Macromolecular Crystal Structures Research Articles

Refinement of Atomic Structures Against cryo-EM Maps.

This review describes some of the methods for atomic structure refinement (fitting) against medium/high-resolution single-particle cryo-EM reconstructed maps. Some of the tools developed for macromolecular X-ray crystal structure analysis, especially those encapsulating prior chemical and structural information can be transferred directly for fitting into cryo-EM maps. However, despite the similarities, there are significant differences between data produced by these two techniques; therefore, different likelihood functions linking the data and model must be used in cryo-EM and crystallographic refinement. Although tools described in this review are mostly designed for medium/high-resolution maps, if maps have sufficiently good quality, then these tools can also be used at moderately low resolution, as shown in one example. In addition, the use of several popular crystallographic methods is strongly discouraged in cryo-EM refinement, such as 2Fo-Fc maps, solvent flattening, and feature-enhanced maps (FEMs) for visualization and model (re)building. Two problems in the cryo-EM field are overclaiming resolution and severe map oversharpening. Both of these should be avoided; if data of higher resolution than the signal are used, then overfitting of model parameters into the noise is unavoidable, and if maps are oversharpened, then at least parts of the maps might become very noisy and ultimately uninterpretable. Both of these may result in suboptimal and even misleading atomic models.

Using support vector machines to improve elemental ion identification in macromolecular crystal structures.

In the process of macromolecular model building, crystallographers must examine electron density for isolated atoms and differentiate sites containing structured solvent molecules from those containing elemental ions. This task requires specific knowledge of metal-binding chemistry and scattering properties and is prone to error. A method has previously been described to identify ions based on manually chosen criteria for a number of elements. Here, the use of support vector machines (SVMs) to automatically classify isolated atoms as either solvent or one of various ions is described. Two data sets of protein crystal structures, one containing manually curated structures deposited with anomalous diffraction data and another with automatically filtered, high-resolution structures, were constructed. On the manually curated data set, an SVM classifier was able to distinguish calcium from manganese, zinc, iron and nickel, as well as all five of these ions from water molecules, with a high degree of accuracy. Additionally, SVMs trained on the automatically curated set of high-resolution structures were able to successfully classify most common elemental ions in an independent validation test set. This method is readily extensible to other elemental ions and can also be used in conjunction with previous methods based on a priori expectations of the chemical environment and X-ray scattering.

Stock-based detection of protein oligomeric states in jsPISA.

A new version of the popular software PISA for the analysis of macromolecular interfaces and identification of biological assemblies (complexes) from macromolecular crystal structures is presented. The new web server jsPISA has a substantially improved user interface, based on modern JavaScript technologies, and also new elements of analysis: assembly stock and interaction radar. The new elements help interpretation of PISA results in difficult and ambiguous cases, for example, when the oligomeric state depends on protein concentration, or when the biologically relevant interaction is weak and cannot be easily discriminated from superficial crystal contacts. jsPISA is maintained by CCP4 at http://www.ccp4.ac.uk/pisa. There are no login requirements for using the server.

Enabling X-ray free electron laser crystallography for challenging biological systems from a limited number of crystals.

There is considerable potential for X-ray free electron lasers (XFELs) to enable determination of macromolecular crystal structures that are difficult to solve using current synchrotron sources. Prior XFEL studies often involved the collection of thousands to millions of diffraction images, in part due to limitations of data processing methods. We implemented a data processing system based on classical post-refinement techniques, adapted to specific properties of XFEL diffraction data. When applied to XFEL data from three different proteins collected using various sample delivery systems and XFEL beam parameters, our method improved the quality of the diffraction data as well as the resulting refined atomic models and electron density maps. Moreover, the number of observations for a reflection necessary to assemble an accurate data set could be reduced to a few observations. These developments will help expand the applicability of XFEL crystallography to challenging biological systems, including cases where sample is limited.

