The inorganic phosphate/inorganic pyrophosphate axis in the mineralization of cartilage and bone

Proteomic characterization of biogenesis and functions of matrix vesicles released from mineralizing human osteoblast-like cells

Matrix vesicles (MVs), released by budding from apical microvilli of osteoblasts during bone formation and development, are involved in the initiation of mineralization by promoting the formation of hydroxyapatite in their lumen. To gain additional insights into MV biogenesis and functions, MVs and apical microvilli were co-isolated from mineralizing osteoblast-like Saos-2 cells and their proteomes were characterized using LC-ESI-MS/MS and compared. In total, 282 MV and 451 microvillar proteins were identified. Of those, 262 were common in both preparations, confirming that MVs originate from apical microvilli. The occurrence of vesicular trafficking molecules (e.g. Rab proteins) and of the on-site protein synthetic machinery suggests that cell polarization and apical targeting are required for the incorporation of specific lipids and proteins at the site of MV formation. MV release from microvilli may be driven by actions of actin-severing proteins (gelsolin, cofilin 1) and contractile motor proteins (myosins). In addition to the already known proteins involved in MV-mediated mineralization, new MV residents were detected, such as inorganic pyrophosphatase 1, SLC4A7 sodium bicarbonate cotransporter or sphingomyelin phosphodiesterase 3, providing additional insights into MV functions.

Sinomenine, theophylline, cysteine, and levamisole: Comparisons of their kinetic effects on mineral formation induced by matrix vesicles

The effects of sinomenine (SIN, an alkaloid extracted from the Chinese medicinal plant Sinomenium acutum used for centuries to treat rheumatic disease, including rheumatoid arthritis) on apatitic nucleation and matrix vesicle (MV)-induced mineral formation were compared with those of cysteine, levamisole, and theophylline. We found that SIN was not an inhibitor of tissue non-specific alkaline phosphatase (TNAP), a marker of biological mineralization, but confirmed that cysteine, levamisole, and theophylline were. Further, none of these four molecules directly affected the nucleation of hydroxyapatite (HA) formation, in contrast to pyrophosphate (PP(i)) which did. Incubation of 0.25-1.0mM cysteine, theophylline, or levamisole with MVs in synthetic cartilage lymph (SCL) containing AMP and Ca(2+), but not inorganic phosphate (P(i)), prolonged the induction time of mineral formation, apparently by inhibiting TNAP activity. SIN at the same levels neither inhibited TNAP activity nor affected the induction time of MV mineral formation. However, SIN did markedly delay MV-induced mineral formation in SCL containing P(i) (instead of AMP) in a manner similar to theophylline, but to a lesser extent than levamisole. Cysteine did not delay, in fact it slightly accelerated MV-induced mineral formation in Pi-containing SCL. These findings suggest that levamisole, SIN and theophylline may directly affect Ca(2+) and/or P(i) accretion during mineral formation; however, TNAP was not directly involved. The possible roles of annexins and other ion transporters, such as proteins of the solute carrier family implicated in Ca(2+) and P(i) influx are discussed.

Dimethyl sulfoxide-induced hydroxyapatite formation: a biological model of matrix vesicle nucleation to screen inhibitors of mineralization

To elucidate the inhibition mechanisms of hydroxyapatite (HA), a biological model mimicking the mineralization process was developed. The addition of 4% (v/v) dimethyl sulfoxide (DMSO) in synthetic cartilage lymph (SCL) medium containing 2 mM calcium and 3.42 mM inorganic phosphate (P(i)) at pH 7.6 and 37 degrees C produced HA as matrix vesicles (MVs) under physiological conditions. Such a model has the advantage of monitoring the HA nucleation process without interfering with other processes at the cellular or enzymatic level. Turbidity measurements allowed us to follow the process of nucleation, whereas infrared spectra and X-ray diffraction permitted us to identify HA. Mineral formation induced by DMSO and by MVs in the SCL medium produced crystalline HA in a similar manner. The nucleation model served to evaluate the inhibition effects of ATP, GTP, UTP, ADP, ADP-ribose, AMP, and pyrophosphate (PP(i)). Here 10 microM PP(i), 100 microM nucleotide triphosphates (ATP, GTP, UTP), and 1 mM ADP inhibited HA formation directly, whereas 1 mM ADP-ribose and 1 mM AMP did not. This confirmed that the PP(i) group is a potent inhibitor of HA formation. Increasing the PP(i) concentration from 100 microM to 1 mM induced calcium pyrophosphate dihydrate. We propose that DMSO-induced HA formation could serve to screen putative inhibitors of mineral formation.