Hidden partners: Using cross‐docking calculations to predict binding sites for proteins with multiple interactions

Protein‐protein interactions control a large range of biological processes and their identification is essential to understand the underlying biological mechanisms. To complement experimental approaches, in silico methods are available to investigate protein‐protein interactions. Cross‐docking methods, in particular, can be used to predict protein binding sites. However, proteins can interact with numerous partners and can present multiple binding sites on their surface, which may alter the binding site prediction quality. We evaluate the binding site predictions obtained using complete cross‐docking simulations of 358 proteins with 2 different scoring schemes accounting for multiple binding sites. Despite overall good binding site prediction performances, 68 cases were still associated with very low prediction quality, presenting individual area under the specificity‐sensitivity ROC curve (AUC) values below the random AUC threshold of 0.5, since cross‐docking calculations can lead to the identification of alternate protein binding sites (that are different from the reference experimental sites). For the large majority of these proteins, we show that the predicted alternate binding sites correspond to interaction sites with hidden partners, that is, partners not included in the original cross‐docking dataset. Among those new partners, we find proteins, but also nucleic acid molecules. Finally, for proteins with multiple binding sites on their surface, we investigated the structural determinants associated with the binding sites the most targeted by the docking partners.     

Mapping of ligand‐binding cavities in proteins

The complex interactions between proteins and small organic molecules (ligands) are intensively studied because they play key roles in biological processes and drug activities. Here, we present a novel approach to characterize and map the ligand‐binding cavities of proteins without direct geometric comparison of structures, based on Principal Component Analysis of cavity properties (related mainly to size, polarity, and charge). This approach can provide valuable information on the similarities and dissimilarities, of binding cavities due to mutations, between‐species differences and flexibility upon ligand‐binding. The presented results show that information on ligand‐binding cavity variations can complement information on protein similarity obtained from sequence comparisons. The predictive aspect of the method is exemplified by successful predictions of serine proteases that were not included in the model construction. The presented strategy to compare ligand‐binding cavities of related and unrelated proteins has many potential applications within protein and medicinal chemistry, for example in the characterization and mapping of “orphan structures”, selection of protein structures for docking studies in structure‐based design, and identification of proteins for selectivity screens in drug design programs. Proteins 2010. © 2009 Wiley‐Liss, Inc.     

Molecular simulations of β‐lactoglobulin complexed with fatty acids reveal the structural basis of ligand affinity to internal and possible external binding sites

The interaction of saturated fatty acids of different length (C8:0 to C18:0) with β‐lactoglobulin (βLG) was investigated by molecular dynamics simulation and docking approaches. The results show that the presence of such ligands in the hydrophobic central cavity of βLG, known as the protein calyx, determines an enhancement of atomic fluctuations compared with the unliganded form, especially for loops at the entrance of the binding site. Concerted motions are evidenced for protein regions that could favor the binding of ligands. The mechanism of anchoring of fatty acids of different length is similar for the carboxylate head‐group, through electrostatic interactions with the side chains of Lys60/Lys69. The key protein residues to secure the hydrocarbon chain are Phe105/Met107, which adapt their conformation upon ligand binding. In particular, Phe105 provides an additional hydrophobic clamp only for the tail of the two fatty acids with the longest chains, palmitic, and stearic acid, which are known to bind βLG with a high affinity. The search of additional external binding sites for fatty acids, distinct from the calyx, was also carried out for palmitic acid. Two external sites with a lower affinity were identified as secondary sites, one consisting in a hydrophobic cavity allowing two distinct binding modes for the fatty acid, and the other corresponding to a surface crevice close to the protein α‐helix. The overall results provide a comprehensive picture of the dynamical behavior of βLG in complex with fatty acids, and elucidate the structural basis of the binding of these physiological ligands. Proteins 2014; 82:2609–2619. © 2014 Wiley Periodicals, Inc.     

