Effects of a sulfated exopolysaccharide produced by Altermonas infernus on bone biology.

The growth and differentiation of bone cells is controlled by various factors, which can be modulated by heparan sulfates. Here, we investigated the effects of an oversulfated exopolysaccharide (OS-EPS) on the bone. We compared the effect of this compound with that of a native EPS. Long-term administration of OS-EPS causes cancellous bone loss in mice due, in part, to an increase in the number of osteoclasts lining the trabecular bone surface. No significant difference in cancellous bone volume was found between EPS-treated mice and age-matched control mice, underlying the importance of sulfation in trabecular bone loss. However, the mechanism sustaining this osteoporosis was unclear. To clarify OS-EPS activities, we investigated the effect of OS-EPS on osteogenesis. Our results demonstrated that OS-EPS inhibited osteoclastogenesis in two cell models. Using the surface plasmon resonance technique, we revealed that OS-EPS can form a hetero-molecular complex OS-EPS/receptor activator of NF-κB ligand (RANKL)/RANK and that RANK had a higher affinity for RANKL pre-incubated with OS-EPS than for RANKL alone, which would be in favor of an increase in bone resorption. However, in vitro, OS-EPS inhibited the early steps of osteoclast precursor adhesion and therefore inhibited the cell fusion step. In addition, we showed that OS-EPS reduced proliferation and accelerated osteoblastic differentiation, leading to strong inhibition of mineralized nodule formation, which would be in favor of an increase in bone resorption. Taken together, these data show different levels of bone resorption regulation by EPSs, most of them leading to proresorptive effects.

Proteases and bone remodelling.

Bone remodelling is regulated by osteogenic cells which act individually through cellular and molecular interaction. These interactions can be established either through a cell-cell contact, involving molecules of the integrin family, or by the release of many polypeptidic factors and/or their soluble receptor chains. Proteolytic shedding of membrane-associated proteins regulates the physiological activity of numerous proteins. Proteases located on the plasma membrane, either as transmembrane proteins or anchored to cell-surface molecules, serve as activators or inhibitors of different cellular and physiological processes. This review will focus on the role of the proteases implicated in bone remodelling either through the proteolytic degradation of the extracellular matrix or through their relations with osteogenic factors. Their implication in bone tumor progression will be also considered.

Key roles of the OPG-RANK-RANKL system in bone oncology.

Osteoprotegerin (OPG)-receptor activator of nuclear factor-kappaB (RANK) and RANK ligand (RANKL) have been identified as members of a ligand-receptor system that directly regulates osteoclast differentiation and osteolysis. RANKL may be a powerful inducer of bone resorption through its interaction with RANK, and OPG is a soluble decoy receptor that acts as a strong inhibitor of osteoclastic differentiation. Any dysregulation of their respective expression leads to pathological conditions. Furthermore, recent data demonstrate that the OPG-RANK-RANKL system modulates cancer cell migration, thus controlling the development of bone metastases. This review describes the most recent knowledge on the OPG-RANK-RANKL system, its involvement in bone oncology and the new therapeutic approaches based on this molecular triad.

Characterization of osteoprotegerin binding to glycosaminoglycans by surface plasmon resonance: role in the interactions with receptor activator of nuclear factor kappaB ligand (RANKL) and RANK.

Osteoprotegerin (OPG) is a decoy receptor for receptor activator of nuclear factor kappaB ligand (RANKL), a key inducer of osteoclastogenesis via its receptor RANK. We previously showed that RANK, RANKL, and OPG are able to form a tertiary complex and that OPG must be also considered as a direct effector of osteoclast functions. As OPG contains a heparin-binding domain, the present study investigated the interactions between OPG and glycosaminoglycans (GAGs) by surface plasmon resonance and their involvement in the OPG functions. Kinetic data demonstrated that OPG binds to heparin with a high-affinity (KD: 0.28 nM) and that the pre-incubation of OPG with heparin inhibits in a dose-dependent manner the OPG binding to the complex RANK-RANKL. GAGs from different structure/origin (heparan sulfate, dermatan sulfate, and chondroitin sulfate) exert similar activity on OPG binding. The contribution of the sulfation pattern and the size of the oligosaccharide were determined in this inhibitory mechanism. The results demonstrated that sulfation is essential in the OPG-blocking function of GAGs since a totally desulfated heparin loses its capacity to bind and to block OPG binding to RANKL. Moreover, a decasaccharide is the minimal structure that totally inhibits the OPG binding to the complex RANK-RANKL. Western blot analysis performed in 293 cells surexpressing RANKL revealed that the pre-incubation of OPG with these GAGs strongly inhibits the OPG-induced decrease of membrane RANKL half-life. These data support an essential function of the related glycosaminoglycans heparin and heparan sulfate in the activity of the triad RANK-RANKL-OPG.

