Noncollagenous matrix proteins (NCPs) of dental hard tissues (dentin, cementum) are involved, both temporally and spatially, in the mineralization of their collagen matrices. Two of the NCPs thought to initiate mineral nucleation and control crystal growth in dentin, are dentin phosphoproteins (DPP) and dentin sialoprotein (DSP). Control of crystal growth would depend on the binding capacity of these two molecules, which may be related to the charge domains on the crystals and/or the phosphorylation of the protein. Phosphophoryn (a highly phosphorylated DPP) and DSP were isolated, purified, and characterized from the immature root apicies of human teeth. Dephosphorylation of phosphophoryn was carried out using bovine intestinal alkaline phosphatase. Enamel crystals were prepared from the maturation stage of developing rat incisor enamel. Protein-coated crystals were prepared for viewing in an atomic force microscope fluid cell using tapping mode. Desorption of the proteins was achieved using a phosphate buffer and surface roughness measurements were obtained from all specimens. Time-lapsed images of the crystals showed "nanospheres" of protein distributed along the crystals but only the phosphophoryn-coated crystals showed a distinctive banding pattern, which was still visible after the phosphate desorption experiments. The surface roughness measurements were statistically greater (P <0.01) when compared to the control for only the phosphophoryn-coated specimens. It is hypothesized that the phosphophoryn binding may be associated with charge arrays on the crystal surface and its phosphorylation. Also, based on its affinity for the crystalsurfaces, phosphophoryn seems the most likely candidate for controlling dentin crystal growth and morphology.

Dental fluorosis is an increasing problem, yet the precise mechanism by which fluoride exerts its effects remains obscure. In the present study, we have used atomic force microscopy to image and quantitate surface features of enamel crystals isolated from specific developmental stages of fluorotic and control rat incisors. The results showed a significant decrease in crystal surface roughness with development in control tissue. Crystals from fluorotic tissue were significantly rougher than controls at all stages of development, did not decrease in roughness during the later stages of their development and had many morphological abnormalities. These data clearly demonstrate an effect for fluoride on enamel crystal surfaces which could reflect changes in the nature and distribution of growth sites and/or in mineral–matrix interactions. These would be expected to affect crystal growth during maturation, resulting in the characteristic porous appearance of fluorotic lesions in mature teeth.