Status of the crystallography beamlines at Diamond Light Source

X-ray crystallography is one of the most widely supported areas of science at Diamond Light Source and is the core technique on a third of the beamlines, accounting for almost a half of user publications across the whole facility. The life science community can currently call on six macromolecular crystallography beamlines (I02, I03, I04, I04-1, I23 and I24) to satisfy the increasing demand for the determination of macromolecular crystal structures from both the academic and pharmaceutical communities. In the physical sciences, there are four beamlines (I11, I15, I16 and I19) that cater for crystallographic studies ranging from the structure solution of small-molecule systems, the determination of structure at extreme conditions and the detailed analysis of the weak diffraction signal from magnetic structures. The physical science beamlines use both single-crystal and powder-diffraction techniques. All the crystallography beamlines at Diamond ensure they undergo a continuous programme of development to enable new technologies and the science they support. In this article, the current status of all the crystallography beamlines at Diamond will be discussed along with some examples of the science they have supported since they became operational.

Crystals on the cover 2015

This issue inaugurates the second decade of publication of Acta Crystallographica Section F and, in keeping with the custom of the last decade, the cover features an artful arrangement of the best photomicrograph images of crystals published in the journal in the previous year. The custom has served as an annual testament to the commitment of Acta Cryst. F to serve the community of scientists whose focus includes the production of crystalline samples for macromolecular crystal structure determination. The crystals featured are as follows: keto-deoxy-d-galactarate (KDG) dehydratase from Agrobacterium tumefaciens, Taberman et al., 2014 ▶; glucose-1-phosphate uridylyltransferase (GalU) from Erwinia amylovora, Toccafondi et al., 2014 ▶; human farnesyl pyrophosphate synthase complexed with risedronate, Yokoyama et al., 2014 ▶; Xaa-Pro dipeptidase from Xanthomonas campestris, Kumar et al., 2014 ▶; Imp3 in complex with an Mpp10 peptide involved in yeast ribosome biogenesis, Zheng & Ye, 2014 ▶; C-terminal domain of guanylate kinase-associated protein from Rattus nor­vegicus, Tong et al., 2014 ▶; virus-like particles from a piscine betanodavirus, Luo et al., 2014 ▶; FMN-bound and FMN-free forms of aromatic acid decarboxylase (CpsUbiX) from the psychrophilic bacterium Colwellia psychrerythraea 34 H, Do et al., 2014 ▶; XoGroEL chaperonin from Xanthomonas oryzae pv. oryzae, Tran et al., 2014 ▶; an artificial molten-globular-like triosephosphate isomerase protein of mixed phylogenetic origin, Goyal et al., 2014 ▶; great cormorant (Phalacrocorax carbo) haemoglobin, Jagadeesan et al., 2014 ▶; FliP, an integral membrane component of the bacterial flagellar type III protein-export apparatus, Fukumura et al., 2014 ▶; feast/famine regulatory protein (Rv2779c) from Mycobacterium tuberculosis H37Rv, Dey & Ramachandran, 2014 ▶; Gos1p, a yeast SNARE protein, Cheng et al., 2014 ▶; d-lactate dehydrogenase from Lactobacillus jensenii, Kim et al., 2014 ▶; YfcM: an important factor for EF-P hydroxylation, Kobayashi et al., 2014 ▶; a novel β-l-arabinofuranosidase (HypBA1) from Bifidobacterium longum, Zhu et al., 2014 ▶; a dye-decolourizing peroxidase from Dictyostelium discoideum, Rai et al., 2014 ▶; a Fic protein from Clostridium difficile, Welner et al., 2014 ▶; human IL-18 and its extracellular complexes, Kimura et al., 2014 ▶; a surface mutant of the middle domain of PB2 from human influenza A (H1N1) virus, Tsurumura et al., 2014 ▶; the (S)-3-hydroxybutyryl-CoA dehydrogenase PaaH1 from Ralstonia eutropha H16, Kim & Kim, 2014 ▶; the putative type VI secretion immunity protein Tli5 (PA5088) from Pseudomonas aeruginosa, Chen et al., 2014 ▶; and a fungal type III polyketide synthase that produces the csypyrones, Yang et al., 2014 ▶. As indicated in the Editorial in the December 2014 issue (Hunter & Weiss, 2014 ▶), this new decade also inaugurates a new regime for articles reporting macromolecular crystallizations. We hope that our authors will respond to the new challenges presented by the changes. We are convinced that the changes will encourage the generation of articles of higher scientific import and of greater general interest. With this cover, and with this month’s article in our continuing International Year of Crystallography series, we emphasize our commitment to continue publication of macromolecular crystallization articles and welcome those that meet the new standards.