RANKL/RANK/OPG: new therapeutic targets in bone tumours and associated osteolysis.

The emergence of the molecular triad osteoprotegerin (OPG)/Receptor Activator of NF-kB (RANK)/RANK Ligand (RANKL) has helped elucidate a key signalling pathway between stromal cells and osteoclasts. The interaction between RANK and RANKL plays a critical role in promoting osteoclast differentiation and activation leading to bone resorption. OPG is a soluble decoy receptor for RANKL that blocks osteoclast formation by inhibiting RANKL binding to RANK. The OPG/RANK/RANKL system has been shown to be abnormally regulated in several malignant osteolytic pathologies such as multiple myeloma [MM, where enhanced RANKL expression (directly by tumour cells or indirectly by stromal bone cells or T-lymphocytes)] plays an important role in associated bone destruction. By contrast, production of its endogenous counteracting decoy receptor OPG is either inhibited or too low to compensate for the increase in RANKL production. Therefore, targeting the OPG/RANK/RANKL axis may offer a novel therapeutic approach to malignant osteolytic pathologies. In animal models, OPG or soluble RANK was shown both to control hypercalcaemia of malignancy and the establishment and progression of osteolytic metastases caused by various malignant tumours. To this day, only one phase I study has been performed using a recombinant OPG construct that suppressed bone resorption in patients with multiple myeloma or breast carcinoma with radiologically confirmed bone lesions. RANK-Fc also exhibits promising therapeutic effects, as revealed in animal models of prostate cancer and multiple myeloma. If the animal results translate to similar clinical benefits in humans, using RANK-Fc or OPG may yield novel and potent strategies for treating patients with established or imminent malignant bone diseases and where standard therapeutic regimens have failed.

Potential synergies between matrix proteins and soluble factors on resorption and proteinase activities of rabbit bone cells.

Human growth hormone (GH) has recently been found to stimulate osteoclastic resorption, cysteine-proteinase and metalloproteinase activities (MMP-2 and MMP-9) in vitro via insulin-like growth factor-I (IGF-I) produced by stromal cells. The present study investigated the effects of two extracellular matrix components (vitronectin and type-I collagen) on hGH- and hIGF-1-stimulated osteoclastic resorption and proteinase activities in a rabbit bone cell model. After 4 days of rabbit bone cell culture on dentin slices with vitronectin coating, hGH and hIGF-1 stimulated bone resorption and hIGF-1 upmodulated cysteine-proteinase activities. MMP-2 expression (but not resorption, cathepsin or MMP-9 activities) was upmodulated by hGH and hIGF-1 on dentin slices coated with type I collagen as compared to those without coating. Then, vitronectin was synergistic with hIGF-1 in the regulation of cysteine-proteinase production whereas collagen showed synergy with hGH and hIGF-1 in the regulation of MMP-2 production. Anti-alphavbeta3 totally abolished the effects of hGH and hIGF-1 on metalloproteinase release, but had no influence on cathepsin release. The results suggest that cysteine-proteinase modulation is not mediated by alphavbeta3 integrin (strongly expressed on osteoclastic surface) whereas the resorption process and metalloproteinase modulation are clearly mediated by this integrin. Our finding about the collagen coating also suggests that hGH- and hIGF-1-stimulated MMP-2 activity are mediated, along with alphavbeta3 integrin, by another adhesion molecule.

