Rational Structure-Based Drug Design

G protein-coupled receptors (GPCRs) are a superfamily of transmembrane proteins that play crucial roles in cell physiology. Being key drug targets in many diseases, drug discovery efforts harness structural information about these receptors to rationally design drugs in the field of structure-based drug design (SBDD). This thesis presents the development of computational tools to manage and assess large GPCR structural model datasets, applies these tools to define key features linked with the best performing models in SBDD and integrates this knowledge into a novel computational workflow that predicts new GPCR models, which may offer new opportunities in SBDD programs.

Machine learning (ML) has revolutionised the field of structure-based drug design (SBDD) in recent years. During the training stage, ML techniques typically analyse large amounts of experimentally determined data to create predictive models in order to inform the drug discovery process. Deep learning (DL) is a subfield of ML, that relies on multiple layers of a neural network to extract significantly more complex patterns from experimental data, and has recently become a popular choice in SBDD. This review provides a thorough summary of the recent DL trends in SBDD with a particular focus on de novo drug design, binding site prediction, and binding affinity prediction of small molecules.

Machine learning (ML) has revolutionised the field of structure-based drug design (SBDD) in recent years. During the training stage, ML techniques typically analyse large amounts of experimentally determined data to create predictive models in order to inform the drug discovery process. Deep learning (DL) is a subfield of ML, that relies on multiple layers of a neural network to extract significantly more complex patterns from experimental data, and has recently become a popular choice in SBDD. This review provides a thorough summary of the recent DL trends in SBDD with a particular focus on de novo drug design, binding site prediction, and binding affinity prediction of small molecules.

Computational approaches are widely used to help discover and develop new drugs, in particular, to understand how these molecules interact with their biomolecular targets. There are well-optimised and validated force field parameters available to describe interactions of common biomolecules, such as proteins, lipids and nucleotides. However, these force fields are not designed to represent heteromolecular ligands such as substrates, inhibitors, co-factors and potential drug molecules. Errors in parameters may result in incorrect predictions of ligand structure, orientation and conformation, which in turn can lead to the failure of computational drug design efforts. With over 25% of all the structures in the Protein Data Bank (PDB) containing ligands and over a million ligand molecules of potential interest in drug design in other databases, ligand validation and parameterisation represents a significant scientific challenge. While several automated parameterisation protocols have been proposed to generate the parameters, none of the current procedures are properly validated. As a response to the high demand for interaction parameters for ligands compatible with the GROMOS force field, a web accessible Automated force field Topology Builder (ATB; http://atb.uq.edu.au/) and Repository was developed. The ATB and Repository is intended to facilitate the development of molecular force fields and generates parameters that can be used in X-ray refinement, structure-based drug design and study of biomolecule-ligand complexes. In this thesis, molecular dynamics (MD) simulations were used to calculate the thermodynamic properties for the validation and refinement of ATB force fields parameters. A fully automated validation protocol of the ATB force field parameters for small organic molecules based on thermodynamic and structural information was developed and incorporated into the ATB. A novel integration and convergence protocol which increases the efficiency of the TI method for free energy calculations was also proposed. The validation has shown good overall agreement between the experimental and computed values and indicated problematic functional groups with parameters to be refined. The generation of parameters for novel molecules that are compatible with a given biomolecular force field can be tedious, time-consuming and error-prone. Here, a novel method to refine parameters in classical force fields in an automated manner that can be extended to the parameterisation of all other atom types is also presented. Single-step perturbation protocols were developed for small halogenated molecules to establish alternative van der Waals parameters which describe the interactions of these molecules with high precision. Parameters were successfully refined against experimental hydration free energies, densities and heats of vaporisation. As a result, alternative van der Waals interactions parameters for chlorine and bromine were proposed. Properties calculated with these proposed parameters closely matched the experiment data. The quality of the ATB force field parameters in validating the X-ray structures and investigating binding of ligands to the protein endothiapepsin was also investigated. Analysis of the MD trajectories showed that the protein structure remained relatively stable after 10 ns of simulations. The results for the validation of the ligands were mixed. While in some cases the simulations reproduced the binding mode and the interactions between the protein and the ligand with high accuracy suggesting the ATB parameters performed well, in other cases the complexes were unstable. This was despite the fact that the electron density for all ligands was well defined and the parameters for the ligand were generated using the same procedure. Endothiapepsin contains a catalytic dyad consisting of two aspartates that are generally assumed to share a proton. We demonstrated that the protonation of both the ligand and the residues with the binding site were critical to the ability to reproduce the crystal complex. To conclude, the work presented in this thesis provided insight into the development and validation of the force field parameters for drug-like molecules and their applications in structure-based drug design.

