A- and B-subunit variant distribution in the holoprotein variants of protein toxin abrin: variants of abrins I and III have constant toxic A subunits and variant lectin B subunits.

The cytotoxic lectin abrin shows more than 30 variant forms (R. Hegde, T. K. Maiti, and S. K. Podder, 1991, Anal. Biochem. 194, 101-109). The lectin B subunit as cause for variance in abrins I and III was detected by a combination of one- and two-dimensional electrophoresis and Western blotting. Intriguingly, in abrin I but not in abrin III, association of a single A subunit with the variant B subunits shifts the holoprotein pI toward the alkaline side indicating that the subunit association involves neutralization of few negative charges. The B-subunit variants of abrins I and III overlap in their pI, and the A-subunit association gives the holoproteins a distinctness on isoelectric focusing gel. The results were also confirmed by analyzing the pH titration curves. These differences in the subunit association pattern between abrins I and III are in corroboration with the previously observed differences in the kinetics of protein synthesis inactivation and accessibility of the disulfide bridge to reducing agents in the presence or absence of putative receptor (R. Hegde, A. Karande, and S. K. Podder, 1993 Eur. J. Biochem. 215, 411-419). Further, the genetic origin of variance was confirmed by peptide mapping of the individual subunit variants. Considering a theoretical value of 0.1 to 0.2 pI/charge, a 15-17 charge difference could be predicted between the variants of two extreme pIs. The fact that the A subunits are not shared between the groups was taken to interpret that the protein synthesized as prepro form is processed posttranslationally and the processing takes place only after the disulfide bond formation between A and B subunits. The N-terminal 16 amino acids of A subunits of abrins I and III showed 26% dissimilarity. The A subunits of abrins I and III did not react with concanavalin A, indicating that the heterogeneity in the molecular weight is because of differential processing but not because of glycosylation.