Determination Of Three-dimensional Protein Structures Research Articles

A Divide-and-Conquer Strategy for the Prediction of Protein Contact Map

The prediction of contact maps in protein is a challenging topic for the determination of three-dimensional protein structures. In this paper, we introduce Forest of Decision Trees, a methodology for the prediction of protein contact maps based on (1) a divide-and-conquer approach to analyze the prediction problem; (2) a codification vector that combines the information obtained from the target amino acids neighborhood, and the sub-sequence between them; (3) an ensemble of classifiers that employs a hybrid of Genetic Algorithms and Decision Trees as base classifiers; and (4) a rulebased interpretation mechanism. The comparison against the top sequence-based methods in CASP10 showed that our predictor is very competitive, showing a high reliability. Their main advantage is its capability to generate a humancomprehensible rule-based interpretation mechanism, giving the specialist some clues to find an easier and interpretable solution for the protein-folding recognition and the prediction of unknown structures. Keywords: CASP10, contact maps prediction, decision trees, genetic algorithms, multiple classifier systems, protein structure prediction.

CNNcon: Improved Protein Contact Maps Prediction Using Cascaded Neural Networks

BackgroundsDespite continuing progress in X-ray crystallography and high-field NMR spectroscopy for determination of three-dimensional protein structures, the number of unsolved and newly discovered sequences grows much faster than that of determined structures. Protein modeling methods can possibly bridge this huge sequence-structure gap with the development of computational science. A grand challenging problem is to predict three-dimensional protein structure from its primary structure (residues sequence) alone. However, predicting residue contact maps is a crucial and promising intermediate step towards final three-dimensional structure prediction. Better predictions of local and non-local contacts between residues can transform protein sequence alignment to structure alignment, which can finally improve template based three-dimensional protein structure predictors greatly.MethodsCNNcon, an improved multiple neural networks based contact map predictor using six sub-networks and one final cascade-network, was developed in this paper. Both the sub-networks and the final cascade-network were trained and tested with their corresponding data sets. While for testing, the target protein was first coded and then input to its corresponding sub-networks for prediction. After that, the intermediate results were input to the cascade-network to finish the final prediction.ResultsThe CNNcon can accurately predict 58.86% in average of contacts at a distance cutoff of 8 Å for proteins with lengths ranging from 51 to 450. The comparison results show that the present method performs better than the compared state-of-the-art predictors. Particularly, the prediction accuracy keeps steady with the increase of protein sequence length. It indicates that the CNNcon overcomes the thin density problem, with which other current predictors have trouble. This advantage makes the method valuable to the prediction of long length proteins. As a result, the effective prediction of long length proteins could be possible by the CNNcon.

Rapid, robotic, small‐scale protein production for NMR screening and structure determination

Three-dimensional protein structure determination is a costly process due in part to the low success rate within groups of potential targets. Conventional validation methods eliminate the vast majority of proteins from further consideration through a time-consuming succession of screens for expression, solubility, purification, and folding. False negatives at each stage incur unwarranted reductions in the overall success rate. We developed a semi-automated protocol for isotopically-labeled protein production using the Maxwell-16, a commercially available bench top robot, that allows for single-step target screening by 2D NMR. In the span of a week, one person can express, purify, and screen 48 different (15)N-labeled proteins, accelerating the validation process by more than 10-fold. The yield from a single channel of the Maxwell-16 is sufficient for acquisition of a high-quality 2D (1)H-(15)N-HSQC spectrum using a 3-mm sample cell and 5-mm cryogenic NMR probe. Maxwell-16 screening of a control group of proteins reproduced previous validation results from conventional small-scale expression screening and large-scale production approaches currently employed by our structural genomics pipeline. Analysis of 18 new protein constructs identified two potential structure targets that included the second PDZ domain of human Par-3. To further demonstrate the broad utility of this production strategy, we solved the PDZ2 NMR structure using [U-(15)N,(13)C] protein prepared using the Maxwell-16. This novel semi-automated protein production protocol reduces the time and cost associated with NMR structure determination by eliminating unnecessary screening and scale-up steps.

