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.

SPINE workshop on automated X-ray analysis: a progress report.

The Structural Proteomics In Europe (SPINE) consortium contained a workpackage to address the automated X-ray analysis of macromolecules. The aim of this workpackage was to increase the throughput of three-dimensional structures while maintaining the high quality of conventional analyses. SPINE was able to bring together developers of software with users from the partner laboratories. Here, the results of a workshop organized by the consortium to evaluate software developed in the member laboratories against a set of bacterial targets are described. The major emphasis was on molecular-replacement suites, where automation was most advanced. Data processing and analysis, use of experimental phases and model construction were also addressed, albeit at a lower level.

NMR trial models: experiences with the colicin immunity protein Im7 and the p85alpha C-terminal SH2-peptide complex.

Two cases of successful molecular replacement using NMR trial models are presented. One is the crystal structure of the Escherichia coli colicin immunity protein Im7; the other is a heretofore unreported crystal structure of a specific PDGF receptor-derived peptide complex of the carboxy-terminal SH2 domain from the p85alpha subunit of human phosphatidylinositol 3-OH kinase. In both cases, molecular replacement was non-trivial. Success was achieved using trial models that consisted of an ensemble of NMR structures from which the more flexible portions had been excised. Use of maximum-likelihood refinement proved critical to be able to refine the poor starting models. The challenges typical of the use of NMR trial models in molecular replacement are discussed.

Structure of the Bacillus cell fate determinant SpoIIAA in phosphorylated and unphosphorylated forms.

BACKGROUND: The asymmetric cell division during sporulation in Bacillus subtilis gives rise to two compartments: the mother cell and the forespore. Each follow different programs of gene expression coordinated by a succession of alternate RNA polymerase sigma factors. The activity of the first of these sigma factors, sigmaF, is restricted to the forespore although sigmaF is present in the predivisional cell and partitions into both compartments following the asymmetric septation. For sigmaF to become active, it must escape from a complex with its cognate anti-sigma factor, SpoIIAB. This relief from SpoIIAB inhibition requires the dephosphorylation of the anti-sigma factor antagonist, SpoIIAA. The phosphorylation state of SpoIIAA is thus a key determinant of sigmaF activity and cell fate. RESULTS: We have solved the crystal structures of SpoIIAA from Bacillus sphaericus in its phosphorylated and unphosphorylated forms. The overall structure consists of a central beta-pleated sheet, one face of which is buried by a pair of alpha helices, while the other is largely exposed to solvent. The site of phosphorylation, Ser57, is located at the N terminus of helix alpha2. The phosphoserine is exceptionally well defined in the 1.2 A electron density maps, revealing that the structural changes accompanying phosphorylation are slight. CONCLUSIONS: Comparison of unphosphorylated and phosphorylated SpoIIAA shows that covalent modification has no significant effect on the global structure of the protein. The phosphoryl group has a passive role as a negatively charged flag rather than the active role it plays as a nucleus of structural reorganization in many eukaryotic signaling systems.

X-ray structure of bovine pancreatic phospholipase A2 at atomic resolution.

Using synchrotron radiation and a CCD camera, X-ray data have been collected from wild-type bovine pancreatic phospholipase A(2) at 100 K to 0.97 A resolution allowing full anisotropic refinement. The final model has a conventional R factor of 9.44% for all reflections, with a mean standard uncertainty for the positional parameters of 0.031 A as calculated from inversion of the full positional least-squares matrix. At 0.97 A resolution, bovine pancreatic phospholipase A(2) reveals for the first time that its rigid scaffolding does not preclude flexibility, which probably plays an important role in the catalytic process. Functionally important regions (the interfacial binding site and calcium-binding loop) are located at the molecular surface, where conformational variability is more pronounced. A cluster of 2-methyl-2,4-pentanediol molecules is present at the entrance of the hydrophobic channel that leads to the catalytic site and mimics the fatty-acid chains of a substrate analogue. Bovine pancreatic phospholipase A(2) at atomic resolution is compared with previous crystallographic structures and with models derived from nuclear magnetic resonance studies. Given the high structural similarity among extracellular phospholipases A(2) observed so far at lower resolution, the results arising from this structural analysis are expected to be of general validity for this class of enzymes.

