West Point, PA, USA
This study examined the upper gastrointestinal (UGI) tolerability of alendronate (ALN) and placebo (PBO) in postmenopausal women with osteoporosis who had previously discontinued ALN due to an UGI adverse Experience (AE) and were subsequently rechallenged with either ALN or PBO.
This study was a double-blind, randomized, parallel-group, placebo-controlled administration of either ALN 10 mg daily or matching alendronate PBO daily (1:1 ratio) for 8 weeks. Thirty-nine sites enrolled 172 postmenopausal women with osteoporosis who had previously discontinued ALN therapy due to an UGI AE. An UGI AE was defined as any upper gastrointestinal event with the exception of severe esophagitis or esophageal ulcer due to previous alendronate therapy. Clinical evaluations (physical exam, vital signs), and AE reporting were performed.
Greater than 80% of the prior history of UGI AE due to ALN therapy were due to acid regurgitaion, dyspepsia, abominal pain, nausea, and gastoesopheageal reflux. These were also the primary causes of discontinuation when rechallenged. No prior UGI AE due to alendronate therapy was considered serious. Upon rechallenge,
the cumulative incidence of patients who discontinued due to UGI AEs (primary endpoint) was 14.5% (95% CI:6.9%, 22.1%) in the ALN group (n=88) and 17.3% (95%CI:9.1, 25.6%) in the PBO group (n=84): difference (ALN-PBO)= -2.8% (95%CI:-14%, 8.1%; p=0.66). No patient discontinued in either group due to a serious UGI AE. There was one serious UGI AE in the PBO group (acute gastritis) that did not result in discontinuation of study medication and was not considered drug related.
In this study, when postmenopausal women with osteoporosis who previously discontinued alendronate therapy due to an UGI AE are rechallenged with ALN or PBO, the incidence of discontinuation due to an UGI AE is similar. In this patient population, we conclude that a significant majority of postmenopausal women with osteoporosis who discontinue ALN due to an UGI AE will tolerate rechallenge with ALN, and that the percentage of patients that have recurrent UGI AE upon rechallenge is similar in PBO rechallenge patients.
MECHANISM-BASED LOCAL IRRITATION BY NITROGEN-BISPHOSPHONATES
A. A. Reszka*, J. E. Fisher, G. A. Rodan
Merck Research Laboratories, West Point, PA, USA
The nitrogen-bisphosphonates (N-BPs) act directly on the osteoclast by inhibiting the intracellular cholesterol biosynthetic enzyme, farnesyl diphosphate (FPP) synthase. Though bisphosphonates also have the potential to irritate both the esophagus and skin, the mechanism for irritation has not been elucidated. Our recent in vitro studies of keratinocytes established mechanism-based N-BP suppression of cell growth, associated with suppression of cdk phosphorylation of the growth-regulatory protein, retinoblastoma (Reszka et al. Mol. Pharmacol., In Press, February, 2001). In this study we examined two new models for local irritation.
In vitro we used Ch1.Es esophageal fibroblasts to model the deeper tissue of the esophagus. Twenty four hour treatment with the N-BPs alendronate or risedronate, at 300 microM, a concentration that might be found in the esophagus after reflux, induced apoptosis. This effect was associated with caspase cleavage of the pro-apoptotic kinase Mst1 and activation of p38 and Jnk kinases. A loss of Erk1 and Erk2 kinase activities was also observed. Consistent with mechanism-based action, all apoptotic and growth effects were mimicked by lovastatin or a selective inhibitor of protein geranylgeranylation. Furthermore, co-incubation with geranylgeraniol completely blocked apoptosis induced by the bisphosphonates at 300 microM or by lovastatin. Moreover, by modifying culture conditions or through pharmacological intervention, we enhanced intracellular levels of both FPP and geranylgeranyl diphosphate (GGPP), and increased the expression of the enzyme HMG-CoA reductase. These changes resulted in a complete suppression of apoptosis induced by any of these agents.
In vivo we examined, in the rat, localized skin irritation at the site of subcutaneous injection of alendronate, risedronate or simvastatin. All treatments caused quantifiable (volumetric) skin thickening characterized by disruption of cell-cell contact between and infiltration into the stratum basale. Pharmacological intervention (as noted above) to enable intracellular accumulation of FPP and GGPP significantly suppressed skin irritation in volumetric quantitative analyses (P<0.01). This suggests that local irritation produced by N-BPs occurs through suppression of FPP synthase and provides in vivo evidence for this mechanism-based action.
BONE MINERAL DENSITY AND THE RISK OF BREAST CANCER: THE ROTTERDAM STUDY
M. van der Klift1,2,3*, H. A. P. Pols2,3, A. Hofman3, C. E. D. H. de Laet1
1Institute for Medical Technology Assessment, Erasmus University Rotterdam, The Netherlands
2Department of Internal Medicine, Erasmus University Rotterdam, the Netherlands
3Department of Epidemiology and Biostatistics, Erasmus University Rotterdam, The Netherlands
Bone mineral density is thought to be a marker for lifetime oestrogen exposure, with trabecular bone being more susceptible to oestrogen than cortical bone. Also, the incidence of breast cancer is associated with serum oestrogen levels. Some studies have suggested that a higher bone mass is associated with an increase in breast cancer risk. In this study we investigated the association between bone mineral density measured at several sites and the risk of breast cancer in women aged 55 or over from the Rotterdam Study, a large population-based cohort study in the Netherlands. Information on baseline bone mineral density and cancer incidence was available for 3228 women. The National Cancer Registry provided information on follow-up of incident cancer. After an average follow-up time of 5.6 years, 67 new breast cancer cases occurred. The incidence in our study was similar to that observed in the general Dutch population. On average, breast cancer cases were three years younger than non-cases. Lumbar spine and femoral neck BMD were divided into tertiles. We computed risk estimates for breast cancer in tertiles of BMD by Cox's proportional hazards model, using the middle tertile as a reference. For lumbar spine BMD, breast cancer risk in the upper tertile was twice as high as the reference after adjustment for age, weight and age at menopause (RR = 2.2 [1.2-4.0]), whereas the lower tertile did not significantly differ from the reference (RR = 1.2 [0.6-2.4]). A low femoral neck BMD, was associated with a decreased risk of breast cancer, although not statistically significant (RR = 0.6 [0.3-1.2]), whereas here the upper tertile did not differ from the reference (RR = 0.9 [0.5-1.5]). The results of this study suggest that in elderly women an association between lumbar spine BMD and incident breast cancer exists. This might be due to the anabolic effect of oestrogen on trabecular bone.
MODEST ALCOHOL INTAKE REDUCES BONE LOSS IN PERI AND EARLY POSTMENOPAUSAL SCOTTISH WOMEN: AN EFFECT DEPENDENT ON ESTROGEN RECEPTOR GENOTYPE?
