Resolution Electron Density Map Research Articles

The structure of melittin. I. Structure determination and partial refinement.

Melittin is the principal protein component of bee venom and is thought to function as a lytic agent. Despite its predominantly hydrophobic character, melittin is soluble as a tetramer in aqueous salt solutions. We report here on the determination of the crystal structure of tetrameric melittin at 2.8-A resolution by the method of multiple isomorphous replacement, followed by partial atomic refinement at 2.0-A resolution. The melittin tetramer contains a noncrystallographic 2-fold axis of symmetry in addition to a crystallographic 2-fold axis, so that the four polypeptide chains have nearly identical structures. The noncrystallographic 2-fold axis was utilized twice during the determination of the structure. The multiple isomorphous replacement electron density map was averaged over this 2-fold axis before model building and strict noncrystallographic symmetry was assumed during the initial stages of atomic refinement. The 2.8-A resolution electron density map suggests that the melittin monomer contains two alpha-helical regions separated by a non-alpha-helical segment at residues 11 and 12. Difference maps at 2.0-A resolution tend to confirm this structure and reveal that at least six solvent molecules are bound to the melittin tetramer in the crystal. The relatively high occupancies of four of these suggest that they are ions of crystallization rather than water molecules.

Structure and heme environment of beef liver catalase at 2.5 A resolution.

Most of the amino acid side chains of beef liver catalase were clearly identifiable in the 2.5 A resolution electron-density map, and the results are in good agreement with the sequence [Schroeder, W. A., Shelton, J. R., Shelton, J. B., Roberson, B. & Apell, G. (1969) Arch. Biochem. Biophys. 131, 653-655]. The tertiary structure of one subunit consists of a large antiparallel beta-pleated sheet domain with helical insertions, followed by a smaller domain containing four alpha-helices. The heme group is buried at least 20 A below the molecular surface and is accessible by a channel lined with hydrophobic residues. The proximal ligand is tyrosine-357, while histidine-74 and asparagine-147 re the important residues on the distal side of the heme. The inhibitor 3-amino-1,2,4-triazole, which has been shown to covalently bond to histidine-74, can be built into the heme cavity with its N(2) atom coordinated to the heme iron.

A model-building procedure with particular application to proteins

A procedure for fitting a molecular model with fixed bond lengths and bond angles to a set of Cartesian coordinates has been developed. This procedure is designed to fit coordinates obtained from an X-ray electron-density map, although it can also fit coordinate sets from other sources. It differs from other model-building methods which use fixed bond lengths and bond angles in that it takes cognizance of certain information present in the electron-density map, which is ignored in other methods. The inherent uncertainty in the values of the coordinates determined from the map influences the fitting; i.e., if the computed location of an atom of the model is within the uncertainty assigned to the corresponding coordinate derived from the map, the fit of the atom of the model is assigned as exact. Also, the fitting process may begin anywhere along the molecular chain, in contrast to other procedures which commence fitting at or near one end of the chain. Thus, the present procedure can avoid errors that might arise if the starting point were chosen in a region where the atoms are poorly defined, as they often are at the ends of the chain. The procedure is designed to be used with a large amount of operator intervention, making it fairly flexible. Complete mathematical details of the method are given. A Fortran IV computer program using the method to fit a polypeptide model to a set of Cartesian coordinates has been written. The program has been used to fit a model of bovine pancreatic trypsin inhibitor to Cartesian coordinates derived from the 2.5 Å resolution electron-density map. The r.m.s. deviation between the model and the weighted coordinates from the map was 0.49 Å. As a preliminary step in a refinement of the 2.5 Å structure by potential-energy-constrained model building, the model obtained here was subjected to energy minimization with the atomic coordinates constrained to remain 'close' to the original guide points.

Iron-sulfur clusters in Azotobacter ferredoxin at 2.5 A resolution.

