Identification of potential small-molecule protein-protein inhibitors of cancer metastasis by 3D epitope-based computational screening.

In cancer cells exposed to extracellular pressure or shear stress, AKT1-FAK interaction drives focal adhesion kinase (FAK) phosphorylation, leading to force-activated cancer cell adhesion and metastasis. Blocking the AKT1-FAK interaction is therefore an attractive target for cancer therapy, avoiding the side effects of global FAK inhibition. Starting with our previous identification of a short FAK peptide that binds AKT1, we identified a series of small-molecule inhibitor candidates using a novel approach for inhibiting protein-protein interactions. Using a 3D structural fragment of the FAK peptide as the query, millions of drug-like, commercially available molecules were screened to identify a subset mimicking the volume and chemistry of the FAK fragment to test for their ability to block pressure-sensitive FAK phosphorylation by AKT1. Two compounds reduced the stimulation of FAK phosphorylation in response to extracellular pressure in human SW620 colon cancer cells without affecting basal FAK phosphorylation. Thus, using a 3D protein interaction epitope as a novel query for ligand-based virtual screening can successfully identify small-molecules that show promise in modulating cancer cell adhesion and metastasis.

A statistical technique for predicting membrane protein structure.

A statistical technique has been developed for predicting the transmembrane segments of membrane proteins from their amino acid sequences. A protein's amino acid sequence is represented by a sequence of membrane propensity values derived from the frequency of occurrence of the amino acids in a number of putative transmembrane segments. A running average over this numeric sequence yields a membrane propensity profile from which transmembrane segments may be chosen. When this method is applied to a pool of ten putative membrane proteins, the predicted intra- and extramembrane regions agree 93.6% on a residue-by-residue basis with previously suggested structures. Predictions of transmembrane segments in cytochrome c oxidase subunits I, II and III and cytochrome b from several species are given, and structural homology between species is examined using membrane propensity profiles. Conclusions are then made about the functionality of several regions in these proteins.

The role of structure in antibody cross-reactivity between peptides and folded proteins.

Peptides have the potential for targeting vaccines against pre-specified epitopes on folded proteins. When polyclonal antibodies against native proteins are used to screen peptide libraries, most of the peptides isolated align to linear epitopes on the proteins. The mechanism of cross-reactivity is unclear; both structural mimicry by the peptide and induced fit of the epitope may occur. The most effective peptide mimics of protein epitopes are likely to be those that best mimic both the chemistry and the structure of epitopes. Our goal in this work has been to establish a strategy for characterizing epitopes on a folded protein that are candidates for structural mimicry by peptides. We investigated the chemical and structural bases of peptide-protein cross-reactivity using phage-displayed peptide libraries in combination with computational structural analysis. Polyclonal antibodies against the well-characterized antigens, hen eggwhite lysozyme and worm myohemerythrin, were used to screen a panel of phage-displayed peptide libraries. Most of the selected peptide sequences aligned to linear epitopes on the corresponding protein; the critical binding sequence of each epitope was revealed from these alignments. The structures of the critical sequences as they occur in other non-homologous proteins were analyzed using the Sequery and Superpositional Structural Assignment computer programs. These allowed us to evaluate the extent of conformational preference inherent in each sequence independent of its protein context, and thus to predict the peptides most likely to have structural preferences that match their protein epitopes. Evidence for sequences having a clear structural bias emerged for several epitopes, and synthetic peptides representing three of these epitopes bound antibody with sub-micromolar affinities. The strong preference for a type II beta-turn predicted for one peptide was confirmed by NMR and circular dichroism analyses. Our strategy for identifying conformationally biased epitope sequences provides a new approach to the design of epitope-targeted, peptide-based vaccines.

Predicting conserved water-mediated and polar ligand interactions in proteins using a K-nearest-neighbors genetic algorithm.

