Joined: May 2002
Curr Opin Genet Dev 2002 Dec;12(6):711-8
Conflict begets complexity: the evolution of centromeres.
Malik HS, Henikoff S.
Centromeres mediate the faithful segregation of eukaryotic chromosomes. Yet they display a remarkable range in size and complexity across eukaryotes, from approximately 125 bp in budding yeast to megabases of repetitive satellites in human chromosomes. Mapping the fine-scale structure of complex centromeres has proven to be daunting, but recent studies have provided a first glimpse into this unexplored bastion of our genomes and the evolutionary pressures that shape it. Evolutionary studies of proteins that bind centromeric DNA suggest genetic conflict as the underlying basis of centromere complexity, drawing interesting parallels with the myriad selfish elements that employ centromeric activity for their own survival.
Despite these difficulties, recent studies have begun to provide `evolutionary snapshots' of the centromere. They suggest that different sequence variants jockey for evolutionary dominance, even as homogeneous arrays of satellite repeats are destroyed by the insertion of a variety of mobile elements. Parallel studies of centromere-binding proteins also suggest that competition may drive the sequence complexity at centromeres, and may be responsible for rapidly changing karyotypes throughout evolution.
Female meiotic success as a major evolutionary force
Another means to turn the tables on `centromere-drive' would be to alter the meiotic tetrad at female meiosis, in effect switching the preferred position in the tetrad to an unpreferred position. One case where centromeres exploit female meiosis is evident in the relative ability of Robertsonian fusions –– when acrocentrics (chromosomes in which the centromere is towards one end) fuse at their centromeres to form a metacentric (chromosomes in which the centromere is in the middle) –– to survive female meiosis relative to its two acrocentric ancestors. In humans and chicken, Robertsonian fusions do better than acrocentrics in female meiosis, whereas the reverse is true in mice [37 and 38]; there is no difference in male meiotic transmission. A survey of karyotype evolution in mammals reveals that genomes have a high proportion of all acrocentric (e.g. mouse) or all metacentric (e.g. human) karyotypes with a distinct paucity of `mixed' karyotypes. This suggests that the switch in female meiotic `preference' has occurred frequently in mammalian evolution and can quickly reshape karyotypes once it happens ( Fig. 4). No other selective force would be expected to make such a rapid impact on karyotype evolution .
In yeast that have symmetric meioses, centromere competition is not expected to occur at all; removal of this genetic conflict may have allowed the optimal co-evolution of centromeric histones and centromeres, along with the gradual simplification of the centromeric sequences themselves. Under this model, S. cerevisiae centromeres, which are believed to consist of one nucleosome each, represent the ultimate stage of centromere optimization, whereas other genomes, including our own, constantly struggle with the consequences of unfair advantages in female meiosis.
In humans, the bias in favor of transmitting Robertsonian fusions in female meiosis has been documented [37 and 38], but Daniel et al. [48 and 49] also reiterate another dramatic effect of Robertsonians –– reduced male fertility. Among families with Robertsonian arrangements coming to prenatal diagnosis, there are 2.4 fold fewer male parent carriers compared to female parents. This is despite the fact that in their progeny there is an ~1:1 ratio of male:female transmission of Robertsonian rearrangements. This points to a significant decline in fertility in male carriers of Robertsonian fusions, compared to female carriers. This duality (i.e. increased chromosomal transmission in female meiosis offset by lowered male fertility) provides strong support for the centromere-drive model.