Effects of urinary macromolecules on hydroxyapatite crystal formation

Abstract

Particle size analysis was combined with titration data obtained in constant-composition, hydroxyapatite (HA)-seeded, crystal growth assays. With addition of large amounts of HA (250 μg), titration rates were linear, new crystal formation was minimal, and aggregation effects could be detected. With addition of small amounts of HA (62.5 μg), nucleation of new HA was observed. The effects of urinary macromolecules, i.e., osteopontin (OPN), recombinant glutathione-S-transferase-OPN (G-OPN), Tamm-Horsfall protein, chondroitin sulfate, human serum albumin, mixed urinary macromolecules from a stone-former (SFUI), mixed urinary macromolecules from a normal individual (NUI), and polyaspartic acid (PA), were examined in this system. Crystal growth inhibition, as measured by the slope of linear titration curves in this system, was observed with PA, G-OPN, OPN, SFUI, and NUI. All of the macromolecules tested inhibited aggregation, including Tamm-Horsfall protein, which did not inhibit growth. As reflected by the ratio of the final number of particles to the initial number in the 62.5-μg seed addition, the macromolecules that were most effective in inhibiting growth, i.e., OPN, G-OPN, PA, SFUI, and NUI, actually increased secondary nucleation. Recombinant G-OPN demonstrated less inhibitory activity than did OPN isolated from cell culture. Chondroitin sulfate and human serum albumin exhibited no significant effects on the various components of HA crystallization under these conditions. SFUI and NUI slowed growth and increased secondary nucleation to similar degrees, and neither exhibited any measurable effect on aggregation. Therefore, crystal surface sites that participate in nucleation, growth, and aggregation processes are affected independently by macromolecules, presumably because of differences in their structural features. These results illustrate the utility of combining these techniques to provide a much greater understanding of crystallization behavior than that possible with either analysis alone.