Crystal structure of GerE, the ultimate transcriptional regulator of spore formation in Bacillus subtilis.

The small, DNA-binding protein GerE regulates gene transcription in the terminally differentiated mother-cell compartment during late stages of sporulation in Bacillus subtilis. This versatile transcription factor shares sequence homology with the LuxR/FixJ/UhpA family of activators and modulates the expression of a number of genes, in particular those encoding the components of the coat that surrounds the mature spore. GerE orchestrates the final stages of coat deposition and maturation that lead to a spore with remarkable resistance properties but that must be responsive to low levels of germination signals. As this germination process is largely passive and can occur in the absence of de novo protein synthesis, the correct assembly of germination machinery, including germinant receptors and energy storage compounds, is crucial to the survival of the cell. The crystal structure of GerE has been solved at 2.05 A resolution using multi-wavelength anomalous dispersion techniques and reveals the nature of the GerE dimer. Each monomer comprises four alpha-helices, of which the central pair forms a helix-turn-helix DNA-binding motif. Implications for DNA-binding and the structural organisation of the LuxR/FixJ/UhpA family of transcription activator domains are discussed.

The three-dimensional structure of human S100A12.

The crystal structure of human EF-hand calcium-binding protein S100A12 in its calcium-bound form has been determined to 1.95 A resolution by molecular replacement using the structure of the S100B protein. The S100 family members are homologous to calmodulin and other related EF-hand calcium-binding proteins. Like the majority of S100 proteins, S100A12 is a dimer, with the interface between the two subunits being composed mostly of hydrophobic residues. The fold of S100A12 is similar to the other known crystal and solution structures of S100 proteins, except for the linker region between the two EF-hand motifs. Sequence and structure comparison between members of the S100 family suggests that the target-binding region in S100A12 is formed by the linker region and C-terminal residues of one subunit and the N-terminal residues of another subunit of the dimer. The N-terminal region of the target-binding site includes two glutamates that are conserved in most of the S100 sequences. The comparison also provided a better understanding of the role of the residues important for intra- and inter-subunit hydrophobic interactions. The precise role of S100A12 in cell behaviour is yet undefined, as is the case for the whole family, although it has been shown that the interaction of S100A12 with the RAGE receptor is implicated in inflammatory response.

Use of TLS parameters to model anisotropic displacements in macromolecular refinement.

An essential step in macromolecular refinement is the selection of model parameters which give as good a description of the experimental data as possible while retaining a realistic data-to-parameter ratio. This is particularly true of the choice of atomic displacement parameters, where the move from individual isotropic to individual anisotropic refinement involves a sixfold increase in the number of required displacement parameters. The number of refinement parameters can be reduced by using collective variables rather than independent atomic variables and one of the simplest examples of this is the TLS parameterization for describing the translation, libration and screw-rotation displacements of a pseudo-rigid body. This article describes the implementation of the TLS parameterization in the macromolecular refinement program REFMAC. Derivatives of the residual with respect to the TLS parameters are expanded in terms of the derivatives with respect to individual anisotropic U values, which in turn are calculated using a fast Fourier transform technique. TLS refinement is therefore fast and can be used routinely. Examples of TLS refinement are given for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and a transcription activator GerE, for both of which there is data to only 2.0 A, so that individual anisotropic refinement is not feasible. GAPDH has been refined with between one and four TLS groups in the asymmetric unit and GerE with six TLS groups. In both cases, inclusion of TLS parameters gives improved refinement statistics and in particular an improvement in R and free R values of several percent. Furthermore, GAPDH and GerE have two and six molecules in the asymmetric unit, respectively, and in each case the displacement parameters differ significantly between molecules. These differences are well accounted for by the TLS parameterization, leaving residual local displacements which are very similar between molecules and to which NCS restraints can be applied.

Structure of a slow processing precursor penicillin acylase from Escherichia coli reveals the linker peptide blocking the active-site cleft.