Design, Synthesis, and Performance of NTA-modified Lipids as Templates for Histidine-tagged Protein Crystallization

Abstract Current methods of drug discovery and development typically involve either structure-based drug design approaches or high-throughput screening methods for lead identification. The structure-based discovery process relies heavily upon the labor-intensive determination of three-dimensional protein structures using X-ray diffraction (XRD) and NMR techniques. Although tens of thousands of soluble protein structures have been resolved at the atomic level using these methods, there are many drug targets, particularly integral membrane proteins, whose structures have not been elucidated. In the case of XRD, this limitation often arises from the lack of reliable and predictable methods for protein crystallization. The extensive number of proteins crystallized by the interfacial templating approach suggests that this method may offer great promise for structural biology applications, particularly for proteins that do not readily form X-ray quality crystals from bulk solution. This review highlights the various approaches that have been used for templating of protein crystallization at gas–liquid and solid–liquid interfaces. Special emphasis is placed on the synthesis of NTA- and IDA-conjugated lipids, their use in templating and concentrating proteins at lipid monolayer interfaces, and protein structure elucidations that have been facilitated by these metal-chelating lipids. Current approaches to control the nucleation and growth of protein crystals involving microfluidics, dip pen nanolithography, and non-covalent symmetry-based templating approaches are also discussed. Taken together, the progress in this field suggests that the interfacial templating approach to crystallization can develop into a powerful tool for high-throughput structural analysis by providing a universal, readily controllable method for crystallizing proteins.

Pre-assembled clusters distort crystal nucleation kinetics in supersaturated lysozyme solutions

Efficient determination of three-dimensional protein structures is critical for unraveling structure–function relationships and for supporting targeted drug design. A major impediment to these efforts is our lack of control over the nucleation and growth of high-quality protein crystals for X-ray structure determinations. While basic research on protein crystal growth mechanisms has provided valuable new insights, studies of crystal nucleation have been plagued by inconsistent and outright contradictory results. Using dynamic light scattering and SDS gel electrophoresis, we have investigated possible causes of these inconsistencies. We find that commercial sources of lyophilized hen-egg white lysozyme (HEWL) used in nucleation studies contain significant populations of large (∼ 100 nm), pre-assembled lysozyme clusters that can readily evade standard assays of sample purity. In supersaturated solutions, these clusters act as heterogeneous nucleation centers that enhance the rate of crystal nucleation and significantly deteriorate the quality of macroscopic crystals.

Method To Assess Packing Quality of Transmembrane α-Helices in Proteins. 1. Parametrization Using Structural Data

Integral membrane proteins (MPs) are pharmaceutical targets of exceptional importance. Modern methods of three-dimensional protein structure determination often fail to supply the fast growing field of structure-based drug design with the requested MPs' structures. That is why computational modeling techniques gain a special importance for these objects. Among the principal difficulties limiting application of these methods is the low quality of the MPs' models built in silico. In this series of two papers we present a computational approach to the assessment of the packing "quality" of transmembrane (TM) alpha-helical domains in proteins. The method is based on the concept of protein environment classes, whereby each amino acid residue is described in terms of its environment polarity and accessibility to the membrane. In the first paper we analyze a nonredundant set of 26 TM alpha-helical domains and compute the residues' propensities to five predefined classes of membrane-protein environments. Here we evaluate the proposed approach only by various test sets, cross-validation protocols and ability of the method to delimit the crystal structure of visual rhodopsin, and a number of its erroneous theoretical models. More advanced validation of the method is given in the second article of this series. We assume that the developed "membrane score" method will be helpful in optimizing computer models of TM domains of MPs, especially G-protein coupled receptors.

Protein Structure Determination by High-Resolution Solid-State NMR Spectroscopy: Application to Microcrystalline Ubiquitin

High-resolution solid-state NMR spectroscopy has become a promising method for the determination of three-dimensional protein structures for systems which are difficult to crystallize or exhibit low solubility. Here we describe the structure determination of microcrystalline ubiquitin using 2D (13)C-(13)C correlation spectroscopy under magic angle spinning conditions. High-resolution (13)C spectra have been acquired from hydrated microcrystals of site-directed (13)C-enriched ubiquitin. Inter-residue carbon-carbon distance constraints defining the global protein structure have been evaluated from 'dipolar-assisted rotational resonance' experiments recorded at various mixing times. Additional constraints on the backbone torsion angles have been derived from chemical shift analysis. Using both distance and dihedral angle constraints, the structure of microcrystalline ubiquitin has been refined to a root-mean-square deviation of about 1 A. The structure determination strategies for solid samples described herein are likely to be generally applicable to many proteins that cannot be studied by X-ray crystallography or solution NMR spectroscopy.