ACHESYM: an algorithm and server for standardized placement of macromolecular models in the unit cell.

Despite the existence of numerous useful conventions in structural crystallography, for example for the choice of the asymmetric part of the unit cell or of reciprocal space, surprisingly no standards are in use for the placement of the molecular model in the unit cell, often leading to inconsistencies or confusion. A conceptual solution for this problem has been proposed for macromolecular crystal structures based on the idea of the anti-Cheshire unit cell. Here, a program and server (called ACHESYM; http://achesym.ibch.poznan.pl) are presented for the practical implementation of this concept. In addition, the first task of ACHESYM is to find an optimal (compact) macromolecular assembly if more than one polymer chain exists. ACHESYM processes PDB (atomic parameters and TLS matrices) and mmCIF (diffraction data) input files to produce a new coordinate set and to reindex the reflections and modify their phases, if necessary.

Free kick instead of cross-validation in maximum-likelihood refinement of macromolecular crystal structures.

The refinement of a molecular model is a computational procedure by which the atomic model is fitted to the diffraction data. The commonly used target in the refinement of macromolecular structures is the maximum-likelihood (ML) function, which relies on the assessment of model errors. The current ML functions rely on cross-validation. They utilize phase-error estimates that are calculated from a small fraction of diffraction data, called the test set, that are not used to fit the model. An approach has been developed that uses the work set to calculate the phase-error estimates in the ML refinement from simulating the model errors via the random displacement of atomic coordinates. It is called ML free-kick refinement as it uses the ML formulation of the target function and is based on the idea of freeing the model from the model bias imposed by the chemical energy restraints used in refinement. This approach for the calculation of error estimates is superior to the cross-validation approach: it reduces the phase error and increases the accuracy of molecular models, is more robust, provides clearer maps and may use a smaller portion of data for the test set for the calculation of Rfree or may leave it out completely.

A new generation of CCP4 monomer library based on Crystallography Open Database

The use of prior chemical knowledge such as bond lengths, bond angles about constituent blocks of macromolecules and ligands is an essential part of macromolecular crystal structure analysis. One of the reliable sources of such chemical knowledge is small molecule database where small molecule crystal structures have been analysed against high-resolution, high-quality experimental data. Furthermore, vast amount of data in small molecule database provide comprehensive coverage of flexible chemical environment and enable proper statistical analysis to avoid biased representation of those chemical properties. This presentation describes our work on organization of the data from open-access and daily-updated small molecule database, Crystallography Open Database(COD) [1], into a new generation of CCP4 monomer library (Dictionary), a container of prior chemical knowledge [2]. In order to describe specific environment atoms are in, they are classified into different atomic types based on local graphs and some basic chemical properties of atoms. This scheme can be applied to any small molecule databases. The atom types, and values of bond lengths and bond associated with them, are further clustered into a hierarchical tree and an isomorphism-mapping algorithm is implemented to facilitate fast search among a large number of atom types (typically several millions). This also provides a mechanism to derive reliable values for bond lengths and angles of novel ligands. Metal and non-organic atoms are treated differently with organic ones. The original data in COD are curated using several criteria and further statistical analysis on derived values of bond lengths and angles are allow to extract reliable chemical information from such databanks as COD. There are several software tools associated with new dictionary including 1) generate "ideal" bond lengths and angles for unknown ligand; 2) generate starting coordinates to represent one of the conformation of the ligand under consideration.

