Genomic Relationships Of Avena Species And Genome-wide Association Study For Hullessness In Oat Using Genotyping By Sequencing(GBS)

The genus Avena L.of the family Poaceae contains up to 30 recognized species with diploid,tetraploid.and hexaploid genomes composed of four basic 7-chromosome sub-genomes:A,B,C and D.The various wild species within the genus Avena are considered to be an important gene pool for cultivated oat,and many elite genes for pathogen resistance have been deployed from this gene pool.Further utilization of the rich genetic diversity in wild oat relatives would be facilitated by better knowledge of genome structure and relationships among these species.One of the important traits in oat is the condition of covered(hull adhering)vs hulless(naked).This trait is related to grain yield,quality and end-use.Hence,a deeper understanding of the genetics of hullessness,and the development of molecular markers associated with hullessness,have theoretical and practical value.To address these areas of priority,we first used flow cytometry to estimate the genome size of each Avena species.We then performed SNP analysis,through an array-based method and as well as through GBS,to investigate the genome structure and relationships among 27 Avena species.Following this,we used high-density GBS markers to perform a genome wide association studies(GWAS)for hullessness in oat.The primary results are summarized as follows:1.Genome size is an indicator of evolutionary distance and a metric for genome characterization.Here,we report accurate estimates of genome size in 99 accessions from 26 species of Avena.We demonstrate that the average genome size of C genome diploid species(2C = 10.26 pg)is 15%larger than that of A genome species(2C =8.95 pg),and that this difference likely accounts for a progression of size among tetraploid species,where AB<AC<CC(average 2C = 16.76,18.60,and 21.78 pg,respectively).All accessions from three hexaploid species with the ACD genome configuration had similar genome sizes(average 2C = 25.74 pg).Genome size was mostly consistent within species and in general agreement with current information about evolutionary distance among species.Results also suggest that most of the polyploid species in Avena have experienced genome downsizing in relation to their diploid progenitors.We conclude also that genome size measurements could provide additional quality control for species identification.2.A total of 120 Avena taxa representing 14 diploids,8 tetraploids and 5 hexaploids were genotyped using the Infinium 6K SNP array.To avoid ascertainment bias,the SNP data were analyzed by using normalized hybridization intensities as a continuous variable,instead of converting them into binary scores through genotype calling.For each marker,we investigated four summary statistics of hybridization intensity,including X,Y,X+Y and X/X+Y.The Mantel test showed that all distance matrices converted from various summary statistics were highly correlated with each other(p<1×104),hence we arbitrarily chose the X+Y distance matrix for further analysis.Both PCA and NJ clustering based on X+Y showed that Avena taxa with the same genomic constitutions are clustered together.The C genome diploids showed the most distinct and tight clustering relative to other species.Within the A genome diploids,the As genome was evident as the most recently evolved genome type,with the exception of A.lusitanica,and there was no significant differentiation among As genome diploids.A.lusitanica showed a close relationship with the Ac and Ad genome diploids,hence its genomic constitution should be investigated further.All AB genome tetraploids,with the exception of A.agadiriana,showed a close relationship with the As genome diploids,therefore the As genome diploids remain as the most likely candidate genome donor for AB genome tetraploids.Three AC(DC)genome tetraploids clustered tightly together,and showed close relationships with the hexaploid species.These results support the hypotheses that all AC(DC)tetraploids have a common ancestor,and that all may have been involved in the evolution of hexaploid species.3.Genomic relationships among 27 species of the genus Avena were investigated by using high-density genetic markers revealed by genotyping-by-sequencing(GBS).Two methods of GBS analysis were used:one based ontag-level haplotypes that were previously mapped in cultivated hexaploid oat(A.sativa),and one intended to sample and enumerate tag-level haplotypes originating from all species under investigation.Qualitatively,both methods gave similar predictions regarding the clustering of species and shared ancestral genomes.Furthermore,results were consistent with previous phylogenies of the genus obtained with conventional approaches,supporting the robustness of whole genome GBS analysis.Evidence is presented to justify the final and definitive classification of the tetraploids A.insularis,A.maroccana,and A.murphyi ascontaining D-plus-C genomes,and not A-plus-C genomes,as is most often specified in past literature.Through electronic painting of the 21 chromosome representations in the hexaploid oat consensus map,we show how the relative frequency of matches between mapped hexaploid-derived haplotypes and AC(DC)-genome tetraploids us.A-and C-genome diploids can accurately reveal the genome origin of all hexaploid chromosomes,including the approximate positions of inter-genome translocations.Evidence is provided that supports the continued classification of a diverged B genome in AB tetraploids,and it is conirmed that no extant A-genome diploids,including A.catnariensis,are similar enough to the D genome of tetraploid and hexaploid oat to warrant consideration as a D-genome diploid.4.In this study,hullessness in oat was investigated by sequencing 161 Chinese naked oat accessions and 20 covered oat lines using GBS,and by combining another raw GBS data set including 633 oat lines primarily from North America and Europe.PIC results showed that the oat lines from North America have the richest genetic diversity,whereas the oat lines from Europe have the lowest genetic diversity followed by oat lines from China.PCA and STRUCTURE analysis revealed the presence of three main sub-populations,which roughly corresponded to China,Europe and central and northern North America,southern North America and Oceania.The genetic diversity distribution and population structure are highly related to the speculative oat evolutionary pathway as well as to geographic centres of modern plant breeding.GWAS were performed using a general linear model with correction for population structure.Two robust markers were detected that were highly associated with hullessness,with-logP of 13.09 and 13.05 respectively.Both markers have been mapped on Mrg21(205 and 212 cM)with 7cM distance between them,and are close to the marker cdo482(199.2cM)which was lined to the N1 locus in a biparental population analysis.