PAIRpred: Partner‐specific prediction of interacting residues from sequence and structure

We present a novel partner‐specific protein–protein interaction site prediction method called PAIRpred. Unlike most existing machine learning binding site prediction methods, PAIRpred uses information from both proteins in a protein complex to predict pairs of interacting residues from the two proteins. PAIRpred captures sequence and structure information about residue pairs through pairwise kernels that are used for training a support vector machine classifier. As a result, PAIRpred presents a more detailed model of protein binding, and offers state of the art accuracy in predicting binding sites at the protein level as well as inter‐protein residue contacts at the complex level. We demonstrate PAIRpred's performance on Docking Benchmark 4.0 and recent CAPRI targets. We present a detailed performance analysis outlining the contribution of different sequence and structure features, together with a comparison to a variety of existing interface prediction techniques. We have also studied the impact of binding‐associated conformational change on prediction accuracy and found PAIRpred to be more robust to such structural changes than existing schemes. As an illustration of the potential applications of PAIRpred, we provide a case study in which PAIRpred is used to analyze the nature and specificity of the interface in the interaction of human ISG15 protein with NS1 protein from influenza A virus. Python code for PAIRpred is available at http://combi.cs.colostate.edu/supplements/pairpred/ . Proteins 2014; 82:1142–1155. © 2013 Wiley Periodicals, Inc.     

Relationships Between the Catechol Substrate Binding Site and Amphetamine, Cocaine, and Mazindol Binding Sites in a Kinetic Model of the Striatal Transporter of Dopamine In Vitro

Abstract: Experiments were conducted to determine how (−)-cocaine and S (+)-amphetamine binding sites relate to each other and to the catechol substrate site on the striatal dopamine transporter (sDAT). In controls, m -tyramine and S (+)-amphetamine caused release of dopamine from intracellular stores at concentrations ≥12-fold those observed to inhibit inwardly directed sDAT activity for dopamine. In preparations from animals pretreated with reserpine, m -tyramine and S (+)-amphetamine caused release of preloaded dopamine at concentrations similar to those that inhibit inwardly directed sDAT activity. S (+)-Amphetamine and m -tyramine inhibited sDAT activity for dopamine by competing for a common binding site with dopamine and each other, suggesting that phenethylamines are substrate analogues at the plasmalemmal sDAT. (−)-Cocaine inhibited sDAT at a site separate from that for substrate analogues. This site is mutually interactive with the substrate site ( K _int = 583 n M ). Mazindol competitively inhibited sDAT at the substrate analogue binding site. The results with (−)-cocaine suggest that the (−)-cocaine binding site on sDAT is distinct from that of hydroxyphenethylamine substrates, reinforcing the notion that an antagonist for (−)-cocaine binding may be developed to block (−)-cocaine binding with minimal effects on dopamine transporter activity. However, a strategy of how to antagonize drugs of abuse acting as substrate analogues is still elusive.     

A Hybrid Color Space for Skin Detection Using Genetic Algorithm Heuristic Search and Principal Component Analysis Technique

Color is one of the most prominent features of an image and used in many skin and face detection applications. Color space transformation is widely used by researchers to improve face and skin detection performance. Despite the substantial research efforts in this area, choosing a proper color space in terms of skin and face classification performance which can address issues like illumination variations, various camera characteristics and diversity in skin color tones has remained an open issue. This research proposes a new three-dimensional hybrid color space termed SKN by employing the Genetic Algorithm heuristic and Principal Component Analysis to find the optimal representation of human skin color in over seventeen existing color spaces. Genetic Algorithm heuristic is used to find the optimal color component combination setup in terms of skin detection accuracy while the Principal Component Analysis projects the optimal Genetic Algorithm solution to a less complex dimension. Pixel wise skin detection was used to evaluate the performance of the proposed color space. We have employed four classifiers including Random Forest, Naïve Bayes, Support Vector Machine and Multilayer Perceptron in order to generate the human skin color predictive model. The proposed color space was compared to some existing color spaces and shows superior results in terms of pixel-wise skin detection accuracy. Experimental results show that by using Random Forest classifier, the proposed SKN color space obtained an average F-score and True Positive Rate of 0.953 and False Positive Rate of 0.0482 which outperformed the existing color spaces in terms of pixel wise skin detection accuracy. The results also indicate that among the classifiers used in this study, Random Forest is the most suitable classifier for pixel wise skin detection applications.     

