Identifying the binding mode of a molecular scaffold

We describe a method for docking of a scaffold-based series and present its advantages over docking of individual ligands, for determining the binding mode of a molecular scaffold in a binding site. The method has been applied to eight different scaffolds of protein kinase inhibitors (PKI). A single analog of each of these eight scaffolds was previously crystallized with different protein kinases. We have used FlexX to dock a set of molecules that share the same scaffold, rather than docking a single molecule. The main mode of binding is determined by the mode of binding of the largest cluster among the docked molecules that share a scaffold. Clustering is based on our 'nearest single neighbor' method [J. Chem. Inf. Comput. Sci., 43 (2003) 208-217]. Additional criteria are applied in those cases in which more than one significant binding mode is found. Using the proposed method, most of the crystallographic binding modes of these scaffolds were reconstructed. Alternative modes, that have not been detected yet by experiments, could also be identified. The method was applied to predict the binding mode of an additional molecular scaffold that was not yet reported and the predicted binding mode has been found to be very similar to experimental results for a closely related scaffold. We suggest that this approach be used as a virtual screening tool for scaffold-based design processes.     

Diphenylindane-based proteomimetics reproduce the projection of the i, i+3, i+4, and i+7 residues on an alpha-helix

[structure: see text] The design of a nonpeptidic scaffold based on 4,7-diphenyl-1,6-disubstituted indanes mimicking i, i+3, i+4, and i+7 residues of an alpha-helix has been described, and its synthesis has been accomplished. This strategy makes general approaches possible to helix mimetic scaffolds that could be targeted to different proteins by changing the nature of the substituents.     

Computational fragment-based design of Wee1 kinase inhibitors with tricyclic core scaffolds

Wee1 is cell cycle protein comprising a kinase domain and is a validated cancer target. We have designed molecules with variable tricyclic core scaffolds [6-6-5] system and extended them based on the chemical space available in the active site of Wee1 kinase using de novo drug design. The core scaffolds and linking fragments were extracted from pharmacophore-based virtual screening of ZINC and PubChem databases and Ludi library. These molecules bind the hinge region of kinase active site and form hydrogen bonds as confirmed from molecular docking, molecular dynamics simulations, and MM_PBSA calculations. When compared with reference inhibitors, AZD1775 and PHA-848125, the de novo designed molecules also show good docking scores and stability, retained non-covalent interactions, and high binding free energies contributed from active site residues.

     

A minimalist approach toward protein recognition by epitope transfer from functionally evolved beta-sheet surfaces

New approaches for identifying small molecules that specifically target protein surfaces as opposed to active site clefts are of much current interest. Toward this goal, we describe a three-step methodology: in step one, we target a protein of interest by directed evolution of a small beta-sheet scaffold; in step two, we identify residues on the scaffold that are implicated in binding; and in step three, we transfer the chemical information from the beta-sheet to a small molecule mimic. As a case study, we targeted the proteolytic enzyme thrombin, involved in blood coagulation, utilizing a library of beta-sheet epitopes displayed on phage that were previously selected for conservation of structure. We found that the thrombin-binding, beta-sheet displaying mini-proteins retained their structure and stability while inhibiting thrombin at low micromolar inhibition constants. A conserved dityrosine recognition motif separated by 9.2 A was found to be common among the mini-protein inhibitors and was further verified by alanine scanning. A molecule containing two tyrosine residues separated by a linker that matched the spacing on the beta-sheet scaffold inhibited thrombin, whereas a similar dityrosine molecule separated by a shorter 6 A linker could not. Moreover, kinetic analysis revealed that both the mini-protein as well as its minimalist mimic with only two functional residues exhibited noncompetitive inhibition of thrombin. Thus, this reductionist approach affords a simple methodology for transferring information from structured protein scaffolds to yield small molecule leads for targeting protein surfaces with novel mechanisms of action.     