Steric and thermodynamic limits of design for the incorporation of large unnatural amino acids in aminoacyl‐tRNA synthetase enzymes

Orthogonal aminoacyl‐tRNA synthetase/tRNA pairs from archaea have been evolved to facilitate site specific in vivo incorporation of unnatural amino acids into proteins in Escherichia coli. Using this approach, unnatural amino acids have been successfully incorporated with high translational efficiency and fidelity. In this study, CHARMM‐based molecular docking and free energy calculations were used to evaluate rational design of specific protein–ligand interactions for aminoacyl‐tRNA synthetases. A series of novel unnatural amino acid ligands were docked into the p‐benzoyl‐L ‐phenylalanine tRNA synthetase, which revealed that the binding pocket of the enzyme does not provide sufficient space for significantly larger ligands. Specific binding site residues were mutated to alanine to create additional space to accommodate larger target ligands, and then mutations were introduced to improve binding free energy. This approach was used to redesign binding sites for several different target ligands, which were then tested against the standard 20 amino acids to verify target specificity. Only the synthetase designed to bind Man‐α‐O‐Tyr was predicted to be sufficiently selective for the target ligand and also thermodynamically stable. Our study suggests that extensive redesign of the tRNA synthatase binding pocket for large bulky ligands may be quite thermodynamically unfavorable. Proteins 2010. © 2010 Wiley‐Liss, Inc.     

A method for localizing ligand binding pockets in protein structures

The accurate identification of ligand binding sites in protein structures can be valuable in determining protein function. Once the binding site is known, it becomes easier to perform in silico and experimental procedures that may allow the ligand type and the protein function to be determined. For example, binding pocket shape analysis relies heavily on the correct localization of the ligand binding site. We have developed SURFNET-ConSurf, a modular, two-stage method for identifying the location and shape of potential ligand binding pockets in protein structures. In the first stage, the SURFNET program identifies clefts in the protein surface that are potential binding sites. In the second stage, these clefts are trimmed in size by cutting away regions distant from highly conserved residues, as defined by the ConSurf-HSSP database. The largest clefts that remain tend to be those where ligands bind. To test the approach, we analyzed a nonredundant set of 244 protein structures from the PDB and found that SURFNET-ConSurf identifies a ligand binding pocket in 75% of them. The trimming procedure reduces the original cleft volumes by 30% on average, while still encompassing an average 87% of the ligand volume. From the analysis of the results we conclude that for those cases in which the ligands are found in large, highly conserved clefts, the combined SURFNET-ConSurf method gives pockets that are a better match to the ligand shape and location. We also show that this approach works better for enzymes than for nonenzyme proteins.     

Differences in binding behavior of (−)‐epigallocatechin gallate to β‐lactoglobulin heterodimers (AB) compared to homodimers (A) and (B)

The lipocalin β‐lactoglobulin (β‐LG) exists in different natural genetic variants—of which β‐LG A and B are predominant in bovine milk. At physiological conditions the protein dimerizes—building homodimers of β‐LG A and β‐LG B and heterodimers of β‐LG AB. Although β‐LG is one of the most intensely characterized lipocalins, the interaction behavior of ligands with hetero‐ and homodimers of β‐LG is largely unknown. The present findings revealed significant differences for hetero‐ and homodimers regarding ligand binding capacity as tested with a model ligand (i.e. surface binding (?)‐epigallocatechin gallate (EGCG)). These findings were confirmed using FT‐IR, where the addition of EGCG influenced the β‐sheet backbone of homodimer A and B with significantly higher intensity compared to heterodimer AB. Further, shape analysis by SAXS revealed oligomerization of both types of dimers upon addition of EGCG; however, homodimer A and B produced significantly larger aggregates compared to the heterodimer AB. In summary, the present study revealed that EGCG showed significantly different interaction reactivity (binding sites, aggregation size and conformational changes) to the hetero and homodimers of β‐LG in the order β‐LG A > B > AB. The results suggest that conformational differences between homodimers and heterodimers strongly influence the EGCG binding ability. This may also occur with other polyphenols and ligands of β‐LG and gives not only important information for β‐LG binding studies, but may also apply for polymorphisms of other self‐aggregating lipocalins. Copyright © 2015 John Wiley & Sons, Ltd.     