[In vitro assessment of combining osteogenic cells with macroporous calcium-phosphate ceramics]. / Etude in vitro de l'association de cellules ostéogènes avec une céramique en phosphate de calcium macroporeuse.

PURPOSE OF THE STUDY: Bone grafts or bone substitutes are required to fill bone defects resulting from trauma or surgical resection of tumors. Calcium-phosphate ceramics are synthetic bone substitutes which promote new bone formation by osteoconduction. These ceramics possess osteoconductive properties but have no intrinsic osteoinductive capacity. They are unable to induce new bone formation in extraossesous sites. One solution to develop bone substitutes with osteogenic properties would be to associate biomaterials with osteoprogenitors. MATERIALS AND METHODS: We studied the in vitro osteogenic potential of human bone-marrow cells cultured on macroporous calcium phosphate (CaP) ceramic, examining stromal cell proliferation and differentiation. Osteogenic differentiation was evaluated in terms of alkaline phosphatase activity and immunological characterization of the extracellular fibrillar matrix formed by these cells. The specimens were examined by scanning and transmission electron microscopy. RESULTS: Human bone-marrow cells proliferated on CaP ceramic. The proliferating bone-marrow cells expressed an osteoblastic phenotype as shown by alkaline phosphatase activity and synthesis in ceramic pores of an extracellular matrix composed of fibronectin, osteocalcin and collagen I. In addition, numerous microcrystals of apatite precipitated on the fibrillar matrix, producing a mineralized fibrillar network within the ceramic. CONCLUSION: This study demonstrates that human bone-marrow cells cultured on macroporous CaP ceramic do not lose their osteoblastic phenotype even after 21 days of culture, and that they can induce osteogenesis in a CaP ceramic in vitro. This type of new "hybrid material" appears promising for the future.

Modulation by soluble factors of gelatinase activities released by osteoblastic cells.

This study investigated the ability of normal human osteoblasts (hOb) and osteogenic sarcoma cells (MG-63 and SaOS2) to produce gelatinases and undergo modulation by interleukin 1beta (IL-1beta), interleukin 6 (IL-6), oncostatin M (OSM), leukaemia inhibitory factor (LIF), growth hormone (GH) and insulin-like growth factor-I (IGF-I). Gelatinase activities were determined by zymogaphy, and a quantitative analysis was performed by ELISA. The MMP-2 activities of the three cell lines were significantly increased in the presence of IL-1beta and IL-6, but no modulation of MMP-2 activities was observed in the presence of OSM, LIF and GH. IGF-I increased the activity released by SaOS2 and hOb, but no modulation was detectable in MG-63 cell conditioned medium. An upmodulation of pro-MMP-2 secretion by SaOS2 and hOb was observed for all soluble factors used, whereas an upmodulation of pro-MMP-2 secretion by MG-63 was observed only in the presence of IL-1beta, IL-6 and IGF-I. Thus, osteoblastic cells modulated by cytokines can be involved in bone resorption as a result of the protease activities released.

Inhibition of apatite formation by vitronectin.

This study investigated the concentration-dependent effect of vitronectin (VN), a glycoprotein of the bone matrix, on apatite formation and growth. Precipitation trials in metastable solution and in a pH-controlled solution system showed an inhibition of apatite microcrystal formation by VN. In the presence of biphasic calcium-phosphate ceramic, transmission electron microscopy showed a reduction of precipitated microcrystal size: precipitates were significantly smaller than in ionic simulated body fluid without proteins or in the presence of type I collagen as a negative control. Moreover, the size of the precipitated microcrystals was reduced in a dose-dependent manner. Two indirect methods showed that calcium-phosphate precipitation was inhibited by VN. It would appear that VN prevents apatite formation by inhibiting the growth of apatite crystals rather than by secondary nucleation, as in the case of osteopontin, a bone-specific protein.

Cysteine protease production by human osteosarcoma cells (MG63, SAOS2) and its modulation by soluble factors.