Creating functional artificial proteins

Much is now known about how protein folding occurs, through the sequence analysis of proteins of known folding geometry and the sequence/structural analysis of proteins and their mutants. This has allowed not only the modification of natural proteins but also the construction of de novo polypeptides with predictable folding patterns. Structure/function analysis of natural proteins is used to construct derived versions that retain a degree of biological activity. The constructed versions made of either natural or artificial sequences contain critical residues for activity such as receptor binding. In some cases, the functionality is introduced by incorporating binding sites for other elements, such as organic cofactors or transition metals, into the protein scaffold. While these modified proteins can mimic the function of natural proteins, they can also be constructed to have novel activities. Recently engineered photoactive proteins are good examples of such systems in which a light-induced electron transfer can be established in normally light-insensitive proteins. The present review covers some aspects of protein design that have been used to investigate protein receptor binding, cofactor binding and biological electron transfer.     

Context-based identification of protein-protein interfaces and "hot-spot" residues

Reliable determination of protein-protein interaction sites is of critical importance for structure-based design of small molecules modulating protein function through macromolecular interfaces. We present an alignment-free computational method for prediction of protein-protein interface residues. The method ("iPred") is based on a knowledge-based scoring function adapted from the field of protein folding and small molecule docking. Based on a training set of 394 hetero-dimeric proteins iPred achieves sustained accuracy on an external unbound test set. Prediction robustness was assessed from more than 1500 diverse complexes containing homo- and hetero-dimers. The technique does not rely on sequence conservation, so that rapid interface identification is possible even for proteins for which homologs are unknown or lack conserved residue patterns in interface region. Functional "hot-spot" residues are enriched among the predicted interface residues, rendering the method predestined for macromolecular binding site identification and drug design studies aiming at modulating protein-protein interaction that might influence protein function. For a comparative structural model of peptidase HtrA from Helicobacter pylori, we performed mutation studies for predicted hot-spot residues, which were confirmed as functionally relevant for HtrA activity or oligomerization.     

Covalent CouN7 enzyme intermediate for acyl group shuttling in aminocoumarin biosynthesis

The last stages of assembly of the aminocoumarin antibiotics, clorobiocin and coumermycin A(1), which target the GyrB subunits of bacterial DNA gyrase, involve enzymatic transfer of the pyrrolyl-2-carbonyl acyl group from a carrier protein (CloN1/CouN1) to the 3'-OH of the noviosyl moiety of the antibiotic scaffold. The enzyme, CouN7, will catalyze both the forward and back reaction on both arms of the coumermycin scaffold. This occurs via an O-acyl-Ser(101)-CouN7 intermediate, as shown by transient labeling of the enzyme with [(14)C]acetyl-S-CouN1 as donor and by inactivating mutation of the active site, Ser(101), to Ala. The intermediacy of the pyrrolyl-2-carbonyl-O-CouN7 allows net pyrrole transfer between distinct aminocoumarin scaffolds, for example, between the descarbamoylnovobiocin scaffold and coumermycin A(1) and vice versa. CouN7 also allows shuttling of surrogate acyl groups between noviosyl-aminocoumarin scaffolds to generate new antibiotic variants.     

Systematic assessment of scaffold distances in ChEMBL: prioritization of compound data sets for scaffold hopping analysis in virtual screening

The evaluation of the scaffold hopping potential of computational methods is of high relevance for virtual screening. For benchmark calculations, classes of known active compounds are utilized. Ideally, such classes should have a well-defined content of structurally diverse scaffolds. However, in reported benchmark investigations, the choice of activity classes is often difficult to rationalize. To provide a compendium of well-characterized test cases for the assessment of scaffold hopping potential, structural distances between scaffolds were systematically calculated for compound classes available in the ChEMBL database. Nearly seven million scaffold pairs were evaluated. On the basis of the global scaffold distance distribution, a threshold value for large scaffold distances was determined. Compound data sets were ranked based on the proportion of scaffold pairs with large distances they contained, taking additional criteria into account that are relevant for virtual screening. A set of 50 activity classes is provided that represent attractive test cases for scaffold hopping analysis and benchmark calculations.     

