Centromeres

Abstract

The centromere is the chromosomal site which associates with spindle microtubules and is responsible for equal division of the chromosomes to both daughter cells. Centromeres can often be recognised cytologically as a distinctive narrow region on metaphase chromosomes known as the primary constriction. The terms ‘centromere’ and ‘kinetochore’ are often used synonymously but it is best to consider the centromere as the DNA and chromatin on which the kinetochore is assembled and the kinetochore as the protein machine, including centromeric chromatin, which mediates attachment to the spindle. Centromeres are found near the centre of the chromosome on metacentric chromosomes, but they can also be submetacentric, between the middle and an end; acrocentric, close to an end; and telocentric, at an end. Some organisms such as Caenorhabditis elegans are holocentric and assemble centromeric chromatin and kinetochores along the entire chromosome. Delving further you discover that centromeres are eccentric! The centromere region is also a site of strong cohesion; in higher eukaryotes it is the last point of contact between sister-chromatids at metaphase prior to their separation at anaphase. Cohesion between sister centromeres is intimately associated with kinetochore function as it counteracts forces of microtubule–kinetochore attachment and possibly aids orientation and attachment of sister kinetochores to opposite poles. Centromeres are frequently associated with heterochromatin. Genes are silenced if placed in these regions by accident or design. It is not known why these regions are transcriptionally silent and whether silencing itself is important for centromere function. In fission yeast, at least, it appears that the silent state is a by-product of full kinetochore assembly. Centromeres range in size and complexity from a relatively minute 125 bp in budding yeast, through more complex structures of 40–120 kb in fission yeast, to enormous arrays of relatively simple repeats extending for 250 kb in flies and 5000 kb in humans. Regardless of size, they all assemble kinetochores that segregate chromosomes with awe-inspiring fidelity. The number of microtubules per metaphase kinetochore varies from one in budding yeast, 2–4 in fission yeast and from 4 to over 40 in mammals. DNA sequence is not conserved between centromeres of different organisms. However, all active centromeres have a histone H3-like protein, such as CENP-A, which must contribute to the assembly of a specialised centromere specific chromatin structure. Yes and no. Yes for yeast. But it's more complicated for mammalian centromeres. In humans vast arrays of so-called alphoid satellite repeats are found at centromeres, and mammalian artificial chromosomes have been created that contain alphoid DNA. So, you think, alphoid DNA is the determinant of centromeric activity! But in some dicentric chromosomes, both centromeres contain alphoid DNA, but only one is active, as assayed by the presence of important kinetochore components such as CENP-A and CENP-C. And in certain rare cases centromere activity has been gained by regions which have never before acted as centromeres. These neocentromeres contain no alphoid DNA but have been propagated through up to three human generations. These strange goings-on indicate that centromeres are epigenetically regulated. Presumably there is a meta-stable mark that normally ensures that centromeres and kinetochores do assemble at the same place division after division. Epigenetic phenomena also contribute to centromere function in fission yeast and Drosophila, and probably other organisms. Where can I find out more?