Homology modeling by the ICM method

Five models have been built by the ICM method for the Comparative Modeling section of the Meeting on the Critical Assessment of Techniques for Protein Structure Prediction. The targets have homologous proteins with known three-dimensional structure with sequence identity ranging from 25 to 77%. After alignment of the target sequence with the related three-dimensional structure, the modeling procedure consists of two subproblems: side-chain prediction and loop prediction. The ICM method approaches these problems with the following steps: (1) a starting model is created based on the homologous structure with the conserved portion fixed and the nonconserved portion having standard covalent geometry and free torsion angles; (2) the Biased Probability Monte Carlo (BPMC) procedure is applied to search the subspaces of either all the nonconservative side-chain torsion angles or torsion angles in a loop backbone and surrounding side chains. A special algorithm was designed to generate low-energy loop deformations. The BPMC procedure globally optimizes the energy function consisting of ECEPP/3 and solvation energy terms. Comparison of the predictions with the NMR or crystallographic solutions reveals a high proportion of correctly predicted side chains. The loops were not correctly predicted because imprinted distortions of the backbone increased the energy of the near-native conformation and thus made the solution unrecognizable. Interestingly, the energy terms were found to be reliable and the sampling of conformational space sufficient. The implications of this finding for the strategies of future comparative modeling are discussed.     

Prediction of protein side-chain rotamers from a backbone-dependent rotamer library: a new homology modeling tool

Modeling by homology is the most accurate computational method for translating an amino acid sequence into a protein structure. Homology modeling can be divided into two sub-problems, placing the polypeptide backbone and adding side-chains. We present a method for rapidly predicting the conformations of protein side-chains, starting from main-chain coordinates alone. The method involves using fewer than ten rotamers per residue from a backbone-dependent rotamer library and a search to remove steric conflicts. The method is initially tested on 299 high resolution crystal structures by rebuilding side-chains onto the experimentally determined backbone structures. A total of 77% of chi1 and 66% of chi(1 + 2) dihedral angles are predicted within 40 degrees of their crystal structure values. We then tested the method on the entire database of known structures in the Protein Data Bank. The predictive accuracy of the algorithm was strongly correlated with the resolution of the structures. In an effort to simulate a realistic homology modeling problem, 9424 homology models were created using three different modeling strategies. For prediction purposes, pairs of structures were identified which shared between 30% and 90% sequence identity. One strategy results in 82% of chi1 and 72% chi(1 + 2) dihedral angles predicted within 40 degrees of the target crystal structure values, suggesting that movements of the backbone associated with this degree of sequence identity are not large enough to disrupt the predictive ability of our method for non-native backbones. These results compared favorably with existing methods over a comprehensive data set.     

Application of a self-consistent mean field theory to predict protein side-chains conformation and estimate their conformational entropy

Understanding the relations between the conformation of the side-chains and the backbone geometry is crucial for structure prediction as well as for homology modelling. To attempt to unravel these rules, we have developed a method which allows us to predict the position of the side-chains from the co-ordinates of the main-chain atoms. This method is based on a rotamer library and refines iteratively a conformational matrix of the side-chains of a protein, CM, such that its current element at each cycle CM (ij) gives the probability that side-chain i of the protein adopts the conformation of its possible rotamer j. Each residue feels the average of all possible environments, weighted by their respective probabilities. The method converges in only a few cycles, thereby deserving the name of self consistent mean field method. Using the rotamer with the highest probability in the optimized conformational matrix to define the conformation of the side-chain leads to the result that on average 72% of chi 1, 75% of chi 2 and 62% of chi 1 + 2 are correctly predicted for a set of 30 proteins. Tests with six pairs of homologous proteins have shown that the method is quite successful even when the protein backbone deviates from the correct conformation. The second application of the optimized conformational matrix was to provide estimates of the conformational entropy of the side-chains in the folded state of the protein. The relevance of this entropy is discussed.     

