Isolation of isoelectrically pure cholera toxin for crystallization

Abstract

We have determined that the failure of cholera toxin to crystallize well results from its isoelectric heterogeneity, which is probably due to a post-translational process such as deamidation of its B subunit. Every sample of cholera toxin we have examined from commercial or academic suppliers has been heterogeneous; heterogeneous cholera toxin does not crystallize satisfactorily. We have overcome this problem by using ion-exchange fast protein liquid chromatography (FPLC) to obtain an isoelectrically homogeneous species of cholera toxin. Homogeneous cholera toxin crystallizes readily, forming single, nonmosaic crystals suitable for X-ray diffraction studies. For this process, protein was applied to a MonoQ ion-exchange column, then eluted with an isocratic low salt buffer followed by a linear salt gradient (0–100 mM NaCl). Column fractions were analyzed on isoelectric focusing gels, and those fractions containing the desired homogeneous species were pooled and concentrated. Crystals formed within 24 to 48 h in a MOPS/PEG buffer, which made use of slow isoelectric precipitation to induce crystallization.