IUCr journals

Unusual layered growth of catalase crystals

Atomic force microscopy (AFM) has made visible the most intimate processes of crystallization. A tiny silicon nitride micro-needle hooked to a transducer and drawn with microtome-precision and feather-weight across the face of a growing crystal can produce breathtaking images that record crystal growth at or near molecular resolution. Some of the finest examples of AFM analyses applied to questions of protein crystal growth have come from collaborations of A. Malkin and A. McPherson at UC Irvine. These include studies of lysozyme and thaumatin in which screw dislocations are imaged on growing surfaces, growth-step edges are typically one unit cell in height, and construction of individual unit cells appears to be completed before the construction of new unit cells is begun.

[beef liver catalase]Crystal growth surface of beef liver catalase.
In a paper by Ko, Day, Malkin and McPherson, “Structure of orthorhombic crystals of beef liver catalase” [Acta Cryst. D55 (1999), 1383-1394], quite a different set of observations is described for the growth of catalase crystals. The observations were made on an orthorhombic crystal form of beef liver catalase rather than the original trigonal one. Salt precipitation was substituted for polyethylene glycol to remove complications to the AFM experiment without loss of isomorphism. The orthorhombic crystal structure was solved with some difficulties presented by multiple pseudo-symmetries. The unusual observations related to growth on a <110> growth face are the following. Surprisingly, no screw dislocations are observed for growth in the c direction and the face develops exclusively by two-dimensional propagation perpendicular to c. The height of the propagating layer, the growth step, is one-half the c-axis length, meaning that the growth step corresponds to one-half a unit cell, in this case two of the four asymmetric units. Islands of new half-unit-cell layers nucleate and propagate on the top of other layers, but the new layers have the opposite hand of the layer on which they grow. This can be seen in the figure above. Most of the island layers show a straight side and a curved side, but the sides alternate left and right between successive layers. These observations and others are discussed in the light of the crystal structure of the orthorhombic form.

One final note. A comparison of the molecular structures of catalase in trigonal and orthorhombic crystals shows that, while lattice interactions may have a profound effect on how crystals develop, they have little influence on the details of the structure of the constituent molecules.

Howard Einspahr
Bristol-Myers Squibb Pharm. Res. Inst., Princeton, NJ, USA