W0067

An Alternative to Sparse Matrix Screens. Madeleine Riès-Kautt, L.E.B.S. C.N.R.S Bat. 34. F-91198 Gif-sur-Yvette Cedex, France.
email: ries@lebs.cnrs-gif.fr

Using a sparse matrix screen is checking if a crystallization condition which has already produced crystals with a different protein could work with a new protein. The strategy proposed here is based on physico-chemical considerations, consideringthe net charge of the protei at the pH of crystallization andthe counter-ions , i.e. the behaviour of the protein salt [1].
The search for crystallization conditions is stepwise : 1) estimation of the net charge over a large pH range, and choice of the pH value(s) of crystallization, 2) search of thelimit of nucleation/precipitation by rapid tests to locate therange of crystallizing agent to focus on, 3) refinement of conditions by small increases in the concentration of crystal-lizing agent in order to promote crystals, or at least any solid phase. The nature of crystallizing agents are chosen depending on the protein net charge, and 4) adjustment of the conditions where some solid phase appears (spherulites, micro-crystals, etc.) to obtain large single crystals.
Crystallization of proteins having a positive net charge at the pH of crystallization have been tested with lysozyme and BPTI. It has been used for the crystallization of snake venom toxins[2], of turkey lysozyme[3], and for the optimizationof lysin of red abalone crystals [4].
Crystallization of proteins having a negative net charge was been developed with collagenase from Hypoderma lineatum[5]. It was confirmed for amylase and parvalbumin, and allowed the crystallization of Grb2 [6].
Proteins having a net charge of about zero are at their lowest solubility compared to a higher net charge. The effect of salts on the solubility of a protein near its pI is illus-trated by carboxyhemoglobin[7]. In our hands the effect of ionic strength is poor, and crystallization is more successfulwith organic crystallizing agents, such as PEGs and MPD.However it is worth testing salts because they can act by ionbinding, the net charge then becomes different from zero.
1. Riès-Kautt M. and Ducruix A. Methods in Enzyology 1997, Chap 3, pp 23-59.
2. Ménez R, Ducruix A. J Mol Biol 1993, 232 :997.
2. Howell P.L. Acta Cryst. 1995, D51:654.
3 Diller et al. Acta Cryst. 1994, D50:620.
4. Carbonnaux C, Riès-Kautt M, Ducruix A:. Protein Sci 1995, 4 :2123.
5. Guilloteau JP et al. Proteins: Structure, Function, and Genetics 1996, 25:112.
6. Green AA. J Biol Chem 1932, 95:47.