Water-mediated ligand interactions are essential to biological processes, from product displacement in thymidylate synthase to DNA recognition by Trp repressor, yet the structural chemistry influencing whether bound water is displaced or participates in ligand binding is not well characterized. Consolv, employing a hybrid k-nearest-neighbors classifier/genetic algorithm, predicts bound water molecules conserved between free and ligand-bound protein structures by examining the environment of each water molecule in the free structure. Four environmental features are used: the water molecule's crystallographic temperature factor, the number of hydrogen bonds between the water molecule and protein, and the density and hydrophilicity of neighboring protein atoms. After training on 13 non-homologous proteins, Consolv predicted the conservation of active-site water molecules upon ligand binding with 75% accuracy (Matthews coefficient Cm = 0.41) for seven new proteins. Mispredictions typically involved water molecules predicted to be conserved that were displaced by a polar ligand atom, indicating that Consolv correctly assesses polar binding sites; 90% accuracy (Cm = 0.78) was achieved for predicting conserved active-site water or polar ligand atom binding. Consolv thus provides an accurate means for optimizing ligand design by identifying sites favored to be occupied by either a mediating water molecule or a polar ligand atom, as well as water molecules likely to be displaced by the ligand. Accuracy for predicting first-shell water conservation between independently determined structures was 61% (Cm=0.23). The ability to predict water-mediated and polar interactions from the free protein structure indicates the surprising extent to which the conservation or displacement of active-site bound water is independent of the ligand, and shows that the protein micro-environment of each water molecule is the dominant influence.

The interdependence of protein surface topography and bound water molecules revealed by surface accessibility and fractal density measures.

To characterize water binding to proteins, which is fundamental to protein folding, stability and activity, the relationships of 10,837 bound water positions to protein surface shape and residue type were analyzed in 56 high-resolution crystallographic structures. Fractal atomic density and accessibility algorithms provided an objective characterization of deep grooves in solvent-accessible protein surfaces. These deep grooves consistently had approximately the diameter of one water molecule, suggesting that deep grooves are formed by the interactions between protein atoms and bound water molecules. Protein surface topography dominates the chemistry and extent of water binding. Protein surface area within grooves bound three times as many water molecules as non-groove surface; grooves accounted for one-quarter of the total surface area yet bound half the water molecules. Moreover, only within grooves did bound water molecules discriminate between different side-chains. In grooves, main-chain surface was as hydrated as that of the most hydrophilic side-chains, Asp and Glu, whereas outside grooves all main and side-chains bound water to a similar, and much decreased, extent. This identification of the interdependence of protein surface shape and hydration has general implications for modelling and prediction of protein surface shape, recognition, local folding and solvent binding.

Cluster analysis of consensus water sites in thrombin and trypsin shows conservation between serine proteases and contributions to ligand specificity.

Cluster analysis is presented as a technique for analyzing the conservation and chemistry of water sites from independent protein structures, and applied to thrombin, trypsin, and bovine pancreatic trypsin inhibitor (BPTI) to locate shared water sites, as well as those contributing to specificity. When several protein structures are superimposed, complete linkage cluster analysis provides an objective technique for resolving the continuum of overlaps between water sites into a set of maximally dense microclusters of overlapping water molecules, and also avoids reliance on any one structure as a reference. Water sites were clustered for ten superimposed thrombin structures, three trypsin structures, and four BPTI structures. For thrombin, 19% of the 708 microclusters, representing unique water sites, contained water molecules from at least half of the structures, and 4% contained waters from all 10. For trypsin, 77% of the 106 microclusters contained water sites from at least half of the structures, and 57% contained waters from all three. Water site conservation correlated with several environmental features: highly conserved microclusters generally had more protein atom neighbors, were in a more hydrophilic environment, made more hydrogen bonds to the protein, and were less mobile. There were significant overlaps between thrombin and trypsin conserved water sites, which did not localize to their similar active sites, but were concentrated in buried regions including the solvent channel surrounding the Na+ site in thrombin, which is associated with ligand selectivity. Cluster analysis also identified water sites conserved in thrombin but not trypsin, and vice versa, providing a list of water sites that may contribute to ligand discrimination. Thus, in addition to facilitating the analysis of water sites from multiple structures, cluster analysis provides a useful tool for distinguishing between conserved features within a protein family and those conferring specificity.

Synthesis of 3 beta-aryl-8-azabicyclo[3.2.1]octanes with high binding affinities and selectivities for the serotonin transporter site.

A novel entry to tropane analogs of cocaine was developed based on the reaction of rhodium-stabilized vinylcarbenoids with pyrroles. These analogs were tested in binding to dopamine, serotonin (5-HT), and norepinephrine transporters in membranes from rat striatum and frontal cortex. In all the analogs, the aryl group at the 3 position was directly bound to the tropane ring and an ethyl ketone moiety was present at the 2 position. By appropriate modification of the aryl and nitrogen substituents, highly potent and 5-HT selective tropanes were prepared. The most potent and selective compound was 3 beta-[4-(1-methylethenyl)phenyl]-2 beta-propanoyl-8-azabicyclo[3.2.1]octane (13b) which had a Ki of 0.1 nM at 5-HT transporters and was 150 times more potent at 5-HT vs dopamine transporters and almost 1000 times more potent at 5-HT vs norepinephrine transporters.