Check your metal - not every density blob is a water molecule

Metals play vital roles in both the mechanism and architecture of biological macromolecules, and are the most frequently encountered ligands (i.e. non-solvent heterogeneous chemical atoms) in the determination of macromolecular crystal structures. However, metal coordinating environments in protein structures are not always easy to check in routine validation procedures, resulting in an abundance of misidentified and/or suboptimally modeled metal ions in the Protein Data Bank (PDB). We present a solution to identify these problems in three distinct yet related aspects: (1) coordination chemistry; (2) agreement of experimental B-factors and occupancy; and (3) the composition and motif of the metal binding environment. Due to additional strain introduced by macromolecular backbones, the patterns of coordination of metal binding sites in metal-containing macromolecules are more complex and diverse than those found in inorganic or organometallic chemistry. These complications make a comprehensive library of "permitted" coordination chemistry in protein structures less feasible, and the usage of global parameters such as the bond valence method more practical, in the determination and validation of metal binding environments. Although they are relatively infrequent, there are also cases where the experimental B-factor or occupancy of a metal ion suggests careful examination. We have developed a web-based tool called CheckMyMetal [1](http://csgid.org/csgid/metal_sites/) for the quick validation of metal binding sites. Moreover, the acquired knowledge of the composition and spatial arrangement (motif) of the coordinating atoms around the metal ion may also help in the modeling of metal binding sites in macromolecular structures. All of the studies described herein were performed using the NEIGHBORHOOD SQL database [2], which connects information about all modeled non-solvent heterogeneous chemical motifs in PDB structure by vectors describing all contacts to neighboring residues and atoms. NEIGHBORHOOD has broad applications for the validation and data mining of ligand binding environments in the PDB.

ARP/wARP for crystallographic model building and drug discovery

The ARP/wARP software project combines automated model building and refinement into an unified approach for macromolecular crystal structure determination. The project is based on two decades of extensive research and development in the areas of macromolecular X-ray crystallography, informatics, data mining and statistical pattern recognition. ARP/wARP collects a vast amount of computationally efficient methods and provides easy-to-use pipelines for building models of proteins, nucleotides, ligands, as well as their complexes. All methods are intuitively accessible from the ArpNavigator [1], which grants direct visualisation and real-time interaction with model building results. Structures determined using ARP/wARP include histones, hsp70, viral proteases, an insect antifreeze protein, transferases, deadenylases, synthases, kinases, photolyases and the spliceosome. The novel release of ARP/wARP, version 7.4, comes with notable innovations for determining structures at medium-to-low resolution such as exploitation of non-crystallographic symmetry, improved protocols for model update and estimation of validity of built models. Joint releases with the CCP4 suite improve software development and integration, and make the installation and updates fast and convenient for the user. A novel procedure for the automatic identification of ligands in electron density maps is introduced. It is based on the sparse parameterisation of density clusters and the matching of the pseudo-atomic grids thus created to conformationally variant ligands using mathematical descriptors of molecular shape, size and topology. The integration of the ViCi web-server for in-silico ligand-based drug design and updated stereo-chemical restraints for ligand fitting make ARP/wARP an asset for crystallographic drug discovery pipelines.

ACHESYM: standardized placement of macromolecular models in the unit cell

Unique choice of the unit cell and the asymmetric unit are well defined and described in the International Tables for Crystallography vol. A. Unfortunately, the placement of molecules within the unit cell is not standardized. Since structure solution programs often use random numbers in their algorithms, the selected set of atomic coordinates may be different even with successive runs of the same program. Although formally correct, an arbitrary choice of molecular placement within the unit cell is confusing and may lead to interpretation errors [1]. With the use of the anti-Cheshire unit cell introduced by Dauter [2], for all space groups without inversion symmetry, it is possible to transform the molecular model such that its center of gravity falls within the anti-Cheshire asymmetric unit cell. It means that for macromolecular crystal structures it should be possible to standardize the placement of the molecules within the unit cell. In consequence, it should be easier for crystallographers and non-crystallographers to compare similar or related crystal structures. An implementation of the anti-Cheshire concept has been programmed in Python as a web service, aCHESYM. The aCHESYM program takes a PDB file as input and transforms the macromolecular model into the desired anti-Cheshire region. The program can also handle structure factor CIF files if the transformation used requires reindexing of the reflection data. The unit cell, coordinates and displacement parameters of all atoms after transformation are saved in a new PDB file. All the calculated transformations are reversible, so there is no danger of data loss. Moreover, the program helps the user to find the most compact assembly of the molecules (chains) in the structure when there are several chains in the asymmetric unit.

Flexible torsion-angle noncrystallographic symmetry restraints for improved macromolecular structure refinement.

One of the great challenges in refining macromolecular crystal structures is a low data-to-parameter ratio. Historically, knowledge from chemistry has been used to help to improve this ratio. When a macromolecule crystallizes with more than one copy in the asymmetric unit, the noncrystallographic symmetry relationships can be exploited to provide additional restraints when refining the working model. However, although globally similar, NCS-related chains often have local differences. To allow for local differences between NCS-related molecules, flexible torsion-based NCS restraints have been introduced, coupled with intelligent rotamer handling for protein chains, and are available in phenix.refine for refinement of models at all resolutions.