X-ray crystallographic study of the ferredoxin-like protein (iron-sulfur protein III) from Azotobacter vinelandii has been extended to 2.5-A resolution. A 4.0-A resolution electron density map revealed that the molecule contains two Fe-S clusters of differing size and shape separated by 12 A (Stout, C.D. (1979) Nature 279, 83-84). Recent Mössbauer results by Emptage et al. (Emptage, M.H., Kent, T.A., Huynh, B.H., Rawlings, J., Orme-Johnson, W.H., and Münck, E. (1980) J. Biol. Chem. 255, 1793-1796) have shown that the molecule contains 7 iron atoms as a high potential iron protein-like Fe4 center and a novel 3-Fe center. The 2.5-A electron density map shows two distinctly different Fe-S clusters. The larger cluster consists of a tetranuclear [4Fe-4S] core ligated to the protein at each iron atom. The smaller cluster is distinctly planar and cannot be modeled with [2Fe-2S] or [4Fe-4S] structures. The best model for this cluster is a [3Fe-3S] core. The protein makes six contacts with this cluster.

Amino acid sequences of active-site histidine peptides from rabbit muscle phosphoglycerate mutase

Monophosphogiyc~rate mutase (MPGM) catalyses the interconversion of 3-phosphoglycerate and 2-phosphoglycerate, in a reaction mechanism involving a pl~osphohistidine illter~nediate [ 1 ,I?]. A correlation of the 2.8 _& resolution electron-density map ffom X-ray crystallographic studies ofyeast MPGM with its amino acid sequence shows 2 histidines at the active site [3]. Structural studies of muscle MGP~I are much less complete, although the enzyme has been used extensively in kinetic work, e.g. [4,5]. The muscle and yeast enzymes have been shown to differ in several respects. The muscle enzyme is a dimer with identical subunits of mol. wt 28 000 [6], whereas the yeast enzyme is isolated as a tetramer composed of 4 identical subunits of mol. wt 27 000 [7]. The activity of the muscle enzyme is sensitive to modification of cysteinyl and arginyl residues [8,9], in contrast to the yeast enzyme that has no thiol [lo], and is only moderately sensitive to butanedione [l l]. In addition, a phosphohistidine peptide isolated from chicken breast muscle MGPM has been sequenced [ 121, and shows no homology with any of the 4 histidine sequences of the yeast enzyme (see table 2). We report here the purification and determination of the amino acid sequences of histidine-containing peptides from rabbit muscle MPGM to obtain an estimate of the extent of structural similarity at the active sites of the muscle and yeast enzymes.

Structure of Streptomyces erythraeus lysozyme at 6 A resolution.

A 6 A resolution electron density map has been calculated for a bacterial lysozyme produced by Streptomyces erythraeus. This lysozyme differs from the vertebrate lysozyme in its size, amino acid composition, and specificity. The structure was determined by the method of isomorphous replacement. Three heavy atom derivatives were obtained by soaking crystals of the lysozyme in HgCl2, K2PtCl4, and UO2(NO3)26H2O. The resulting electron density map clearly shows the molecular boundary. The molecule is ellipsoidal in shape with average dimensions 50 A X 35 A X 35 A. High resolution analysis and sequence analysis of the molecule are in progress.

Crystal structure of cat muscle pyruvate kinase at a resolution of 2.6 Å

The structure of pyruvate kinase (EC 2.7.1.40) has been determined from a 2.6 Å resolution electron density map. This map shows more detail than the previous 3.1 Å map (Stammers & Muirhead, 1977) and has enabled a detailed chain folding to be established for two out of the three domains which make up each of the four identical subunits. A provisional chain folding has been established for the third domain. The results have been briefly reported in a previous paper (Levine et al., 1978). Details of the structure determination and a further discussion of the results are presented in this paper. Domain A (the three domains of pyruvate kinase are referred to as A, B and C) can be described in terms of a cylindrical eight-stranded parallel β sheet and an outer coaxial cylinder of eight α helices. The α helices connect adjacent strands of the β sheet. Domain B is made up of a closed anti-parallel β sheet structure. Domain C is a five-stranded β sheet of which the fourth strand is anti-parallel and the rest parallel. These strands are also interconnected by α helices. Domain A can be dissected into eight consecutive β strand—α helix units starting from the N-terminus. The arrangement of these relative to each other can be most simply described by relating them to eight planes, each at 40 ° to the cylinder axis and symmetrically placed around the cylinder. When unit 2 is aligned with one of these planes then units 1, 3, 4, 5 and 8 are also closely aligned with a plane. This analysis is also applied to triosephosphate isomerase and a strikingly similar arrangement is found. A detailed comparison of the two structures is presented. Although the lack of a chemical sequence makes it difficult to identify the amino acid residues of pyruvate kinase, side-chains are clearly visible in the map and this information is correlated with the results of previous 6 Å substrate soaking experiments and with the structure of triosephosphate isomerase. The similarities and differences are discussed in terms of similarities and differences in the reactions catalysed and also of different subunit packing.