H. M. Macdonald1,3*, S. A. New2, F. E. McGuigan3, M. H. N. Golden3, S. H. Ralston3, D. A. Grubb4, D. M. Reid1,3
1Osteoporosis Research Unit, Aberdeen, UK
2University of Surrey, Guildford, UK
3University of Aberdeen, Aberdeen, UK
4Rowett Research Institute, Aberdeen, UK
Whereas excessive alcohol has been shown to be risk factor for osteoporosis there is increasing evidence to show that modest alcohol intake has beneficial effects for bone health. We have examined the interaction between alcohol intake and the PvuII polymorphism of the estrogen receptor gene on bone mineral density (BMD) change over 5-7 years in 729 Scottish women as they move through the menopause (mean age 47.5y at baseline and 54 y at follow-up). BMD of the lumbar spine (LS) and femoral neck (FN) was measured by dual energy X-ray absorptiometry (DXA) and alcohol intake assessed from our ‘validated’ food frequency questionnaire (FFQ). The mean results from baseline and follow-up FFQs were used. The genotype distribution was "PP" 198 (27.2%), "Pp" 336 (46.1%) and "pp" 195 (26.7%). There was no significant difference in baseline BMD, follow up BMD or change in BMD with genotypes. We found that modest alcohol intake (energy adjusted) was associated with reduced bone loss at both sites. This effect was confined to women carrying the "p" allele (n 531 Spearmans r=0.163, p<0.001 for LS and r=0.126, p= 0.004 for FN). There was no association for "PP" homozygotes between alcohol intake and change in BMD (n 198, LS: r=0.065, p=0.361 FN: r=0.015, p=0.838). For women who carry the "p" allele, LS BMD change was greater for the lowest quarter of alcohol intake (q1: <1 unit per week), compared to the highest quarter (q4: 1-2 units/day), for both HRT users and non-HRT users (Table 1). These levels are well within government guidelines but it is possible that there may have been some under-reporting. Alcohol may be involved in stimulating adrenal estrone production and this may be of particular benefit to women carrying the p allele.
| Table 1 Annual percentage bone loss at the lumbar spine for Pp/pp women | |||||||
| Alcohol Quartile 1 | Alcohol Quartile 4 | P | |||||
| N | Mean | SEM | N | Mean | SEM | ||
| All women | 132 | -1.21 | 0.09 | 137 | -0.71 | 0.08 | <0.001 |
| No HRT | 57 | -1.79 | 0.14 | 47 | -1.31 | 0.13 | 0.012 |
| Peri and post no HRT | 65 | -1.63 | 0.14 | 64 | -1.01 | 0.13 | 0.001 |
| HRT users | 42 | -0.63 | 0.12 | 44 | -0.25 | 0.12 | 0.028 |
THE EPIDEMIOLOGY OF FRACTURES IN ENGLAND AND WALES
T. P. van Staa1,2,3, E. M. Dennison2*, H. G. M. Leufkens1, C. Cooper2
1Department of Pharmacoepidemiology and Pharmacotherapy, University of Utrecht, Sorbonnelaan 16, Utrecht, the Netherlands
2Medical Research Council Environmental Epidemiology Unit, Southampton University Hospital, Southampton, UK
3Procter & Gamble Pharmaceuticals, Lovett House, Lovett Road, Staines, UK
The objective of this study was to describe the age- and sex-specific fracture incidence rates in the whole of England and Wales.
Information was obtained from the General Practice Research database in the UK, which contains medical records of general practitioners. The study population consisted of all patients aged 20 years or older with a fracture during 1988 to 1998.
There were 222 369 patients who sustained a fracture over 21.6 million person-years of follow-up. In men, the most common fracture was that of the carpal bones. In females, the most common fracture was that of the radius/ulna. Fractures were slightly more common in women than men (1.1 per 100 person-years in women compared to a rate of 1.0 in men). The lifetime risk of any fracture was 53.2% for a 50-year old woman, the corresponding figure in men was 20.7%. For women, this risk was 16.6% at the radius, 11.4% at the femur/hip and 3.1% at the spine; corresponding figures for men were 2.9%, 3.1% and 1.2%. It is apparent that this risk remains very high even in the ninth decade (lifetime risk 28.6% in women, 9.6% in men), and indeed 10-year risk increases with advancing age. It was found that fractures of the femur/hip and vertebrae were associated with excess mortality up to 5 years in both sexes, but that fractures of the radius/ulna were associated with slight excess mortality in men only.
We have shown the marked increase in many fractures with age, and related these clinical events to increased mortality.
CORTICAL BONE STRENGTH AND INTRACAPSULAR HIP FRACTURE
J. Power1*, N. Loveridge1, K. L. Bell1, N. Crabtree1, N. Rushton1, M. Parker2, H. Kroger3, J. Reeve1
1University of Cambridge, UK
2Peterborough District Hospital, Peterborough, UK
3Kuopio University Hospital, Kuopio, Finland
Preliminary studies highlighted the importance of regional cortical bone width and porosity in the aetiology of intracapsular hip fractures: cortical bone loss was most marked in the infero-anterior region which bears the greatest load during a sideways fall. We have now analysed cortical and cancellous bone mass (Densiscan pQCT) and regional cortical bone width and porosity (histomorphometry) in whole femoral neck biopsies distal to the fracture site from 46 female cases of hip fracture from 3 centres (Cambridge UK, Cam-F: n=21, mean age 77y range 65-91y; Peterborough UK, Pet-F: n=11, 81.8y, 70-95y and Kuopio, Finland, n=14, Kuo-F: 79.5y, 63-88y) and 26 age and sex-matched controls taken at post-mortem in Cambridge (80.3y, 62-100y).
After adjusting for location along the femoral neck (using the ratio of maximum and minimum external diameters) pQCT analysis of the whole biopsy indicated that cortical bone area (A), density (D) and mass (M) was significantly lower (compared to controls, p<0.05) in all 3 centres (Cam-F: A-16%, D-12%, M-26%; Pet-F: A-17%, D-18%, M-31%; Kuo-F: A-14%, D-8%, M-19%) but were not different between centres (p>0.05). Cancellous bone area, density and mass were no different between cases or controls or between centres (p>0.05). Histomorphometric analysis showed that the cortices were thinner in the Pet-F group compared to the Finnish and Cambridge groups while porosity was lowest in the Finnish group. Although the cases had thinner cortices in all regions this was most marked in the infero-anterior (-22%, p=0.002), inferior (-20%, p=0.002) and infero-posterior (-23%, p=0.003) regions. Cortical porosity was elevated in the anterior (+32%, p= 0.04) and infero-anterior (+22%, p=0.02) regions so reducing bone elasticity in these regions. Predicted bone strength (elasticity*cortical width) was significantly reduced in the anterior (-43%), infero-anterior (-40%), inferior (-26%) and infero-posterior (-31%) regions.
In conclusion, these results show that a regional loss of cortical bone is a common feature in the aetiology of hip fracture within different population groups. They also add substantial weight to the hypothesis that the loss of cortical bone strength in regions subjected to the greatest load during a sideways fall, rather than changes in cancellous bone, is a major risk factor for femoral neck fractures.
IS DISTAL FOREARM FRACTURE IN MEN DUE TO OSTEOPOROSIS?
S. P. Tuck1*, G. D. Summers2, J. S. Harrop2, J. N. V. Miles3
1South Cleveland Hospital, Middlesbrough, UK
2Derby City General Hospital, UK
3Institute of Behavioural Sciences, Derby University, UK
In women distal forearm fracture is regarded as a typical manifestation of postmenopausal osteoporosis. Traditionally this has not been the case for men. This is because the incidenceof these fractures does not increase with age as it does in women. However, Cuddihy et al have shown that men have a 2.7 and a 10.7 fold increase in hip and vertebral fracture respectively following a distal forearm fracture. This suggests that such men might in fact have low bone mineral density (BMD) and be at risk of osteoporosis. The aim of this study is to determine whether or not this is indeed the case.