The production of cysteine protease by two human osteosarcoma cell lines (MG-63 and SaOS2) was analyzed, as well as their modulation by interleukin 1beta (hIL-1 beta), interleukin 6 (hIL-6), insulin growth factor-1 (hIGF-1), oncostatin M (hOSM), leukemia inhibitory factor (hLIF) and growth hormone (hGH). Cysteine protease activities were detected using a synthetic substrate. The protease activities (especially cathepsin L activity) of both cell lines were increased significantly in the presence of hIL-1 beta, hIL-6 and hOSM. In contrast, hIGF-1 and hGH decreased these activities, and no effect was detectable in the presence of hLIF. The addition of antibodies against the gp-130 chain of the hIL-6 and hOSM receptors totally inhibited the stimulating effect of these two cytokines on cysteine protease activities. In increasing collagen type I degradation, hIL-1beta, hIL-6 and hOSM could be involved in bone resorption, whereas the inhibitory action of hIGF-1 and hGH on collagen type I degradation suggest that this factor could play a role in bone formation.

[In vitro effects of growth hormone on osteoclastic activity: clinical applications]. / Effets in vitro de l'hormone de croissance sur l'activité des ostéoclastes : intérêt clinique.

PURPOSE OF THE STUDY: This study was designed to investigate the in vitro effects of human growth hormone (hGH) on osteoclastic resorption in a nonfractionated rabbit bone cell model. MATERIAL AND METHODS: Rabbit bone cells were cultured on dentine slices in the presence of parathyroid hormone and vitamin D3. The percentage of dentine slice surface resorbed, number of lacunae per surface unit and mean area of lacunae were compared between cell cultures grown in the presence of graded concentrations of hGH and human insulin-like growth factor-1 (hIGF-1) and controls. RESULTS: After 4 days of culture, rabbit bone cells cultured on dentine slices in the presence of hGH and hIGF-1 showed significantly stimulated osteoclastic resorption activity. When neutralizing anti-hIGF-1 anti-serum (4 microg/l) was added to the starting culture, the stimulatory effects of hIGF-1 and hGH on osteoclastic resorption activity were totally abolished. DISCUSSION: These findings indicate that the effects of hGH stimulation on osteoclastic resorption in vitro are mediated via local hIGF-1 secretion by stromal cells such as osteoblasts. Proteases appear to play a role in the degradation of the organic matrix. Our experiments show that hIGF-1 and hGH stimulate the production of matrix metalloproteinases MMP-9 and MMP-2. Similar to the resorption activity, hGH stimulates protease activity via stromal cell production of hIGF-1. CONCLUSION: This study suggests that natural or synthetic MMP inhibitor modulation of protease activity could reduce the degradation of the organic matrix and then prevent, for example, inflammatory reactions subsequent to prosthetic loosening.

Effects of fibronectin on hydroxyapatite formation.

There is increasing evidence that noncollagenous matrix proteins initiate bone mineralization in vivo. Fibronectin, which is present during the early phases of mineralization, may contribute to this process in bone tissues. In this context, the mineralization potential of fibronectin was tested in an agarose gel precipitation system and a metastable calcium phosphate solution. The protein inhibited the precipitation of calcium phosphate crystals in solution but had no apparent effect in gel. Conversely, fibronectin stimulated crystal formation when apatite powder was used to seed crystal growth in gel. Although these results in vitro do not clearly indicate that fibronectin is involved in the mineralization process, they are consistent with in vivo events. Free fibronectin (e.g. in biological fluids) could inhibit crystal growth but might also activate the mineralization process when absorbed on apatite powder in a bone environment and areas of ectopic mineralization.

Apatite precipitation after incubation of biphasic calcium-phosphate ceramic in various solutions: influence of seed species and proteins.