Penicillin G acylase is a periplasmic protein, cytoplasmically expressed as a precursor polypeptide comprising a signal sequence, the A and B chains of the mature enzyme (209 and 557 residues respectively) joined by a spacer peptide of 54 amino acid residues. The wild-type AB heterodimer is produced by proteolytic removal of this spacer in the periplasm. The first step in processing is believed to be autocatalytic hydrolysis of the peptide bond between the C-terminal residue of the spacer and the active-site serine residue at the N terminus of the B chain. We have determined the crystal structure of a slowly processing precursor mutant (Thr263Gly) of penicillin G acylase from Escherichia coli, which reveals that the spacer peptide blocks the entrance to the active-site cleft consistent with an autocatalytic mechanism of maturation. In this mutant precursor there is, however, an unexpected cleavage at a site four residues from the active-site serine residue. Analyses of the stereochemistry of the 260-261 bond seen to be cleaved in this precursor structure and of the 263-264 peptide bond have suggested factors that may govern the autocatalytic mechanism.

Crystal structure of dodecameric vanadium-dependent bromoperoxidase from the red algae Corallina officinalis.

The three-dimensional structure of the vanadium bromoperoxidase protein from the marine red macroalgae Corallina officinalis has been determined by single isomorphous replacement at 2.3 A resolution. The enzyme subunit is made up of 595 amino acid residues folded into a single alpha+beta domain. There are 12 bromoperoxidase subunits, arranged with 23-point group symmetry. A cavity is formed by the N terminus of each subunit in the centre of the dodecamer. The subunit fold and dimer organisation of the Cor. officinalis vanadium bromoperoxidase are similar to those of the dimeric enzyme from the brown algae Ascophyllum nodosum, with which it shares 33 % sequence identity. The different oligomeric state of the two algal enzymes seems to reflect separate mechanisms of adaptation to harsh environmental conditions and/or to chemically active substrates and products. The residues involved in the vanadate binding are conserved between the two algal bromoperoxidases and the vanadium chloroperoxidase from the fungus Curvularia inaequalis. However, most of the other residues forming the active-site cavity are different in the three enzymes, which reflects differences in the substrate specificity and stereoselectivity of the reaction. A dimer of the Cor. officinalis enzyme partially superimposes with the two-domain monomer of the fungal enzyme.

Solution of the structure of the cofactor-binding fragment of CysB: a struggle against non-isomorphism.

The elucidation of the structure of CysB(88-324) by multiple isomorphous replacement (MIR) techniques was seriously delayed by problems encountered at every stage of the analysis. There was extensive non-isomorphism both between different native crystals and between native and heavy-atom-soaked crystals. The heavy-atom substitution was invariably weak and different soaking experiments frequently led to substitution at common sites. These correlated heavy-atom binding sites resulted in an overestimation of the phase information. Missing low-resolution reflections in the native data set, constituting only 2% of the total observations, reduced the power of density modification and phase refinement. Finally, the extensive dimer interface made it difficult to isolate a single molecule in the course of model building into the MIR maps. The power of maximum likelihood refinement (REFMAC) was exploited in solving the structure by means of iterative cycles of refinement of a partial model, initially comprising only 30% of the protein atoms in the final coordinate set. This technique, which uses experimental phases, can automatically discriminate the correct and incorrect parts of electron-density maps and give properly weighted combined phases which are better than the experimental or calculated ones. This allowed the model to be gradually extended by manual building into improved electron-density maps. A model generated in this way, containing just 50% of the protein atoms, proved good enough to find the transformations needed for multi-crystal averaging between different crystal forms. The averaging regime im-proved the phasing dramatically such that the complete model could be built. The problems, final solutions and some possible causes for the observed lack of isomorphism are discussed.

Efficient anisotropic refinement of macromolecular structures using FFT.

This paper gives the equations for the use of fast Fourier transformations in individual atomic anisotropic refinement. Restraints on bonded atoms, on the sphericity of each atom and between non-crystallographic symmetry related atoms are described. These have been implemented in the program REFMAC and its performance with several examples is analysed. All the tests show that anisotropic refinement not only reduces the R value and Rfree but also improves the fit to geometric targets, indicating that this parameterization is valuable for improving models derived from experimental data. The computer time taken is comparable to that for isotropic refinements.

