Identification of the amino acid mutations associated with human erythrocyte spectrin alpha II domain polymorphisms

Four distinct spectrin alpha II domain polymorphisms are known to occur in several nonwhite populations. Type 1 is essentially the only form found in whites and is also the most common form in nonwhites. In contrast to most other spectrin mutations that are single-base substitutions, two of the alpha II domain polymorphisms, types 2 and 3, are particularly unusual because they appear to involve 4-Kd insertions relative to type 1. We have identified the mutations responsible for these polymorphisms using biochemical approaches and a computer database of spectrin-domain peptides separated by two-dimensional gels. The type 3 mutation is characterized by an apparent 4-Kd increase in alpha II domain peptides with no change in pI. This apparent molecular weight increase is a sodium dodecyl sulfate (SDS) gel artifact resulting from an Arg-->His mutation at residue 22 of the domain. The type 4 polymorphism shows a basic charge shift with no apparent change in molecular weight on gels. This charge shift results from a mutation of Thr-->Arg at position 174 of the domain. This mutation appears to be linked to a “silent” mutation at position 130 from an Ile-->Val. Support for possible linkage was obtained from analysis of three unrelated donors with the type 2 polymorphism. The type 2 polymorphism shows both the charge shift characteristic of the type 4 mutation and the apparent size shift that defines the type 3 polymorphism. Analysis of type 2 peptides confirmed that the two mutations described above for type 4 as well as the mutation at residue 22 observed in type 3 occur simultaneously in type 2. The observation that the type 2 polymorphism is a composite of the type 3 and 4 mutations is especially surprising because the type 2 polymorphism occurs far more frequently than either the type 3 or 4 forms. The basis for apparent linkage between the mutations at residues 130 and 174, which are encoded by different exons, is also not clear. Identification of the mutations described here permits design of genetic screening analyses that can be applied to larger populations to evaluate this potential linkage.