A reduced representation of proteins for use in restraint satisfaction calculations

The use of soluble protein structures in modeling helical proteins in a layered membrane

Major advances have been made in the prediction of soluble protein structures, led by the knowledge-based modeling methods that extract useful structural trends from known protein structures and incorporate them into scoring functions. The same cannot be reported for the class of transmembrane proteins, primarily due to the lack of high-resolution structural data for transmembrane proteins, which render many of the knowledge-based method unreliable or invalid. We have developed a method that harnesses the vast structural knowledge available in soluble protein data for use in the modeling of transmembrane proteins. At the core of the method, a set of transmembrane protein decoy sets that allow us to filter and train features recognized from soluble proteins for transmembrane protein modeling into a set of scoring functions. We have demonstrated that structures of soluble proteins can provide significant insight into transmembrane protein structures. A complementary novel two-stage modeling/selection process that mimics the two-stage helical membrane protein folding was developed. Combined with the scoring function, the method was successfully applied to model 5 transmembrane proteins. The root mean square deviations of the predicted models ranged from 5.0 to 8.8 Å to the native structures.

Adaptive torsion-angle quasi-statics: a general simulation method with applications to protein structure analysis and design

MOTIVATION

The cost of molecular quasi-statics or dynamics simulations increases with the size of the simulated systems, which is a problem when studying biological phenomena that involve large molecules over long time scales. To address this problem, one has often to either increase the processing power (which might be expensive), or make arbitrary simplifications to the system (which might bias the study).

RESULTS

We introduce adaptive torsion-angle quasi-statics, a general simulation method able to rigorously and automatically predict the most mobile regions in a simulated system, under user-defined precision or time constraints. By predicting and simulating only these most important regions, the adaptive method provides the user with complete control on the balance between precision and computational cost, without requiring him or her to perform a priori, arbitrary simplifications. We build on our previous research on adaptive articulated-body simulation and show how, by taking advantage of the partial rigidification of a molecule, we are able to propose novel data structures and algorithms for adaptive update of molecular forces and energies. This results in a globally adaptive molecular quasi-statics simulation method. We demonstrate our approach on several examples and show how adaptive quasi-statics allows a user to interactively design, modify and study potentially complex protein structures.

PREDICT modeling and in-silico screening for G-protein coupled receptors

G-protein coupled receptors (GPCRs) are a major group of drug targets for which only one x-ray structure is known (the nondrugable rhodopsin), limiting the application of structure-based drug discovery to GPCRs. In this paper we present the details of PREDICT, a new algorithmic approach for modeling the 3D structure of GPCRs without relying on homology to rhodopsin. PREDICT, which focuses on the transmembrane domain of GPCRs, starts from the primary sequence of the receptor, simultaneously optimizing multiple 'decoy' conformations of the protein in order to find its most stable structure, culminating in a virtual receptor-ligand complex. In this paper we present a comprehensive analysis of three PREDICT models for the dopamine D2, neurokinin NK1, and neuropeptide Y Y1 receptors. A shorter discussion of the CCR3 receptor model is also included. All models were found to be in good agreement with a large body of experimental data. The quality of the PREDICT models, at least for drug discovery purposes, was evaluated by their successful utilization in in-silico screening. Virtual screening using all three PREDICT models yielded enrichment factors 9-fold to 44-fold better than random screening. Namely, the PREDICT models can be used to identify active small-molecule ligands embedded in large compound libraries with an efficiency comparable to that obtained using crystal structures for non-GPCR targets.

Optimal bundling of transmembrane helices using sparse distance constraints

We present a two-step approach to modeling the transmembrane spanning helical bundles of integral membrane proteins using only sparse distance constraints, such as those derived from chemical cross-linking, dipolar EPR and FRET experiments. In Step 1, using an algorithm, we developed, the conformational space of membrane protein folds matching a set of distance constraints is explored to provide initial structures for local conformational searches. In Step 2, these structures refined against a custom penalty function that incorporates both measures derived from statistical analysis of solved membrane protein structures and distance constraints obtained from experiments. We begin by describing the statistical analysis of the solved membrane protein structures from which the theoretical portion of the penalty function was derived. We then describe the penalty function, and, using a set of six test cases, demonstrate that it is capable of distinguishing helical bundles that are close to the native bundle from those that are far from the native bundle. Finally, using a set of only 27 distance constraints extracted from the literature, we show that our method successfully recovers the structure of dark-adapted rhodopsin to within 3.2 A of the crystal structure.

