Transcriptome Assemblies in Polyploid Spartina (Poaceae, Chloridoideae), Duplicate Gene Detection and Molecular Divergence Using Next-Generation Sequencing

Spartina species (Poaceae, Chloridoideae) are perennial grasses exhibiting high ploidy levels (from 4x to 12x), making genomic analyses and gene orthology search challenging, particularly when diploid ancestral genomes are not available. Notorious recent hybridization and genome duplication had important evolutionary consequences in Spartina. Hybridization between the hexaploid S. maritima (2n= 60, European Atlantic Coasts) and the hexaploid S. alterniflora (2n=62, introduced from the American East Coast) in the early 19^th century, resulted in two F1 sterile hybrids, Spartina x townsendii (in England) and Spartina x neyrautii (in France). Genome duplication of S. x townsendii gave rise to the highly fertile and invasive allopolyploid Spartina anglica(2n=12x=120,122,124).

Genomic resources are lacking in Spartinaand more generally in the subfamily Chloridoideae. We seized the opportunity of NGS to build reference transcriptomes of these 5 species of interest to study the evolutionary consequences of hybridization and multiple genome duplications. We developed a bioinformatics approach to assemble and co-assemble Roche-454 and Illumina data, obtaining a total of 110 423 annotated contigs for the 5 species (37 867 non-redundant contigs). Sequences were co-aligned to detect SNPs and different putative duplicates were detected within 286 937 contigs. This method allowed us to search for homoeolog sequences using homoeoSNPs (which discriminate homoelogous sequences) and to evaluate the divergence between duplicated genes.

These results now allow us to identify duplicated copies at any ploidy level in Spartina and to explore the fate of duplicate genes in these complex genomes.