Evolutionary-new centromeres preferentially emerge within gene deserts

Abstract

Background: Evolutionary-new centromeres (ENCs) result from the seeding of a centromere at an ectopic location along the chromosome during evolution. The novel centromere rapidly acquires the complex structure typical of eukaryote centromeres. This phenomenon has played an important role in shaping primate karyotypes. A recent study on the evolutionary-new centromere of macaque chromosome 4 (human 6) showed that the evolutionary-new centromere domain was deeply restructured, following the seeding, with respect to the corresponding human region assumed as ancestral. It was also demonstrated that the region was devoid of genes. We hypothesized that these two observations were not merely coincidental and that the absence of genes in the seeding area constituted a crucial condition for the evolutionary-new centromere fixation in the population.

Results: To test our hypothesis, we characterized 14 evolutionary-new centromeres selected according to conservative criteria. Using different experimental approaches, we assessed the extent of genomic restructuring. We then determined the gene density in the ancestral domain where each evolutionary-new centromere was seeded.

Conclusions: Our study suggests that restructuring of the seeding regions is an intrinsic property of novel evolutionary centromeres that could be regarded as potentially detrimental to the normal functioning of genes embedded in the region. The absence of genes, which was found to be of high statistical significance, appeared as a unique favorable scenario permissive of evolutionary-new centromere fixation in the population.

Figure 1

The phylogenetic relationships of the species under study. Data on OWMs and Hominoidea are from Raaum et al. [22], while those on NWMs are from Schneider et al. [24].

Figure 2

FISH examples. (a) Examples of FISH experiments using species-specific BAC clones yielding duplicated signals around the centromere. The CH250 and CH271 are BAC libraries specific for macaque and gibbon, respectively. The DAPI-stained chromosome without the signal is reported on the left to better show the morphology of the chromosome. (b) FISH experiment using the BAC clone CH250-417O7 (MMU2) on a macaque metaphase, showing pericentromeric signals on several chromosomes.

Figure 3

Gene density simulations. The observed density of (a) genes (Refseq), (b) Refseq exons and (c) expressed sequence tag (Est) exons within the corresponding region of the 14 ENCs were compared against a simulated set of 10,000 regions distributed randomly within the human genome (see Materials and methods). A significant depletion of exons and genes was observed.