Evidence for the bilobal nature of diferric rabbit plasma transferrin.

Plasma transferrin is involved in iron transport within the circulatory system of vertebrates, and provides an iron source for haemoglobin synthesis and other metabolic requirements. However, despite extensive studies by spectroscopic, biochemical and physiological techniques, the nature of iron binding and the mechanisms of uptake and release of iron are not fully understood. Plasma transferrins are monomeric glycoproteins with a molecular weight of approximately 80,000 (ref. 2); they have two similar and very strong binding sites for Fe(III), together with two associated anion binding sites. Fragmentation studies on various transferrins have shown that the polypeptide chain is composed of two domains formed from the N-terminal and C-terminal halves of the polypeptide chain. Each domain contains one metal binding site. The marked sequence similarities which exist between the two halves may reflect a doubling of an ancestral structural gene during the phylogenetic development of the protein. Preliminary crystallographic investigations of diferric rabbit plasma transferrin have been reported from this laboratory. We now report initial studies of the X-ray structure determination of dife-ric rabbit plasma transferrin which have led to a 6-A resolution electron density map.

Structure of a bacteriochlorophyll a-protein from the green photosynthetic bacterium Prosthecochloris aestuarii

The three-dimensional structure of a water-soluble bacteriochlorophyll a-containing protein from the green photosynthetic bacterium Prosthecochloris aestuarii has been determined by X-ray crystallography from a 2.8 Å resolution electron density map based on four isomorphous derivatives. Details of the crystallographic procedures used to obtain the map are presented. The bacteriochlorophyll a-protein is shown to consist of three identical subunits, tightly packed around a 3-fold symmetry axis. Each subunit consists of a core of seven bacteriochlorophyll a molecules enclosed within a “bag” of protein. The polypeptide chain forms an extensive 15-strand β-sheet, which is almost planar in its central region, and twisted at its extremities, and wraps around the chlorophyll core to form an efficient amphipathic layer between the chlorophylls and the aqueous environment. There are extensive contacts between the phytyl chains of the seven bacteriochlorophylls within each subunit. These hydrocarbon chains constitute an inner hydrophobic core of the molecule which may be important in forming the complex. There are also extensive contacts between the protein and both the bacteriochlorophyll head groups and tails, but relatively few contacts between the respective head groups. The seven magnesiums all appear to be five co-ordinated. In five cases the presumed ligand is a histidine side-chain, in one case a polypeptide carbonyl oxygen, and in the other case a water molecule. At low temperature, both the absorption and circular dichroism spectra of the bacteriochlorophyll a-protein show splitting which can be interpreted in general terms as due to exciton interactions between the seven chromophores, but calculations of the expected splitting based on the bacteriochlorophyll co-ordinates determined crystallographically are in poor agreement with the observed spectra. Furthermore, the observed red shift of the Q y absorption band of bacteriochlorophyll a, from about 770 nm in organic solvents to 809 nm in the bacteriochlorophyll a-protein, is not explained by the exciton calculations. It seems likely that the red shift is due to perturbations of the spectra of the individual bacteriochlorophylls by the protein environment, but, pending the determination of the amino acid sequence, it is not possible at this time to define in detail all the proteinchlorophyll interactions. It is suggested that the bacteriochlorophyll a-protein serves as a good model for the organization of chlorophyll in vivo, and that the types of interaction seen here between chlorophyll and protein are likely to be found in other chlorophyll proteins.

