Riassunto analitico
Histone deacetylases (HDACs) are a family of 11 enzymes that remove acetyl groups from lysine residues on histone and non-histone proteins. Due to their important role in regulating many cellular processes, HDACs have emerged as key therapeutic drug targets. Current inhibitors targeting the orthosteric site of Zn-dependent HDACs are composed of a pharmacophore that contains a zinc-binding group (ZBG), a linker moiety, and a hydrophobic cap that extends toward the outer surface of the enzyme. Despite their therapeutic potential, current HDAC inhibitors are often associated with significant toxicity and side effects. These are mainly attributed to the poor inhibitor selectivity across different HDAC isoforms, and to the inherent toxicity and mutagenicity of orthosteric ligands that contain hydroxamate ZBGs. To mitigate these concerns, several efforts have been undertaken toward the development of isoform-selective modulators. Therefore, other classes of HDAC inhibitors have been obtained that contain alternative ZBGs and display better isoform selectivity. However, these compounds still face several hurdles, which are mainly ascribed to the presence of the ZBG. Based on this rationale, our hypothesis is that an allosteric inhibitor, that would not require a ZBG for binding to HDACs, would provide higher selectivity and lower toxicity compared to compounds that bind at the orthosteric site. As a matter of fact, allosteric inhibition could bypass the need for a ZBG and achieve selectivity by targeting less conserved regions. To the best of our knowledge, this approach was never explored so far. Therefore, this thesis represents the first effort in this direction.
My studies focused on HDAC6, a cytosolic isoform of HDAC that is highly relevant in cancer and neurodegeneration. Given the lack of known allosteric modulators for HDAC6, this work initially focused on identifying cryptic binding sites through structural comparisons of 3D structures available in the Protein Data Bank (PDB). My interest was mainly focused on finding possible allosteric sites located in the outer surface of HDAC6. Importantly, an allosteric ligand that binds to these regions would disrupt the interaction of HDAC6 with counterparts as important as Hsp90 and Tubulin, thus modulating their activity and providing an alternative means of controlling downstream events.
Afterwards, molecular dynamics (MD) simulations were carried out a selected crystal structure of human HDAC6 with the Amber24 software, in order to investigate the conformational variability near the catalytic site of the protein, and to identify transient cryptic pockets potentially targetable by small molecules. The generated MD simulations were then analyzed by means of a custom workflow integrating binding site estimations with SiteMap for pockets prediction, followed by evaluations of the pockets conformations through principal component analysis and a density-based approach. Subsequently, a library of commercially available compounds were filtered to retain only those with favorable drug-like properties, and docked into these representative structures of the identified cryptic site using the Glide docking program; the binding poses were re-ranked by binding free energies calculated with Prime MM-GBSA. A visual inspection of the results allowed the identification of a set compounds lacking a ZBG that can bind favorably to HDAC6, making them promising candidates for further development. Future studies will evaluate the stability of these ligands within the pocket and to assess experimentally their potential as allosteric modulators of HDAC6.
|
