DEPENDENCE OF LYSOZYME GROWTH-KINETICS ON STEP SOURCES AND IMPURITIES

Interferometric microscopy was used to investigate the growth morpholo gy and kinetics of {110} and {101} faces of tetragonal lysozyme crysta ls. Solutions were prepared from as-received Sigma and Seikagaku mater ial, and Seikagaku lysozyme further purified by cation exchange liquid chromatography under salt-free conditions. The protein composition of the solutions was characterized by sodium dodecyl sulphate (SDS) elec trophoresis with silver staining. We found that on crystals smaller th an about 150 mu m, 2D nucleation sites were randomly distributed over the faces. With increasing crystal size, surface nucleation became res tricted to facet edges and, eventually, to facet corners. This reflect s the higher interfacial supersaturation at these locations. However, on some crystals, we observed 2D nucleation at preferred non-corner si tes presumably associated with defects. Upon abrupt temperature decrea ses, dislocation step sources formed on faces that previously had none . Within groups of dislocations, the dominating step source changed fr equently. Depending on the activity of the dislocation groups, growth rates of different crystals differed by up to a factor of five during the same experiment. On facets with dislocation step sources, step gen eration by 2D nucleation became dominant above a critical supersaturat ion sigma. In the absence of dislocations, nucleation-induced growth set in at sigma < sigma. In solutions with higher impurity concentrat ions, the density of the steps generated by 2D nucleation was higher a nd sigma was lower. Hence, it appears that impurity adspecies are act ive in surface nucleation. The presence of less than 1% of protein imp urities with molecular weight (MW) greater than or equal to 30 kD had significant effects on the crystallization kinetics. Step motion was i mpeded even at high sigma, presumably through blocking of kink sites. In solutions without these high MW impurities, facets containing step sources did not grow below sigma = In(C/C-sat) < 0.5. In the less pure solutions such a ''dead zone'' was not observed. Hence, it appears th at in lysozyme dead zones are caused by non-protein impurities. In gro wth from the highly purified material no growth sector boundaries were visible, in contrast to the as-received lysozyme, and striae formatio n on growth temperature changes appeared drastically reduced.