The centromere is a specialized locus of each chromosome consisting of a special chromatin structure which promotes the assembly of the kinetochore (the protein structure that serves as an interface between the microtubules and the chromosome). Therefore, the centromere/kinetochore is a fundamental piece of the chromosome segregation apparatus. In budding yeast, the centromere is called a point centromere because it is specified by a sequence of only 125 base pairs, while in most eukaryotes the centromeric region spans megabases of DNA, and is defined by repetitive DNA sequences that lack such sequence specificity.
The specific properties of the centromeric region are conferred by a specific chromatin configuration; in the point centromere of budding yeast the centromeric chromatin is anchored by a single nucleosome that contains a specific variant of histone H3 called Cse4 (called CENP-A or CenH3 in other eukaryotes). Although the budding yeast point centromere is composed of just one specific histone, it nucleates the cohesin complex binding for at least a 4 kb region around centromeres, called pericentric chromatin, and also facilitates the assembly of inner kinetochore (CBF3) complexes that extend over several nucleososmes beyond the Cse4 nucleosome. In fact, it has been proposed that centromeres are “signaling hubs” that influences the chromatin configuration over a much larger region.
Even given their obvious differences, the overall structure of the centromere/kinetochore complex is relatively conserved, being the budding yeast a simplified version, able to attach only one single microtubule. It has been visualized in mitosis that centromeric regions loop out from chromosome axes. Pericentric cohesins are important for the conformation of these loops, probably not through its “conventional” role of linking the two sister chromatids but through the formation of intramolecular links. Although there are no studies yet in meiosis about this structure, the centromeric region is also enriched in cohesin in meiosis.
Besides associating with the microtubules through the kinetochore, centromeres can also form associations with other centromeres in a still not well characterized way. All experiments performed on this thesis try to provide a better understanding of the centromeric associations that take place in meiotic prophase. In a purely descriptive way, when centromeres are observed in meiotic prophase, they are usually either forming clusters or pairs.
Most species start meiosis with the centromeres clustered, forming what is known as the “Rabl” orientation. The name is after Carl Rabl, who in 1885, looking at interphase chromosomes of salamander larvae, observed that the centromeres were all clustered toward the periphery of the nucleus.
The first description of a pairwise association dated from 1976, from the observation of the onion Allium fistulosum. In this organism centromeres can be easily visualized at all stages of the meiotic cell cycle using the electron microscope. A carefully characterization of centromere distribution showed that the centromeres form associations of different natures at different stages of prophase:
Premeiotic S-phase: single centromeres and associations composed from 2 to about 7 centromeres. (Allium fistulosum have 8 pairs of chromosomes).
Zygotene: single centromeres or associations composed of 2 centromeres.
Pachytene: associations composed of 2 (they conclude that the association is between homologous centromeres since the SC traverses the centromeres uninterrupted).
Subsequent studies of wheat and budding yeast suggested that the formation of centromeric associations is a conserved phenomenon showing for the first time that non-homologous centromeres are able to form pairwise associations in early prophase.
The end of meiotic prophase is characterized by the SC disassembly, causing de-synapsis of the homologous chromosomes, which had been traditionally believed to remain linked only at the sites of crossing over (chiasmata). However, a persistence of the pairing of homologous centromeres after pachytene, partially dependent on SC components, has been recently shown in a variety of organisms like budding yeast, Drosophila and mouse.
Therefore, at least conceptually, we can differentiate 4 different centromere conformations during meiotic prophase:
1. Centromeres forming clusters in or before pre-meiotic S-phase.
2. Centromeres forming pairwise associations, mostly between nonhomologous chromosomes (thereafter called centromere coupling) in early prophase.
3. Centromeres forming pairwise associations between homologous chromosomes (as well as the rest of the chromosome) during pachytene synapsis.
4. Centromeres forming pairwise associations between homologous chromosomes at the end of meiotic prophase, after SC disassembly (thereafter called centromere pairing).
Most species start meiosis with the Rabl orientation. The centromeres are clustered next to the microtubule nucleating center, the SPB in yeast, while the arms extend and the telomeres are attached to the nuclear envelope at the opposite pole.
This configuration is believed to be a remnant of the last mitotic division, in which all the centromeres have been pulled together towards the SPB while the rest of the chromosome and the telomeres are dragging behind. However, the maintenance of this polarization seems also to be mediated by microtubule/kinetochore interactions, at least in interphase cells, since the disruption of either of the two creates a more dispersed centromeric phenotype. It makes sense to think that while the original polarization might well be determined by the anaphase pulling forces, persistent microtubulekinetochore and telomere-nuclear envelope attachments are necessary to avoid chromosome position randomization by Brownian motion. Experiments described on this thesis (Appendix 2) further support the idea of centromere clustering being actively maintained and regulated at the start of meiosis.
The observation of budding yeast centromeres, using fluorescence in situ hybridization probes, have shown that the centromere clusters disintegrate before the appearance of the earliest precursors of the SC (short axial elements and short synapsed segments). More recent immunofluorescence studies confirm these data, clearly showing the centromere cluster to be released in between the end of pre-meiotic S-phase and the initial formation of the first SC stretches containing Zip1.