Accurate determination of local defocus and specimen tilt in electron microscopy

Accurate knowledge of defocus and tilt parameters is essential for the determination of three-dimensional protein structures at high resolution using electron microscopy. We present two computer programs, CTFFIND3 and CTFTILT, which determine defocus parameters from images of untilted specimens, as well as defocus and tilt parameters from images of tilted specimens, respectively. Both programs use a simple algorithm that fits the amplitude modulations visible in a power spectrum with a calculated contrast transfer function (CTF). The background present in the power spectrum is calculated using a low-pass filter. The background is then subtracted from the original power spectrum, allowing the fitting of only the oscillatory component of the CTF. CTFTILT determines specimen tilt parameters by measuring the defocus at a series of locations on the image while constraining them to a single plane. We tested the algorithm on images of two-dimensional crystals by comparing the results with those obtained using crystallographic methods. The images also contained contrast from carbon support film that added to the visibility of the CTF oscillations. The tests suggest that the fitting procedure is able to determine the image defocus with an error of about 10 nm, whereas tilt axis and tilt angle are determined with an error of about 2° and 1°, respectively. Further tests were performed on images of single protein particles embedded in ice that were recorded from untilted or slightly tilted specimens. The visibility of the CTF oscillations from these images was reduced due to the lack of a carbon support film. Nevertheless, the test results suggest that the fitting procedure is able to determine image defocus and tilt angle with errors of about 100 nm and 6°, respectively.

Functional genomics and proteomics--the role of nuclear medicine.

Now that the sequencing of the human genome has been completed, the basic challenges are finding the genes, locating their coding regions and predicting their functions. This will result in a new understanding of human biology as well as in the design of new molecular structures as potential novel diagnostic or drug discovery targets. The assessment of gene function may be performed using the tools of the genome program. These tools represent high-throughput methods used to evaluate changes in the expression of many or all genes of an organism at the same time in order to investigate genetic pathways for normal development and disease. This will lead to a shift in the scientific paradigm: In the pre-proteomics era, functional assignments were derived from hypothesis-driven experiments designed to understand specific cellular processes. The new tools describe proteins on a proteome-wide scale, thereby creating a new way of doing cell research which results in the determination of three-dimensional protein structures and the description of protein networks. These descriptions may then be used for the design of new hypotheses and experiments in the traditional physiological, biochemical and pharmacological sense. The evaluation of genetically manipulated animals or newly designed biomolecules will require a thorough understanding of physiology, biochemistry and pharmacology and the experimental approaches will involve many new technologies, including in vivo imaging with single-photon emission tomography and positron emission tomography. Nuclear medicine procedures may be applied for the determination of gene function and regulation using established and new tracers or using in vivo reporter genes such as enzymes, receptors, antigens or transporters. Pharmacogenomics will identify new surrogate markers for therapy monitoring which may represent potential new tracers for imaging. Also, drug distribution studies for new therapeutic biomolecules are needed, at least during preclinical stages of drug development. Finally, new biomolecules will be developed by bioengineering methods which may be used for isotope-based diagnosis and treatment of disease.

New Crystallization Systems Envisioned for Microgravity Studies

Laboratory-based systems have been constructed to demonstrate two methods which will allow for dynamic control of protein-crystal growth. The technologies developed in these systems will be incorporated into future flight hardware for use in microgravity studies. The first method uses a precisely controlled vapordiffusion approach to monitor and control proteincrystal growth. This approach utilizes a humidity sensor and various interfaces under computer control to effect virtually any evaporation rate from up to 40 different growth solutions simultaneously. A static laser-lightscattering sensor can be used to detect aggregation events and trigger a change in the evaporation rate for a growth solution. The second method exploits the varying solubility of proteins versus temperature to control the growth of protein crystals. This approach utilizes miniature thermo-electric devices under microcomputer control which change temperature as needed to grow crystals of a given protein. Complex temperature ramps are possible using this approach. A static laser-light-scattering probe is also included in this system as a noninvasive probe for detection of aggregation events. The systems constructed demonstrate significant advances in the ability of researchers to gain control of the protein-crystal growth process and will provide tremendous opportunities for microgravity research. crystals is often the rate-limiting step for the determination of three-dimensional protein structures. Many different techniques have been used in the attempt to grow high-quality protein crystals. The most widely used protein-crystal growth technique, vapor diffusion, utilizes a growth solution containing protein and a precipitating agent. A popular vapor-diffusion configuration, typically described as the hanging-drop method (McPherson, 1982) (Fig. 1), uses a reservoir solution containing precipitant and a buffered protein/precipitant solution which 'hangs' from a sealed coverslip positioned over the reservoir. The initial solution conditions are such that water vapor diffuses from the protein solution into the reservoir solution, thereby increasing the concentration of the protein beyond its solubility point. One significant limitation of the traditional vapor-diffusion technique is that the evaporation of water from the growth solution (within a particular geometry) is fixed by the starting concentrations of the solution components (Fig. 2). Thus, the rate at which the approach to supersaturation of the growth solution occurs is unchangeable by the experimenter, even if modification of this evaporation rate is desirable. While vapor diffusion has been (and still is) a very popular technique, it has not always proven to be the best method for a given protein and hence a wide range of other approaches have been used as a means to obtain high-quality protein crystals. Another protein

Automated combined assignment of NOESY spectra and three-dimensional protein structure determination.