All men aged 40 to 80 years who suffered a distal forearm fracture between 1996 and 1998 were identified from the Accident and Emergency records of Derbyshire Royal Infirmary. They were then sent questionnaires to determine their demographic characteristics, risk factors and degree of trauma involved. The subjects were invited to attend for dual energy x-ray absorptiometry (DXA) scans of the femoral neck, total femur and anteroposterior spine. These were compared with the DXA scans of 198 age-matched controls selected from a pre-existing local database of healthy men without fractures. Of the 202 men identified with distal forearm fractures, 55 were excluded because they had subsequently died or lived outside the area, 102 (70%) responded and 67 agreed to undergo DXA scans.
There was no significant difference between the fracture or control group in terms of age or body mass index. However, the fracture group had a significantly lower BMD compared to controls at all sites (p<0.0005). The proportion with osteoporosis (T-score<-2.5) was significantly higher in the fracture group (41.8%) compared with the controls (10.3%) p<0.0005.
This study has shown for the first time that distal forearm fractures in men are associated with significantly lower BMD than the normal population and an increased incidence of osteoporosis. Our results indicate a 7% decrease in lumbar spine BMD in men with forearm fractures, which is of similar magnitude to that in women with Colles' fracture. These findings suggest that men suffering distal forearm fractures should undergo DXA scans and be screened for risk factors relating to osteoporosis and falls.
PREDICTION OF DISTAL RADIUS FAILURE LOAD WITH MICRO-FE MODELS BASED ON 3D-PQCT SCANS
W. Pistoia1*, B. van Rietbergen2, F. Eckstein3, C. Lill4, E. M. Lochmüller3, P. Ruegsegger1
1Inst. Biomedical Eng., Univ. and ETH Zürich, Switzerland
2Dept. of Biomedical Eng., Eindhoven University of Technology, The Netherlands
3Inst. of Anatomy, LMU, München, Germany
4AO Research Inst., Davos, Switzerland
INTRODUCTION The purpose of the present study is to investigate if a tissue failure criterion based on the effective strain distribution, in combination with micro-FE models generated from 3D-pQCT images, can predict the failure load of the human radius. For this purpose, we created micro-FE models for a set of human cadaver arms and compared predicted failure loads with experimentally measured values.
METHODS 70 human cadaver forearms were measured with a special purpose 3D-pQCT system. The distal 4 cm of each forearm was covered with a stack of 240 CT images and transformed to a 3D voxel grid with cubic voxels of 165 micrometers side length. To obtain realistic load transfer, a layer of elements representing cartilage was modelled at the distal end. Micro-FE models of the radii were created by converting bone and cartilage voxels to bricks elements. Material properties were chosen isotropic and linear-elastic. The model was loaded by a distributed force Fext=1000 N applied to the cartilage layer while the proximal end was fully constrained. We assumed different models of failure criteria: these models assume that bone failure is initiated if a significant part of the bone tissue has yielded. Since these are linear FE-analyses, a scaling factor f for the tissue level strains in each model could be determined such that the criterion was just met. The predicted bone failure load is then calculated as f x Fext. The forearms were subjected to a compression test to determine their failure load.
RESULTS For all arms, the compression test produced a Colles’ type fracture in the distal 4-cm of the radius. The best failure criteria are chosen for the prediction of the failure loads calculated from the FE. The failure loads predicted from the FE-analyses correlated well with those measured in the experiments. The coefficient of determination R2 was 0.75 (p<0.001).
CONCLUSION Micro-FE models based on 3D-pQCT images and the chosen failure criterion are adequate for a quantitative prediction of bone failure load. A major advantage of the mechanistic approach introduced here over stochastic models for bone failure prediction, such as those based on DXA data, is the fact that micro-FE models directly account for the mechanical aspects of the full, patient-specific, bone architecture, making it potentially a more precise and versatile approach.
FAMILY BASED STUDIES OF COLIA1 SP1 POLYMORPHISM AND BMD USING THE QUANTITATIVE TRANSMISSION DISEQUILIBRIUM TEST
F. E. A. McGuigan1*, R. Smith1, D. M. Reid1, M. C. deVernejoul2, S. H. Ralston1
1Aberdeen University, Scotland
2INSERM, Paris, France
Previous studies in several populations have reported allelic association between a polymorphic Sp1 binding site in the COLIA1 gene, low bone mineral density (BMD) and osteoporotic fracture. A potential problem with population based association studies is a false positive result due to population admixture. In this study we looked for evidence of an association between COLIA1 alleles and BMD using a pedigree based test of transmission disequilibrium of quantitative traits (qTDT; Abecasis et al, Eur J Hum Genet 2000). We studied a series of 624 individuals from 82 families (52 from the UK and 30 from France) who were recruited on the basis that at least one family member had osteoporosis on the diagnosis of vertebral fractures or low BMD (Z-score <-2.0).
Genotype data was available for 329 individuals and the genotype frequencies were; Aberdeen: "SS" 69 (48.9%), "Ss" 61 (43.3%) and "ss" 11 (7.8%). Paris "SS" 104 (59.7%), "Ss" 66 (37.9%) and "ss" 4 (2.3%). These frequencies are in accordance with the Hardy-Weinberg distribution. Analysis of the data by qTDT showed significant evidence of association between Sp1 alleles and bone density at the femoral neck (p=0.047). Segregation analysis showed evidence of a polygenic influence on BMD at the spine (p<0.001) and the hip (p=0.016) and variance components analysis suggested that the genetic component accounted for 71.6% of the variance in spine BMD and 29.4% of the variance in hip BMD. There was strong evidence of an association between COLIA1 Sp1 genotype and BMD at the spine (p<0.0.031) such that BMD Z-score values were about 0.6 Z score units lower in those who carried the "s" allele when compared with "SS" homozygotes (p=0.031). We were unable to determine if there was an association with fracture due to the small numbers of fractures (n=14) in offspring of heterozygous parents. In conclusion, this study supports the hypothesis that the association between COLIA1 alleles and osteoporosis is real, rather than the result of confounding factors such as population admixture.
THE ROLE OF OSTEOCLAST FORMATION IN THE PATHOGENESIS OF OSTEOPOROSIS IN RHEUMATOID ARTHRITIS
T. Hirayama, L. Danks*, A. Sabokbar, N. A. Athanasou
University of Oxford, Oxford, UK
Rheumatoid arthritis (RA) is a systemic inflammatory disorder in which there is destruction of articular cartilage and bone. RA may be complicated by osteoporosis, but the cause of this generalised bone loss is not known. Bone resorption is carried out by osteoclasts which are formed from mononuclear precursors that circulate in the monocyte fraction. We determined whether there was an increase in the number of circulating osteoclast precursors in RA patients relative to age/sex matched osteoarthritis (OA) controls. We also examined the sensitivity of these precursors to known factors required for osteoclastogenesis. Peripheral blood mononuclear cells (PBMCs) from 10 RA patients and 10 age/sex matched OA controls were cultured on dentine slices and glass coverslips for up to 21 days in the presence of various concentrations of soluble RANKL (sRANKL) and M-CSF, or in the presence of RANKL-expressing UMR106 cells with various concentrations of M-CSF and 1,25(OH)2 vitamin D3. Cells were cultured with and without dexamethasone (Dex) at 10-8M. As assessed by expression of the osteoclast markers, TRAP and lacunar resorption, PBMC serial dilution studies showed no difference in the number of circulating osteoclast precursors in RA patients and OA controls. However, osteoclasts formed from PBMCs of RA patients were capable of substantially more resorption than osteoclasts formed in OA patients (p=0.008). PBMCs from RA patients exhibited increased sensitivity to the osteoclastogenic effect of M-CSF when cultured with sRANKL (p=0.004) or with UMR106 cells (p=0.05). When cultured with UMR106 cells, an increase in sensitivity to 1,25(OH)2D3 (p=0.02) was noted in RA patients compared to OA controls. PBMCs from RA patients also showed increased sensitivity to the effect of sRANKL. PBMCs from both RA (p=0.001) and OA (p=0.001) patients cultured in the presence of Dex exhibited significantly more lacunar resorption than cultures to which no Dex was added. These findings suggest that the increase in generalised bone loss seen in RA results not from an increase in the absolute number of circulating osteoclast precursors but to an increase in the sensitivity of these precursors to local/systemic factors (i.e. M-CSF, RANKL, 1,25(OH)2D3 and corticosteroids) that influence osteoclast formation and bone resorbing activity.
