Abstract

FROM LAPTOP TO TOP-LAB: A COMBINED IN SILICO IN VITRO APPROACH TO IDENTIFY NOVEL MICROTUBULE STABILIZERS AS POTENTIAL DRUG CANDIDATES


Florian Fischer1, Daniela Schuster1, Veronika Temml1
1Institute of Pharmacy, Department of Pharmaceutical and Medicinal Chemistry, Paracelsus Medical University, Salzburg, Austria
Objective: Microtubules are integral components of eukaryotic cell structures, providing essential support and facilitating various cellular processes. Composed of α- and β-tubulin subunits, they form dynamic filaments crucial for structural integrity, intracellular transport, and cell division. Through dynamic polymerization and depolymerisation of these tube-like aggregates, these proteins represent an important element in the life cycle of eukaryotic cells. Disruption of microtubule dynamics can induce cell death, particularly in rapidly dividing cells like cancer or immune cells, by promoting apoptosis. One strategy involves inhibiting continuous depolymerization using microtubule stabilizers from the taxane family, such as paclitaxel and docetaxel. However, resistance to these potent chemotherapeutics can develop in cancer cells. Consequently, there is considerable scientific interest in identifying new compound classes and scaffolds with potential antitumor activity through microtubule stabilization, offering promising avenues for cancer therapeutics.[1, 2]
Methods: Our approach combines innovative in silico techniques such as pharmacophore modeling with in vitro studies to discover new enhancers for the β-tubulin subunit. High-quality pharmacophore models were generated using LigandScout version 4.4.5 [3]. Pharmacophore modeling identifies key physicochemical features of protein-ligand interactions, facilitating the screening of molecules with similar interaction patterns. In this study, a combination of structure- and ligand-based modeling was employed. Structure-based models utilized 3D crystal structures, while ligand-based pharmacophore models were developed by aligning active molecules from the literature. The optimized and theoretically validated pharmacophore models were used to screen the SPECS database, which contains > 200,000 commercially available compounds (april2023). Selected Hits identified by our models were further investigated in vitro using a fluorescence-based tubulin polymerization assay kit (Cytoskeleton, BK011P).
Results: A comprehensive literature search was conducted to compile microtubule stabilizers targeting the taxane binding site of the beta-tubulin subunit, resulting in a curated set of 56 active compounds. Additionally, a decoy set of 1907 compounds was generated for comparison. Using LigandScout 4.4.5, databases were converted into multi-conformational screening datasets, facilitating the training of pharmacophore models. A dataset consisting of 56 biologically tested microtubule stabilizers (MSTs) and 1907 decoy compounds was transformed into a 3D dataset. One structure-based (SB) and three ligand-based (LB) pharmacophore models were utilized to identify active substances from the SPECS database. Among 3616 candidates, 22 were screened in vitro at a concentration of 30 µM, with 8 exhibiting significantly higher polymerization rates compared to the blank. This integrated in silico and in vitro approach led to the discovery of novel microtubule stabilizers, which will undergo further investigation using cancer cell models.
Conclusion: Our highly optimized structure-based and ligand-based pharmacophore models have the capability to predict substances that also exhibit the desired activity in an in vitro assay with a success rate of 36%. To demonstrate the potency of the top-ranked biologically tested compounds, further concentration-dependent screenings are now required.
References:
1. Cao YN, Zheng LL, Wang D, Liang XX, Gao F, Zhou XL. Recent advances in microtubule-stabilizing agents. Eur J Med Chem. 2018 Jan 01;143:806-28.
2. Lei L, Wang XJ, Tang SC. Novel taxanes in development: Hopes or hypes? Crit Rev Oncol Hematol. 2022 Aug;176:103727.
3. Wolber G, Langer T. LigandScout: 3-D pharmacophores derived from protein-bound ligands and their use as virtual screening filters. J Chem Inf Model. 2005;45(1):160-9.