Prediction of ligand-binding sites of proteins by molecular docking calculation for a random ligand library

A new approach to predicting the ligand-binding sites of proteins was developed, using protein-ligand docking computation. In this method, many compounds in a random library are docked onto the whole protein surface. We assumed that the true ligand-binding site would exhibit stronger affinity to the compounds in the random library than the other sites, even if the random library did not include the ligand corresponding to the true binding site. We also assumed that the affinity of the true ligand-binding site would be correlated to the docking scores of the compounds in the random library, if the ligand-binding site was correctly predicted. We call this method the molecular-docking binding-site finding (MolSite) method. The MolSite method was applied to 89 known protein-ligand complex structures extracted from the Protein Data Bank, and it predicted the correct binding sites with about 80-99% accuracy, when only the single top-ranked site was adopted. In addition, the average docking score was weakly correlated to the experimental protein-ligand binding free energy, with a correlation coefficient of 0.44.     

CaMELS: In silico prediction of calmodulin binding proteins and their binding sites

Due to Ca²⁺‐dependent binding and the sequence diversity of Calmodulin (CaM) binding proteins, identifying CaM interactions and binding sites in the wet‐lab is tedious and costly. Therefore, computational methods for this purpose are crucial to the design of such wet‐lab experiments. We present an algorithm suite called CaMELS (CalModulin intEraction Learning System) for predicting proteins that interact with CaM as well as their binding sites using sequence information alone. CaMELS offers state of the art accuracy for both CaM interaction and binding site prediction and can aid biologists in studying CaM binding proteins. For CaM interaction prediction, CaMELS uses protein sequence features coupled with a large‐margin classifier. CaMELS models the binding site prediction problem using multiple instance machine learning with a custom optimization algorithm which allows more effective learning over imprecisely annotated CaM‐binding sites during training. CaMELS has been extensively benchmarked using a variety of data sets, mutagenic studies, proteome‐wide Gene Ontology enrichment analyses and protein structures. Our experiments indicate that CaMELS outperforms simple motif‐based search and other existing methods for interaction and binding site prediction. We have also found that the whole sequence of a protein, rather than just its binding site, is important for predicting its interaction with CaM. Using the machine learning model in CaMELS, we have identified important features of protein sequences for CaM interaction prediction as well as characteristic amino acid sub‐sequences and their relative position for identifying CaM binding sites. Python code for training and evaluating CaMELS together with a webserver implementation is available at the URL: http://faculty.pieas.edu.pk/fayyaz/software.html#camels .     

Crystallographic studies of the complexes of antiviral protein griffithsin with glucose and N-acetylglucosamine

Crystal structures of complexes of an antiviral lectin griffithsin (GRFT) with glucose and N-acetylglucosamine were solved and refined at high resolution. In both complexes, all six monosaccharide-binding sites of GRFT were occupied and the mode of binding was similar to that of mannose. In our previous attempts to obtain a complex with N-acetylglucosamine by soaking, only a single site was occupied; thus, cocrystallization was clearly superior despite lower concentration of the ligand. Isothermal titration calorimetric experiments with N-acetylglucosamine, glucose, and mannose provided enthalpic evidence of distinct binding differences between the three monosaccharides. A comparison of the mode of binding of different monosaccharides is discussed in the context of the antiviral activity of GRFT, based on specific binding to high-mannose-containing complex carbohydrates found on viral envelopes.     

