Sequence Assembly, Evolution and Functional Application of the Cotton Genomes

We used the highly homozygous allotetraploid G. hirsutum (AD₁, 2n = 52) genetic standard line TM-1 for DNA sequencing. To successfully assemble the AD genome, we combined 180-fold whole-genome shotgun (WGS) sequences with a total of 100,187 complete BAC-to-BAC sequences . We also constructed a high-density genetic map using 167 RIL individuals derived from a cross between TM-1 and the allotetraploid G. barbadense (AD₂, 2n = 52) genetic standard line 3-79. Prior to that, we had successfully assembled G. arboreum (A₂, 2n = 26) and G. raimondii (D₅, 2n = 26), two diploid descendants of the probable AD subgenome progenitors. Repeat sequences accounted for 68.5% of the A genome, 57% of the D genome, and 67.2% of the AD genome. The tetraploid genome displayed high degrees of conserved gene orders with the two diploids. Insertions of long terminal repeats (LTR), especially Gypsy-typeretrotransposons, seemed to be responsible for the two-fold size differences between the D (775.2 Mb) and A (1,694 Mb) genomes. Transposable elements (TEs) originated from D subgenome are more active than those from A subgenome. In-depth genome-wide analysis helped to resolve the puzzle with regard to gossypol biosynthesis and ethylene production in various cottons. The expression patterns of a large number of functional genes, especially both primary cell wall-specific and secondary cell wall-specific CesA or CesA-like (CSL) genes, are significantly different in the two subgenomes, suggesting that genome variation, caused likely by TE, especially by retrotransposon insertion, may play a crucial role during the concerted evolution of cotton genomes.

Key words: cotton genomes, allotetraploid, retrotransposons, genome evolution, functional studies