The 10 megabases of Arabidopsis centromeric sequence behave as a single locus over evolutionary time (#2)
Eukaryotic centromeres are comprised of several megabases of short tandem repeats that adopt a higher order structure required for proper chromosomal segregation. Generally, the centromeric repeats within one species are more similar to each other than to related repeats in neighbouring species, consistent with a process of concerted evolution.
Making use of the full sequence from the 1001 Arabidopsis genome project we show that this within species homogenisation has been absolute. As a result, every centromeric repeat within the population shares a common ancestor more recently than the species origin. Essentially, while centromere position and function has been maintained between species, their underlying sequence has undergone complete replacement.
We follow the subsequent birth and spread of new repeat variants within and between centromeres using linkage disequilibrium in the natural population, and segregation in F2 crosses. This rapid sequence evolution is not easily explained by current models of repeat turnover, and has interesting population genetic consequences.