G protein-coupled receptors: in silico drug discovery in 3D

The application of structure-based in silico methods to drug discovery is still considered a major challenge, especially when the x-ray structure of the target protein is unknown. Such is the case with human G protein-coupled receptors (GPCRs), one of the most important families of drug targets, where in the absence of x-ray structures, one has to rely on in silico 3D models. We report repeated success in using ab initio in silico GPCR models, generated by the predict method, for blind in silico screening when applied to a set of five different GPCR drug targets. More than 100,000 compounds were typically screened in silico for each target, leading to a selection of <100 "virtual hit" compounds to be tested in the lab. In vitro binding assays of the selected compounds confirm high hit rates, of 12-21% (full dose-response curves, Ki < 5 microM). In most cases, the best hit was a novel compound (New Chemical Entity) in the 1- to 100-nM range, with very promising pharmacological properties, as measured by a variety of in vitro and in vivo assays. These assays validated the quality of the hits as lead compounds for drug discovery. The results demonstrate the usefulness and robustness of ab initio in silico 3D models and of in silico screening for GPCR drug discovery.

Using experimental information to produce a model of the transmembrane domain of the ion channel phospholamban

Molecular models of the transmembrane domain of the phospholamban pentamer have been generated by a computational method that uses the experimentally measured effects of systematic single-site mutations as a guiding force in the modeling procedure. This method makes the assumptions that 1) the phospholamban transmembrane domain is a parallel five-helix bundle, and 2) nondisruptive mutation positions are lipid exposed, whereas 3) disruptive or partially disruptive mutations are not. Our procedure requires substantially less computer time than systematic search methods, allowing rapid assessment of the effects of different experimental results on the helix arrangement. The effectiveness of the approach is investigated in test calculations on two helix-dimer systems of known structure. Two independently derived sets of mutagenesis data were used to define the restraints for generating models of phospholamban. Both resulting models are left-handed, highly symmetrical pentamers. Although the overall bundle geometry is very similar in the two models, the orientation of individual helices differs by approximately 50 degrees, resulting in different sets of residues facing the pore. This demonstrates how differences in restraints can have an effect on the model structures generated, and how the violation of these restraints can identify inconsistent experimental data.

Ambiguous distance data in the calculation of NMR structures

The use of molecular dynamics for simulated annealing optimization of structures calculated from NMR data is reviewed. I focus on ways of directly using and automatically assigning ambiguous peaks from nuclear Overhauser enhancement experiments during the structure calculation.

Relationship between protein structure and geometrical constraints

We evaluate to what extent the structure of proteins can be deduced from incomplete knowledge of disulfide bridges, surface assignments, secondary structure assignments, and additional distance constraints. A cost function taking such constraints into account was used to obtain protein structures using a simple minimization algorithm. For small proteins, the approximate structure could be obtained using one additional distance constraint for each amino acid in the protein. We also studied the effect of using predicted secondary structure and surface assignments. The constraints used in this approach typically may be obtained from low-resolution experimental data. When using a cost function based on distances, half of the resulting structures will be mirrored, because the resulting structure and its mirror image will have the same cost. The secondary structure assignments were therefore divided into chirality constraints and distance constraints. Here we report that the problem of mirrored structures, in some cases, can be solved by using a chirality term in the cost function.

Automated method for modeling seven-helix transmembrane receptors from experimental data

A rule-based automated method is presented for modeling the structures of the seven transmembrane helices of G-protein-coupled receptors. The structures are generated by using a simulated annealing Monte Carlo procedure that positions and orients rigid helices to satisfy structural restraints. The restraints are derived from analysis of experimental information from biophysical studies on native and mutant proteins, from analysis of the sequences of related proteins, and from theoretical considerations of protein structure. Calculations are presented for two systems. The method was validated through calculations using appropriate experimental information for bacteriorhodopsin, which produced a model structure with a root mean square (rms) deviation of 1.87 A from the structure determined by electron microscopy. Calculations are also presented using experimental and theoretical information available for bovine rhodopsin to assign the helices to a projection density map and to produce a model of bovine rhodopsin that can be used as a template for modeling other G-protein-coupled receptors.

Protein structure prediction: recognition of primary, secondary, and tertiary structural features from amino acid sequence

This review attempts a critical stock-taking of the current state of the science aimed at predicting structural features of proteins from their amino acid sequences. At the primary structure level, methods are considered for detection of remotely related sequences and for recognizing amino acid patterns to predict posttranslational modifications and binding sites. The techniques involving secondary structural features include prediction of secondary structure, membrane-spanning regions, and secondary structural class. At the tertiary structural level, methods for threading a sequence into a mainchain fold, homology modeling and assigning sequences to protein families with similar folds are discussed. A literature analysis suggests that, to date, threading techniques are not able to show their superiority over sequence pattern recognition methods. Recent progress in the state of ab initio structure calculation is reviewed in detail. The analysis shows that many structural features can be predicted from the amino acid sequence much better than just a few years ago and with attendant utility in experimental research. Best prediction can be achieved for new protein sequences that can be assigned to well-studied protein families. For single sequences without homologues, the folding problem has not yet been solved.