Stabilizing bound O2 in myoglobin by valine68 (E11) to asparagine substitution.

The isopropyl side chain of valine68 in myoglobin has been replaced by the acetamide side chain of asparagine in an attempt to engineer higher oxygen affinity. The asparagine replacement introduces a second hydrogen bond donor group into the distal heme pocket which could further stabilize bound oxygen. The Val68 to Asn substitution leads to approximately 3-fold increases in oxygen affinity and 4-6-fold decreases in CO affinity. As a result, the M-value (KCO/KO2) is lowered 15-20-fold to a value close to unity. An even larger enhancement of O2 affinity is seen when asparagine68 is inserted into H64L sperm whale myoglobin which lacks a distal histidine. The overall rate constants for oxygen and carbon monoxide binding to the single V68N myoglobin mutants are uniformly lower than those for the wild-type protein. In contrast, the overall rate constant for NO association is unchanged. Analyses of time courses monitoring the geminate recombination of ligands following nanosecond and picosecond flash photolysis of MbNO and MbO2 indicate that the barrier to ligand binding from within the heme pocket has been raised with little effect on the barrier to diffusion of the ligand into the pocket from the solvent. The crystal structures of the aquomet, deoxy, oxy, and carbon monoxy forms of the V68N mutant have been determined to resolutions ranging from 1.75 to 2.2 A at 150 K. The overall structures are very similar to those of the wild-type protein with the principal alterations taking place within and around the distal heme pocket. In all four structures the asparagine68 side chain lies almost parallel to the plane of the heme with its amide group directed toward the back of the distal heme pocket. The coordinated water molecule in the aquomet form and the bound oxygen in the oxy form can form hydrogen-bonding interactions with both the Asn68 amide group and the imidazole side chain of His64. Surprisingly, in the carbon monoxy form of the V68N mutant, the histidine64 side chain has swung completely out the distal pocket, its place being taken by two ordered water molecules. Overall, these functional and structural results show that the asparagine68 side chain (i) forms a strong hydrogen bond with bound oxygen through its -NH2 group but (ii) sterically hinders the approach of ligands to the iron from within the distal heme pocket.

Validation tools: can they indicate the information content of macromolecular crystal structures?

The explosive increase in the number of published three-dimensionsal structures of macromolecules determined by X-ray analysis places a responsibility on experimentalists, referees and curators of databases to ensure correspondence between the structure parameters and data. Validation tools will evolve as more appropriate statistical techniques and new information, such as that from proteins analysed at atomic resolution, becomes available.

Incorporation of prior phase information strengthens maximum-likelihood structure refinement.

The application of a maximum-likelihood analysis to the problem of structure refinement has led to striking improvements over the traditional least-squares methods. Since the method of maximum likelihood allows for a rational incorporation of other sources of information, we have derived a likelihood function that incorporates experimentally determined phase information. In a number of different test cases, this target function performs better than either a least-squares target or a maximum-likelihood function lacking prior phases. Furthermore, this target gives significantly better results compared with other functions incorporating phase information. When combined with a procedure to mask 'unexplained' density, the phased likelihood target also makes it possible to refine very incomplete models.

The structure of the cofactor-binding fragment of the LysR family member, CysB: a familiar fold with a surprising subunit arrangement.