Studies on the binding sites of IgG2 monoclonal antibodies recognized by terpyridine‐based affinity ligands

This investigation has examined the origin of the molecular recognition associated with the interaction of monoclonal IgG2's with terpyridine‐based ligands immobilized onto agarose‐derived chromatographic adsorbents. Isothermal titration calorimetric (ITC) methods have been employed to acquire thermodynamic data associated with the IgG2‐ligand binding. These ITC investigations have documented that different enthalpic and entropic processes are involved depending on the nature of the chemical substituents in the core structure of the terpyridinyl moiety. In addition, molecular docking studies have been carried out with IgG2 structures with the objective to identify possible ligand binding sites and key interacting amino acid residues. These molecular docking experiments with the different terpyridine‐based ligands have shown that all of the examined ligands can potentially undergo favorable interactions with a site located within the Fab region of the IgG2. However, another favorable binding site was also identified from the docking poses to exist within the Fc region of the IgG2 for some, but not all, of the ligands studied. These investigations have provided a basis to elucidate the unique binding properties and chromatographic behaviors shown by several substituted terpyridine ligands in their interaction with IgGs of different isotype. Copyright © 2016 John Wiley & Sons, Ltd.     

Molecular insights into the binding selectivity of a synthetic ligand DAAG to Fc fragment of IgG

Affinity chromatography with synthetic ligands has been focused as the potential alternative to protein A‐based chromatography for antibody capture because of its comparable selectivity and efficiency. Better understanding on the molecular interactions between synthetic ligand and antibody is crucial for improving and designing novel ligands. In this work, the molecular interaction mechanism between Fc fragment of IgG and a synthetic ligand (DAAG) was studied with molecular docking and dynamics simulation. The docking results on the consensus binding site (CBS) indicated that DAAG could bind to the CBS with the favorable orientation like a tripod for the top‐ranked binding complexes. The ligand‐Fc fragment complexes were then tested by molecular dynamics simulation at neutral condition (pH 7.0) for 10 ns. The results indicated that the binding of DAAG on the CBS of Fc fragment was achieved by the multimodal interactions, combining the hydrophobic interaction, electrostatic interaction, hydrogen bond, and so on. It was also found that multiple secondary interactions endowed DAAG with an excellent selectivity to Fc fragment. In addition, molecular dynamics simulation conducted at acidic condition (pH 3.0) showed that the departure of DAAG ligand from the surface of Fc fragment was the result of reduced interaction energies. The binding modes between DAAG and CBS not only shed light on the molecular mechanisms of DAAG for antibody purification but also provide useful information for the improvement of ligand design. Copyright © 2014 John Wiley & Sons, Ltd.     

Pre-docking filter for protein and ligand 3D structures

Virtual drug screening using protein-ligand docking techniques is a time-consuming process, which requires high computational power for binding affinity calculation. There are millions of chemical compounds available for docking. Eliminating compounds that are unlikely to exhibit high binding affinity from the screening set should speed-up the virtual drug screening procedure. We performed docking of 6353 ligands against twenty-one protein X-ray crystal structures. The docked ligands were ranked according to their calculated binding affinities, from which the top five hundred and the bottom five hundred were selected. We found that the volume and number of rotatable bonds of the top five hundred docked ligands are similar to those found in the crystal structures and corresponded with the volume of the binding sites. In contrast, the bottom five hundred set contains ligands that are either too large to enter the binding site, or too small to bind with high specificity and affinity to the binding site. A pre-docking filter that takes into account shapes and volumes of the binding sites as well as ligand volumes and flexibilities can filter out low binding affinity ligands from the screening sets. Thus, the virtual drug screening procedure speed is increased.     

