Ligand Binding Site Residues Research Articles

Comparison of Structure-based Tools for the Prediction of Ligand Binding Site Residues in Apo-structures

In this paper, a group of structure-based algorithms used in the prediction of ligand binding sites in proteins were tested on apo-structures, and their results are analyzed and discussed. In particular, we focused on the accuracy of predicting the residues at the binding sites as the main criterion for comparison. The data set used for the testing was compiled using the LigASite database and consisted entirely of apo-structures. The tools compared are LIGSITE, SURFNET, DEPTH, GHECOM, SiteHound, and Fpocket.

FunFOLD: an improved automated method for the prediction of ligand binding residues using 3D models of proteins

BackgroundThe accurate prediction of ligand binding residues from amino acid sequences is important for the automated functional annotation of novel proteins. In the previous two CASP experiments, the most successful methods in the function prediction category were those which used structural superpositions of 3D models and related templates with bound ligands in order to identify putative contacting residues. However, whilst most of this prediction process can be automated, visual inspection and manual adjustments of parameters, such as the distance thresholds used for each target, have often been required to prevent over prediction. Here we describe a novel method FunFOLD, which uses an automatic approach for cluster identification and residue selection. The software provided can easily be integrated into existing fold recognition servers, requiring only a 3D model and list of templates as inputs. A simple web interface is also provided allowing access to non-expert users. The method has been benchmarked against the top servers and manual prediction groups tested at both CASP8 and CASP9.ResultsThe FunFOLD method shows a significant improvement over the best available servers and is shown to be competitive with the top manual prediction groups that were tested at CASP8. The FunFOLD method is also competitive with both the top server and manual methods tested at CASP9. When tested using common subsets of targets, the predictions from FunFOLD are shown to achieve a significantly higher mean Matthews Correlation Coefficient (MCC) scores and Binding-site Distance Test (BDT) scores than all server methods that were tested at CASP8. Testing on the CASP9 set showed no statistically significant separation in performance between FunFOLD and the other top server groups tested.ConclusionsThe FunFOLD software is freely available as both a standalone package and a prediction server, providing competitive ligand binding site residue predictions for expert and non-expert users alike. The software provides a new fully automated approach for structure based function prediction using 3D models of proteins.

Binary image representation of a ligand binding site: its application to efficient sampling of a conformational ensemble.

BackgroundModelling the ligand binding site of a protein is an important component of understanding protein-ligand interactions and is being actively studied. Even if the side chains are restricted to rotamers, a set of commonly-observed low-energy conformations, the exhaustive combinatorial search of ligand binding site conformers is known as NP-hard. Here we propose a new method, ROTAIMAGE, for modelling the plausible conformers for the ligand binding site given a fixed backbone structure.ResultsROTAIMAGE includes a procedure of selecting ligand binding site residues, exhaustively searching rotameric conformers, clustering them by dissimilarities in pocket shape, and suggesting a representative conformer per cluster. Prior to the clustering, the list of conformers generated by exhaustive search can be reduced by pruning the conformers that have near identical pocket shapes, which is done using simple bit operations. We tested our approach by modelling the active-site inhibitor binding pockets of matrix metalloproteinase-1 and -13. For both cases, analyzing the conformers based on their pocket shapes substantially reduced the 'computational complexity' (10 to 190 fold). The subsequent clustering revealed that the pocket shapes of both proteins could be grouped into approximately 10 distinct clusters. At this level of clustering, the conformational space spanned by the known crystal structures was well covered. Heatmap analysis identified a few bit blocks that combinatorially dictated the clustering pattern. Using this analytical approach, we demonstrated that each of the bit blocks was associated with a specific pocket residue. Identification of residues that influenced the shape of the pocket is an interesting feature unique to the ROTAIMAGE algorithm.ConclusionsROTAIMAGE is a novel algorithm that was efficient in exploring the conformational space of the ligand binding site. Its ability to identify 'key' pocket residues also provides further insight into conformational flexibility with specific implications for protein-ligand interactions.

Assessment of predictions submitted for the CASP7 function prediction category

Here we present a full overview of the Critical Assessment of Protein Structure Prediction (CASP7) function prediction category. Predictions were submitted for Gene Ontology molecular function terms, Enzyme Commission numbers, and ligand binding site residues. The first two categories were difficult to assess because very little new functional information becomes available after the experiment. The majority of the known Gene Ontology terms and all the Enzyme Commission numbers were available a priori to predictors before the experiment, so prediction for these two categories was not blind. Nevertheless, for Gene Ontology terms we were able to demonstrate that some groups made better predictions than others. In the binding residue category, the predictors did not know in advance which ligands were bound and therefore blind evaluation was possible, but there were disappointingly few predictions in this category. After CASP 6 and 7 the need to organize a more effective blind function prediction category is obvious, even if it means focusing on binding site prediction as the only category that can be truly assessed in the CASP spirit.