The dissolution-precipitation process for calcium-phosphate ceramics in contact with biological fluid was studied by incubating blocks of biphasic calcium phosphate composed of hydroxyapatite (HA) and beta-tricalcium phosphate (beta-TCP) in different solutions: ionic simulated body fluid (SBF) without protein or SBF that contained various proteins and macromolecules separately (fibronectin, vitronectin, albumin, and poly-L-glutamic acid). Transmission electron microscopy studies revealed that apatite-precipitated microcrystals appeared around ceramic crystals as a result of secondary nucleation; microcrystals were in continuity with the lattice planes of the HA crystals but in a different direction from that of beta-TCP; the size of the precipitates was smaller when fibronectin, vitronectin, and poly-(L-glutamic acid) were present in SBF as compared to SBF without protein; and fibronectin and vitronectin initiated crystal nucleation in the void spaces between the ceramic crystals.

Characterization of a strictly specific acid beta-galactosidase from Achatina achatina.

An acid beta-galactosidase was isolated from the digestive juice of Achatina achatina and purified to homogeneity by anion exchange, gel-filtration and hydroxyapatite chromatographies. This enzyme is soluble, as are the cytosolic beta-galactosidases, functions at acid pH like the lysosomal enzymes but differs from the other soluble animal beta-galactosidases in that it is highly specific for the beta-D-galactosyl residue. In addition, it cleaves the beta1-4 linkage much faster than the beta1-3 and beta1-6 linkages. The enzyme is a monomeric glycoprotein with a molecular mass of 120-125 kDa and the carbohydrate moiety makes up approximately 6% (w/w) of the protein. The amino acid composition displays an important amount of acidic/amide and hydroxy amino acid residues and a low content of basic residues. The enzyme activity is markedly affected by the ionic strength of the medium and the rate-pH curve was shifted towards higher pH values in the presence of added salt. Acid beta-galactosidase is capable of catalysing transgalactosylation reactions. The yields of galactosylation of hydroxy amino acid derivatives, catalysed by the enzyme in the presence of lactose as the glycosyl donor, were higher than those reported previously with conventional sources of beta-galactosidases. In addition, the pH optimum is different for the hydrolysis (pH 3.2) and transgalactosylation (pH 5.0) reactions. On the basis of this work, the enzyme could be used as a tool in the structural analysis of D-galactose-containing oligosaccharide chains, as well as for the synthesis of glycoconjugates.

Interleukin-8 processing by neutrophil elastase, cathepsin G and proteinase-3.

Activated neutrophils secrete two forms of IL-8 with 77 and 72 amino acids, IL-8(77) and IL-8(72), along with proteinases that could process these cytokines. Significant conversion of IL-8(77) to more potent, N-terminally truncated forms was observed upon incubation with neutrophil granule lysates and purified proteinase-3. IL-8(72) was considerably more resistant to proteolytic processing than IL-8(77). The present observations indicate that neutrophil proteinases released in inflamed tissues convert IL-8 to more active forms and therefore tend to conserve or enhance, rather than decrease IL-8 activity.

Oxidized and Met358-->Leu mutated alpha 1-proteinase inhibitor as substrates of Pseudomonas aeruginosa elastase.

This paper investigates the catalytic activity of Pseudomonas aeruginosa elastase using the bait region of the alpha 1-proteinase inhibitor as a substrate. The bacterial enzyme cleaves the Pro357-Met358 bond of the wild-type inhibitor and the recombinant Met358 inhibitor and the Pro357-Leu358 bond of the recombinant Met358-->Leu inhibitor with kcat/Km values of 9 x 10(4) M-1 s-1, 1.4 x 10(5) M-1 s-1 and 3.5 x 10(5) M-1 s-1, respectively. In contrast, the N-chlorosuccinimide-oxidized inhibitor (Met351 and Met358 = methionine sulfoxides) is cleaved at the Glu354-Ala355 position with a significantly lower rate (kcat/Km = 10(4) M-1 s-1). The pH optimum for the cleavage of the native, the oxidized, the Met358-->Leu mutated inhibitor, and 2-aminobenzoyl-Ala-Gly-Leu-Ala-4-nitrobenzylamide, a synthetic Pseudomonas elastase substrate are, 6.0, 7.0, 6.5 and 5.8, respectively. We conclude that P. aeruginosa elastase readily hydrolyzes substrates with P'1 methionine or alanine residues and that its pH optimum is not as alkaline as usually thought.

