Centromeres can run, but they can’t hide (from kinetochores)

2000 
Centromeres are the specialized regions of eukaryotic chromatin responsible for the physical association of chromosomes with the mitotic spindle. In particular, a large proteinaceous structure known as the kinetochore assembles onto the centromeric regions of chromosomes and mediates their attachment to spindle microtubules. While it has long been appreciated that these attachments are essential for proper chromosome segregation during mitosis, the process by which kinetochore proteins associate exclusively with centromeres (as opposed to associating non-specifically with bulk chromatin) is not well understood.The simplest explanation for this association is that there exist ‘kinetochore-anchoring’ factors that bind to specific DNA sequences found exclusively at centromeres. For instance, in the budding yeast Saccharomyces cerevisiae, a unique ∼125-bp CEN sequence (derived from yeast centromere DNA) is sufficient to confer proper chromosome segregation activity to episomal DNA (e.g. plasmids). However, a similarly active centromeric sequence element has not been demonstrated in higher eukaryotes.Nonetheless, it has been demonstrated that there exist unique, highly repetitive sequences at the centromeres of various higher eukaryotes. For example, human centromeres are characterized by long (up to 5 Mb) repeats of a ∼171-bp sequence known as α-satellite DNA. These α-satellite sequences can associate with known kinetochore proteins in vitro and in vivo, but DNA fragments encoding α-repeats are not sufficient to confer spindle attachment reliably to artificial chromosomes. Thus, kinetochore assembly at centromeres could be an epigenetic phenomenon in higher metazoans.Consistent with this possibility, mutant chromosomes (presumably arising from mitotic recombination events) have been discovered in which kinetochore components assemble onto regions of chromatin that are not at the normal centromeric position. Most significantly, these ‘neocentromeres’ lack detectable α-satellite DNA sequences but allow microtubule attachment and confer proper chromosome segregation during mitosis.In this study, the authors used a combination of fluorescence in-situ hybridization (FISH) and immunofluorescence to detect cytologically the presence of 22 different kinetochore proteins on wild-type and neocentromere-containing human chromosomes1xHuman centromeres and neocentromeres show identical distribution patterns of >20 functionally important kinetochore-associated proteins. Saffery, R. et al. Hum. Mol. Genet. 2000; 9: 175–185Crossref | PubMedSee all References1. Remarkably, all of the proteins except the DNA-binding protein CENP-B (which is not essential for viability in mice) were found to associate with neocentromeres in a manner indistinguishable from their association with wild-type centromeres. Thus, neocentromeres are competent to recruit bona fide kinetochore structures regardless of their lack of typical centromeric DNA sequence.Perhaps the major implication of these findings is the suggestion that higher-order chromatin structure, rather than sequence, is the principal characteristic of centromeric DNA that renders it competent for kinetochore binding. Furthermore, the observation that specific α-satellite DNA sequences do not seem to be necessary for kinetochore assembly suggests that some aspect of chromatin structure might be a heritable characteristic that distinguishes centromeres from bulk chromatin through generations of cell divisions.In sum, the results described herein provide the valuable insight that a functionally complete kinetochore can assemble in the absence of the standard centromeric DNA sequence. As such, there likely exist epigenetic mechanisms by which kinetochore proteins are recruited consistently to centromeric regions of chromosomes. The characterization of such mechanisms should not only offer profound insights into the process of chromosome segregation but might eventually lead to the development of centromeric vectors for stably introducing therapeutic genes into human tissues.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    1
    References
    0
    Citations
    NaN
    KQI
    []