Dual Function for a Unique Site within the β2I Domain of Integrin αMβ2

Integrin activation has been postulated to occur in part via conformational changes in the I domain of the beta subunit (the betaI domain), especially near the F-alpha(7) loop, in response to "inside-out" signaling. However, direct evidence for a role of the F-alpha(7) loop in ligand binding and activity modulation is still lacking. Here, we report our finding that the F-alpha(7) loop (residues 344-358) within the beta(2)I domain has dual functions in ligand binding by alpha(M)beta(2). On the one hand, it supports intercellular adhesion molecule 1 (ICAM-1) binding to alpha(M)beta(2) directly as part of a recognition interface formed by five noncontiguous segments (Pro(192)-Glu(197), Asn(213)-Glu(220), Leu(225)-Leu(230), Ser(324)-Thr(329), and Glu(344)-Asp(348)) on the apex of the beta(2)I domain. On the other hand, it controls the open and closed conformation of the alpha(M)beta(2) receptor, thereby indirectly affecting alpha(M)beta(2) binding to other ligands. Switching the five constituent sequences of the ICAM-1-binding site within the beta(2)I domain to their beta(1) counterparts destroyed ICAM-1 binding but had no effect on the gross conformations of the receptor. Of the five ICAM-1 binding-defective mutants, four had normal or even stronger interaction with Fg and C3bi, as reported in our previous study. Synthetic peptides derived from the identified site inhibited alpha(M)beta(2)-ICAM-1 interaction and supported direct binding to ICAM-1. Most importantly, perturbation of the F-alpha(7) loop caused conformational changes within the beta(2)I domain, which was further propagated to other regions of alpha(M)beta(2). Altogether, our data demonstrate that inside-out signaling could modulate ligand binding directly by changing the ligand-binding pocket per se and/or indirectly by inducing multiple conformational changes within the receptor.

Structure-based phenotyping predicts HIV-1 protease inhibitor resistance.

Mutations in HIV-1 drug targets lead to resistance and consequent therapeutic failure of antiretroviral drugs. Phenotypic resistance assays are time-consuming and costly, and genotypic rules-based interpretations may fail to predict the effects of multiple mutations. We have developed a computational procedure that rapidly evaluates changes in the binding energy of inhibitors to mutant HIV-1 PR variants. Models of WT complexes were produced from crystal structures. Mutant complexes were built by amino acid substitutions in the WT complexes with subsequent energy minimization of the ligand and PR binding site residues. Accuracy of the models was confirmed by comparison with available crystal structures and by prediction of known resistance-related mutations. PR variants from clinical isolates were modeled in complex with six FDA-approved PIs, and changes in the binding energy (DeltaE(bind)) of mutant versus WT complexes were correlated with the ratios of phenotypic 50% inhibitory concentration (IC(50)) values. The calculated DeltaE(bind) of five PIs showed significant correlations (R(2) = 0.7-0.8) with IC(50) ratios from the Virco Antivirogram assay, and the DeltaE(bind) of six PIs showed good correlation (R(2) = 0.76-0.85) with IC(50) ratios from the Virologic PhenoSense assay. DeltaE(bind) cutoffs corresponding to a four-fold increase in IC(50) were used to define the structure-based phenotype as susceptible, resistant, or equivocal. Blind predictions for 78 PR variants gave overall agreement of 92% (kappa = 0.756) and 86% (kappa = 0.666) with PhenoSense and Antivirogram phenotypes, respectively. The structural phenotyping predicted drug resistance of clinical HIV-1 PR variants with an accuracy approaching that of frequently used cell-based phenotypic assays.

High Affinity Ligand Binding by Integrins Does Not Involve Head Separation

Conformational change in the integrin extracellular domain is required for high affinity ligand binding and is also involved in post-ligand binding cellular signaling. Although there is evidence to the contrary, electron microscopic studies showing that ligand binding triggers alpha- and beta-subunit dissociation in the integrin headpiece have gained popularity and support the hypothesis that head separation activates integrins. To test directly the head separation hypothesis, we enforced head association by introducing disulfide bonds across the interface between the alpha-subunit beta-propeller domain and the beta-subunit I-like domain. Basal and activation-dependent ligand binding by alpha(IIb)beta(3) and alpha(V)beta(3) was unaffected. The covalent linkage prevented dissociation of alpha(IIb)beta(3) into its subunits on EDTA-treated cells. Whereas EDTA dissociated wild type alpha(IIb)beta(3) on the cell surface, a ligand-mimetic Arg-Gly-Asp peptide did not, as judged by binding of complex-specific antibodies. Finally, a high affinity ligand-mimetic compound stabilized noncovalent association between alpha(IIb) and beta(3) headpiece fragments in the presence of SDS, indicating that ligand binding actually stabilized subunit association at the head, as opposed to the suggested subunit separation. The mechanisms of conformational regulation of integrin function should therefore be considered in the context of the associated alphabeta headpiece.