Computer-aided drug design plays a vital role in drug discovery and development and has become an indispensable tool in the pharmaceutical industry. Computational medicinal chemists can take advantage of all kinds of software and resources in the computer-aided drug design field for the purposes of discovering and optimizing biologically active compounds. This article reviews software and other resources related to computer-aided drug design approaches, putting particular emphasis on structure-based drug design, ligand-based drug design, chemical databases and chemoinformatics tools.

Chapter 3 - Ligand-based drug design (LBDD)

INTRODUCTION. Malignant glioma is one of the most lethal diseases in adulthood and remains refractory to the current therapies including chemoradiotherapies. The cancer stem cell theory dictates that brain tumor stem-like cells (BTSC) account for such intractability of malignant glioma. Recently, several molecularly-targeted therapies have demonstrated a varying degree of success in treatment of cancers including leukemia and breast cancer. In this study, we sought to determine the key kinases that regulate survival, self-renewal, and proliferation of BTSC, and to design novel small molecules by using the structure-based computantional modeling system. RESULTS. With the neurosphere cultures derived from our glioma patients, we identified 3 kinases that play vital roles in survival and/or proliferation of BTSC in vitro; Fibroblast Growth Factor Receptor 1, Aurora kinase B, and Maternal Embryonic Leucine-zipper Kinase. Structure-based computational modeling allowed us to perform in silico docking of various small molecule candidates to these kinases. As a result, four novel small molecules demonstrated highly specific targeting of the key structural elements of these kinases. We then synthesized these identified candidate compounds and confirmed their in vitro inhibition of the targeting kinase activities. Finally, we will also present our data about the treatment effect with these compounds on the in vitro BTSC growth and the in vivo tumor growth. CONCLUSION. Structure-based drug design will likely open up a new avenue to discover novel anti-cancer stem cell agents that can selectively target the key kinases in BTSC, leading to the growth arrest of malignant glioma. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3298. doi:10.1158/1538-7445.AM2011-3298

Computer-aided drug design (CADD) approaches are playing an increasingly important role in understanding the fundamentals of ligand-receptor interactions and helping medicinal chemists design therapeutics. About 5years ago, we presented a chapter devoted to an overview of CADD methods and covered typical CADD protocols including structure-based drug design (SBDD) and ligand-based drug design (LBDD) approaches that were frequently used in the antibiotic drug design process. Advances in computational hardware and algorithms and emerging CADD methods are enhancing the accuracy and ability of CADD in drug design and development. In this chapter, an update to our previous chapter is provided with a focus on new CADD approaches from our laboratory and other peers that can be employed to facilitate the development of antibiotic therapeutics.

In the current era of high-throughput technology, where enormous amounts of biological data are generated day by day via various sequencing projects, thereby the staggering volume of biological targets deciphered. The discovery of new chemical entities and bioisosteres of relatively low molecular weight has been gaining high momentum in the pharmacopoeia, and traditional combinatorial design wherein chemical structure is used as an initial template for enhancing efficacy pharmacokinetic selectivity properties. Once the compound is identified, it undergoes ADMET filtration to ensure whether it has toxic and mutagenic properties or not. If the compound has no toxicity and mutagenicity is either considered a potential lead molecule. Understanding the mechanism of lead molecules with various biological targets is imperative to advance related functions for drug discovery and development. Notwithstanding, a tedious and costly process, taking around 10–15 years and costing around $4 billion, cascaded approached of Bioinformatics and Computational biology viz., structure-based drug design (SBDD) and cognate ligand-based drug design (LBDD) respectively rely on the availability of 3D structure of target biomacromolecules and vice versa has made this process easy and approachable. SBDD encompasses homology modelling, ligand docking, fragment-based drug design and molecular dynamics, while LBDD deals with pharmacophore mapping, QSAR, and similarity search. All the computational methods discussed herein, whether for target identification or novel ligand discovery, continuously evolve and facilitate cost-effective and reliable outcomes in an era of overwhelming data.