A procedure for automated protein structure determination is presented that is based on an iterative procedure during which the NOESY peak list assignment and the structure calculation are performed simultaneously. The input consists of a list of NOESY peak positions and a list of chemical shifts as obtained from sequence-specific resonance assignment. For the present applications of this approach the previously introduced NOAH routine was implemented in the distance geometry program DIANA. As an illustration, experimental 2D and 3D NOESY cross-peak lists of six proteins have been analyzed, for which complete sequence-specific 1H assignments are available for the polypeptide backbone and the amino acid side chains. The automated method assigned 70-90% of all NOESY cross peaks, which is on average 10% less than with the interactive approach, and only between 0.8% and 2.4% of the automatically assigned peaks had a different assignment than in the corresponding manually assigned peak lists. The structures obtained with NOAH/DIANA are in close agreement with those from manually assigned peak lists, and with both approaches the residual constraint violations correspond to high-quality NMR structure determinations. Systematic comparisons of the bundles of conformers that represent corresponding automatically and interactively determined structures document the absence of significant bias in either approach, indicating that an important step has been made towards automation of structure determination from NMR spectra.

6] Determination of three-dimensional protein structures in solution by nuclear magnetic resonance: An overview

Publisher Summary This chapter discusses the survey of the method, and describes more specialized contributions into perspective. Furthermore, the range of potential applications for the nuclear magnetic resonance (NMR) method of protein structure determination, and the type of results anticipated from it, are briefly discussed. At present, NMR is the only approach, besides diffraction techniques with single crystals, that is available for protein structure determination. Its introduction is of fundamental interest, because NMR can provide data that are in many ways complementary to those obtained from X-ray crystallography, and thus promises to widen the view of protein molecules for a better grasp of the relationships between structure and function. The method of protein structure determination by NMR involves several steps: data acquisition, spectral analysis, and structural interpretation of the experimental data. The NMR method is of prime interest for those working with polypeptides and proteins having molecular weights up to approximately 20,000. For molecules in this class the method provides stringent tests for the purity of the protein preparations the amino acid composition, and the amino acid sequence, and can yield a three-dimensional structure in solution at high resolution.

Application of molecular dynamics with interproton distance restraints to three-dimensional protein structure determination: A model study of crambin

The applicability of restrained molecular dynamics for the determination of threedimensional protein structures on the basis of short interproton distances (<4 Å) that can be realistically determined from nuclear magnetic resonance measurements in solution is assessed. The model system used is the 1.2 Å resolution crystal structure of the 46 residue protein crambin, from which a set of 240 approximate distance restraints, divided into three ranges (2.5±0.5, 3.0 −1.0 +0.5 and 4±1 Å), is derived. This interproton distance set comprises 159 short-range (|i−j|≤ 5) and 56 (|i−j|>5) long-range inter-residue distances and 25 intra-residue distances. Restrained molecular dynamics are carried out using a number of different protocols starting from two initial structures: a completely extended β-strand; and an extended structure with two α-helices in the same positions as in the crystal structure (residues 7 to 19, and 23 to 30) and all other residues in the form of extended β-strands. The root-mean-square (r.m.s.) atomic differences between these two initial structures and the crystal structure are 43 Å and 23 Å, respectively. It is shown that, provided protocols are used that permit the secondary structure elements to form at least partially prior to folding into a tertiary structure, convergence to the correct final structure, both globally and locally, is achieved. The r.m.s. atomic differences between the converged restrained dynamics structures and the crystal structure range from 1.5 to 2.2 Å for the backbone atoms and from 2.0 to 2.8 Å for all atoms. The r.m.s. atomic difference between the X-ray structure and the structure obtained by first averaging the co-ordinates of the converged restrained dynamics structures is even smaller: 1.0 Å for the backbone atoms and 1.6 Å for all atoms. These results provide a measure with which to judge future experimental results on proteins whose crystal structures are unknown. In addition, from an examination of the dynamics trajectories, it is shown that the convergence pathways followed by the various simulations are different.