Characterization of the high affinity heparin binding site of the Alzheimer's disease βA4 amyloid precursor protein (APP) and its enhancement by zinc(II)

The Alzheimer's disease βA4 amyloid precursor protein (APP) has been shown to be involved in a diverse set of biological protein precursor-like proteins (APLP1 and APLP2) belong to a superfamily of proteins that are probably functionally related. In order to characterize the cell adhesion properties of APP the brain specific isoform APP₆₉₅ was purified and used to assess the binding to herparin, a structural and functional analogue of the glycosaminoglycan heparan sulfate. We show that APP binds in a time dependent and saturable manner to heparin. The salt concentration of 620 mM at which APP elutes from heparin Sepharose is greater than physiological. Tha apparent equilibrium constant for dissociation was determined to be 300 pM for APP binding to heparin Sepharose. A high affinity heparin binding site was identified within a region conversed in rodent and human APP, APLP1 and APLP2. This binding site was located between residues 316-337 of APP₆₉₅ which is within the carbohydrate domain of APP. We also demonstrate an interaction between this heparin binding site and the zinc(II) binding site which is conserved in all members of the APP superfamily. We show by using an automated surface plasmon resonance biosensor (BIAcore, Pharmacia) that the affinity for heparin is increased two- to four-fold in the presence of micromolar zinc(II). The identification of zinc-enhanced binding of APP to heparin sulfate side chains of proteoglycans offers a molecular link between zinc(II), as a putative environmental toxin for Alzheimer's disease, and aggregation of amyloid βA4 protein.     

Structural basis of neurophysin hormone specificity: Geometry, polarity, and polarizability in aromatic ring interactions

The structural origins of the specificity of the neurophysin hormone-binding site for an aromatic residue in peptide position 2 were explored by analyzing the binding of a series of peptides in the context of the crystal structure of liganded neurophysin. A new modeling method for describing the van der Waals surface of binding sites assisted in the analysis. Particular attention was paid to the unusually large (5 kcal/mol) difference in binding free energy between Phe and Leu in position 2, a value representing more than three times the maximum expected based on hydrophobicity alone, and additionally remarkable since modeling indicated that the Leu side chain was readily accommodated by the binding pocket. Although evidence was obtained of a weak thermodynamic linkage between the binding interactions of the residue 2 side chain and of the peptide alpha-amino group, two factors are considered central. (1) The bound Leu side chain can establish only one-third of the van der Waals contacts available to a Phe side chain. (2) The bound Phe side chain appears to be additionally stabilized relative to Leu by more favorable dipole and induced dipole interactions with nonaromatic polar and sulfur ligands in the binding pocket, as evidenced by examination of its interactions in the pocket, analysis of the detailed energetics of transfer of Phe and Leu side chains from water to other phases, and comparison with thermodynamic and structural data for the binding of residue 1 side chains in this system. While such polar interactions of aromatic rings have been previously observed, the present results suggest their potential for significant thermodynamic contributions to protein structure and ligand recognition.     

FTMove: A Web Server for Detection and Analysis of Cryptic and Allosteric Binding Sites by Mapping Multiple Protein Structures

Protein mapping distributes many copies of different molecular probes on the surface of a target protein in order to determine binding hot spots, regions that are highly preferable for ligand binding. While mapping of X-ray structures by the FTMap server is inherently static, this limitation can be overcome by the simultaneous analysis of multiple structures of the protein. FTMove is an automated web server that implements this approach. From the input of a target protein, by PDB code, the server identifies all structures of the protein available in the PDB, runs mapping on them, and combines the results to form binding hot spots and binding sites. The user may also upload their own protein structures, bypassing the PDB search for similar structures. Output of the server consists of the consensus binding sites and the individual mapping results for each structure - including the number of probes located in each binding site, for each structure. This level of detail allows the users to investigate how the strength of a binding site relates to the protein conformation, other binding sites, and the presence of ligands or mutations. In addition, the structures are clustered on the basis of their binding properties. The use of FTMove is demonstrated by application to 22 proteins with known allosteric binding sites; the orthosteric and allosteric binding sites were identified in all but one case, and the sites were typically ranked among the top five. The FTMove server is publicly available at https://ftmove.bu.edu.     