De Novo Design of Xeno‐Metallo Coiled Coils

Bioinorganic chemists aspire to achieve the same exquisite and highly controlled inorganic chemistry featured in biology. An exciting mimetic approach involves the use of miniature artificial protein scaffolds designed de novo (often based on the coiled coil (CC) scaffold), for reproducing native metal ion sites and their function. Recently, there is increased interest, instead, in the design of xeno‐metal sites within CC assemblies. This involves incorporating either non‐biological metal ions, cofactors or non‐proteinogenic amino acid ligands for metal ion coordination, whilst retaining a minimal CC protein scaffold. Using this approach, one should be able to create functional designs with unique and unusual properties, which combine the advantages of both biology and ‘traditional’ non‐biological inorganic chemistry. It is the recent progress with respect to the design of xeno‐metallo CCs which will be discussed in this Focus Review.     

Combining horizontal and vertical substructure relationships in scaffold hierarchies for activity prediction

For a systematic exploration of structural relationships between molecular scaffolds, ～24,000 unique scaffolds were extracted from 458 different target sets. Substructure relationships between these scaffolds were systematically determined. The scaffold tree data structure was utilized to study structural relationships between original scaffolds and derivative scaffolds obtained by rule-based decomposition. Leaf-to-root substructure relationships that resulted from rule-based decomposition were compared to leaf-to-leaf relationships between original scaffolds most of which were not part of the scaffold tree hierarchy. Decomposed scaffolds not contained in active target set compounds were prioritized on the basis of hierarchical scaffold patterns and additional substructure relationships. For high-priority virtual scaffolds, activity predictions were carried out, and these scaffolds were often found in external test compounds having the predicted activity. Taken together, our results suggest that leaf-to-root substructure relationships in scaffold trees should best be complemented with additional substructure relationships to determine high-priority virtual scaffolds for activity prediction.     

Metalloprotein and metallo-DNA/RNAzyme design: current approaches, success measures, and future challenges

Specific metal-binding sites have been found in not only proteins but also DNA and RNA molecules. Together these metalloenzymes consist of a major portion of the enzyme family and can catalyze some of the most difficult biological reactions. Designing these metalloenzymes can be both challenging and rewarding because it can provide deeper insights into the structure and function of proteins and cheaper and more stable alternatives for biochemical and biotechnological applications. Toward this goal, both rational and combinatorial approaches have been used. The rational approach is good for designing metalloenzymes that are well characterized, such as heme proteins, while the combinatorial approach is better at designing those whose structures are poorly understood, such as metallo-DNA/RNAzymes. Among the rational approaches, de novo design is at its best when metal-binding sites reside in a scaffold whose structure has been designed de novo (e.g., alpha-helical bundles). Otherwise, design using native scaffolds can be equally effective, allowing more choices of scaffolds whose structural stability is often more resistant to multiple mutations. In addition, computational and empirical designs have both enjoyed successes. Because of the limitation in defining structural parameters for metal-binding sites, a computational approach is restricted to mostly metal-binding sites that are well defined, such as mono- or homonuclear centers. An empirical approach, even though it is less restrictive in the metal-binding sites to be designed, depends heavily on one's knowledge and choice of templates and targets. An emerging approach is a combination of both computational and empirical approaches. The success of these approaches can be measured not only by three-dimensional structural comparison between the designed and target enzymes but also by the total amount of insight obtained from the design process and studies of the designed enzymes. One of the biggest advantages of designed metalloenzymes is the potential of placing two different metal-binding sites in the same protein framework for comparison. A final measure of success is how one can utilize the insight gained from the intellectual exercise to design new metalloenzymes, including those with unprecedented structures and functions. Future challenges include designing more complex metalloenzymes such as heteronuclear metal centers with strong nanomolar or better affinities. A key to meeting this challenge is to focus on the design of not only primary but also secondary coordination spheres using a combination of improved computer programs, experimental design, and high-resolution crystallography.     

