UC Berkeley

Microtubules (MTs) are an essential component of the eukaryotic cytoskeleton formed by the polymerization of tubulin into hollow, cylindrical polymers and are involved in a diverse array of cellular functions. The diversity of MT functionality is achieved in part by the specialization of MTs for certain activities through the accumulation of post-translational modifications (PTMs), and the binding of MT associated proteins (MAPs) that help organize and regulate the MT network. I used cryo-electron microscopy (cryo-EM) to directly visualize these important biological assemblies in their native states. The first part of my work was to determine a role for the acetylation of tubulin. This PTM is unique in that it is located on the luminal surface of the MT, away from the exterior of the MT where MAPs bind. I sought to gain insight into the function of acetylation by looking for structural changes in MTs that occur upon acetylation. No significant changes were observed in protofilament distributions or MT helical lattice parameters. Furthermore, no clear differences in tubulin structure were detected when comparing subnanometer reconstructions of deacetylated or acetylated MTs. Our results indicate that the effect of acetylation must be highly localized and directly affect interactions with proteins that bind to the luminal surface. I also wanted to understand how the enzyme that carries out this modification, αTAT1, gains access to the lumen of the MT. I found that αTAT1 interacts with the outside of the MT, possibly serving as a funneling mechanism to deliver the enzyme to the lumen through lattice defects or transient openings in the MT wall. My data show that αTAT1 does not rely on open ends of MTs to access the lumen.

I also visualized the structures of engineered kinesins on MTs. The motors had been modified to directly test design principles of biological motors as well as respond to external light stimulation. By changing the length of the lever arm it was possible to alter the velocity of these motors. The cryo-EM reconstructions confirmed the geometry of the engineered lever arms. Incorporation of a light sensitive (LOV) domain into the lever arm made the motor photoactivatable. However, the LOV domain introduced greater flexibility into the engineered domain, preventing us from visualizing the full lever arm.

Yeast tubulin mutants offer the possibility to directly test molecular mechanisms of dynamic instability. I performed near atomic resolution reconstructions of WT and mutant yeast microtubules to understand the structural basis for the mutant phenotypes. I found that, unlike in mammalian MTs, the tubulin dimer size in WT yeast MTs is the same for dynamic lattice and those stabilized with drugs or GMPCPP. These yeast lattices are all expanded. However, the lattice compacts with the addition of the plus-end tracking protein, Bim1, or when it is assembled with GTPγS. The greatest differences for the mutant tubulin were seen outside of the microtubule lattice where the length and curvature of tubulin oligomers in solution were greater. All yeast oligomers were longer and straighter than those formed by mammalian tubulin.

Finally, the atomic level details about the interaction between PRC1, a MT crosslinker that stabilizes antiparallel MT arrays, and MTs was determined. The reconstruction shows that the spectrin domain of PRC1 binds using a loop that remained disordered in previous crystal structures. The binding site of PRC1 on MTs overlaps with that used by motor proteins at the intra-dimer interface. Using cryo-EM to examine MTs in their natural state at near atomic resolution has provided important information about their function and how they interact with MAPs.