Crystal structure of beta1,4-galactosyltransferase complex with UDP-Gal reveals an oligosaccharide acceptor binding site

The crystal structure of the catalytic domain of bovine beta1,4-galactosyltransferase (Gal-T1) co-crystallized with UDP-Gal and MnCl(2) has been solved at 2.8 A resolution. The structure not only identifies galactose, the donor sugar binding site in Gal-T1, but also reveals an oligosaccharide acceptor binding site. The galactose moiety of UDP-Gal is found deep inside the catalytic pocket, interacting with Asp252, Gly292, Gly315, Glu317 and Asp318 residues. Compared to the native crystal structure reported earlier, the present UDP-Gal bound structure exhibits a large conformational change in residues 345-365 and a change in the side-chain orientation of Trp314. Thus, the binding of UDP-Gal induces a conformational change in Gal-T1, which not only creates the acceptor binding pocket for N-acetylglucosamine (GlcNAc) but also establishes the binding site for an extended sugar acceptor. The presence of a binding site that accommodates an extended sugar offers an explanation for the observation that an oligosaccharide with GlcNAc at the non-reducing end serves as a better acceptor than the monosaccharide, GlcNAc. Modeling studies using oligosaccharide acceptors indicate that a pentasaccharide, such as N-glycans with GlcNAc at their non-reducing ends, fits the site best. A sequence comparison of the human Gal-T family members indicates that although the binding site for the GlcNAc residue is highly conserved, the site that binds the extended sugar exhibits large variations. This is an indication that different Gal-T family members prefer different types of glycan acceptors with GlcNAc at their non-reducing ends.     

Molecular docking using surface complementarity

A method is described to dock a ligand into a binding site in a protein on the basis of the complementarity of the inter-molecular atomic contacts. Docking is performed by maximization of a complementarity function that is dependent on atomic contact surface area and the chemical properties of the contacting atoms. The generality and simplicity of the complementarity function ensure that a wide range of chemical structures can be handled. The ligand and the protein are treated as rigid bodies, but displacement of a small number of residues lining the ligand binding site can be taken into account. The method can assist in the design of improved ligands by indicating what changes in complementarity may occur as a result of the substitution of an atom in the ligand. The capabilities of the method are demonstrated by application to 14 protein–ligand complexes of known crystal structure. © 1996 Wiley Liss, Inc.     

Flavylium salts as in vitro precursors of potent ligands to brain GABA-A receptors

The synthesis of a series of derivatized flavylium cations was undertaken and the affinity to the benzodiazepine binding site of the GABA-A receptor evaluated. The observed high affinity for some derivatives (sub-μM range) was explained by an in vitro transformation of the flavylium cations into the corresponding trans-retrochalcones, components which are proposed to be the active species in this series.     

Computational analysis of multivalency in lectins: structures of garlic lectin-oligosaccharide complexes and their aggregates

Multivalency in lectins is a phenomenon that has been discussed at considerable length. The structural basis for the role of multivalency in garlic lectin has been investigated here through computational studies. Biochemical studies have shown that the binding affinity of garlic lectin for high mannose oligosaccharides is orders of magnitude greater than that for mannose. Modeling and energy calculations clearly indicate that such increase in affinity cannot be accounted for by binding of these oligosaccharides at any of the six sites of a garlic lectin dimer. These studies also indicate that a given oligosaccharide cannot bind simultaneously to more than one binding site on a lectin dimer. The possibility of a given oligosaccharide simultaneously binding to and hence linking two or more lectin molecules was therefore explored. This study showed that trimannosides and higher oligomers can cross-link lectin dimers, amplifying the protein-oligosaccharide interactions severalfold, thus explaining the role of multivalency in enhancing affinity. A comprehensive exploration of all possible cross-links posed a formidable computational problem. Even a partial exploration involving a carefully chosen region of the conformational space clearly showed that a given dimer pair can be cross-linked not only by a single oligosaccharide molecule but also simultaneously by two oligosaccharides. The number of such possible double cross-links, including those forming interesting tetrameric structures, generally increases with the size of the oligosaccharide, correlating with the biochemical data. In addition to their immediate relevance to garlic lectin, these studies are of general interest in relation to lectin-oligosaccharide interactions.     