Oxidized alpha 1-proteinase inhibitor: a fast-acting inhibitor of human pancreatic elastase.

Unlike human neutrophil elastase or porcine and rat pancreatic elastases, human pancreatic elastase is rapidly inhibited by oxidized alpha 1-proteinase inhibitor. The second-order association-rate constant for the reaction of the oxidized inhibitor with this enzyme (kass = 10(5) M-1 s-1) is only 8-fold lower than that measured with native alpha 1-proteinase inhibitor. Elastase releases faster from its complex with the oxidized inhibitor (t1/2 approximately 0.7 days) than from its complex with the native inhibitor (t1/2 approximately 5 days). Oxidized alpha 1-proteinase inhibitor is as efficient as the native inhibitor in inhibiting the elastolytic activity of elastase. Oxidized alpha 1-proteinase inhibitor may thus be considered as a physiological inhibitor of human pancreatic elastase which may prevent degradation of blood vessel elastin during acute hemorrhagic pancreatis.

Elastin decreases the efficiency of neutrophil elastase inhibitors.

Elastase inhibitors are potential drugs for the control of lung emphysema. Since neutrophils may release elastase in the lung interstitium, elastin and inhibitors may complete locally for the binding of enzyme. To better evaluate the potential activity of antielastases, we have run experiments that mimic this in vivo competition. Elastase was added to mixtures of human lung elastin and inhibitor, and the solubilization of the fibrous substrate was measured as a function of time. Controls in which a synthetic substrate was used instead of elastin were run under identical conditions. We show that the rate constants for the irreversible inhibition of elastase by methoxysuccinyl-Ala2-Pro-Val-chloromethylketone and L-657,229, a substituted beta lactam, are 28- and 63-fold lower with elastin than with a synthetic substrate, respectively. The rate constant decreases with increasing concentrations of elastin, indicating that the inhibition is competitive. Elastin also impairs the potency of the following reversible inhibitors: trifluoroacetyl-Lys-Ala-NH-C6H4-p-C6H11, trifluoroacetyl-Lys-Ala-NH-C6H4-pN(C2H5)2, methoxysuccinyl-Ala2-Pro-Boro-Val-OH, and mucus proteinase inhibitor whose Ki values are 29- to 127-fold higher with elastin than with a synthetic substrate. Again the inhibition is competitive. We conclude that association rate constants of irreversible inhibitors and Ki values of reversible ones may be measured accurately using elastin as a substrate. The kinetic constants measured with elastin and not those determined with synthetic substrates should be used to decide whether a given inhibitor is potent enough to be a physiologic antielastase or a potential antielastase drug.

Pseudomonas aeruginosa elastase does not inactivate alpha 1-proteinase inhibitor in the presence of leukocyte elastase.

Pseudomonas aeruginosa elastase rapidly inactivates alpha 1-proteinase inhibitor by splitting its Pro-357-Met-358 peptide bond. The present study was aimed at testing whether this reaction takes place in the presence of leukocyte elastase. To this end was added alpha 1-proteinase inhibitor to a mixture of the two elastases, and we performed the following assays: (i) measurement of the residual leukocyte elastase activity, (ii) sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and (iii) immunoassay of the leukocyte elastase-alpha 1-proteinase inhibitor complex. These experiments were done with various concentrations of the three proteins. All experiments gave the same result: leukocyte elastase was fully inhibited by alpha 1-proteinase inhibitor in the presence of P. aeruginosa elastase even when the bacterial enzyme was 10-fold more concentrated than the neutrophil enzyme. We also measured the initial rate of the P. aeruginosa elastase-catalyzed inactivation of alpha 1-proteinase inhibitor as a function of the inhibitor concentration. The kcat/Km value derived from this experiment was 9 x 10(4) M-1 s-1, a value much lower than the rate constant for the leukocyte elastase-inhibitor association (kass, 1.7 x 10(7) M-1 s-1). This rationalizes the above results. In conclusion, when alpha 1-proteinase inhibitor is faced with its target enzyme, leukocyte elastase, it will perform its physiologic antielastase function even if the bacterial elastase is present in excess.