Dentin matrix protein-1 (DMP1) is a mineralized tissue matrix protein synthesized by osteoblasts, hypertrophic chondrocytes, and ameloblasts as well as odontoblasts. DMP1 is believed to have multiple in vivo functions, acting both as a signaling molecule and a regulator of biomineralization. Using a cell-free system in vitro, we evaluated the action of DMP1 in the regulation of hydroxylapatite (HA) formation and crystal growth. The non-phosphorylated recombinant protein acted as an HA nucleator, increasing the amount of mineral formed in a gelatin gel HA growth system relative to protein-free controls. The recombinant protein phosphorylated in vitro had no detectable effect on HA formation and growth. In contrast, phosphorylated bovine DMP1 expressed in marrow stromal cells with an adenovirus vector containing 29.7 phosphates/mol was an effective inhibitor of HA formation and growth. The native full-length protein appeared to be absent or present in only small amounts in the extracellular matrix of bones and teeth. However, two highly phosphorylated fragments representing the N-and C-terminal portions of DMP1 have been identified, apparently arising from proteolytic cleavage of four X-Asp bonds. The highly phosphorylated C-terminal 57-kDa fragment (containing 42 phosphates/mol), like the non-phosphorylated DMP1, was an HA nucleator. These data suggest that, in its native form, DMP1 inhibits mineralization, but when cleaved or dephosphorylated, it initiates mineralization. These in vitro data are consistent with the findings in the DMP1 knockout mouse.Dentin matrix protein-1 (DMP1) 1 is an acidic, phosphorylated, integrin-binding extracellular matrix protein first identified by screening a rat cDNA library (1). Northern blot analysis in the rat originally suggested that the DMP1 message was odontoblast-specific, with in situ hybridization showing DMP1 mRNA expression to be restricted to those fully differentiated odontoblasts engaged in active dentin matrix formation, with transient expression by ameloblasts (1, 2). More recent in situ hybridization studies show a much broader pattern of expression of DMP1, with expression associated with a number of mineralizing tissues, including bone and cementum (3). In rat and chicken bone, immunohistochemical detection showed an association of DMP1 with osteocytes, but not with osteoblasts (4). In the mouse, hypertrophic chondrocytes express more DMP1 than other cell types (5). DMP1 was detected at high levels by Northern analysis in fetal bovine brain and cultured long bone as well as in odontoblasts (6). Most recently, DMP1 was localized by immunohistochemistry in human lung cancer tissue (7).DMP1 is a small protein that has been postulated to have a high affinity for hydroxylapatite (HA) (8,9). This coupled with its localization in mineralized tissues suggested that DMP1 could have an important role in mineralization. DMP1 is a member of the SIBLING (small integrin-binding ligand, Nlinked glycoprotein) family of proteins (10, 11). Like other members of the SIBLING family...

Sialoproteins such as bone sialoprotein (BSP) and dentin sialoprotein (DSP) accumulate at the mineralization fronts in bone and dentin, respectively, suggesting they have some function in the mineralization process. BSP, a highly phosphorylated protein rich in polyglutamate repeats, is an effective nucleator of hydroxyapatite (HA) formation in vitro. The present study examines the effect of DSP, a low phosphorylated but related sialoprotein, on the formation and growth of HA. In vitro, in a gelatin gel diffusion system, DSP at low concentrations (<25 microg/ml) slightly increased the yield of HA formed at 3.5 and 5 days, while at higher concentrations (50-100 microg/ml) it slightly inhibited accumulation. Fewer mineral crystals were formed in the presence of high concentrations of DSP but they tended to aggregate (making them appear larger by electron microscopic analysis) than those formed in DSP-free gels. X-ray diffraction line broadening analysis failed to show significant changes in c-axis crystal dimensions with increasing DSP concentration. When HA-seed crystals were coated with DSP before inclusion in the gelatin gel there was a reduction in mineral accumulation relative to HA-seeds which had not been coated with DSP, but the extent of inhibition was significantly less than that seen in this system with other mineralized tissue matrix sialoproteins, such as osteopontin or BSP. The low affinity of DSP for well-characterized seed crystals and the limited effect of this protein on HA formation and growth suggest that the role of DSP in dentin is not primarily that of a mineralization regulator.