SERUM OSTEOPROTEGERIN IS A MAJOR DETERMINANT OF BONE DENSITY AND VERTEBRAL FRACTURE STATUS IN PATIENTS AFTER CARDIAC TRANSPLANTATION
H. Dobnig1*, A. Fahrleitner1, G. Prenner2, K. H. Tscheliessnig2, C. Piswanger-Sölkner1, G. Leb1
1Department of Internal Medicine
2Department of Surgery, Karl-Franzens University, Graz, Austria
Background: The pathophysiology of transplantation osteoporosis is complex and little understood especially since fractures frequently occur in patients with normal bone density. Osteoprotegerin (OPG) is the key "downstream" regulator inducing osteoclastogenesis and stimulating osteoclast activity. The purpose of this study was to evaluate whether OPG levels in cardiac transplant patients would help explain the magnitude of bone loss and find the patients at risk for vertebral fractures.
Methods: Fifty-seven cardiac transplant recipients (age 57±8, SD) were enrolled in this cross-sectional study. They were 1 to 10 years (median 3.2) post transplantation and were on no bone specific treatment. Standardized thoracic and lumbar spine X-rays as well as hip bone density measurements were performed in all patients. Serum OPG was determined using a commercially available ELISA.
Findings: Vertebral fractures were present in 56% and WHO defined densitometric osteoporosis in 39% of all patients. There was a highly significant increase in mean serum creatinine (r=0.67) and intact parathyroid hormone levels (r=0.41) with increasing time interval to transplantation. Mean femoral z-score decreased by 1 SD between year 2 and year 7 (p=0.01). There was also a marked decrease in serum 25OH Vitamin D3 (-0.42, p=0.001) and testosterone (-0.42, p=0.001) over time. All femoral neck subregions were highly correlated to serum OPG concentration (r-values between 0.52 and 0.68, all p<0.0005). OPG levels were low in almost all individuals with osteoporosis (sensitivity 97%, specificity 47%) and together with the time interval since transplantation accounted for 76.5% of the variance in femoral neck z-scores. OPG did not change over time, was not related to kidney function or associated with current cortisone dosage.
Interpretation: Even late in the posttransplantation period bone status further deteriorates in untreated patients. Although many cardiac recipients develop frank renal hyperparathyroidism, vitamin D deficiency and hypogonadism serum OPG is by far the best indicator of bone mass status.
POTENT ANTIRETROVIRAL THERAPY CAN INDUCE BONE MINERAL LOSS IN HIV-INFECTED CHILDREN BY ENHANCING BONE RESORPTION
S. Mora1*, N. Sala2, D. Bricalli2, G. Zuin2, G. Chiumello1, A. Viganò2
1Laboratory of Pediatric Endocrinology, IRCCS S. Raffaele, University of Milan, Italy
2Chair of Pediatrics, Ospedale L. Sacco, Milan, Italy
Lipodystrophy and glucose/lipid metabolism abnormalities have been recently recognized as a consequence of highly active antiretroviral therapy (HAART) in HIV-infected patients. Osteoporosis is a newer additional complication described in HIV-infected adults; its etiology remains undefined and pediatric data are lacking.
Bone mineral density (BMD) and bone metabolism indexes were measured in 36 HIV-infected children (HIV+) (age range 6.2 – 17.0 yrs): 33 on HAART and 3 naïve to any antiretroviral treatment. Six out of 33 HAART-treated patients had signs of lipodystrophy. BMD was measured at the lumbar spine (L2-L4) and in the whole skeleton by dual-energy x-ray absorptiometry (DPX-L, Lunar Corp.). Bone formation was assessed by measurements of N-terminal propeptide of type I procollagen (PINP, Orion Diagnostica, Finland), and bone resorption was estimated by urine measurements of N-terminal telopeptide of type I collagen (NTx, Osteomark, Ostex, USA). BMD and bone metabolism indexes were also measured in 314 healthy controls (age range 4.9 – 18.5 yrs). Differences between patients and controls were assessed by multivariate analyses after controlling for confounding variables (age, sex, and anthropometric measurements). Bone density values were significantly lower in HIV+ than in HC: mean spinal BMD 0.803±0.205 g/cm2 in HIV+ vs. 0.858±0.206 g/cm2 in controls (p=0.004); mean whole body BMD 0.915±0.129 g/cm2 in HIV+ vs. 0.987±0.130 g/cm2 (p<0.0001). Children with lipodystrophy had BMD values remarkably lower compared to those of untreated patients, while patients on HAART but without lipodystrophy had intermediate values. However, differences were significant only for the whole skeleton (p=0.028). PINP serum concentrations were similar in HIV+ and in control subjects (524.0±266.1 microg/L vs. 526.5±211.7 microg/L). On the contrary, NTx urine levels were significantly increased in HIV+ vs. controls (441.2±268.9 nM BCE/mM creatinine vs. 244.6±135.2 nM BCE/mM creatinine) (p<0.0001). In conclusion, bone density in HIV-infected children is markedly reduced, and the severity of osteopenia appears to be related with adipose tissue maldistribution. Enhancement of bone resorption seems to induce bone demineralization in HIV-infected children.
ASSESSMENT OF EFFECTS OF LY333334 [RECOMBINANT HUMAN PARATHYROID HORMONE (1-34)] ON CORTICAL BONE USING DIGITAL X-RAY RADIOGRAMMETRY.
L. Hyldstrup1*, J. T. Jorgensen2, G. Gaich3
1Dept of Endocrinology, Hvidovre Hospital, University of Copenhagen, DK
2Pronosco A/S, Vedbaek, DK
3Eli Lilly and Company, Indianapolis IN, USA
LY333334 [rhPTH(1-34)] has recently been shown to substantially reduce vertebral and nonvertebral fragility fracture risk by similar amounts, although the increases in bone mineral density (BMD) are much larger at the spine than at nonvertebral sites (Neer et al. Endocrine Society 2000). Digital X-ray radiogrammetry (DXR) technology can measure the cortical thickness of the radius, ulna and the three middle metacarpals at a resolution of 85 microns. A pilot study was performed to determine if DXR could detect changes in cortical thickness that might contribute to bone strength independent of BMD.
DXR was performed at baseline and at 1 year in 40 postmenopausal women with osteoporosis, who participated in a large double-blind randomized multicentre study. Twelve patients received placebo and 28 received either rhPTH(1-34) 20 or 40 microgram once daily by self-injection. The effects of both doses were similar, so results were pooled.