Exploring the structural requirements in multiple chemical scaffolds for the selective inhibition of Plasmodium falciparum calcium-dependent protein kinase-1 (PfCDPK-1) by 3D-pharmacophore modelling,and docking studies

Current research on antimalarial protein kinases has provided an opportunity to design kinase-based antimalarial drugs. We have developed a common feature-based pharmacophore model from a set of multiple chemical scaffolds including derivatives of 3,6-imidazopyridazines, pyrazolo[2,3-d]pyrimidines and imidazo[1,5-a]pyrazines, in order to incorporate the maximum structural diversity information in the model for the Plasmodium falciparum calcium-dependent protein kinase-1 (PfCDPK-1) target. The best pharmacophore model (Hypo-1) with the essential features of two hydrogen bond donors (HBD), one hydrophobic aromatic (HYAr) and one ring aromatic (RA) showed the classification accuracies of 86.27%, 78.43% and 100.00% in labelling the training and test set (test set-1 and test set-2) compounds into more active and less active classes. In order to identify the crucial interaction between multiple scaffold ligands and the target protein, we first developed the homology model using a template structure of P. bergheii (PbCDPK1; PDB ID: 3Q5I), and thereafter performed the docking studies. The residues such as Lys85, Phe147, Tyr148, Leu198, Val211, and Asp212 were found to be the most important interacting residues for possessing PfCDPK-1 inhibitory activity.     

QMOD: physically meaningful QSAR

Computational methods for predicting ligand affinity where no protein structure is known generally take the form of regression analysis based on molecular features that have only a tangential relationship to a protein/ligand binding event. Such methods have utility in retrospective rationalization of activity patterns of substituents on a common scaffold, but are limited when either multiple scaffolds are present or when ligand alignment varies significantly based on structural changes. In addition, such methods generally assume independence and additivity of effect from scaffold substituents. Collectively, these non-physical modeling assumptions sharply limit the utility of widely used QSAR approaches for prospective prediction of ligand activity. The recently introduced Surflex-QMOD approach, by virtue of constructing physical models of binding sites, comes closer to a modeling approach that is congruent with protein ligand binding events. A set of congeneric CDK2 inhibitors showed that induced binding pockets can be quite congruent with the enzyme’s active site but that model predictivity within a chemical series does not necessarily depend on congruence. Muscarinic antagonists were used to show that the QMOD approach is capable of making accurate predictions in cases where highly non-additive structure activity effects exist. The QMOD method offers a means to go beyond non-causative correlations in QSAR analysis.     

De novo design by pharmacophore-based searches in fragment spaces

De novo ligand design supports the search for novel molecular scaffolds in medicinal chemistry projects. This search can either be based on structural information of the targeted active site (structure-based approach) or on similarity to known binders (ligand-based approach). In the absence of structural information on the target, pharmacophores provide a way to find topologically novel scaffolds. Fragment spaces have proven to be a valuable source for molecular structures in de novo design that are both diverse and synthetically accessible. They also offer a simple way to formulate custom chemical spaces. We have implemented a new method which stochastically constructs new molecules from fragment spaces under consideration of a three dimensional pharmacophore. The program has been tested on several published pharmacophores and is shown to be able to reproduce scaffold hops from the literature, which resulted in new chemical entities.     

