Phenotypic characterization and candidate gene analysis of a short kernel and brassinosteroid insensitive mutant from...

In an ethyl methanesulfonate oat (Avena sativa) mutant population we have found a mutant with striking differences to the wild-type (WT) cv. Belinda. We phenotyped the mutant and compared it to the WT. The mutant was crossed to the WT and mapping-by-sequencing was performed on a pool of F2 individuals sharing the mutant phenotype, and variants were called. The impacts of the variants on genes present in the reference genome annotation were estimated. The mutant allele frequency distribution was combined with expression data to identify which among the affected genes was likely to cause the observed phenotype. A brassinosteroid sensitivity assay was performed to validate one of the identified candidates. A literature search was performed to identify homologs of genes known to be involved in seed shape from other species. The mutant had short kernels, compact spikelets, altered plant architecture, and was found to be insensitive to brassinosteroids when compared to the WT. The segregation of WT and mutant phenotypes in the F2 population was indicative of a recessive mutation of a single locus. The causal mutation was found to be one of 123 single-nucleotide polymorphisms (SNPs) spanning the entire chromosome 3A, with further filtering narrowing this down to six candidate genes. In-depth analysis of these candidate genes and the brassinosteroid sensitivity assay suggest that a Pro303Leu substitution in AVESA.00010b.r2.3AG0419820.1 could be the causal mutation of the short kernel mutant phenotype. We identified 298 oat proteins belonging to orthogroups of previously published seed shape genes, with AVESA.00010b.r2.3AG0419820.1 being the only of these affected by a SNP in the mutant. The AVESA.00010b.r2.3AG0419820.1 candidate is functionally annotated as a GSK3/SHAGGY-like kinase with homologs in Arabidopsis, wheat, barley, rice, and maize, with several of these proteins having known mutants giving rise to brassinosteroid insensitivity and shorter seeds. The substitution in AVESA.00010b.r2.3AG0419820.1 affects a residue with a known gain-of function substitution in Arabidopsis BRASSINOSTEROID-INSENSITIVE2. We propose a gain-of-function mutation in AVESA.00010b.r2.3AG0419820.1 as the most likely cause of the observed phenotype, and name the gene AsGSK2.1. The findings presented here provide potential targets for oat breeders, and a step on the way towards understanding brassinosteroid signaling, seed shape and nutrition in oats.