An analysis of the data from all sites combined (five inner and five outer bone diameter measurements per patient) showed that PTH increased the outer diameter (p=0.016) and decreased the inner diameter (p=0.08), compared with placebo. There were consistent trends toward decreased cortical thickness in the placebo group and increased cortical thickness in the PTH group at the individual sites, although this was statistically significant at only one site (see table). These observations are in contrast to the changes seen in patients with primary hyperparathyroidism where both an expansion of the outer and inner bone diameters occur.
In conclusion, DXR can detect PTH-related increases in cortical thickness that may contribute to an increase in bone strength independent of BMD. The study was small, however, and the results need to be confirmed in a larger study.
| Placebo | PTH(1-34) 20 or 40 microg | p-value | |
| N | 12 | 28 | |
| Radius (mm) | -0.01187 | 0.03122 | 0.159 |
| Ulna (mm) | -0.02395 | 0.00415 | 0.487 |
| Metacarpal 2 (mm) | -0.00676 | 0.00087 | 0.627 |
| Metacarpal 3 (mm) | -0.03377 | -0.00295 | 0.117 |
| Metacarpal 4 (mm) | -0.01628 | 0.01905 | 0.026 |
TRANSFORMING GROWTH FACTOR-BETA2 REVERSES THE INCREASED ADIPOCYTE GENE EXPRESSION INDUCED BY SKELETAL UNLOADING IN RAT MARROW STROMAL CELLS.
S. Ahdjoudj1, F. Lasmoles1, X. Holy2, E. Zerath2, P. J. Marie1*
1INSERM U349 affiliated CNRS, Paris, France
2IMASSA-CERMA, Dept of Aerospatial Physiology, Brétigny sur Orge, France
Skeletal unloading induced by hindlimb suspension in rats induces osteopenia by inhibiting bone formation. We previously showed that TGF-beta mRNA levels are reduced in unloaded metaphyseal bone and that TGF-beta2 administration increases osteoprogenitor cell recruitment, osteoblast marker gene expression and bone formation in unloaded rats. In this study, we determined the effects of unloading and TGF-beta2 on adipocyte differentiation in the bone marrow stroma. Unloaded adult Wistar male rats were infused with rhTGF-beta2 (2 µg/kg b.w., Novartis, Switzerland) during 0-7 days or only at 4-7 days of suspension. Unloaded and loaded control rats were sacrificed at days 0, 4 and 7, and tibias and femurs were processed for histomorphometric evaluation. RT-PCR analysis of osteoblast and adipocyte gene expression corrected for GAPDH was conducted separately in metaphyseal bone and marrow stroma. Bone volume, osteoblast and osteoid surfaces were decreased in unloaded metaphyseal tibia at 7 days compared to control rats. This was associated with decreased expression of type I collagen mRNA at 4-7 days and decreased Osf2 and osteocalcin mRNA levels at 7 days of suspension. Treatment with rhTGF-beta2 at 4-7 days corrected the abnormal expression of Osf2, osteocalcin and type I collagen mRNAs, and normalized bone formation in unloaded bone. Analysis of adipocyte gene expression in unloaded bone marrow stroma showed increased expression of PPARgamma2 at 4-7 days associated with increased expression of lipoprotein lipase (LPL) and aP2 mRNA levels at 7 days, indicating increased adipogenic differentiation. Treatment with rhTGF-beta2 rapidly decreased PPARgamma2 and CEBPalpha mRNA levels at 4 days and corrected the elevated aP2 and LPL mRNA levels at 7 days. In addition, treatment with rhTGF-beta2 (0-4 days) transiently increased c-fos mRNA in unloaded marrow stroma, indicating increased cell proliferation. The results show for the first time that continuous or transient treatments with TGF-beta2 not only correct the defective osteoblast differentiation and bone formation induced by unloading, but also reverse the increased adipocyte differentiation of marrow stromal cells, by acting on differentiation-specific gene expression. This provides novel mechanisms of action of TGF-beta2 on bone marrow adipogenesis induced by unloading in vivo.
MOLECULAR MECHANISM OF CELL DIFFERENTIATION AND APOPTOSIS INDUCED BY A-RING ANALOGUES OF 1ALPHA,25-DIHYDROXYVITAMIN D3
T. Okano1*, K. Nakagawa1, K. Ozono2, N. Kubodera3, K. Mikami4
1Kobe Pharmaceutical University, Kobe, Japan
2Osaka Medical Center for Maternal and Child Health, Osaka, Japan
3Chugai Pharmaceutical Co., Ltd., Tokyo, Japan
4Tokyo Institute of Technology, Tokyo, Japan
1alpha,25-dihydroxyvitamin D3 [1alpha,25(OH)2D3] exhibits antiproliferative, prodifferentiating, and apoptosis-inducing effects on many malignant cells. These properties have raised the possibility of its use as a therapeutic agent in cancer. We have reported the biological activities of several series of 1alpha,25(OH)2D3 analogues with respect to modulation of vitamin D-dependent target gene expressions and regulation of cell differentiation, proliferation and apoptosis in human promyelocytic leukemia (HL-60) cells (Chem & Biol 2000; 7:173, Biochem Pharmacol 2000; 59:691, ibid, in press). It was interesting to note that the analogues could induce differentiation or apoptosis of HL-60 cells on the basis of the stereochemistry of both hydroxy groups at positions 1 and 3 of the A-ring. To further elucidate the possible roles of both the hydroxy groups in regulating cell differentiation and apoptosis, we have synthesized all possible diastereomers of the A-ring of 1alpha,25(OH)2D3 and evaluated their biological activities; (1) at the molecular level, transactivation on rat 25-hydroxyvitamin D3-24-hydroxylase and human osteocalcin gene promoters, vitamin D receptor (VDR) / retinoid X receptor complex formation, VDR / coactivator (SRC-1) interaction in transfected cells, (2) at the cellular level, modulation of differentiation, cell cycle and apoptosis of HL-60 cells. 1alpha,25(OH)2D3 stimulated VDR-mediated gene expressions and induced cell differentiation dose-dependently. Epimerization of either 1alpha- or 3beta-hydroxy group or both remarkably reduced the biological activities. 1alpha,25(OH)2D3 and its 3-epimer (3alpha) induced apoptosis only in a low concentration in the medium. In contrast, both the 1-epimer (1beta) and 1,3-epimer (1beta,3alpha) of 1alpha,25(OH)2D3 induced apoptosis dose-dependently through stimulating caspase-9 and -3 activities with increased Baxalpha gene expression and changes of mitochondrial membrane potential, but failed to stimulate VDR-mediated actions at the molecular and cellular levels. These results suggest that 1alpha,25(OH)2D3 induces apoptosis and differentiation of HL-60 cells in a low concentration and in a high concentration, respectively. Using diastereomers of 1alpha,25(OH)2D3, we clearly identified for the first time the structural motifs inducing differentiation and apoptosis on the basis of the stereochemistry of the C-1 and C-3 hydroxy groups in the A-ring of 1alpha,25(OH)2D3. These findings provide useful information not only for structure-function studies of 1alpha,25(OH)2D3 analogues but also for the development of therapeutic agents for the treatment of cancer.
