Stability and Function of Hybrid Protein Domains of Dystrophin Created by Exon Shuffling

Abstract

The dystrophin gene is the largest human gene, at ~2.5 Mbp, and one of the most complex. Its 79 exons are processed to a ~14 kbp mRNA resulting in the 427 kDa full length dystrophin. This protein is largely composed of 24 tandem copies of the spectrin-type-repeat, STR a ~110 amino acid motif that forms a triple α-helical rodshaped bundle. Tandem placement of many STR in this protein produces a long rod. The exon structure of this gene is such that every second exon junction very nearly coincides with STR junctions, while intervening exon junctions fall ~0.43 through each. Each STR is thus very nearly coded for by two exons. Recently it has been shown that dystrophin mRNA is alternatively processed in a complex, tissue dependent fashion. Many of these result in the excision of multiple exons, and the most common type of deletion observed start and end at junctions in the middle of STRs, rather than at junctions coincident with STR boundaries. If translated into protein (something that has not been demonstrated) this would result in novel, hybrid domains made up of two heterologous halves fused together. It is unclear whether such shuffling would result in viable protein structures. We have tested this in one case, reproducing an observed brain-specific deletion in a recombinant dystrophin fragment, and determining the resulting hybrid's stability by thermal and urea denaturation. We find that this hybrid structure, while slightly less stable than the native structure thermodynamically, is well folded and forms a viable and stable protein. Other measures of biological stability, including resistance to proteolysis, also show that is it stable, and in fact more so than the full-length structure.