Main-chain conformational features at different conformations of the side-chains in proteins

An analysis of the known protein structures has shown that the main-chain torsion angles, phi and psi of a residue can be affected by the side-chain torsion angle, chi1. The (chi1, psi) plot of all residues (except Gly, Ala and Pro) show six distinct regions where points are concentrated-although some of these regions are nearly absent in specific cases. The mean of these clusters can show a shift along the psi axis by as much as 30 degrees as chi1 is changed from around 180 to -60 to 60 degrees. Because of the lesser steric constraint points are more diffused along the psi axis when chi1 is approximately -60 degrees. Although points are more spread out along the phi axis in the (chi1, phi) plot, the dependence of phi on chi1 shows up in a shortened phi range (by about 30 degrees) when chi1 is around -60 degrees, and a distinct tendency of clustering of points into two regions when chi1 is approximately equal to 60 degrees, especially for the aromatic residues. Based on the dependence of the backbone conformation on its side-chain the 17 amino acids can be grouped into five classes: (i) aliphatic residues branched at the Cbeta position (although Thr is atypical), (ii) Leu (branched at the Cgamma position), (iii) aromatic residues (Trp can show some deviations), (iv) short polar residues (Asp and Asn), and (v) the remaining linear-chain residues, mainly polar. Ser and Thr have the highest inclination to occur with two different orientations of the side-chain that can be located through crystallography. Such residues exhibiting two chi1 angles have their phi and psi angles in a region that is common to the Ramachandran plots at the two different chi1 angles. The dependence of phi and psi angles on chi1 can be used to understand the helical propensities of some residues. Moreover, the average phi, psi values in the alpha-helices vary with the side-chain conformation.     

Prediction and evaluation of side-chain conformations for protein backbone structures

A common approach to protein modeling is to propose a backbone structure based on homology or threading and then to attempt to build side chains onto this backbone. A fast algorithm using the simple criteria of atomic overlap and overall rotamer probability is proposed for this purpose. The method was first tested in the context of exhaustive searches of side chain configuration space in protein cores and was then applied to all side chains in 49 proteins of known structure, using simulated annealing to sample space. The latter procedure obtains the correct rotamer for 57% and the correct chi 1 value for 74% of the 6751 residues in the sample. When low-temperature Monte-Carlo simulations are initiated from the results of the simulated-annealing processes, consensus configurations are obtained which exhibit slightly more accurate predictions. The Monte-Carlo procedure also allows converged side chain entropies to be calculated for all residues. These prove to be accurate indicators of prediction reliability. For example, the correct rotamer is obtained for 79% and the correct chi 1 value is obtained for 84% of the half of the sample residues exhibiting the lowest entropies. Side chain entropy and predictability are nearly completely uncorrelated with solvent-accessible area. Some precedents for and implications of this observation are discussed.     

Dynamic simulations of the molecular conformations of wild type and mutant xanthan polymers suggest that conformational differences may contribute to observed differences in viscosity