Automated identification of elemental ions in macromolecular crystal structures.

Many macromolecular model-building and refinement programs can automatically place solvent atoms in electron density at moderate-to-high resolution. This process frequently builds water molecules in place of elemental ions, the identification of which must be performed manually. The solvent-picking algorithms in phenix.refine have been extended to build common ions based on an analysis of the chemical environment as well as physical properties such as occupancy, B factor and anomalous scattering. The method is most effective for heavier elements such as calcium and zinc, for which a majority of sites can be placed with few false positives in a diverse test set of structures. At atomic resolution, it is observed that it can also be possible to identify tightly bound sodium and magnesium ions. A number of challenges that contribute to the difficulty of completely automating the process of structure completion are discussed.

Combining Crystallographic and Structure-Modeling Approaches in Macromolecular Crystallography

Molecular replacement is a powerful method for determining macromolecular crystal structures. The method remains limited by the requirement for a template structure that is similar to the structure to be determined. Here we describe two approaches for carrying out molecular replacement with distant homology models, each approach bringing in new information about the structure to be determined.The first approach combines the power of Rosetta structure modeling with macromolecular structure determination. Rosetta and other tools from the structure-modeling field have optimized force fields that are focused on obtaining a physically plausible model. In contrast, crystallographic model-building tools focus on interpretation of patterns of density in an electron density map. As these sources of information are nearly orthogonal, putting them together can increase the range of cases where crystal structures of macromolecules can be determined. An easy-to-use tool has been developed in the Phenix crystallographic software system for carrying out extensive molecular replacement searches followed by model-building using both Rosetta and crystallographic tools. This mr_rosetta software is able to find solutions to challenging molecular replacement problems.The second approach takes advantage of the observation that many pairs of proteins have local structures that can be superimposed much more closely than can their complete structures. For example, a β-sheet in one protein may be locally similar to a sheet in another, but when the two structures are superimposed using all atoms, the sheets may be translated relative to each other. We have developed a method for applying local distortions to a structure (morphing it) using an electron density map to guide the distortions. This procedure allows local similarity to be maintained while global structure is changed. We show that morphing molecular replacement templates after placing them in their approximate locations in the crystallographic unit cell can greatly improve subsequent model-building.

X-ray Constrained Extremely Localized Molecular Orbitals: Theory and Critical Assessment of the New Technique.

Following the X-ray constrained wave function approach proposed by Jayatilaka, we have devised a new technique that allows to extract molecular orbitals strictly localized on small molecular fragments from sets of experimental X-ray structure factors amplitudes. Since the novel strategy enables to obtain electron distributions that have quantum mechanical features and that can be easily interpreted in terms of traditional chemical concepts, the method can be also considered as a new useful tool for the determination and the analysis of charge densities from high-resolution X-ray experiments. In this paper, we describe in detail the theory of the new technique, which, in comparison to our preliminary work, has been improved both treating the effects of isotropic secondary extinctions and introducing a new protocol to halt the fitting procedure against the experimental X-ray scattering data. The performances of the novel strategy have been studied both in function of the basis-sets flexibility and in function of the quality of the considered crystallographic data. The tests performed on four different systems (α-glycine, l-cysteine, (aminomethyl)phosphonic acid and N-(trifluoromethyl)formamide) have shown that the achievement of good statistical agreements with the experimental measures mainly depends on the quality of the crystal structures (i.e., geometry positions and thermal parameters) used in the X-ray constrained calculations. Finally, given the reliable transferability of the obtained Extremely Localized Molecular Orbitals (ELMOs), we envisage to exploit the novel approach to construct new ELMOs databases suited to the development of linear-scaling methods for the refinement of macromolecular crystal structures.

MAIN software for density averaging, model building, structure refinement and validation.