Molecular ConceptorTM (http://www.molecular-conceptor.com) is a drug design e-courseware that has been developed as a reference, a source of inspiration for medicinal chemists and a database of case studies in drug discovery. Synergix has developed Molecular Conceptor, as drug design e-courseware that can be used as a reference by medicinal chemists as a source of inspiration in drug design, and as an expanding compendium of knowledge in drug discovery. Molecular Conceptor includes 40 volumes and chapters that cover a wide range of topics such as structure-based drug design, ligand-based drug design, cheminformatics, molecular similarity, 3D-QSAR (three-dimensional Quantitative Structure-Activity Relationships), peptidomimetics, molecular docking and many others. In this report, we have chosen ‘molecular mimicry’ as a representative example to illustrate how it appears in Molecular Conceptor. This topic appears in many sections of the e-courseware, including: (i) three chapters devoted to pharmacophore-based drug design and (ii) several chapters on molecular similarity, 3D database searching, principles of rational drug design and library design. Two-dimensional (2D) representations are often misleading for understanding SAR, and one of the key features of Molecular Conceptor is its ability to visualize molecules in 3D, which is crucial to grasping the subtle recognition and discrimination processes of molecular machinery. In the example illustrated in Figure 1, moving the nitrogen atom in the pyrimidine heterocycle from A to B turns an inactive molecule into a very potent one. This is extremely difficult to see in the 2D formulas. A cogent explanation is given in Molecular Conceptor that can only be derived by viewing the molecules in 3D. Molecules A and B differ solely in the position of the nitrogen atom in the pyrimidine ring and exhibit very different biological activities (left). A possible explanation for this difference is visualized in three-dimension (right). Molecular mimicry is extremely useful in the early stages of drug discovery. When little knowledge is available, such as for example, one active reference compound, this is enough and the project can start. The aim of many such projects was described by terms such as ‘non-peptides, non-nucleotides, non-steroids, non-tricyclic mimetics’ and so forth. Snapshots of some pages associated with this topic are available in Figure 2 and Supplementary Material to this report. Simple and more complex case studies are presented, decomposed and assessed so that methodological rules can be deduced. The core topic is presented in three successive chapters: the first discusses ways to analyse the project (principles of analyses, control of molecular geometries, misuse of structural information, managing hypotheses); the second addresses the design issue (e.g. the four design methods) and the third discusses success stories from 13 different therapeutic areas. Other chapters such as the ones dealing with 3D-QSAR, molecular similarity, library design and ADME (Absorption, Distribution, Metabolism, Elimination) properties highlight the importance of molecular mimicry in these different areas. When molecular mimicry is applied to peptides, it is referred to as ‘peptidomimetics’. This type of approach is mandatory in drug research when dealing with a peptide molecule that cannot be developed as a drug. This topic is presented in detail in two different chapters: (i) Introduction to Peptidomimetics and (ii) Peptidomimetic Examples. However, many other examples are presented for diverse purposes in other sections of the software. For instance, the successful design of Aliskiren (a renin inhibitor now in late stages of development) is presented in both the chapter on examples of peptidomimicry as well as in the chapter on case histories in drug discovery. Snapshot of a page from Molecular Conceptor. In conclusion, the concept of molecular similarity has nurtured the imagination and creativity of several generations of medicinal chemists. Because of the fuzziness of the concept, medicinal chemists live with the contradiction that similar compounds simultaneously have and do not have similar activities. The key concepts and features of thousands of scientific articles devoted to this topic are summarized in Molecular Conceptor. The essence of millions of hours of scientific work, human creativity, investment in human resources and expenditures in R & D are presented on a friendly and efficient platform. Molecular Conceptor is rapidly becoming a prime reference for medicinal chemists, a source of inspiration in drug design and can also be used to train newly recruited medicinal chemists in the industry (1-4). Figure S1: Snapshots of pages from Molecular Conceptor dealing with molecular mimicry. For clarity, both the navigation tree and the text of the pages have been omitted Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Structure-based drug design is emerging as one of the key components in drug discovery, with many approved drugs tracing, at least part of their origins, to the use of structural information from X-ray, NMR, surface plasmon resonance, differential thermal denaturation, fluorescence polarization and other techniques for analysis of protein targets and their ligand-bound complexes. Furthermore, in silico structure-based drug design approach has enabled millions of possible structures for a given protein sequence to be evaluated rapidly helping to fast forward drug discovery as well as reduce drug discovery costs. Structure-based drug design is now arguably an essential contributor to addressing the need to improve research and development productivity faced by the pharmaceutical industry. The purpose of this meeting is to highlight the impact of the intersection of structural biology with chemistry and biology particularly on how the structures of relevant drug targets can serve as a starting point for drug design and development and provide the maximal synergy between target validation, structure determination, and hit-to-lead development. Some thoughts will be proposed with regard to the future of structure-based drug design and where emphasis could be placed to further increase the utilization of this approach on drug discovery.

