It is intuitive that cell division is fundamentally important to the continuation of life and that two cells produced from the division process must be genetically identical to their parent. Despite this, our understanding of cell division mechanics remains incomplete (Mitchison and Salmon, 2001), a fact reflected by our inability to explain why the majority (~70%) of human cancers are associated with flawed divisions (Weaver and Cleveland, 2006). Partitioning of the replicated genome, which is packed into chromosomes, is accomplished by an apparatus termed the mitotic spindle. Our work focuses on the biochemical mechanisms underlying chromosome segregation, with emphasis on protein factors that influence the dynamics of microtubules, the polymer from which the mitotic spindle is constructed.
Figure 1. Left, an African clawed frog, Xenopus laevis. Right, Tons of laid, dejellied eggs!
Figure 2. Chromosome segregation in vivo, in Xenopus S3 cells, and in vitro using spindles assembled in Xenopus egg extracts. The S3 cell is imaged using phase contrast optics. Chromatin is phase dark, and the mitotic spindle is apparent. Microtubules in the egg extract spindle is labeled with rhodamine-tubulin and the kinetochores with fluorescent anti-CENP-A antibodies. The spindle was triggered to enter anaphase wtih a pulse of calcium. The bottom image shows the extent of chromatid separation that occurs ~10 minutes after the induction of anaphase.
Our approach to studying mitosis and meiosis is multidisciplinary, with reductionist strategies used to reconstitute processes observed in live cells. In addition to live cell imaging, we use a powerful in vitro system, extract prepared from the laid eggs of the African clawed frog Xenopus laevis, to image spindle morphogenesis and chromosome behaviour. The in vitro nature of Xenopus egg extracts allows the complexity of mitosis to be dissected biochemically, a key feature which facilitates the assignment of key aspects of spindle assembly and its dynamics to the activity of specific protein factors.
Our most recent work has focused on the contribution of two microtubule depolymerizing Kinesin-13s (MCAK and XKIF2) to spindle assembly (Ohi et al., 2007). Although both proteins have similar enzymatic activities in vitro, as assessed by their abilities to depolymerize stabilized microtubules, MCAK is specifically required to promote spindle assembly. The reason for this is not clear and further work is likely to reveal interesting differences between the two motors. Through this work, we also discovered that spindle length is exquisitely sensitive to MCAK concentration in extract. Excess MCAK reduces spindle length whereas too little MCAK causes the formation of long spindles. We proposed that this occurs through regulation of microtubule plus end dynamics (ie., dynamic instability), a notion which predicts that the average length of microtubules may influence overall spindle length. Ongoing studies are examining whether this assertion is true.
Currently, we are interested in mechanisms that promote disassembly of kinetochore-microtubules that are improperly attached. We are searching for novel substrates of the Aurora B kinase, a key component of the error correction machinery, and are trying to better understand how Aurora B affects the dynamics of kinetochore-microtubules.
Figure 3. Our current view of extract spindle microtubule dynamics (left) and spindle microtubule dynamics in somatic cells. For details, see Ohi et al., 2007.