MAIN is software that has been designed to interactively perform the complex tasks of macromolecular crystal structure determination and validation. Using MAIN, it is possible to perform density modification, manual and semi-automated or automated model building and rebuilding, real- and reciprocal-space structure optimization and refinement, map calculations and various types of molecular structure validation. The prompt availability of various analytical tools and the immediate visualization of molecular and map objects allow a user to efficiently progress towards the completed refined structure. The extraordinary depth perception of molecular objects in three dimensions that is provided by MAIN is achieved by the clarity and contrast of colours and the smooth rotation of the displayed objects. MAIN allows simultaneous work on several molecular models and various crystal forms. The strength of MAIN lies in its manipulation of averaged density maps and molecular models when noncrystallographic symmetry (NCS) is present. Using MAIN, it is possible to optimize NCS parameters and envelopes and to refine the structure in single or multiple crystal forms.

Ce-SAD Phasing of Glucose Isomerase and Thermolysin Using Cu <i>Kα</i> Radiation

Current structural genomics projects aim to solve a large number of selected protein structures as fast as possible. High degree of automation and standardization is required at every step of the whole process to speed up protein structure determination. Phase problem is a bottleneck in macromolecular structure determination and also in model building which is a time-consuming task. The simplest approach to phasing macromolecular crystal structures is the use of a SAD signal. SAD data can be collected using the in-house copper (1.54 A) wavelength source. Data collected using copper wavelength with the incorporation of anomalously scattering heavy metal atoms may serve as a powerful tool for structural biologists to solve novel protein structures as well where synchrotron beam line is not available. A short soak of protein crystals in heavy metal solution or by incorporating heavy atoms into the protein drop while crystallizing the protein (co-crystallization) leads to incorporation of these heavy metal ions into the ordered solvent shell around the protein surface. The present work aims to determine whether cerium ion can be successfully incorporated into the protein crystal through quick-soaking method while maintaining the isomorphism. The study also aims in understanding whether this metal ion can be used for phasing purpose. The intensity data are collected and analyzed for anomalous signal, substructure solution and the binding sites.

A Multibody Atomic Statistical Potential for Predicting Enzyme-Inhibitor Binding Energy

Accurate prediction of enzyme-inhibitor binding energy has the capacity to speed drug design and chemical genomics efforts by helping to narrow the focus of experiments. Here a non-redundant set of three hundred high-resolution crystallographic enzyme-inhibitor structures was compiled for analysis, complexes with known binding energies (ΔG) based on the availability of experimentally determined inhibition constants (ki). Additionally, a separate set of over 1400 diverse high-resolution macromolecular crystal structures was collected for the purpose of creating an all-atom knowledge-based statistical potential, via application of the Delaunay tessellation computational geometry technique. Next, two hundred of the enzyme-inhibitor complexes were randomly selected to develop a model for predicting binding energy, first by tessellating structures of the complexes as well as the enzymes without their bound inhibitors, then by using the statistical potential to calculate a topological score for each structure tessellation. We derived as a predictor of binding energy an empirical linear function of the difference between topological scores for a complex and its isolated enzyme. A correlation coefficient (r) of 0.79 was obtained for the experimental and calculated ΔG values, with a standard error of 2.34 kcal/mol. Lastly, the model was evaluated with the held-out set of one hundred complexes, for which structure tessellations were performed in order to calculate topological score differences, and binding energy predictions were generated from the derived linear function. Calculated binding energies for the test data also compared well with their experimental counterparts, displaying a correlation coefficient of r= 0.77 with a standard error of 2.50 kcal/mol.

Studying and polishing the PDB's macromolecules

Macromolecular crystal structures are among the best of scientific data, providing detailed insight into these complex and biologically important molecules with a relatively low level of error and subjectivity. However, there are two notable problems with getting the most information from them. The first is that the models are not perfect: there is still opportunity for improving them, and users need to evaluate whether the local reliability in a structure is up to answering their question of interest. The second is that protein and nucleic acid molecules are highly complex and individual, inherently handed and three-dimensional, and the cooperative and subtle interactions that govern their detailed structure and function are not intuitively evident. Thus there is a real need for graphical representations and descriptive classifications that enable molecular 3D literacy. We have spent our career working to understand these elegant molecules ourselves, and building tools to help us and others determine and understand them better. The Protein Data Bank (PDB) has of course been vital and central to this undertaking. Here we combine some history of our involvement as depositors, illustrators, evaluators, and end-users of PDB structures with commentary on how best to study and draw scientific inferences from them.