PTHrP AND PTH1 RECEPTOR EXPRESSION AND TONIC EFFECTS OF PERIPHERAL RECEPTOR GENE DELIVERY IN RENAL VESSELS OF SPONTANEOUSLY HYPERTENSIVE RATS (SHR)
T. Massfelder*, N. Taesch, A. Eichinger, S. Fritsch, B. Escande, M. Barthelmebs, J. J. Helwig
University Louis Pasteur Medical School, E 0015 INSERM, Strasbourg, France
Several reasons prompted us to explore the PTHrP system in renal vessels of SHR: first, the kidney is a major target of the vasodilatory effect of PTHrP; second, genetically determined renal mechanisms, including high renal vascular resistance (RVR), play a major role in the development of primary hypertension; third, PTHrP is upregulated in the aorta of SHR. Finally, we reported earlier that the vasodilatory effect of PTHrP is reduced in the isolated perfused kidney (IPK) of SHR.
We now report that in intrarenal arteries of 12-week-old SHR, immunoreactive PTHrP expression was increased by ~40%, while PTH1 receptor (PTH1R) mRNA and protein were decreased by ~60-80%, as compared to age-matched normotensive rats. We then asked whether downregulation of PTH1R could be responsible for the high RVR. For that purpose, we delivered naked pcDNA1.1 vector (0.5 mg), containing the full length of human PTH1R DNA, under the control of the CMV promoter, into 9-week-old SHR, via a single intravenous injection. Control-treated SHR received 0.5 mg empty vector. Three weeks after plasmid delivery, the expression of hPTH1R was identified in heart, brain, aorta, liver and lungs by RT-PCR. In kidney, the transgene was preferentially localized in intrarenal arteries as compared to total kidney. The vasodilation induced by 10 nM PTHrP(1-36) in IPK perfused at constant flow and preconstricted with 10 microM phenylephrine, increased from 19±11% in control-treated SHR (n=5, p>0.05), to 54±8% (n=5, p<0.05) in hPTH1R-transfected animals. PTH1R gene delivery, also decreased basal RVR of the IPK from 7.2±0.2 (n=9) in control-treated to 5.2±0.3 mmHg min g ml-1 (n=13) in hPTH1R-transfected (p<0.05). Importantly, the decrease in IPK tone was due to endogenous vasodilatory PTHrP, since infusion of 40 nM PTHrP(3-36), a PTH1R antagonist, increased basal RVR by 13% (n=4, p<0.05) in PTH1R-transfected animals.
In conclusion, endogenous PTHrP regulates renal tone in vitro in SHR: downregulation of PTH1R contributes to the high RVR, despite the upregulation of its endogenous vasodilatory ligand.
THE CAPACITY OF OSTEOBLASTIC CELLS TO RESPOND TO PTH STIMULATION IS MODULATED BY CONNEXIN43-MEDIATED GAP-JUNCTIONAL COMMUNICATION
G. D'Ippolito1,2, P. Hernandez1, G. A. Howard1,2,3, B. A. Roos1,2,3, P. C. Schiller1,2*
1GRECC and Research Service, VA Medical Center, Miami, FL, USA
2University of Miami School of Medicine, Miami, FL, USA
3State of Florida Teaching Nursing Home at Miami Jewish Home & Hospital for the Aged, Miami, FL, USA
Connexin43-mediated gap-junctional communication (Cx43/GJC) is a strong determinant of the osteoblastic phenotype in vitro and in vivo. Cx43/GJC increases as a function of osteoblast maturation in an in vitro model. PTH treatment of cultured osteoblastic cells can increase their capacity to produce a mineralized extracellular matrix (ECM), an effect consistent with the in vivo bone anabolic effects of PTH. However, the PTH stimulation of ECM mineralization is dependent on the maturational stage of the osteoblastic cells, it is observed at the time when Cx43/GJC is highest, and inhibited if GJC is blocked. To begin investigating the mechanism by which Cx43/GJC may modulate the capacity of osteoblastic cells to respond to PTH we transfected osteoblastic cells with plasmids encoding either a YFP-tagged functional Cx43 (pCx43YFP, provided by Dr. DW Laird) or a dominant-negative mutant of Cx43 (pDNCx43; provided by Dr. VA Krutovskikh). While the first construct increased GJC in poorly coupled rat UMR 106.01 osteoblastic cells, pDNCx43 decreased GJC in UMR 106.01 and in the highly coupled rat ROS17/2.8 osteoblastic cells. To evaluate the effect of the specific genetic modulation of Cx43/GJC on the responsiveness of osteocalcin (OC) promoter, the prototype osteoblastic gene, to PTH stimulation; the pCx43YFP or pDNCx43 expression vectors were cotransfected with pOClux, a luciferase reporter gene driven by cloned 1.1 Kb rat OC promoter sequences. In highly coupled ROS 17/2.8 cells, OC promoter activity was very high and the PTH treatment did not increase its activity. Co-expression of pDNCx43 decreased basal OC promoter activity by 50-60%. In poorly coupled UMR106.01 cells basal OC promoter activity was low and PTH treatment had a 1.5-fold stimulatory effect. Co-expression of pDNCx43 decreased promoter activity to background levels, while co-expression of pCx43YFP not only increased the basal promoter activity but also enhanced the PTH responsiveness of the promoter by 20-50 fold. All experiments were normalized to ß-galactosidase activity from a reporter construct used as internal control. In conclusion, modulation of Cx43/GJC has a profound effect on how osteoblastic cells respond to PTH stimulation. Cx43/GJC may modulate the signal transduction mechanisms induced by PTH binding to its receptor.
MICE WITH A TARGETED GENE DELETION IN ESTROGEN RECEPTOR-BETA SHOW ENHANCED OSTEOGENIC RESPONSIVENESS TO 17BETA-ESTRADIOL
K. E. McDougall*, M. J. Perry, J. H. Tobias
University of Bristol, Bristol, UK
Though alpha and beta isoforms of the estrogen receptor (ERalpha and ERbeta) are both known to be expressed by bone cells, whether these play distinct roles in regulating skeletal responses to estrogen is currently unknown. To determine if ERbeta plays a significant role in mediating the stimulatory action of estrogen on bone formation, we examined whether this response is impaired in mice homozygous for a targeted gene deletion in ERbeta (BERKO mice), as assessed by measurement of bone mineral density (BMD). A breeding colony of mice heterozygous for a deletion of the ERbeta gene was established in our institution (Krege et al, Proc Natl Acad Sci USA 1998, 95:15677). Twelve-week-old male and female BERKO animals, and age-matched wild-type litter mates, were administered vehicle or 17beta-estradiol (E2) 4, 40, 400 or 4000 mircog/kg/day for 28 days (2-4 animals per group). Femoral and tibial BMD were subsequently measured using a Lunar PIXImus. No significant differences in BMD were observed between vehicle-treated WT and BERKO animals. Though E2 led to a dose-responsive increase in femoral BMD overall (p=0.0001), this response was found to plateau at a ten-fold lower concentration in BERKO mice. For example, in mice receiving an intermediate dose of E2 (i.e. 40 microg/kg/day) (E40), whole femoral BMD (mg/cm2) was found to be significantly greater in BERKO compared to WT mice, particularly in males (p=0.02 by two-way ANOVA) (mean±SEM):-
BMD was also higher in BERKO mice given E40 compared to WT controls, as assessed at the distal femoral metaphysis, whole tibia, and proximal tibial metaphysis (p=0.005, p=0.002, p=0.01 respectively, by two-way ANOVA). Our findings suggest that ERbeta exerts a negative influence on bone formation, and does not mediate the stimulatory action of estrogen on bone formation.