Studying the roles of W86, E202, and Y337 in binding of acetylcholine to acetylcholinesterase using a combined molecular dynamics and multiple docking approach

A combined molecular dynamics simulation and multiple ligand docking approach is applied to study the roles of the anionic subsite residues (W86, E202, Y337) in the binding of acetylcholine (ACh) to acetylcholinesterase (AChE). We find that E202 stabilizes docking of ACh via electrostatic interactions. However, we find no significant electrostatic contribution from the aromatic residues. Docking energies of ACh to mutant AChE show a more pronounced effect because of size/shape complementarity. Mutating to smaller residues results in poorer binding, both in terms of docking energy and statistical docking probability. Besides separating out electrostatics by turning off the partial charges from each residue and comparing it with the native, the mutations in this study are W86F, W86A, E202D, E202Q, E202A, Y337F, and Y337A. We also find that all perturbations result in a significant reduction in binding of extended ACh in the catalytically productive orientation. This effect is primarily caused by a small shift in preferred position of the quaternary tail.     

Josef Rudinger Memorial Lecture 2002. The chemistry of the opioid receptor binding sites.

Since the discovery of the opioid receptors and their endogenous ligands an immense research work has been devoted to the exploration of their specificity, the mechanism of ligand binding and ligand-receptor interactions. One of the main goals has been the location and characterization of the binding sites. The present review compiles the results achieved in this field in the last quarter of a century, and puts some questions concerning the success of these efforts.     

[3H]Neurokinin B and 125I-Bolton Hunter Eledoisin Label Identical Tachykinin Binding Sites in the Rat Brain

[3H]Neurokinin B ([3H]NKB) of high specific activity (75 Ci/mmol) was synthesized for study of its binding to crude synaptosomes from the rat cerebral cortex. The specific binding of [3H]NKB (75% of total binding) was temperature dependent, saturable, and reversible. Scatchard analyses and Hill plots showed the existence of a single population of noninteracting binding sites (KD = 4.3 nM; Bmax = 123 fmol/mg of protein). Competition studies indicated the following rank order of potencies among tachykinins: NKB greater than eledoisin (E) greater than kassinin greater than physalaemin greater than neurokinin A (NKA) greater than substance P (SP), a result suggesting that NKB might be the endogenous ligand for [3H]NKB binding sites. It is of interest that 127I-Bolton Hunter (BH) NKA (127I-BHNKA) was much more potent than NKA in inhibiting the specific binding of [3H]NKB, which raises certain questions concerning the use of 125I-BHNKA as a ligand for NKA binding sites in the brain. These results, as well as those obtained with different SP analogues, show a close similarity to those obtained previously with 125I-BHE binding to cortical synaptosomes. This suggested that the two ligands labeled identical binding sites. In addition, using either [3H]NKB or 125I-BHE as ligands, similar displacement curves were obtained with increasing concentrations of NKB and 127I-BHE. The similarity of the [3H]NKB and 125I-BHE binding sites was further confirmed by comparison of their localization on rat brain sections by autoradiography. The distribution of binding sites for [3H]NKB and 125I-BHE was identical throughout the brain, and the highest density of binding sites for the two ligands was found in layers IV and V of the cerebral cortex, the paraventricular nucleus of the hypothalamus (magnocellular part), and the ventral tegmental area.     

Improving accuracy and efficiency of blind protein-ligand docking by focusing on predicted binding sites

The use of predicted binding sites (binding sites calculated from the protein structure alone) is evaluated here as a tool to focus the docking of small molecule ligands into protein structures, simulating cases where the real binding sites are unknown. The resulting approach consists of a few independent docking runs carried out on small boxes, centered on the predicted binding sites, as opposed to one larger blind docking run that covers the complete protein structure. The focused and blind approaches were compared using a set of 77 known protein-ligand complexes and 19 ligand-free structures. The focused approach is shown to: (1) identify the correct binding site more frequently than blind docking; (2) produce more accurate docking poses for the ligand; (3) require less computational time. Additionally, the results show that very few real binding sites are missed in spite of focusing on only three predicted binding sites per target protein. Overall the results indicate that, by improving the sampling in regions that are likely to correspond to binding sites, the focused docking approach increases accuracy and efficiency of protein ligand docking for those cases where the ligand-binding site is unknown. This is especially relevant in applications such as reverse virtual screening and structure-based functional annotation of proteins.     