Xanthan gum is an exopolysaccharide secreted by the bacterium Xanthamonas campestris whose ability to make solutions viscous at low concentrations and over a pH and temperature range have generated much interest in both academic and industrial environments. Mutant Xanthamonas strains have been derived that produce xanthan gums with an altered or variant subunit chemical structure and different measured viscosities when compared with the wild type (wt) form of the polymer. Two variant gums were targeted as potentially interesting in this study, these being the nonacetylated tetramer (natet) and the acetylated tetramer (atet), which both lack a side-chain terminal mannose residue and in one case (natet) lacks an acetate group on an internal mannose residue. Solutions of these tetrameric gums possess viscosities higher (natet) and lower (atet) than the wt gum, and therefore we have attempted to determine whether these molecules possess unique conformational preferences when compared with the wt and with each other. In this manner we can initiate an understanding of how a polysaccharide's conformation contributes to its solution properties. The GEGOP software permits a sampling of the static and dynamic equilibrium states of carbohydrate molecules, and this software was employed to calculate equilibrium states of representative oligosaccharides with chemical structures representative of xanthan-like molecules. Energy minimization techniques revealed similar local minima for all three molecules. Some of these minima are comprised of elongate backbone conformations (A type) in which side chains fold onto backbone surfaces. Other minima with A backbones possessed side chains in less intimate backbone contact especially when calculations were performed with a low dielectric constant. This phenomenon was particularly pronounced in the wt molecule where an increased number of negatively charged side-chain residues experience charge repulsion resulting in reduced side-chain-backbone contact. Metropolis Monte Carlo (MMC) dynamic simulations performed with an elevated temperature factor (1000 K) allowed a better qualitative representation of conformational space than 300 K simulations. Employing a nonhierarchical cluster analysis method (population density profile: PDP) coupled with a classification scheme, it was possible to partition resulting MMC data sets into conformational families. This analysis revealed that in simulations performed with different dielectric constant values (10, 25, and infinity) all molecules possessed primarily A-type backbones. Less elongate, more open helical backbone forms (B, C, D, J, and Flat-a) did occur during the simulations but were populated to a lesser extent. In the natet molecule significantly open helical backbones existed (E, F, G, H, and I) that did not occur in the lower viscosity wt and atet molecules. PDP clustering methods and subsequent conformational classification applied to the first residue (mannose) of the side chain permitted a determination of side-chain orientation. Comparison of all three molecules indicated a larger population of side-chain conformational families in less direct backbone contact for the wt molecule than either of the variant molecules (natet/atet) suggesting that the side chains in the wt are more flexible. Thus, a major conformational difference between the high viscosity natet and the lower viscosities of the wt/atet is the increased amount of open helical backbone in the natet. In addition, the significant difference between the higher viscosity wt and the lower viscosity atet is the increase side-chain flexibility in the wt. We hypothesize that conformational differences of this kind could form a partial explanation of the observed differences in viscosity between these xanthan-like polymers.     

Conformational analysis of carboxylate and carboxamide side-chains bound to cations

The known structures, both small as well as macromolecules, as stored in the respective databases, provide a wealth of information that, when properly rationalized, can be used in the design of new molecules. The engineering of metal-binding sites in proteins requires an understanding of the effect of such a binding on the ligand conformation. Here we present an analysis of the side-chain conformations of aspartic and glutamic acids, asparagine and glutamine bound to cations, in proteins. The most populated state of the chi 1 torsion angle for Asx (aspartate and asparagine) residues is g- (around 64 degrees) and is occupied by all groups that have another ligand two residues ahead of them. Co-ordinating residues that are sequentially well separated from other ligands (and most Glx (glutamate and glutamine) belong to this category) show a preference for the g+ or t state (dihedral angle near -60 and 180 degrees, respectively) as is normally observed. A chi 2 value close to, but less than 180 degrees, offers the minimum energy conformation for Glx, but another ligand closely in the sequence can force the torsion to be in a gauche form. A survey of small molecule structures involving Asx and Glx fragments show the terminal torsion to be centred at 0 degree. This observation is imitated by ligand Asx groups, whereas for Glx the angle veers towards the negative side. This statistical preference becomes less prominent when the cation is held in the less commonly observed anti geometry, and is lost completely when the residues are involved in anion binding. Cations exhibit an absolute preference to bind the oxygen that is on the same side as C alpha for the shorter side-chain, and the one eclipsing C beta for the longer chain. The shortest amino acid with a charged side-chain, Asp, shows very subtle conformational variations. For example, the distribution of chi 1 is not symmetrical about 180 degrees, and the g+ state (at -68(+/- 5) degrees) is the most stable of all on the basis of both steric as well as electrostatic grounds. Besides, the magnitude and the sign of chi 2 show strong dependence on the chi 1 values. Even for the longer Glu side-chain, a chi 2-angle in the g+ state necessitates the chi 1 also to reside in the g+ conformation. These mutual dependences of torsion angles in small molecule structures are also retained in proteins.(ABSTRACT TRUNCATED AT 400 WORDS)     

Fold assembly of small proteins using monte carlo simulations driven by restraints derived from multiple sequence alignments