Scaffold hopping in drug discovery using inductive logic programming

In chemoinformatics, searching for compounds which are structurally diverse and share a biological activity is called scaffold hopping. Scaffold hopping is important since it can be used to obtain alternative structures when the compound under development has unexpected side-effects. Pharmaceutical companies use scaffold hopping when they wish to circumvent prior patents for targets of interest. We propose a new method for scaffold hopping using inductive logic programming (ILP). ILP uses the observed spatial relationships between pharmacophore types in pretested active and inactive compounds and learns human-readable rules describing the diverse structures of active compounds. The ILP-based scaffold hopping method is compared to two previous algorithms (chemically advanced template search, CATS, and CATS3D) on 10 data sets with diverse scaffolds. The comparison shows that the ILP-based method is significantly better than random selection while the other two algorithms are not. In addition, the ILP-based method retrieves new active scaffolds which were not found by CATS and CATS3D. The results show that the ILP-based method is at least as good as the other methods in this study. ILP produces human-readable rules, which makes it possible to identify the three-dimensional features that lead to scaffold hopping. A minor variant of a rule learnt by ILP for scaffold hopping was subsequently found to cover an inhibitor identified by an independent study. This provides a successful result in a blind trial of the effectiveness of ILP to generate rules for scaffold hopping. We conclude that ILP provides a valuable new approach for scaffold hopping.     

Learning from the data: mining of large high-throughput screening databases

High-throughput screening (HTS) campaigns in pharmaceutical companies have accumulated a large amount of data for several million compounds over a couple of hundred assays. Despite the general awareness that rich information is hidden inside the vast amount of data, little has been reported for a systematic data mining method that can reliably extract relevant knowledge of interest for chemists and biologists. We developed a data mining approach based on an algorithm called ontology-based pattern identification (OPI) and applied it to our in-house HTS database. We identified nearly 1500 scaffold families with statistically significant structure-HTS activity profile relationships. Among them, dozens of scaffolds were characterized as leading to artifactual results stemming from the screening technology employed, such as assay format and/or readout. Four types of compound scaffolds can be characterized based on this data mining effort: tumor cytotoxic, general toxic, potential reporter gene assay artifact, and target family specific. The OPI-based data mining approach can reliably identify compounds that are not only structurally similar but also share statistically significant biological activity profiles. Statistical tests such as Kruskal-Wallis test and analysis of variance (ANOVA) can then be applied to the discovered scaffolds for effective assignment of relevant biological information. The scaffolds identified by our HTS data mining efforts are an invaluable resource for designing SAR-robust diversity libraries, generating in silico biological annotations of compounds on a scaffold basis, and providing novel target family specific scaffolds for focused compound library design.     

Target family-directed exploration of scaffolds with different SAR profiles

The scaffold concept is widely applied in chemoinformatics and medicinal chemistry to organize bioactive compounds according to common core structures or associate compound classes with specific biological activities. A variety of scaffold analyses have been carried out to derive statistics for scaffold distributions, generate structural organization schemes, or identify scaffolds that preferentially occur in given compound activity classes. Herein we further extend scaffold analysis by identifying scaffolds that display defined SAR profiles consisting of multiple properties. A structural relationship-based scaffold network has been designed as the basic data structure underlying our analysis. From network representations of scaffolds extracted from compounds active against 32 different target families, scaffolds with different SAR profiles have been extracted on the basis of decision trees that capture structural and functional characteristics of scaffolds in different ways. More than 600 scaffolds and 100 scaffold clusters were assigned to 10 SAR profiles. These scaffold sets represent different activity and target selectivity profiles and are provided for further SAR investigations including, for example, the exploration of alternative analog series for a given target of target family or the design of novel compounds on the basis of scaffold(s) with desired SAR profiles.     

Computational Design and Study of Artificial Selenoenzyme with Controllable Activity Based on an Allosteric Protein Scaffold

The establishment of new enzymatic function in an existing scaffold is a great challenge for protein engineers. In previous work, a highly efficient artificial selenoenzyme with controllable activity was constructed, based on a Ca²⁺-responsive recoverin (Rn) protein. In this study, a design strategy combining docking, molecular dynamics, and MM-PBSA is presented, to predict the catalytically active site of glutathione peroxidase (GPx) on the allosteric domain of Rn. The energy contributions of the binding hot spot residues are evaluated further by energy decomposition analysis to determine the detailed substrate recognition mechanism of Rn, which provides clear guidance for artificial enzyme design for improved substrate binding (Michaelis–Menten constant, K_m).