Diversity of sequences encoded by the Gsp-1 genes in wheat and other grass species

The Gsp-1 genes of wheat encode two components, the “grain softness protein” (whose role in determining texture has not been substantiated) and a 15 residue arabinogalactan peptide (AGP) sequence which is O-glycosylated and is also of unknown function. We have determined genomic Gsp-1 sequences from 29 species within the Triticeae tribe and an additional 12 species from the major subfamilies of the Poaceae (Anomochlooideae, Bambusoideae, Ehrhartoideae, Chloridoideae and Panicoideae). Twelve new AGP sequence types were identified with forms present in Agropyron mongolicum, Secale cereale, Oryza sativa subsp. japonica and Sorghum bicolor containing an extra ten amino acids within the AGP sequence. Phylogenetic analysis showed distinct groupings of AGP/GSP sequence types which had no apparent relationship to the species or even the genus. However, individual forms of AGP forms were associated with specific groups of GSP sequences, providing no evidence that the AGP and GSP-1 parts of the protein have diverged at different rates or in different ways.

► The distribution of the Gsp-1 gene is extended to cover the major grass subfamilies.
► It encodes the grain softness protein and an arabinogalactan peptide, which is O-glycosylated.
► Twelve new forms of the arabinogalactan peptide, a 15 residue sequence (AGPs) are reported.
► A novel extended AGP (25 residue) form which contains a sequence repeat is also reported.
► Using phylogenetic tools we have shown that AGP motifs are linked to specific GSP-1 sequences.