The feasibility of predicting the global fold of small proteins by incorporating predicted secondary and tertiary restraints into ab initio folding simulations has been demonstrated on a test set comprised of 20 non-homologous proteins, of which one was a blind prediction of target 42 in the recent CASP2 contest. These proteins contain from 37 to 100 residues and represent all secondary structural classes and a representative variety of global topologies. Secondary structure restraints are provided by the PHD secondary structure prediction algorithm that incorporates multiple sequence information. Predicted tertiary restraints are derived from multiple sequence alignments via a two-step process. First, seed side-chain contacts are identified from correlated mutation analysis, and then a threading-based algorithm is used to expand the number of these seed contacts. A lattice-based reduced protein model and a folding algorithm designed to incorporate these predicted restraints is described. Depending upon fold complexity, it is possible to assemble native-like topologies whose coordinate root-mean-square deviation from native is between 3.0 A and 6.5 A. The requisite level of accuracy in side-chain contact map prediction can be roughly 25% on average, provided that about 60% of the contact predictions are correct within +/-1 residue and 95% of the predictions are correct within +/-4 residues. Precision in tertiary contact prediction is more critical than absolute accuracy. Furthermore, only a subset of the tertiary contacts, on the order of 25% of the total, is sufficient for successful topology assembly. Overall, this study suggests that the use of restraints derived from multiple sequence alignments combined with a fold assembly algorithm holds considerable promise for the prediction of the global topology of small proteins.     

A simple and sensitive experiment for measurement of JCC couplings between backbone carbonyl and methyl carbons in isotopically enriched proteins

A simple 2D difference experiment is described that allows quantitative measurement of 13C-13C J couplings between backbone carbonyl and side-chain carbons. Precise 3JCC values were measured from data recorded in just 2 h for a 1-mM solution of the 20-kD complex between the protein calmodulin and a 26-residue synthetic peptide. The J couplings aid in determining the chi 1 angles of valine, isoleucine and threonine residues, and in making stereospecific assignments of the Val C gamma methyl groups. Error analysis indicates that the uncertainty in the derived J couplings is generally less than ca. 0.3 Hz.     

Cross-strand side-chain interactions versus turn conformation in beta-hairpins

A series of designed peptides has been analyzed by 1H-NMR spectroscopy in order to investigate the influence of cross-strand side-chain interactions in beta-hairpin formation. The peptides differ in the N-terminal residues of a previously designed linear decapeptide that folds in aqueous solution into two interconverting beta-hairpin conformations, one with a type I turn (beta-hairpin 4:4) and the other with a type I + G1 beta-bulge turn (beta-hairpin 3:5). Analysis of the conformational behavior of the peptides studied here demonstrates three favorable and two unfavorable cross-strand side-chain interactions for beta-hairpin formation. These results are in agreement with statistical data on side-chain interactions in protein beta-sheets. All the peptides in this study form significant populations of the beta-hairpin 3:5, but only some of them also adopt the beta-hairpin 4:4. The formation of beta-hairpin 4:4 requires the presence of at least two favorable cross-strand interactions, whereas beta-hairpin 3:5 seems to be less susceptible to side-chain interactions. A protein database analysis of beta-hairpins 3:5 and beta-hairpins 4:4 indicates that the former occur more frequently than the latter. In both peptides and proteins, beta-hairpins 3:5 have a larger right-handed twist than beta-hairpins 4:4, so that a factor contributing to the higher stability of beta-hairpin 3:5 relative to beta-hairpin 4:4 is due to an appropriate backbone conformation of the type I + G1 beta-bulge turn toward the right-handed twist usually observed in protein beta-sheets. In contrast, as suggested previously, backbone geometry of the type I turn is not adequate for the right-handed twist. Because analysis of buried hydrophobic surface areas on protein beta-hairpins reveals that beta-hairpins 3:5 bury more hydrophobic surface area than beta-hairpins 4:4, we suggest that the right-handed twist observed in beta-hairpin 3:5 allows a better packing of side chains and that this may also contribute to its higher intrinsic stability.     