Profiling the interaction of 1‐phenylbenzimidazoles to cyclooxygenases

In the design of 1‐phenylbenzimidazoles as model cyclooxygenase (COX) inhibitors, docking to a series of crystallographic COX structures was performed to evaluate their potential for high‐affinity binding and to reproduce the interaction profile of well‐known COX inhibitors. The effect of ligand‐specific induced fit on the calculations was also studied. To quantitatively compare the pattern of interactions of model compounds to the profile of several cocrystallized COX inhibitors, a geometric parameter, denominated ligand‐receptor contact distance (LRCD), was developed. The interaction profile of several model complexes showed similarity to the profile of COX complexes with inhibitors such as iodosuprofen, iodoindomethacin, diclofenac, and flurbiprofen. Shaping of high‐affinity binding sites upon ligand‐specific induced fit mostly determined both the affinity and the binding mode of the ligands in the docking calculations. The results suggest potential of 1‐phenylbenzimidazole derivatives as COX inhibitors on the basis of their predicted affinity and interaction profile to COX enzymes. The analyses also provided insights into the role of induced fit in COX enzymes. While inhibitors produce different local structural changes at the COX ligand binding site, induced fit allows inhibitors in diverse chemical classes to share characteristic interaction patterns that ensure key contacts to be achieved. Different interaction patterns may also be associated with different inhibitory mechanisms.     

Predicting molecular interactions and inducible complementarity: Fragment docking of fab-peptide complexes

Antibody-antigen interactions are representative of a broad class of receptor-ligand interactions involving both specificity and potential inducible complementarity. To test possible mechanisms of antigenantibody recognition and specificity computationally, we have used a Metropolis Monte Carlo algorithm to dock fragments of the epitope Glu-Val-Val-Pro-His-Lys-Lys to the X-ray structures of both the free and the complexed Fab of the antibody B13I2 (raised against the C-helix of myohemerythri). The fragments Pro-His and Val-Pro-His, which contain residues experimentally identified as important for binding, docked correctly to both structures, but all tetrapeptide and larger fragments docked correctly only to the complexed Fab, even when torsional flexibility was added to the ligand. However, only tetrapeptide and larger fragments showed significantly more favorable energies when docked to the complexed Fab coordinates than when docked to either the free Fab or a non-specific site remote from the combining site. Comparison of the free and complexed B13I2 structures revealed that atoms within 5 Å of Val-Pro-His showed little movement upon peptide binding, but atoms within 5 Å of the other four epitope residues showed greater movements. These results computationally distinguish recognition and binding processes with practical implications for drug design strategies. Overall, this new fragment docking approach establishes distinct roles for the “lock-and-key” (recognition) and the “handshake” (binding) paradigms in antibody-antigen interaction, suggests an incremental approach to incorporating flexibility in computational docking, and identifies critical regions within receptor binding sites for ligand recognition. © 1994 Wiley-Liss, Inc.     

Binding of nonsteroidal anti‐inflammatory drugs to Aβ fibril

Nonsteroidal anti‐inflammatory drugs are considered as potential therapeutic agents against Alzheimer's disease. Using replica exchange molecular dynamics and atomistic implicit solvent model, we studied the mechanisms of binding of naproxen and ibuprofen to the Aβ fibril derived from solid‐state NMR measurements. The binding temperature of naproxen is found to be almost 40 K higher than of ibuprofen implicating higher binding affinity of naproxen. The key factor, which enhances naproxen binding, is strong interactions between ligands bound to the surface of the fibril. The naphthalene ring in naproxen appears to provide a dominant contribution to ligand‐ligand interactions. In contrast, ligand‐fibril interactions cannot explain differences in the binding affinities of naproxen and ibuprofen. The concave fibril edge with the groove is identified as the primary binding location for both ligands. We show that confinement of the ligands to the groove facilitates ligand‐ligand interactions that lowers the energy of the ligands bound to the concave edge compared with those bound to the convex edge. Our simulations appear to provide microscopic rationale for the differing binding affinities of naproxen and ibuprofen observed experimentally. Proteins 2010. © 2010 Wiley‐Liss, Inc.     