BACKGROUND: CysB is a tetrameric protein of identical subunits (M(r) = 36,000) which controls the expression of genes associated with the biosynthesis of cysteine in bacteria. CysB is both an activator and a repressor of transcription whose activity is responsive to the inducer N-acetylserine; thiosulphate and sulphide act as anti-inducers. CysB is a member of the LysR family of prokaryotic transcriptional regulatory proteins which share sequence similarities over approximately 280 residues including a putative helix-turn-helix DNA-binding motif at their N terminus. The aims of the present study were to explore further the complex molecular biology and curious ligand binding properties of CysB and to provide structural insights into the LysR family of proteins. RESULTS: The crystal structure of a dimeric chymotryptic fragment of Klebsiella aerogenes CysB comprising residues 88-324, has been solved by multiple isomorphous replacement and multi-crystal averaging and refined against data extending to 1.8 A resolution. The protein comprises two alpha/beta domains (I and II) connected by two short segments of polypeptide. The two domains enclose a cavity lined by polar sidechains, including those of two residues whose mutation is associated with constitutive expression of the cysteine regulon. A sulphate anion and a number of well ordered water molecules have been modelled into discrete electron-density peaks within this cavity. In the dimer, strands beta B from domain I and strands beta G from domain II come together so that a pair of antiparallel symmetry-related 11-stranded twisted beta-pleated sheets is formed. CONCLUSIONS: The overall structure of CysB(88-324) is strikingly similar to those of the periplasmic substrate-binding proteins. A similar fold has also been observed in the cofactor-binding domain of Lac repressor, implying a structural relationship between the Lac repressor and LysR families of proteins. In contrast to Lac repressor, in CysB the twofold axis of symmetry that relates the monomers in the dimer is perpendicular rather than parallel to the long axis of the cofactor-binding domain. This seems likely to place the DNA-binding domains at opposite extremes of the molecule possibly accounting for CysB's extended DNA footprints.

Refinement of macromolecular structures by the maximum-likelihood method.

This paper reviews the mathematical basis of maximum likelihood. The likelihood function for macromolecular structures is extended to include prior phase information and experimental standard uncertainties. The assumption that different parts of a structure might have different errors is considered. A method for estimating sigma(A) using 'free' reflections is described and its effects analysed. The derived equations have been implemented in the program REFMAC. This has been tested on several proteins at different stages of refinement (bacterial alpha-amylase, cytochrome c', cross-linked insulin and oligopeptide binding protein). The results derived using the maximum-likelihood residual are consistently better than those obtained from least-squares refinement.

Structure of the heme d of Penicillium vitale and Escherichia coli catalases.

A heme d prosthetic group with the configuration of a cis-hydroxychlorin gamma-spirolactone has been found in the crystal structures of Penicillium vitale catalase and Escherichia coli catalase hydroperoxidase II (HPII). The absolute stereochemistry of the two heme d chiral carbon atoms has been shown to be identical. For both catalases the heme d is rotated 180 degrees about the axis defined by the alpha-gamma-meso carbon atoms, with respect to the orientation found for heme b in beef liver catalase. Only six residues in the heme pocket, preserved in P. vitale and HPII, differ from those found in the bovine catalase. In the crystal structure of the inactive N201H variant of HPII catalase the prosthetic group remains as heme b, although its orientation is the same as in the wild type enzyme. These structural results confirm the observation that heme d is formed from protoheme in the interior of the catalase molecule through a self-catalyzed reaction.

Crystal structure of an extracellular fragment of the rat CD4 receptor containing domains 3 and 4.

BACKGROUND: CD4 is a transmembrane protein on the surface of T lymphocytes that interacts with MHC class II proteins at the surface of accessory cells, and is involved in the triggering of the lymphocytes by foreign antigens. It is also the major receptor for the human immunodeficiency virus. The extracellular portion of CD4 was predicted to contain four immunoglobulin superfamily domains and this has been confirmed by X-ray crystallography, but no detailed structure of domains 3 and 4 has been available. RESULTS: We now report the expression of a form of rat CD4 containing only domains 3 and 4, its crystallization, and the refinement and analysis of its structure by X-ray crystallography with 2.6 A spacing data. Both domains show variations in core residues when compared with immunoglobulin domains. Features of the structure are discussed with respect to the structure of the complete extracellular part of CD4 and its function. CONCLUSIONS: Domains 3 and 4 of CD4 show considerable similarity to domains 1 and 2, although there is a 25 degrees rotation in the relative positions of the domains with respect to one another. The absence of the disulphide bond in domain 3 is associated with an alteration in the packing of the beta-sheets, which may be important for interactions with domain 2 in the overall receptor structure. The location of N-linked glycosylation on one face of domain 3 appears to preclude the dimerization that is observed in antibodies.