Recently a theory for protein crystallization was put forward by ten Wolde and Frenkel [1, 2, 3] which if correct suggests a procedure for preparing a protein solution most conducive to crystallization. The central point of this theory is that nucleation of a crystal is enhanced in the vicinity of a metastable critical point between two metastable liquid phases (Fig. 1). The physical intuition behind this model is that the critical nucleus is wetted by the dense protein solution which lowers the surface tension between the liquid & crystal thereby reducing the free energy barrier to nucleation. Based on this model, a systematic approach to crystallization consists of first mapping the phase diagrams of various proteins as a function of temperature, ionic strength, and added polymer. These physical parameters effect the range and strength of the protein interparticle potential.

The phase diagram of a typical protein with short range attraction has a metastable liquid-liquid line (dilute protein solution co-existing with a dense protein solution) underlying a liquid crystal boundary referred to as the liquidus line (Fig. 1b). When the attraction is long-ranged (Fig. 1a) the liquid-liquid boundary becomes stable [4, 5, 6, 7, 8]. The theory predicts that when the critical point is slightly metastable (Fig. 1b) the nucleation rate of proteins would increase dramatically due to a decrease in the free energy of the nucleation barrier in the vicinity of the critical point. These conditions are ideal for would be ideal for protein crystallization; there is a high nucleation rate, and simultaneously the crystal growth rate is slow because the solution is only slightly supersaturated.

Next Page: The Depletion Model
Previous Page: The Problem of Protein Crystallization