Conformations of arginine and lysine side chains in association with anions

The side chain conformations shown by arginine and lysine in amino-acid and peptide crystal structures and bound to oxyanions in proteins have been analyzed in an attempt to understand the behaviour of these long-chain amino acids in an ionic environment. Except for chi 1, torsions have a preference for the trans conformation. However, for arginine in protein structures, chi 3 and chi 4 appear to be flexible and can be tuned for optimal anion binding. For chi 4, values in the range -80 to 80 degrees are excluded for steric reasons; the remaining region in conformational space is accessible. This orientational variety exhibited by chi 4 has not been hitherto appreciated. Factors that can forbid a chi-angle to be in the trans geometry are the simultaneous binding of the anion by the main- and side-chain atoms, or the sharing of the anion between two different molecules in the crystal structure. Small molecules containing arginine have a distinct tendency to crystallize with two molecules in the asymmetric unit. This may be a general phenomenon for all extended molecules which have hydrogen-bond donors (or acceptors) embedded in a rigid set-up.     

PAPQMD parametrization of molecular systems with cyclopropyl rings: conformational study of homopeptides constituted by 1-aminocyclopropane-1-carboxylic acid

The suitability of ab initio, semiempirical and density functional methods as sources of stretching and bending parameters has been explored using the PAPQMD (Program for Approximate Parametrization from Quantum Mechanical Data) strategy. Results show that semiempirical methods provide parameters comparable to those compiled on empirical force fields. In this respect the AMI method seems to be a good method to obtain parameters at a minimum computational cost. On the other hand, harmonic force fields initially developed for proteins and DNA have been extended to include compounds containing highly strained three-membered rings, like 1-aminocyclopropane-1-carboxylic acid. For this purpose the cyclopropyl ring has been explicitly parametrized at the AMI level considering different chemical environments. Finally, the new set of parameters has been used to investigate the conformational preferences of homopeptides constituted by 1-aminocyclopropane-1-carboxylic acid. Results indicate that such compounds tend to adopt a helical conformation stabilized by intramolecular hydrogen bonds between residues i and i + 3. This conformation allows the arrangement of the cyclic side chains without steric clashes.     

Variability of the canonical loop conformations in serine proteinases inhibitors and other proteins

Canonical loops of protein inhibitors of serine proteinases occur in proteins having completely different folds. In this article, conformations of the loops have been analyzed for inhibitors belonging to 10 structurally different families. Using deviation in Calpha-Calpha distances as a criterion for loop similarity, we found that the P3-P3' segment defines most properly the length of the loop. When conformational differences among loops of individual inhibitors were compared using root mean square deviation (rmsd) in atomic coordinates for all main chain atoms (deltar method) and rmsd operating in main chain torsion angles (deltat method), differences of up to 2.1 A and 72.3 degrees, respectively, were observed. Such large values indicate significant conformational differences among individual loops. Nevertheless, the overall geometry of the inhibitor-proteinase interaction is very well preserved, as judged from the similarity of Calpha-Calpha distances between Calpha of catalytic Ser and Calpha of P3-P3' residues in various enzyme-inhibitor complexes. The mode of interaction is very well preserved both in the chymotrypsin and subtilisin families, as distances calculated for subtilisin-inhibitor complexes are almost always within the range of those for chymotrypsin-inhibitor complexes. Complex formation leads to conformational changes of up to 160 degrees for chi1 angle. Side chains of residue P1 and P2' adopt in different complexes a similar orientation (chi1 angle = -60 degrees and -180 degrees, respectively). To check whether the canonical conformation can be found among non-proteinase-inhibitor Brookhaven Protein Data Bank structures, two selection criteria--the allowed main chain dihedral angles and Calpha-Calpha distances for the P3-P3' segment--were applied to all these structures. This procedure detected 132 unique hexapeptide segments in 121 structurally and functionally unrelated proteins. Partial preferences for certain amino acids occurring at particular positions in these hexapeptides could be noted. Further restriction of this set to hexapeptides with a highly exposed P1 residue side chain resulted in 40 segments. The possibility of complexes formation between these segments and serine proteinases was ruled out in molecular modeling due to steric clashes. Several structural features that determine the canonical conformation of the loop both in inhibitors and in other proteins can be distinguished. They include main chain hydrogen bonds both within the P3-P3' segment and with the scaffold region, P3-P4 and P3'-P4' hydrophobic interactions, and finally either hydrophobic or polar interactions involving the P1' residue.     

