Background: Alpha-gliadins form a multigene protein family encoded by multiple alpha-gliadin (Gli-2) genes at three genomic loci, Gli-A2, Gli-B2 and Gli-D2, respectively located on the homoeologous wheat chromosomes 6AS, 6BS, and 6DS. These proteins contain a number of important celiac disease (CD)-immunogenic domains. The alpha-gliadins expressed from the Gli-B2 locus harbour fewer conserved CD-epitopes than those from Gli-A2, whereas the Gli-D2 gliadins have the highest CD-immunogenic potential. In order to detect differences in the highly CD-immunogenic alpha-gliadin fraction we determined the relative expression level from the homoeologous Gli-2 loci in various tetraploid and hexaploid wheat genotypes by using a quantitative pyrosequencing method and by analyzing expressed sequence tag (EST) sequences.
Results: We detected large differences in relative expression levels of alpha-gliadin genes from the three homoeologous loci among wheat genotypes, both as relative numbers of expressed sequence tag (EST) sequences from specific varieties and when using a quantitative pyrosequencing assay specific for Gli-A2 genes. The relative Gli-A2 expression level in a tetraploid durum wheat cultivar ('Probstdorfer Pandur') was 41%. In genotypes derived from landraces, the Gli-A2 frequency varied between 12% and 58%. In some advanced hexaploid bread wheat cultivars the genes from locus Gli-B2 were hardly expressed (e.g., less than 5% in 'Lavett') but in others they made up more than 40% (e.g., in 'Baldus').
Conclusion: Here, we have shown that large differences exist in relative expression levels of alpha-gliadins from the homoeologous Gli-2 loci among wheat genotypes. Since the homoelogous genes differ in the amount of conserved CD-epitopes, screening for differential expression from the homoeologous Gli-2 loci can be employed for the pre-selection of wheat varieties in the search for varieties with very low CD-immunogenic potential. Pyrosequencing is a method that can be employed for such a 'gene family-specific quantitative transcriptome profiling'.