| Male | Female | |||
| WT | BERKO | WT | BERKO | |
| Vehicle | 56.2±2.9 | 53.3±2.1 | 52.5±1.0 | 54.1±0.8 |
| E40 | 55.0±1.3 | 64.5±2.2 | 55.1±2.7 | 59.5±3.0 |
| %Increase | -2.1 | +22.9 | +5.0 | +10.0 |
ESTROGEN RECEPTOR SPECIFICITY FOR THE REGULATION OF THE FEMALE SKELETON
M. Lindberg1*, N. Andersson1, N. Hellberg1, J-Å. Gustafsson2, C. Ohlsson1
1Division of Endocrinology, Department of Internal Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden
2Department of Medical Nutrition, Karolinska Institute, Novum, Huddinge, Sweden
Estrogen is important for the regulation of longitudinal bone growth and trabecular bone mineral density (tBMD) in females. There are two known estrogen receptors, estrogen receptor-alpha (ERalpha) and estrogen receptor-beta (ERbeta), which are suggested to mediate the effects of estrogen. The aim of the present study was to compare longitudinal bone growth in female mice lacking ERalpha (ERKO), ERbeta (BERKO), or both receptors (DERKO) and to investigate the effect of estrogen treatment on the tBMD in ovariectomized mice. The length of femur was decreased in ERKO, increased in BERKO while it was intermediate in DERKO females, demonstrating opposing effects of ERalpha and ERbeta. These effects on longitudinal bone growth were correlated with similar effects on serum levels of IGF-I. After ovariectomy, the mice were treated for three weeks with 2.3µg/mouse/ day of 17beta-estradiol or vehicle. The tBMD was measured using peripheral Quantitative Computerized Tomography (pQCT) on the distal part of the femur. Treatment with estrogen resulted in a pronounced increase of the tBMD in wildtype and BERKO mice while no significant effect was seen in ERKO and DERKO mice, demonstrating that the effects of estrogen on tBMD is predominantly mediated via ERalpha.
In summary, ERalpha and ERbeta exert opposing effects regarding longitudinal bone growth in female mice. In contrast, the estrogenic effects on tBMD are mediated via ERalpha in female mice.
ER-ALPHA, BUT NOT ER-BETA, MEDIATES THE REGULATION OF THE INSULIN–LIKE GROWTH FACTOR 1 GENE BY SELECTIVE ESTROGEN RECEPTOR MODULATORS
B. Fournier1*, S. Gutzwiller1, T. Dittmar1, G. Matthias2, P. Steenbergh3, P. Matthias2
1Arthritis & Bone Metabolism Therapeutic Area, Novartis Pharma Research, 4002 Basel, Switzerland
2Friedrich Miescher Institute, 4002 Basel, Switzerland
3Department of Physiological Chemistry, Utrecht University, The Netherlands.
The importance of insulin-like growth factor 1 (IGF-1) on maintenance of skeletal integrity has been widely recognized. Although osteoblasts secrete some IGF-1, the liver is the primary endocrine source for IGF-1. We have studied the regulation of the human IGF-1 promoter in the hepatocytic cell line Hep-3B. We have shown that the IGF-1 promoter, when cotransfected together with an estrogen receptor-alpha (ER alpha)-expression vector, was transcriptionally regulated by raloxifene or raloxifene-like molecules, but not by 17 beta-estradiol or 4(OH)-tamoxifen. The induction mediated by raloxifene is antagonized by 17 beta-estradiol and is mediated selectively by ER alpha, not by ER beta. Transfer of IGF-1 promoter sequences from -733 to -65 or -375 to -65 to a minimal Fos promoter resulted in a comparable responsiveness to raloxifene. This region contains two CCAAT Enhancer Binding Protein (C/EBP) sites and an AP-1 site, both of which have been shown to be involved in estrogen receptor mediated transactivation. When the C/EBP sites were mutated in a construct bearing the sequence from –375 to –65 in front of the minimal Fos promoter, raloxifene induction was reduced, while mutation of the other elements did not affect induction. In addition, using chimeric proteins we delineated the domains of ER alpha which confer transactivation abilities on the IGF-1 promoter that are not exhibited by ER beta. These data shed new light on the mechanism of action of selective estrogen receptor modulators (SERM) and might help explain, at least in part, the bone protective effects observed for some SERMs in ovariectomized animals.
SIGNALLING NETWORKS IN THE MEMBRANE ACTION OF ANDROGENS IN MALE OSTEOBLASTS
M. Lieberherr*, Y. Zagar
CNRS UPR 1524, INRA, Jouy-en-Josas, France
Testosterone (T) and 5 alpha-dihydrotestosterone (DHT) from 1 pM to 10 nM increased within 5 s the intracellular calcium concentration ([Ca2+]i) by mobilizing Ca2+ from the endoplasmic reticulum. T and DHT enhanced within 5 s the formation of inositol trisphosphate (IP3) and diacylglycerol (DG). Androgens had no effect on cAMP whereas cAMP was increased by 1 nM PTH or 100 nM forskoline. Incubation of the cells with 100 ng/ml pertussis toxin (PTX) inhibited the increase in [Ca2+]i, IP3 and DG. We then characterized the isoform of PLC, the subunits of heterotrimeric G-protein and the isoform of conventional PKC involved in the rapid effects of androgens. We used specific antibodies against the various types of PLC (PLC beta 1, 2, 3 and 4; PLC gamma 1 and 2), the subunits alpha of various G-proteins (Gs, Gi, and Gq/11) and the different beta subunits (1, 2, 3 and 4). Cells were saponified for 5 min to allow the entrance of the antibody, well washed to remove saponin, then incubated for 60 min with the antibody and loaded the last 20 min with 1 µM fura-2/AM. Only the PLC beta 2 and the subunit beta 4 of a PTX-sensitive G-protein were involved in the calcium mobilization induced by T and DHT. Western immunoblotting with specific antibodies against PKC alpha and PKC beta 1 showed a translocation of both PKC within 10 s from the cytosol to the membrane. Osteoblasts were incubated for 20 min with specific inhibitors of these PKC before adding the androgen at 100 pM. This enhanced the calcium response to androgen showing a feeback control by these PKC on the PLC involved.
In conclusion, this finding that a heterotrimeric G-protein coupled to a PLC beta is involved in the membrane signalling of androgens may be an important step towards understanding the process of these rapid actions. Moreover, the activation of PKC may induce to the rapid phosphorylation of different MAPkinases leading to the activation of early genes involved in the proliferation and differentiation by bypassing the nuclear androgen receptor. Finally, the membrane effects of androgens start at concentrations as low as 1 pM.
