COMPUTER-AIDED MOLECULAR DESIGN AND MODELING OF TUBULIN TARGETING AGENTS

My research activity is in the frame of the TubInTrain European Joint Doctorate (EJD), which is a European research program formed by 13 early stage researchers (ESRs) from different disciplines and countries within the fields of biology and chemistry focused on the study of the breakdown of microtubules (MTs) associated with neurodegenerative diseases and neurotoxicity. Tubulin protein is the building block of MTs and is the target protein on which the work described in this Ph.D. thesis is focused. Within the TubInTrain program, the objective of my project is to apply a series of computer-aided molecular design techniques to thoroughly develop new tubulin binding agents and to explore at an atomic level the dynamics of their interactions when in complex with tubulin. This will facilitate further studies of MTs and MT-associated proteins (MAPs) related to neurodegenerative diseases. Chapter I provides an introduction of the biological system that is the main subject of this thesis: the MT and its building block tubulin. The structures and function of these important components of the cell are covered in detail, emphasizing their significance as therapeutic targets for the treatment of various diseases, such as cancer and neurodegeneration. The chapter also discusses the known tubulin binding sites and the effect of tubulin targeting agents on MT dynamics when bound to their tubulin site. Chapter II is an introductory chapter to the theoretical principles of the applied computer-aided molecular design techniques, including docking techniques and molecular dynamics simulations. The utilized docking software S4MPLE is given special attention, as it was the main computational tool employed during the present Ph.D. work. Chapter III details the investigation carried out on the recently discovered todalam site of α-tubulin. This chapter describes how we employed computer-aided molecular design approaches to efficiently search for and identify alternative chemotypes that target the todalam site. To test and characterize these new molecules, we established an interdisciplinary pipeline, which also included synthesis, X-ray crystallography, and in vitro assays of the computationally identified potential todalam-site binders. Through virtual screening we found several hits that were experimentally validated. This enabled further structure-based design of putative covalent todalam site binders predicted to target αCys4 of α-tubulin. We conducted a virtual screening of building blocks with cysteine-binding warheads from commercial libraries and selected the best candidates based on the results of covalent and non-covalent docking experiments, as well as molecular dynamics simulations of the computationally designed compounds. In Chapter IV is described the study performed on the maytansine binding site of tubulin. We computationally designed long-chain maytansinoids and maytansinoid conjugates. Using molecular docking, we predicted the binding mode of short-chain maytansinoids within the maytansine site. Our study reports the tolerance of the C3 position of maytansinol to the addition of bulky substituents without being sterically hindered. These findings provide a solid foundation for further exploration of maytansinoids as molecular tools for investigating MT dynamics. Chapter V describes the work carried out on the taxane site of tubulin. In this chapter, molecular docking was performed to predict the binding mode of flutax-2 to β-tubulin and to identify four novel paclitaxel derivatives from a virtual library of 50 compounds, which were selected based on their chemical feasibility, reagent availability, and likelihood of exhibiting a binding mode similar to flutax-2. These derivatives offer an excellent opportunity to explore the impact of bulky substituents at the C7-OH site of paclitaxel and investigate whether there are steric requirements that affect MT lattice.