Crystal structure of farnesyl protein transferase complexed with a CaaX peptide and farnesyl diphosphate analogue

The crystallographic structure of acetyl-Cys-Val-Ile-selenoMet-COOH and alpha-hydroxyfarnesylphosphonic acid (alphaHFP) complexed with rat farnesyl protein transferase (FPT) (space group P61, a = b = 174. 13 A, c = 69.71 A, alpha = beta = 90 degrees, gamma = 120 degrees, Rfactor = 21.8%, Rfree = 29.2%, 2.5 A resolution) is reported. In the ternary complex, the bound substrates are within van der Waals contact of each other and the FPT enzyme. alphaHFP binds in an extended conformation in the active-site cavity where positively charged side chains and solvent molecules interact with the phosphate moiety and aromatic side chains pack adjacent to the isoprenoid chain. The backbone of the bound CaaX peptide adopts an extended conformation, and the side chains interact with both FPT and alphaHFP. The cysteine sulfur of the bound peptide coordinates the active-site zinc. Overall, peptide binding and recognition appear to be dominated by side-chain interactions. Comparison of the structures of the ternary complex and unliganded FPT [Park, H., Boduluri, S., Moomaw, J., Casey, P., and Beese, L. (1997) Science 275, 1800-1804] shows that major rearrangements of several active site side chains occur upon substrate binding.     

High-dose-rate brachytherapy as the primary treatment of medically inoperable stage I-II endometrial carcinoma

A protein fold recognition method was tested by the blind prediction of the structures of a set of proteins. The method evaluates the compatibility of an amino acid sequence with a three-dimensional structure using the four evaluation functions: side-chain packing, solvation, hydrogen-bonding, and local conformation functions. The structures of 14 proteins containing 19 sequences were predicted. The predictions were compared with the experimental structures. The experimental results showed that 9 of the 19 target sequences have known folds or portions of known folds. Among them, the folds of Klebsiella aerogenes urease beta subunit (KAUB) and pyruvate phosphate dikinase domain 4 (PPDK4) were successfully recognized; our method predicted that KAUB and PPDK4 would adopt the folds of macromomycin (Ig-fold) and phosphoribosylanthranilate isomerase:indoleglycerol-phosphate synthase (TIM barrel), respectively, and the experimental structure revealed that they actually adopt the predicted folds. The predictions for the other targets were not successful, but they often gave secondary structural patterns similar to those of the experimental structures.     

Update on protein structure prediction: results of the 1995 IRBM workshop

Computational tools for protein structure prediction are of great interest to molecular, structural and theoretical biologists due to a rapidly increasing number of protein sequences with no known structure. In October 1995, a workshop was held at IRBM to predict as much as possible about a number of proteins of biological interest using ab initio prediction of fold recognition methods. 112 protein sequences were collected via an open invitation for target submissions. 17 were selected for prediction during the workshop and for 11 of these a prediction of some reliability could be made. We believe that this was a worthwhile experiment showing that the use of a range of independent prediction methods and thorough use of existing databases can lead to credible and useful ab initio structure predictions.     

