Coffee Trees Genome Size Variation: Its Relation With Retrotransposons Copy Number

A de Kochko, A Boualem, J Sarry, V Poncet, P Hamon and M Noirot

Génomique des Caféiers, UMR DGPC, IRD, Montpellier, France
Back to Kews Plant Genome Size Meeting Papers

Abstract


The genus Coffea contains over 80 species and taxa, among which only C. arabica is allotetraploid. The genome size varies from 2C = 1.03 to 1.76 pg, with a constant chromosome number, 2n = 22. The genome size varies globally, from East Africa, with the smallest genomes, towards Central and West Africa where the bigger genomes are found. Coffea species supposedly originated in East Africa. A hypothesis states that genome size differences resulted from an increase due to the presence of different types and copy numbers of repeated sequences among which retrotransposons. Gene duplication is also to be considered, but it may not be responsible for such a wide difference among species. Duplication of long chromosomal regions, instead of single genes, is also possible. Reduction of genome size cannot be discounted but it is a more complicated phenomenon, with more dramatic consequences which can hardly account for the present situation in the genus.

Changes in genome size are certainly accompanied by speciation, as the success of interspecific crosses is correlated with the difference in the genome size of the parental species. We are currently analyzing repeated sequences present in these genomes. We are estimating the copy numbers of different retrotransposons already identified. Paradoxically, one of the biggest genomes, that of C. heterocalyx (2C = 1.73 pg), has an apparently low copy number of the gypsy/Ty3 retrotransposon type, lower than that of C. pseudozanguebariae, which has a small genome (2C = 1.13 pg). The copia/Ty1 retrotransposon type seems to be more frequent than the gypsy/Ty3 type in Coffea genomes. Nevertheless, increase of genome sizes may depend not only on the copy number of retrotransposons but also on their genomic ubiquitous presence, which might be at the origin of non-reciprocal recombination due to the recognition of homologous sequences dispersed all over the genome.