PENTOXIFYLLINE ACTIVATES MAPK CASCADES AND PROMOTES OSTEOBLAST DIFFERENTIATION BY A MECHANISM INDEPENDENT OF PROTEIN KINASE A ACTIVATION
G. Rawadi1*, F. Ferrer1, S. Spinelle-Jaegle1, S. Roman-Roman1, Y. Bouali1, R. Baron1,2
1Bone Disease Group, Hoechst Marion Roussel, Aventis Pharma., France
2Yale University School of Medicine, New Haven, USA
Phosphodiesterase (PDE) inhibitors have been described to affect osteoblastogenesis and osteoclastogenesis in vitro but little is known about the molecular mechanisms by which PDE inhibitors affect bone cells. In the present study we have investigate the effect of pentoxyfilline (PeTx), a non selective phosphodiesterase inhibitor, on osteoblastic differentiation in vitro by using two pluripotent mesenchymal cell lines, C3H10T1/2 and C2C12, able to acquire the osteoblastic phenotype in the presence of bone morphogenetic protein-2 (BMP-2).PeTx induces the osteoblastic markers, osteocalcin (OC) and Osf2/Cbfa1, in C3H10T1/2 and C2C12 cells. Moreover, PeTx enhances the BMP-2-induced expression of OC, Osf2/Cbfa1 and alkaline phosphatase (ALP). This activity can be at least partially attributed to the fact that PeTx is able to enhance BMP-2-induced Smad1 transcriptional activity. Although PeTx clearly stimulates PKA in the cells used here, neither pretreatment of cells with the PKA inhibitor H89 nor transfection with the specific PKA inhibitor PKI prevented the induction or enhancement of osteoblast markers by PeTx demonstrating that these effects are independent of PKA activation. Interestingly PeTx induces the activation of ERK1/2 and p38 kinase pathways independently of the activation of PKA. Selective inhibitors of these MAPK cascades prevented the acquisition of osteoblastic markers in cells treated with PeTx suggesting that activation of these two pathways plays a role in the effect of PeTx on osteoblastic differentiation. In conclusion, PeTx has the potential to commit undifferentiated mesenchymal cells into the osteoblastic lineage by a mechanism involving MAPK but independent of the activation of PKA.
INVOLVEMENT OF THE MAP-KINASE AND SRC-KINASE PATHWAYS IN THE NUCLEAR TRANSLOCATION OF THE OF THE AP-1 FAMILY MEMBERS IN ROS17/2.8 OSTEOBLAST-LIKE CELLS SUBMITTED TO CHANGES IN MECHANICAL ENVIRONMENT
C. Granet*, A. Guignandon, O. Akhouayri, C. Alexandre, M. H. Lafage-Proust
INSERM EM9901, J Monnet University, St-Etienne, France
Bone loss occurs in weight-bearing bones in microgravity whereas bone mass increases after skeletal loading. This adaptation is mediated through alterations of osteoblastic behavior, whose control remains partially unknown. Some of the members of the AP-1 transcription factor family play a specific role in bone metabolism including its response to strain.
We studied in ROS17/2.8 osteoblast-like cells, gene expression and nuclear translocation of all the members of the AP-1 family, in a simulated microgravity model, the clinostat 2D, and in a model of mechanical strain, the Flexcell. We performed RT-PCR and immuno-cytochemistry analyses at baseline and 10, 20, 30, 45, 60, 90 and 120min after the end of stimulation (either a mitogenic regimen i.e, 10min 1%-stretch, 0.05Hz or 10-min clinorotation, 50rpm). Beside the already known c-fos activation by mechanical strain, gene expression of all the other AP-1 members was induced in both models, except fosB in the clinostat, with differences in their kinetics. Nuclear translocation of all the members was observed in both models (from 30 to 45min for all the AP-1 members in stretched cells and a prolonged expression of Fra-1 in clino-rotated cells). Neither down-regulation of PKC by PMA long-term treatment nor indomethacin COX-inhibition nor inhibition of ERK1/2, p38MAPK and src-kinases by PD98059, SB203580 and PP2, respectively, abolished AP-1 gene expression in both models. In contrast, in clino-rotated cells, nuclear translocation of c-fos, FosB, Fra-2 was abolished by PD98059, SB203580 and PP2 while Fra-1 and Jun-B were present regardless of the treatment. Results obtained on the Flexcell are summarised in Table1.
Thus, changes in the osteobblastic mechanical environment induced a dramatic induction of all the AP-1 members with specific kinetics and involvement of MAP-kinases and src-kinases which differed whether the cells were stretched or submitted to simulated microgravity. Knowing the critical importance of AP-1 members in osteoblastic proliferation and maturation, these results might be the first step to further understand bone response to its mechanical environment.
| Flexcell | ERK1/2 | p38 | src kinases |
| c- fos | + | 0 | 0 |
| FosB | + | + | + |
| Fra-1 | 0 | 0 | 0 |
| Fra- 2 | + | + | + |
| c- jun | + | + | + |
| JunB | 0 | + | 0 |
| JunD | 0 | 0 | 0 |
| +: abolition, 0 no effect | |||
PROSTAGLANDINS REGULATE FGF-2 AND FGF RECEPTOR EXPRESSION AND SUBCELLULAR LOCALIZATION IN OSTEOBLASTS
M. G. Sabbieti1*, L. Marchetti1, M. Menghi1, S. Materazzi2, G. Menghi1, L. G. Raisz3, M. M. Hurley3
1Dept. of Comparative Morphological & Biochemical Sciences, Univ. of Camerino, Camerino, Italy
2Dept. of Chemistry, Univ. "La Sapienza", Rome, Italy
3Div. of Endocrinology & Metabolism, Univ. of Connecticut School of Medicine, Farmington, Connecticut, USA
We previously found that prostaglandins (PGs) stimulated FGF-2 production in bone cells and postulated that some effects of PGs on bone remodeling might be mediated by endogenous FGF-2. Recent data support distinct biological functions depending on the subcellular location of FGF-2 protein. We examined the effects of PGs on FGF-2 and FGFR2 expression and localization in osteoblastic Py1a cells. Confluent cells were serum deprived for 24h before treatment with prostaglandin E2 (PGE2), prostaglandin F2 (PGF2a) or the PGF2a agonist fluprostenol (Flup) for 2 to 24h. FGF-2 and FGFR2 mRNA expression was determined by Northern blot analysis. In the absence of serum, Py1a cells expressed very low levels of 4 and 6kb FGF-2 mRNA transcripts. PGE2 increased the 6kb FGF-2 mRNA transcript within 2h with maximal increase at 4h. PGF2a and Flup induced the 6kb FGF-2 mRNA transcript after 4h and maximal increase occurred between 4 and 6h. PGE2 reduced FGFR2 mRNA expression by 30% and 80% at 4 and 24h respectively. PGF2a caused a 40% and Flup a 60% reduction in FGFR2 mRNA levels at 4h and both effectors caused an 80% reduction after 24h. Parallel experiments were conducted to investigate insitu localization of FGF-2 and FGFR2 protein binding patterns examined by confocal microscopy. Py1a cells expressed low levels of cytoplasmic and nuclear staining for FGF-2 and FGFR2 protein. However, following 24h of treatment with PGE2, PGF2a or Flup, FGF-2 was increased inside the nucleus. Although FGFR2 protein was increased in the nucleus, maximum binding was concentrated near the nuclear envelope. Cells incubated with cycloheximide and subsequently stimulated by PGE2, PGF2a or Flup did not show nuclear binding for FGF-2 suggesting new protein synthesis was required. To begin exploring the signaling pathway by which PGs regulate FGF-2 nuclear accumulation; we examined the effect of PMA, an activator of the protein kinase C (PKC) pathway. PMA pretreatment before administration of PGF2a prevented the nuclear accumulation of FGF-2. We conclude that PGs stimulate nuclear accumulation of FGF-2 by a PKC dependent pathway. We speculate that interaction of FGF-2 and its receptor may be important in the biologic responses to PGs in bone.