Search for low-energy conformations of a neurotoxic protein by means of predictive rules, tests for hard-sphere overlaps, and energy minimization

A method to obtain models for the three-dimensional structure of the neurotoxin alpha from Naja nigricollis from its amino acid sequence is explored here. Empirical predictive rules were used to estimate the positions of helices, extended structures and bends; advantage was taken of the availability of 14 homologous sequences for the neurotoxins in an attempt to increase the reliability of these predictions. Unassigned residues were allowed to take up several possible conformations determined from the frequencies of occurrence of each type of conformation of that residue in x-ray structures of many proteins. The conformational space of the molecule was explored initially by testing for hard-sphere overlaps and approximate closure of disulfide loops with the aid of a computer; this procedure yielded a limited number of conformations, whose conformational energies were then determined and minimized by optimizing the backbone and side-chain dihedral angles of each residue. Five compact conformations with low energy were found for this neurotoxin. The procedure used here provides an illustration as to how empirical protein algorithms may be used to limit the conformational space, in which energy minimization has to be carried out.     

Stress and strain in staphylococcal nuclease

Protein molecules generally adopt a tertiary structure in which all backbone and side chain conformations are arranged in local energy minima; however, in several well-refined protein structures examples of locally strained geometries, such as cis peptide bonds, have been observed. Staphylococcal nuclease A contains a single cis peptide bond between residues Lys 116 and Pro 117 within a type VIa beta-turn. Alternative native folded forms of nuclease A have been detected by NMR spectroscopy and attributed to a mixture of cis and trans isomers at the Lys 116-Pro 117 peptide bond. Analyses of nuclease variants K116G and K116A by NMR spectroscopy and X-ray crystallography are reported herein. The structure of K116A is indistinguishable from that of nuclease A, including a cis 116-117 peptide bond (92% populated in solution). The overall fold of K116G is also indistinguishable from nuclease A except in the region of the substitution (residues 112-117), which contains a predominantly trans Gly 116-Pro 117 peptide bond (80% populated in solution). Both Lys and Ala would be prohibited from adopting the backbone conformation of Gly 116 due to steric clashes between the beta-carbon and the surrounding residues. One explanation for these results is that the position of the ends of the residue 112-117 loop only allow trans conformations where the local backbone interactions associated with the phi and psi torsion angles are strained. When the 116-117 peptide bond is cis, less strained backbone conformations are available. Thus the relaxation of the backbone strain intrinsic to the trans conformation compensates for the energetically unfavorable cis X-Pro peptide bond. With the removal of the side chain from residue 116 (K116G), the backbone strain of the trans conformation is reduced to the point that the conformation associated with the cis peptide bond is no longer favorable.     

Knowledge-based structure prediction of MHC class I bound peptides: a study of 23 complexes

BACKGROUND: The binding of T-cell antigenic peptides to MHC molecules is a prerequisite for their immunogenicity. The ability to identify binding peptides based on the protein sequence is of great importance to the rational design of peptide vaccines. As the requirements for peptide binding cannot be fully explained by the peptide sequence per se, structural considerations should be taken into account and are expected to improve predictive algorithms. The first step in such an algorithm requires accurate and fast modeling of the peptide structure in the MHC-binding groove. RESULTS: We have used 23 solved peptide-MHC class I complexes as a source of structural information in the development of a modeling algorithm. The peptide backbones and MHC structures were used as the templates for prediction. Sidechain conformations were built based on a rotamer library, using the 'dead end elimination' approach. A simple energy function selects the favorable combination of rotamers for a given sequence. It further selects the correct backbone structure from a limited library. The influence of different parameters on the prediction quality was assessed. With a specific rotamer library that incorporates information from the peptide sidechains in the solved complexes, the algorithm correctly identifies 85% (92%) of all (buried) sidechains and selects the correct backbones. Under cross-validation, 70% (78%) of all (buried) residues are correctly predicted and most of all backbones. The interaction between peptide sidechains has a negligible effect on the prediction quality. CONCLUSIONS: The structure of the peptide sidechains follows from the interactions with the MHC and the peptide backbone, as the prediction is hardly influenced by sidechain interactions. The proposed methodology was able to select the correct backbone from a limited set. The impairment in performance under cross-validation suggests that, currently, the specific rotamer library is not satisfactorily representative. The predictions might improve with an increase in the data.