Microtubule-Kinetochore Attachment Interface
In the cell the dynamics of microtubules is regulated by their interaction with different factors. Of special interest is the coupling of microtubules to the kinetochore, a process where microtubule dynamics reaches its "climax".
In collaboration with Georjana Barnes and David Drubin we are studying how the yeast kinetochore
complex Dam1 binds and tracks microtubule plus ends. The complex self-assembles into rings and helices
around the microtubule wall (Westermann et al., Mol Cell 2005). The binding is enhanced for GMPCPP microtubules,
supporting a preferential binding to growing microtubule ends. The interaction is via the charged
C-terminal of tubulin and allows for ring sliding along microtubules, which can be coupled to
microtubule disassembly to allow for the movement of Dam1 as it remains attached to a depolymerizing end
(Westermann et al.,
Nature 2006). This explains how this ring structure is able to track the depolymerizing ends of
microtubules (as during anaphase), without requiring energy!
Defining the architecture of the Dam1 complex and its microtubule-driven self-assembly is essential to understanding the mechanisms by which rings couple processive movement to microtubule disassembly and thus contribute to the end-on attachment of chromosomes to the mitotic spindle. It is also crucial to determining how the assembly of the ring is regulated and how the ring attaches to other components of the kinetochore. We have used EM-based single-particle and helical analyses to obtain initial structures of the Dam1 complex before and after its oligomerization around microtubules. This work has allowed us to define the architecture of the Dam1 complex and the self-assembly mechanism (Wang et al., 2007, Nat Struct Mol Biol. 14, 721-6). Ring oligomerization seems to be facilitated by a conformational change upon binding to microtubules, suggesting that the Dam1 ring is not preformed, but self-assembles around kinetochore microtubules. The C terminus of the Dam1p protein, where most of the Aurora kinase Ipl1 phosphorylation sites reside, is in a strategic location to affect oligomerization and interactions with the microtubule. One of Ipl1's roles might be to fine-tune the coupling of the microtubule interaction with the conformational change required for oligomerization, with phosphorylation resulting in ring breakdown.
In addition to the Dam1 complex, we are investigating how the highly conserved Ndc80 complex binds dynamic microtubule ends to generate force and motion. The complex acts an elongated coiled-coil arm which reaches out of kinetochores to grab on to the microtubule surface through the combined action of a globular head, composed of 2 CH domains from the Ndc80 and Nuf2 subunits, and a disordered tail, the 80 amino-acid amino terminus of Ndc80. Phosphorylation of the tail by the Aurora B kinase inhibits microtubule binding, which is thought to act as part of the error-correction machinery which ensures proper chromosome biorientation during metaphase.
We recently generated a sub-nanometer resolution cryo-EM reconstruction of the microtubule-bound complex, allowing us to generate a pseudo-atomic model of the primary conserved kinetochore-microtubule interface (Alushin et al. Nature 2010). During microtubule polymerization, protofilaments bend at the interfaces between tubulin subunits, both within and between tubulin dimers. Our cryo-EM structural analysis delineated a surprisingly small Ndc80 binding site on the microtubule surface that spans the interface between two tubulin monomers. This site would be disrupted upon tubulin bending during microtubule depolymerization, and a we found that Ndc80 binding to bent tubulin is strongly disfavored.
Furthermore, using electron tomography, we found that under substoichiometric conditions, the complex forms small clusters along individual protofilaments, in contrast to the ring structure formed by the Dam1 complex. The formation of clusters is dependent on the Ndc80 tail and is downregulated by phosphorylation, as is microtubule binding, suggesting that microtubule binding and clustering are linked. In the cryo-EM map we observed density peaks between neighboring Ndc80 molecules that appeared to be connecting them. The docked crystal structures of Ndc80 (which lacked the tail) and tubulin could not explain these densities, and thus we attributed them to the tail.
We propose that the formation of Ndc80 clusters is a phosphoregulated mechanism of the kinetochore recognizing and securing correct attachments, as cluster formation should be favored when sister kinetochores are under tension. It has previously been shown that Ndc80 can couple movement of a load to microtubule depolymerization via biased diffusion (Powers et al. Cell 2008), and we propose that the formation of clusters along protofilaments could be consistent with this mechanism, although it remains to be seen if clusters can diffuse. Moreover, Ndc80's preferential binding to straight tubulin polymers suggests that loss of tubulin subunits from the depolymerizing end is not required to bias diffusion: the curling of protofilaments would suffice. Current efforts are focused on obtaining higher-resolution reconstructions and dissecting the specific roles of the Ndc80 tail to inform a detailed mechanistic understanding of Ndc80's interaction with dynamic microtubules.