Generating the three-dimensional (3D) structure of small molecules is crucial in both structure- and ligand-based drug design. Structure-based drug design needs bioactive conformations of compounds for lead identification and optimization. Ligand-based drug design techniques, such as 3D shape similarity search, 3D pharmacophore model, 3D-QSAR, etc., all require high-quality small-molecule ligand conformations to obtain reliable results. Although predicting a small molecular bioactive conformer requires information from the receptor, a crystal structure of the molecule is a proper approximation to its bioactive conformer in a specific receptor because the binding pose of a small molecule in its receptor's binding pockets should be energetically close to the crystal structures. This study presents a de novo small molecular structure predictor (dMXP) with graph attention networks based on crystal data derived from the Cambridge Structural Database (CSD) combined with molecular electrostatic information calculated by density-functional theory (DFT). Two featuring strategies (topological and atomic partial change features) were employed to explore the relation between these features and the 3D crystal structure of a small molecule. These features were then assembled to construct the holistic 3D crystal structure of a molecule. Molecular graphs were encoded using a graph attention mechanism to deal with the issues of the inconsistencies of local substructures contributing to the entire molecular structure. The root-mean-square deviation (RMSDs) of approximately 80% dMXP predicted structures and the native binding poses within receptors are less than 2.0 Å.

Chapter 3 - Advances in structure-based drug design

The conventional drug discovery approach is an expensive and time-consuming process, but its limitations have been overcome with the help of mathematical modeling and computational drug design approaches. Previously, finding a small molecular candidate as a drug against a disease was very costly and required a long time to screen a compound against a specific target. The development of novel targets and small molecular candidates against different diseases including emerging and reemerging diseases remains a major concern and necessitates the development of novel therapeutic targets as well as drug candidates as early as possible. In this regard, computational and mathematical modeling approaches for drug development are advantageous due to their fastest predictive ability and cost-effectiveness features. Computer-aided drug design (CADD) techniques utilize different computer programs as well as mathematics formulas to comprehend the interaction of a target and drugs. Traditional methods to determine small-molecule candidates as a drug have several limitations, but CADD utilizes novel methods that require little time and accurately predict a compound against a specific disease with minimal cost. Therefore, this review aims to provide a brief insight into the mathematical modeling and computational approaches for identifying a novel target and small molecular candidates for curing a specific disease. The comprehensive review mainly focuses on biological target prediction, structure-based and ligand-based drug design methods, molecular docking, virtual screening, pharmacophore modeling, quantitative structure–activity relationship (QSAR) models, molecular dynamics simulation, and MM-GBSA/MM-PBSA approaches along with valuable database resources and tools for identifying novel targets and therapeutics against a disease. This review will help researchers in a way that may open the road for the development of effective drugs and preventative measures against a disease in the future as early as possible.

Structural optimization and bioactivity evaluation of 2-(Methylcarbonylamino) thiazole derivatives as novel PDE4B inhibitors