Synthesis of 2beta-acyl-3beta-(substituted naphthyl)-8-azabicyclo[3.2.1]octanes and their binding affinities at dopamine and serotonin transport sites.

A series of 3beta-naphthyltropane derivatives were synthesized and found to show high affinity at both the dopamine and serotonin transporter sites, leading to some of the most potent inhibitors known based on the tropane structure. Comparative molecular field analysis (CoMFA) models were developed for both dopamine and serotonin transporter binding data. These models provide insights into those factors that influence binding at the two transporters.

Diabetic treatment and primary ventricular fibrillation in acute myocardial infarction.

The relation between mode of therapy and mortality rate and incidence of primary ventricular fibrillation was studied in 265 patients with diabetes mellitus and acute myocardial infarction. Sixty patients were being treated with diet only, 54 were receiving insulin and 151 were taking oral hypoglycemic agents. Fourteen patients (5.3 percent) had primary ventricular fibrillation, and all but one died. No statistically significant association was found between the incidence of primary ventricular fibrillation and the type of treatment for diabetes mellitus. Sixty-four (24.2 percent) of the 265 patients died during hospitalization. Mortality was greater among diabetic patients receiving oral therapy. However, after adjusting for age and sex, the difference among these three treatment regimens did not reach the P less than 0.05 level of significance.

Protein flexibility and dynamics using constraint theory.

A new approach is presented for determining the rigid regions in proteins and the flexible joints between them. The short-range forces in proteins are modeled as constraints and we use a recently developed formalism from graph theory to analyze flexibility in the bond network. Forces included in the analysis are the covalent bond-stretching and bond-bending forces, salt bridges, and hydrogen bonds. We use a local function to associate an energy with individual hydrogen bonds, which then can be included or excluded depending on the bond strength. Colored maps of the rigid and flexible regions provide a direct visualization of where the motion of the protein can take place, consistent with these distance constraints. We also define a flexibility index that quantifies the local density of flexible or floppy modes, in terms of the dihedral angles that remain free to rotate in each flexible region. A negative flexibility index provides a measure of the density of redundant bonds in rigid regions. A new application of this approach is to simulate the maximal range of possible motions of the flexible regions by introducing Monte Carlo changes in the free dihedral angles, subject to the distance constraints. This is done using a method that maintains closure of the rings formed by covalent and hydrogen bonds in the flexible parts of the protein, and van der Waals overlaps between atoms are avoided. We use the locus of the possible motions of HIV protease as an example: movies of its motion can be seen at http://www.pa.msu.edu/~lei.

Cytochrome oxidase: structural insights from electron microscopy and from secondary structure prediction.

Electron microscopic images of selectively contrasted cytochrome oxidase dimer crystals are interpreted in a manner consistent with the structure of monomers determined by Fuller et al. (J. Molec. Biol. 134, 305-327). The arms of the y-shaped monomers lie within and perpendicular to the lipid bilayer protruding approximately 25 A on the matrix side of the membrane. The cytoplasmic-side tails of two monomers spread apart in a dimer forming a large cleft. Decoration of the exposed matrix side of vesicle crystals with antisubunit IV antibody fragments indicates that subunit IV lies along the a-crystal axis roughly 20 A from the center of the dimer. A membrane propensity algorithm applied to the sequences of cytochrome oxidase subunits predicts a total of 19 transmembrane alpha-helices per monomer.

Knowledge discovery in medical and biological datasets using a hybrid Bayes classifier/evolutionary algorithm.

A key element of bioinformatics research is the extraction of meaningful information from large experimental data sets. Various approaches, including statistical and graph theoretical methods, data mining, and computational pattern recognition, have been applied to this task with varying degrees of success. Using a novel classifier based on the Bayes discriminant function, we present a hybrid algorithm that employs feature selection and extraction to isolate salient features from large medical and other biological data sets. We have previously shown that a genetic algorithm coupled with a k-nearest-neighbors classifier performs well in extracting information about protein-water binding from X-ray crystallographic protein structure data. The effectiveness of the hybrid EC-Bayes classifier is demonstrated to distinguish the features of this data set that are the most statistically relevant and to weight these features appropriately to aid in the prediction of solvation sites.