NOVEL MICROTUBULE TARGETING AGENTS: SYNTHESIS AND STRUCTURAL DETERMINATION OF THEIR INTERACTIONS

This thesis, performed within the EU-funded Innovative Training Network - European Joint Doctorate TubInTrain, via the cotutelle agreement for a double degree between University of Milan and University of Basel, is focused on the microtubules dynamics and its breakdown associated to neurodegenerative diseases. Microtubule-targeting agents constitute a large group of chemically diverse molecules that bind tubulin and perturb the stability and functions of microtubules. While microtubule-targeting agents constitute a successful class of cancer drugs, the way in which tubulin binders can face neurodegeneration and neurotoxicity has yet to be determined. We aimed to develop a set of tubulin binders acting as chemical probes to explore different ligand-tubulin interactions modes and to potentially study how they affect the behaviour of microtubules in cell free systems. The chosen compounds are further studied in neuronal models, in order to clarify how small molecules affect microtubule dynamics in neurons. In Chapter I, the introduction on microtubules, their structure, and functions, and in particular the microtubule-targeting agents is described. The chemical structures of the representative tubulin binders, their respective bindings sites and therapeutic applications are addressed. Furthermore, the aims and objectives of the thesis are discussed. Within the workflow, the main objectives of my project included the synthesis of new tubulin binders, performed at the University of Milan. This was complemented by the X-ray crystallography structure determination of individual tubulin-compound complexes, which was performed in the course of my one-year secondment period at the Paul Scherrer Institute. Based on that, Chapter II covers the synthetic part of the work, while Chapter III covers structural determination of ligands-tubulin complexes. More precisely, Chapter II includes four different projects. Section 2.1 focuses on the functionalization of a natural product maytansinol. Maytansinoids represent a successful class of natural and semisynthetic tubulin binders, known for their potent cytotoxic activity. Their wider application as cytotoxins and chemical probes to study tubulin dynamics was set back due to the complexity of natural product chemistry and lack of SAR studies. Natural product maytansinol is a valuable precursor for the preparation of maytansine derivatives. Thus, one of the main objectives was to establish synthetic routes for the functionalization of maytansinol and to synthesize a set of maytansine analogues with different side chain substituents in C3-position, in order to investigate the effect of bulky substituents in the C3-position on the binding mode and activity of maytansinoids. Design, synthesis and biological evaluation of short-chain, long-chain and maytansinoid conjugates is described. Section 2.2 focuses on the synthetic efforts towards the total synthesis of a natural product glycybridin B. We aimed to identify and synthesize new maytansine site binders, structurally less complex than the family of maytansinoids. A natural product glycybridin B, isolated for the first time in 2017 from the herbal medicine Glycyrrhiza glabra, has been chosen for synthesis based on the computational screening. It attracted our interest due to the following factors: a) its total synthesis has not yet been reported in literature; b) different biological studies attribute it a variety of biological activities; c) it has been predicted to bind to maytansine site in computational studies. Therefore, synthetic efforts to synthesize the desired glycybridin B are described. Section 2.3 and 2.4 both focus on the covalent-targeting strategy to target tubulin. Covalent probes are characterized by prolonged duration of action, improved potency, and exquisite selectivity. Moreover, they could be used e.g., to immobilize the protein of interest, tubulin. In Section 2.3, the synthesis of simplified biotinylated pironetin analogues is described. Pironetin, a natural product that covalently binds alfa-tubulin, and the possibility of its biotinylation to use then biotin capturing on streptavidin for protein immobilization. Taking into consideration the low availability and time-consuming synthesis of natural pironetin, we designed and synthesized its simpler biotinylated analogues. In its turn, Section 2.4 comprises Todalam site ligands. Todalam targets the most recently discovered eighth small molecule binding site on tubulin and displays a unique molecular mechanism of action. Using the time- and cost-efficient synthetic approaches, we intended to develop novel, different from Todalam scaffolds and then, based on the structural data, rationally optimize them with the idea of site-specific covalent targeting. I was involved both in the chemistry and structural biology aspects of this project. I aimed to obtain and characterize the crystal structures of todalam analogues synthesized in our laboratory in complex with tubulin, to then design (in collaboration with the computational groups) and synthesize a library of optimized warhead-bearing ligands for the covalent targeting. Section 2.4 covers the synthetic work performed on this project, while the structural work including the obtain crystal structures of the developed todalam site ligands in complex with tubulin is described in detail in Section 3.3.5. Chapter III contains the introduction to the methods used to determine protein-ligand interactions, with particular focus on X-ray crystallography, discussed in Section 3.1. In the following Section 3.2, tubulin crystallization systems and the experimental data regarding the used tubulin crystallization system T2R-TTL is described. Finally, the determined tubulin-ligands crystal structures are discussed in detail in Section 3.3. Two training projects, which allowed to acquire skills in protein expression and purification; crystallization, crystal harvesting and cryo-cooling, data collection at synchrotron, model building and refinement, are contained in Sections 3.3.1 and 3.3.2. Both training projects focus on the colchicine site ligands: the analogues of combretastatin A4 and the stilbene-derived derivatives. Section 3.3.3 addresses the results of the crystallographic screening of other sixty-eight compounds derived from synthesis or virtual screening of Enamine libraries. As a result of the screening campaign, we managed to identify another colchicine site ligand, specifically binding to β3 isotype of tubulin. Isotype-specific ligands present an advantage as they can target certain isotypes, overexpressed in specific cells and therefore can be selectively used for certain types of diseases (e.g. beta3 tubulin present mostly in neurons). The relevant results and the successfully obtained crystal structure of the hit are described in this Section.