Dynamics and Regulation of Cytoskeletal Proteins

Abstract

In this dissertation, I apply molecular dynamics (MD) simulations to improve our understanding of the dynamics, and hence, function and regulation of cytoskeletal proteins. Microtubules and kinesin motor proteins play a critical role in the cytoskeleton of the cell. They provide structural support, facilitate cellular transport, and are involved in beating of cilia and flagella, and in separation of chromosomes during the cell cycle. The importance of tubulin as a vital therapeutic target is exemplified by the widely prescribed paclitaxel (Taxol), an anti-cancer drug that prevents cancer cells from undergoing cell division by arresting tubulin dynamics. Furthermore, the importance of understanding the structural, dynamical, and functional aspects of kinesin motor domains and their modifications is demonstrated by efforts in developing small-molecule inhibitors as antimitotic therapeutic agents in various cancers. However, despite strong conservation of the motor domain across the kinesin superfamily, how various kinesins have tailored their motility characteristics to best meet their functional needs in cells remains unclear.

Detailed comparison of structures from large heterogeneous protein families, such as kinesin motors, can inform on structural dynamic mechanisms critical for protein function including ligand binding, enzymatic catalysis, allosteric regulation and bimolecular recognition. However, existing tools for quantitative analyses of their sequence, structure and dynamics often require significant computational expertise and typically remain accessible only to expert users with relevant programming skills. In the first section of my dissertation, I describe the development of Bio3D-web, a free and open-source online application for interactive investigation of protein sequence-structure-dynamic relationships. Bio3D-web requires no programming knowledge and thus decreases the entry barrier to performing advanced comparative structural bioinformatics analyses.

In the second part, I discuss a method for analyzing experimental structures and dynamical data generated with MD simulations. The ensemble distance difference matrix method (eDDM) analyzes changes in residue-residue distances in protein structures and dynamical data to identify residues critical for protein regulation and function. I apply eDDM to three families of kinesin motor proteins in the following case studies: First, I elucidate the effect of a posttranslational modification in kinesin 5 mitotic motor Eg5. I show that acetylation of residue K146 in Eg5 alters its mechanochemical properties, wherein it acts as a “brake” during spindle separation in cells during mitosis. Second, I identify residues critical for force generation in kinesin 1 transport motor KIF5C. Mutating these residues in two important structural elements—A5G and S8G in the cover strand and N334A in the neck linker—severely cripple the ability of motors in ensemble to generate force during intracellular transport. Third, I characterize the allosteric effects of disease-associated variants in kinesin 3 neuronal transport motor KIF1A. KIF1A-associated neurological disorder (KAND) is associated with cognitive disability, spasticity, and cerebellar atrophy, typically with a progressive course.

In the third part, I highlight the divergent mechanism of tubulin polymerization in C. elegans. Through comparative analysis of MD simulations of C. elegans and B. taurus tubulin dimers, I found that sequence changes in the C. elegans tubulin lead to additional secondary structure formation in the lateral contact loops, and this changes the polymerization behavior as well as the structure of the microtubule. Finally, I also map the inter-conformer relationships of experimentally determined structures of tubulin through principal component analysis (PCA), enabling comparison of the intrinsic dynamics of tubulin heterodimers, such as different isoforms, nucleotide states, and disease-associated mutations.