New insights into centromere organization and evolution from the white-cheeked gibbon and marmoset

Abstract

The evolutionary history of alpha-satellite DNA, the major component of primate centromeres, is hardly defined because of the difficulty in its sequence assembly and its rapid evolution when compared with most genomic sequences. By using several approaches, we have cloned, sequenced, and characterized alpha-satellite sequences from two species representing critical nodes in the primate phylogeny: the white-cheeked gibbon, a lesser ape, and marmoset, a New World monkey. Sequence analyses demonstrate that white-cheeked gibbon and marmoset alpha-satellite sequences are formed by units of approximately 171 and approximately 342 bp, respectively, and they both lack the high-order structure found in humans and great apes. Fluorescent in situ hybridization characterization shows a broad dispersal of alpha-satellite in the white-cheeked gibbon genome including centromeric, telomeric, and chromosomal interstitial localizations. On the other hand, centromeres in marmoset appear organized in highly divergent dimers roughly of 342 bp that show a similarity between monomers much lower than previously reported dimers, thus representing an ancient dimeric structure. All these data shed light on the evolution of the centromeric sequences in Primates. Our results suggest radical differences in the structure, organization, and evolution of alpha-satellite DNA among different primate species, supporting the notion that 1) all the centromeric sequence in Primates evolved by genomic amplification, unequal crossover, and sequence homogenization using a 171 bp monomer as the basic seeding unit and 2) centromeric function is linked to relatively short repeated elements, more than higher-order structure. Moreover, our data indicate that complex higher-order repeat structures are a peculiarity of the hominid lineage, showing the more complex organization in humans.

FIG. 1.—

Schematic representation of structure and organization of alphoid sequences (modified from Alexandrov et al. 1993b). Arrows indicate single 171 bp monomers arranged in tandem stretches in (a) monomeric, (b) dimeric, and (c) trimeric structures. The similarity percentage between monomers has been reported on the arrows (see text for details).

FIG. 2.—

Examples of hybridization experiments on gibbon and human with α-PCR products. FISH experiments using probe α27/α30-NLE PCR product on NLE (A) and human metaphases (B). (C) Hybridization signals using α27/α30-HSA amplification product on human chromosomes. No signals were detected on chromosome Y indicated by an arrow. (D) Cohybridization using the clone pGamma_36 (green) and a telomeric probe (telomere PNA FISH, kit Cy3, DAKO) (red) showing colocalization of telomeric and centromeric sequences in gibbon chromosomes. (E) The figure shows the colocalization of NLE α satellite clone pGAMMA_41 with human and gibbon BAC clones that cover NLE EBs reported by Roberto et al. (2007) (see text and supplementary table S9, Supplementary Material online, for details).

FIG. 3.—

Phylogenetic analysis of marmoset α-satellite sequences. We used the Neighbor-Joining method (ClustalW) to construct the phylogenetic tree of alphoid monomers extracted from Callithrix jacchus (CJA), Papio anubis (PAN), Macaca mulatta (MMU), and Homo sapiens (HSA). We selected 417 monomeres from CJA (that cluster into 7 sets) and 10 from PAN, MMU, and HSA each. Bootstrap values (n = 100 replicates) are also indicated on the branches that support the evolutionary separation of marmoset alphoids from the human and Old World monkey α-satellite.

FIG. 4.—

Examples of FISH pattern signals on CJA metaphases. (A–G) FISH results using clones specific for each of the seven clusters, 1–7, respectively, obtained by bioinformatics approach. Clone names have been reported next to the red square. (H) Marmoset karyotype using standard Q banding according to Sherlock et al. (1996).

FIG. 5.—

Restriction analysis and southern blot on CJA genomic DNA. Hybridization of clones CJA 3.1.5 to southern blots of CJA genomic DNA digested with the indicated restriction enzymes. The HaeIII lane in the autoradiography on the right, shows hybridizing bands with a periodicity of 342 bp (the arrow shows the 342 bp monomer). We used the 2-log ladder as marker in the last lane of the gel (left) and on the first lane of the southern autoradiography (right). The sizes of the most representative bands of the marker are indicated in the middle.