Merging chemical and biological space: Structural mapping of enzyme binding pocket space

Structure‐based drug design tries to mutually map pharmacological space populated by putative target proteins onto chemical space comprising possible small molecule drug candidates. Both spaces are connected where proteins and ligands recognize each other: in the binding pockets. Therefore, it is highly relevant to study the properties of the space composed by all possible binding cavities. In the present contribution, a global mapping of protein cavity space is presented by extracting consensus cavities from individual members of protein families and clustering them in terms of their shape and exposed physicochemical properties. Discovered similarities indicate common binding epitopes in binding pockets independent of any possibly given similarity in sequence and fold space. Unexpected links between remote targets indicate possible cross‐reactivity of ligands and suggest putative side effects. The global clustering of cavity space is compared to a similar clustering of sequence and fold space and compared to chemical ligand space spanned by the chemical properties of small molecules found in binding pockets of crystalline complexes. The overall similarity architecture of sequence, fold, and cavity space differs significantly. Similarities in cavity space can be mapped best to similarities in ligand binding space indicating possible cross‐reactivities. Most cross‐reactivities affect co‐factor and other endogenous ligand binding sites. Proteins 2009. © 2008 Wiley‐Liss, Inc.     

Exploration of human serum albumin binding sites by docking and molecular dynamics flexible ligand–protein interactions

Five‐nanosecond molecular dynamics (MD) simulations were performed on human serum albumin (HSA) to study the conformational features of its primary ligand binding sites (I and II). Additionally, 11 HSA snapshots were extracted every 0.5 ns to explore the binding affinity (K_d) of 94 known HSA binding drugs using a blind docking procedure. MD simulations indicate that there is considerable flexibility for the protein, including the known sites I and II. Movements at HSA sites I and II were evidenced by structural analyses and docking simulations. The latter enabled the study and analysis of the HSA–ligand interactions of warfarin and ketoprofen (ligands binding to sites I and II, respectively) in greater detail. Our results indicate that the free energy values by docking (K_d observed) depend upon the conformations of both HSA and the ligand. The 94 HSA–ligand binding K_d values, obtained by the docking procedure, were subjected to a quantitative structure‐activity relationship (QSAR) study by multiple regression analysis. The best correlation between the observed and QSAR theoretical (K_d predicted) data was displayed at 2.5 ns. This study provides evidence that HSA binding sites I and II interact specifically with a variety of compounds through conformational adjustments of the protein structure in conjunction with ligand conformational adaptation to these sites. These results serve to explain the high ligand‐promiscuity of HSA. © 2009 Wiley Periodicals, Inc. Biopolymers 93: 161–170, 2010. This article was originally published online as an accepted preprint. The “Published Online” date corresponds to the preprint version. You can request a copy of the preprint by emailing the Biopolymers editorial office at biopolymers@wiley.com     

Fast and automated functional classification with MED‐SuMo: An application on purine‐binding proteins

Ligand–protein interactions are essential for biological processes, and precise characterization of protein binding sites is crucial to understand protein functions. MED‐SuMo is a powerful technology to localize similar local regions on protein surfaces. Its heuristic is based on a 3D representation of macromolecules using specific surface chemical features associating chemical characteristics with geometrical properties. MED‐SMA is an automated and fast method to classify binding sites. It is based on MED‐SuMo technology, which builds a similarity graph, and it uses the Markov Clustering algorithm. Purine binding sites are well studied as drug targets. Here, purine binding sites of the Protein DataBank (PDB) are classified. Proteins potentially inhibited or activated through the same mechanism are gathered. Results are analyzed according to PROSITE annotations and to carefully refined functional annotations extracted from the PDB. As expected, binding sites associated with related mechanisms are gathered, for example, the Small GTPases. Nevertheless, protein kinases from different Kinome families are also found together, for example, Aurora‐A and CDK2 proteins which are inhibited by the same drugs. Representative examples of different clusters are presented. The effectiveness of the MED‐SMA approach is demonstrated as it gathers binding sites of proteins with similar structure‐activity relationships. Moreover, an efficient new protocol associates structures absent of cocrystallized ligands to the purine clusters enabling those structures to be associated with a specific binding mechanism. Applications of this classification by binding mode similarity include target‐based drug design and prediction of cross‐reactivity and therefore potential toxic side effects.