The Constraints Protein–Protein Interactions Place on Sequence Divergence

Structural analyses on a small number of protein families have shown that residues in protein interfaces are more conserved than average amino acid residues. This is also true of other ligand-binding and active site residues. This raises the question whether protein interactions place additional constraints on sequence divergence beyond this general background of functional restrictions on all different types of proteins. In order to investigate this, the sequence identities of Saccharomyces cerevisiae (SC) proteins to their Schizosaccharomyces pombe (SP) orthologues were used as a measure of sequence divergence. The SC proteins were divided into those in stable complexes, those that participate in transient interactions and the remaining proteins. All types of proteins can undergo extensive divergence: all three sequence identity distributions range from less than 20 to over 90%. However, overall, protein interactions do place additional constraints on sequence divergence and the distributions differ significantly: proteins not known to be involved in interactions have an average sequence identity of 38% while this value is 46% for proteins in stable complexes. Proteins that have transient interactions are intermediate between the two, with an average sequence identity of 41%. This trend is independent of whether the proteins are involved in informational functions (transcription, translation and replication) or not and of protein dispensability.

Topology of the CD2–CD48 cell-adhesion molecule complex: implications for antigen recognition by T cells

The T-lymphocyte cell-surface molecule, CD2, was the first heterophilic cell-adhesion molecule to be discovered and has become an important paradigm for understanding the structural basis of cell adhesion. Interaction of CD2 with its ligands. CD58 (in humans) and CD48 (in mice and rats), contributes to antigen recognition by T cells. CD2, CD48 and CD58 are closely related members of the immunoglobulin superfamily and their extracellular regions are predicted to have very similar structures. The three-dimensional crystal structure of this region of CD2 has been determined, revealing two immunoglobulin domains with the ligand-binding site situated on an exposed beta sheet in the membrane-distal domain. This GFCC'C" beta sheet is also involved in a homophilic 'head-to-head' interaction in the CD2 crystal lattice, which has been proposed to be a model for the interactions of CD2 with its ligands. We show that the CD2-binding site on rat CD48 lies on the equivalent beta-sheet of its membrane-distal immunoglobulin domain. By making complementary mutations, we have shown that two charged residues in the CD48 ligand-binding site interact directly with two oppositely charged residues in CD2's ligand-binding site. These results indicate that the amino-terminal immunoglobulin domains of CD2 and CD48 bind each other in the same orientation as the CD2-CD2 crystal lattice interaction, strongly supporting the suggestion that CD2 interacts head-to-head with its ligand. Modelling CD48 onto the CD2 structure reveals that the CD2-CD48 complex spans approximately the same distance (134 A) as predicted for the complex between the T-cell receptor and the peptide-bound major histocompatibility complex (MHC) molecule. Our results, together with recent structural studies of CD2, provide the first indication of the specific topology of a cell-adhesion molecule complex. The similar dimensions predicted for the CD2-CD48 complex and the complex between the T-cell receptor and the peptide-bound MHC molecule suggest that one of the functions of CD2 may be to position the plasma membranes of the T cell and the antigen-presenting (or target) cell at the optimal distance for the low-affinity interaction between the T-cell receptor and the peptide-bound MHC molecule.

The role of tyrosine at the ligand-binding site of the nicotinic acetylcholine receptor

Identification of the critical residues in a receptor's ligand-binding site provides valuable structural information important for understanding the basis for ligand recognition. The design of specific ligands targeted for receptor action will depend to a great extent on detailed structural knowledge of this kind. Although the nicotinic acetylcholine receptor (nAChR) is perhaps the best characterized of all receptors, the detailed configuration of the ligand-binding site remains unknown. Structural comparisons of nicotinic agonists and antagonists have long predicted a negative subsite on the receptor to interact with the positively charged alkyl-ammonium moiety common to nearly all nicotinic agents. We have used intrinsic fluorescence spectroscopic analyses together with binding studies of selectively modified peptide fragments of the nAChR to suggest that one or two invariant tyrosine residues at positions 190 and 198 on the alpha-subunit provide the critical negative subsite required for ligand binding. Tyrosines may similarly be part of the negative subsite of muscarinic receptors and other neurotransmitter receptors that bind cationic ligands.