Centromere structures highlighted by the 100%-complete Cyanidioschyzon merolae genome

Abstract

Centromere dynamics are largely unknown in lower plants (algae). We have recently identified the centromere-specific histone H3 variant (CENH3) and clarified the dynamic centromere rearrangement at mitosis in the primitive red alga Cyanidioschyzon merolae. We also showed that the CENH3-containing nucleosomes constituted the kinetochore closely interacting with the nuclear envelope. CENH3 visualization during the whole cell cycle suggests that C. merolae centromeres are monocentric and confined to specific loci. We completed 100% no-gap telomereto-telomere sequencing of the C. merolae genome. Interestingly, a single A+T-rich region has been identified on each fully sequenced chromosome. No centromere-like A+T-rich repetitive sequence have been found within these regions, implying that the C. merolae centromeres may be ‘point’ centromeres, or be comprised of nonrepetitive heterogeneous DNA sequences.

Centromere function is evolutionarily conserved in almost all eukaryotes. It is known that centromeric DNAs undergo rapid evolution and have no obvious constraints on their sequence conservation. However, several centromere proteins are conserved at the domain and motif level, suggesting that key protein-protein interactions, retained through centromere evolution, might allow for the functional conservation and DNA sequence diversity of the centromere. Most prominent are centromere-specific histone H3 (CENH3) family proteins, because the histone fold domain of this family is well conserved among all the eukaryotic lineages to assemble the centromeric nucleosomes with other conventional histones.¹

We previously clarified the centromere movement and reconstitution during the cell cycle by tracing the CENH3 in the ultrasmall unicellular red alga Cyanidioschyzon merolae. On the relationship between the kinetochore and the nuclear envelope (NE), which had been poorly understood in red algae, we demonstrated using electron microscopy that they are closely associated at mitosis. Given the cellular characteristics that the chromosomes barely condense and the NE remains intact throughout the cell cycle in C. merolae, we postulate that this kinetochore-NE interaction might produce a ‘signal’ until the uncondensed and lagged chromosome arm regions have been completely segregated and the tension on the NE has been attenuated. Visualization of CENH3 proved that the C. merolae chromosomes are not holocentric (entire chromosomes serve as centromeres) but are likely to be monocentric (one ‘regional’ or ‘point’ centromere occurs on each chromosome).²

Previous C. merolae genome sequencing showed that several chromosomes possess single A+T-rich regions, which are annotated as putative centromeric regions. However, there were many gaps in the whole genome assembly that had not been fully sequenced, and a full picture of the putative centromeric regions was unclear.³ Recently, we finished the 100% complete genome sequencing and obtained all the 20 chromosome assemblies as full telomere-to-telomere sequences.⁴ Although generally the rDNA regions are highly repetitive and structurally unstable in most eukaryotes, the C. merolae complete genome sequence included the complete set of three ‘static’ singlet rDNA clusters scattered across different chromosomal loci, which is one of the most distinguishing structural characteristics.⁵

Figure 1 shows the overall G+C content distribution on the fully sequenced chromosomes. It is interesting to note that a single A+T-rich region is present on each chromosome. We postulate that the single A+T-rich regions are something more than just stochastic distribution, and are likely to play some role in the maintenance of chromosome structure, since these regions show essentially a one-on-one relationship with chromosomes. Although less clear, the genome sequence of the unicellular green alga Ostreococcus tauri (Prasinophyceae) similarly shows a biased A+T distribution pattern.⁶ To determine whether the A+T-rich regions are associated with repetitive sequences like most centromeric regions in other species, we employed Tandem repeats finder, a program used to search for repeat sequences.⁷ However, we have not found any repeat sequence, any common pattern or any particular rule concerning the size, A+T% or sequence of these regions.

Figure 1.

G+C content and distribution on the Cyanidioschyzon merolae chromosomes. Arrowheads indicate the positions of the putative centromeric regions.

Several pioneering works provide useful information on the centromere structures in lower eukaryotes. The centromeres of the human malaria parasite Plasmodium falciparum are composed of extremely A+T-rich repetitive elements.⁸ In the trypanosome parasites, the Trypanosoma brucei chromosomes possess A+T-rich repeats within the centromeric region as identified by etoposide-mediated topoisomerase-II cleavage analysis, while these A+T elements are not found in T. cruzi.⁹ It is also important to note that centromeres in Candida albicans are all comprised of different and unique DNA sequences, and are maintained by an epigenetic mechanism.^10,11

We presume that monocentric C. merolae chromosomes are unlikely to possess ‘regional’ centromeres composed of A+T-rich centromeric repeats, but rather the centromere structure is similar to ‘point’ (approximately 120 bp) centromeres in Saccharomyces cerevisiae.¹ Alternatively, the functional C. merolae centromeres might be non-repetitive, heterogeneous DNA elements, lacking in inter-chromosomal sequence similarities. With the increasing numbers of lower plant genome sequences that are available, comparative analysis of centromeric sequences will help to elucidate the evolution of centromere DNA-protein interactions in the plant kingdom.

Addendum to: Centromere dynamics in the primitive red alga Cyanidioschyzon merolae. Maruyama S, Kuroiwa H, Miyagishima SY, Tanaka K, Kuroiwa T. Plant J. 2007;49:1122–1129.

and

Addendum to: A 100%-complete sequence reveals unusually simple genomic features in the hot-spring red alga Cyanidioschyzon merolae. Nozaki H, Takano H, Misumi O, Terasawa K, Matuzaki M, Maruyama S, Nishida K, Yagisawa F, Yoshida Y, Fujiwara T, Takio S, Tamura K, Chung SJ, Nakamura S, Kuroiwa H, Tanaka K, Sato N, Kuroiwa T. BMC Biol. 2007;5:28.