Effects of Impurities on Calcium Oxalate Crystallization as Measured by Atomic Force Microscopy
Abstract
Calcium oxalate crystals are found in kidney stones as either calcium oxalate monohydrate (COM) or calcium oxalate dihydrate (COD). COM crystals are the most abundant form as they are thermodynamically more stable than COD crystals under physiological conditions. Certain aspartic acid-rich molecules such as osteopontin (OPN) are known to affect stone formation by inhibiting COM and COD growth. We have studied COM {010} and COD {100} faces in the presence of OPN, poly-aspartic acid (poly-ASP) and synthetic peptides derived from OPN to investigate the inhibitor mechanism.
We observed that poly-ASP preferentially inhibits one particular direction of {010} faces on COM crystals, creating “finger-like” features, while growth continues in other directions. We find that these features appear at a threshold concentration of poly-ASP that varies with the length of the polymer. Attempts to model the threshold concentration suggest that this is due to a competition between adsorption of polymer to already-pinned steps and removal of poly-ASP by desorption and possibly coverage by advancing steps between strongly pinned steps.
Growth of {100} faces of COD crystals in the presence of synthetic peptides derived from OPN with different iso-electric points displayed varying levels of inhibition. The two principal directions of the elongated elliptical hillocks on {100} faces were inhibited to different degrees by the peptides. This preferential inhibition is most likely the result of different Ca2+ ion densities and structures of the two different step directions.
The effects of imaging on the growth itself were studied for the case of COM growth in the presence of OPN. In the presence of OPN molecules, a continuously-scanned area showed faster growth relative to a larger area scanned less frequently. This enhanced vertical growth in the continuously-scanned area was found to vary with the OPN concentration and calcium oxalate supersaturation. We modeled this effect using a modified version of the Cabrera-Vermilyea model (1958), in which steps are pinned by adsorbed impurities, resulting in a decreased propagation speed due to the resulting curvature. In our case, we assume adsorption directly to steps, rather than to terraces as in the Cabrera-Vermilyea model.