A holistic molecular docking approach for predicting protein-protein complex structure

A holistic protein-protein molecular docking approach, HoDock, was established, composed of such steps as binding site prediction, initial complex structure sampling, refined complex structure sampling, structure clustering, scoring and final structure selection. This article explains the detailed steps and applications for CAPRI Target 39. The CAPRI result showed that three predicted binding site residues, A191HIS, B512ARG and B531ARG, were correct, and there were five submitted structures with a high fraction of correct receptor-ligand interface residues, indicating that this docking approach may improve prediction accuracy for protein-protein complex structures.     

The binding requirements of monkey brain lysosomal enzymes to their immobilised receptor protein

The lysosomal enzyme binding protein (receptor protein) isolated from monkey brain was immobilised on Sepharose 4B and used to study the binding of brain lysosomal enzymes. The immobilised protein could bind \-D-glucosaminidase, α-D-mannosidase, α-L-fucosidase and2-D-glucuronidase. The bound enzymes could be eluted either at an acid pH of 4.5 or by mannose 6-phosphate but not by a number of other sugars tested. Binding could be abolished by prior treatment of the lysosomal enzymes with sodium periodate. Alkaline phosphatase treatment of the enzymes did not prevent the binding of the lysosomal enzymes to the column but decreased their affinity, as seen by a shift in their elution profile, when a gradient elution with mannose 6-phosphate was employed. These results suggested that an ‘uncovered’ phosphate on the carbohydrate moiety of the enzymes was not essential for binding but can enhance the binding affinity.     

Crystal structure of human pepsin and its complex with pepstatin.

The three-dimensional crystal structure of human pepsin and that of its complex with pepstatin have been solved by X-ray crystallographic methods. The native pepsin structure has been refined with data collected to 2.2 A resolution to an R-factor of 19.7%. The pepsin:pepstatin structure has been refined with data to 2.0 A resolution to an R-factor of 18.5%. The hydrogen bonding interactions and the conformation adopted by pepstatin are very similar to those found in complexes of pepstatin with other aspartic proteinases. The enzyme undergoes a conformational change upon inhibitor binding to enclose the inhibitor more tightly. The analysis of the binding sites indicates that they form an extended tube without distinct binding pockets. By comparing the residues on the binding surface with those of the other human aspartic proteinases, it has been possible to rationalize some of the experimental data concerning the different specificities. At the S1 site, valine at position 120 in renin instead of isoleucine, as in the other enzymes, allows for binding of larger hydrophobic residues. The possibility of multiple conformations for the P2 residue makes the analysis of the S2 site difficult. However, it is possible to see that the specific interactions that renin makes with histidine at P2 would not be possible in the case of the other enzymes. At the S3 site, the smaller volume that is accessible in pepsin compared to the other enzymes is consistent with its preference for smaller residues at the P3 position.     

Protein function annotation by local binding site surface similarity

Hundreds of protein crystal structures exist for proteins whose function cannot be confidently determined from sequence similarity. Surflex‐PSIM, a previously reported surface‐based protein similarity algorithm, provides an alternative method for hypothesizing function for such proteins. The method now supports fully automatic binding site detection and is fast enough to screen comprehensive databases of protein binding sites. The binding site detection methodology was validated on apo/holo cognate protein pairs, correctly identifying 91% of ligand binding sites in holo structures and 88% in apo structures where corresponding sites existed. For correctly detected apo binding sites, the cognate holo site was the most similar binding site 87% of the time. PSIM was used to screen a set of proteins that had poorly characterized functions at the time of crystallization, but were later biochemically annotated. Using a fully automated protocol, this set of 8 proteins was screened against ～60,000 ligand binding sites from the PDB. PSIM correctly identified functional matches that predated query protein biochemical annotation for five out of the eight query proteins. A panel of 12 currently unannotated proteins was also screened, resulting in a large number of statistically significant binding site matches, some of which suggest likely functions for the podorly characterized proteins. Proteins 2014; 82:679–694. © 2013 Wiley Periodicals, Inc.