Three-dimensional structure of thioredoxin induced by bacteriophage T4.

The three-dimensional structure of thioredoxin from bacteriophage T4 has been determined from a 2.8-angstrom resolution electron density map. Phase angles for this map were determined from one heavy atom derivative and anomalous differences from cadmium in the native crystals. The molecule of 87 amino acid residues is built up from two simple folding units; a betaalphabeta unit from the amino end of the chain and a betabetaalpha unit from the carboxyl end. This structure is similar to that of thioredoxin from Escherichia coli in spite of their completely different amino acid sequences. The redox-active S--S bridge is part of a protrusion of the molecule as in E. coli thioredoxin, but with quite different surroundings. The structural differences in this region have been correlated to differences in specificity towards the enzyme ribonucleotide reductase from different species.

Sequencing a protein by X-ray crystallography: I. Interpretation of yeast hexokinase B at 2.5 Å resolution by model building

Because the chemically derived amino acid sequence for yeast hexokinase B is not known, a skeletal model of the 51,000 molecular weight monomer (complexed with the inhibitor, o-toluoylglucosamine) was built from a 2.5 Å resolution electron density map as a first step in an attempt to derive an approximate amino acid sequence, as well as accurate atomic co-ordinates, by crystallographic refinement. The electron density map used in this study was improved over that previously reported (Steitz et al., 1976) by the inclusion of an additional and excellent heavy atom derivative in the phasing, by extension of the resolution to 2.5 Å, and by improvements in the data processing and heavy atom refinement. Although side-chain identities were surmised for most residues from the shape of the electron density, it is unlikely that more than one-fourth of the residues are correct at this stage due to inherent crystal disorder and inaccurate phases. The model contains only 2909 of the approximately 3596 non-hydrogen atoms expected from the amino acid composition, yet it provides an adequate starting point for crystallographic refinement of the amino acid sequence and the coordinates. The model confirms the overall path of the polypeptide backbone presented earlier (Fletterick et al., 1975) but differs most notably in that a short segment of chain thought to be in extended conformation is now seen to be an α-helix. This change eliminates the only left-handed β-chain-strand-β-chain super-secondary structure in our model. Also, the 13 α-helices in hexokinase that are in close contact with each other cross pairwise at angles of about 10 °, 40 ° to 60 °, and 80 ° to 90 °. The sugar moiety of the inhibitor, o-toluoylglucosamine, binds to hexokinase in the C-1 chair equatorial conformation with the 1-hydroxyl axial (α-anomer). Specific interactions between protein side-chains and both the 4-hydroxyl and 3-hydroxyl groups can be seen but it is not clear in this map whether there might also be interactions with the 1- and 6-hydroxyl groups.

A crystallographic model for azurin at 3 Å resolution

The structure of the blue copper protein azurin ( M r 14,000) from Pseudomonas aeruginosa has been determined from a 3.0 Å resolution electron density map computed with phases based on a uranyl derivative to 3 Å resolution and a platinum derivative to 3.7 Å. Interpretation of the somewhat noisy map was based on comparison of the density of the four molecules in the asymmetric unit with their averaged density. The polypeptide chain folds into an eight-strand β barrel with an additional flap containing a short helix. The copper atom is bound at one end and on the inside of the barrel, probably to a cysteine, a methionine, and two histidine residues.

Structure of the lysozyme from bacteriophage T4: An electron density map at 2.4 Å resolution

The three-dimensional structure of the lysozyme from bacteriophage T4 has been determined from a 2.4 Å resolution electron density map. The map, determined using isomorphous replacement and anomalous scattering differences from platinum and mercury-substituted crystals, reveals the approximate conformation of the molecule unambiguously. Details of the structure determination are presented and the conformation of the molecule is described in detail. The observed properties of those mutant lysozymes for which the changes in the amino acid sequence have been determined are discussed in the light of the three-dimensional structure. In most cases the behavior of the mutant lysozymes can be easily rationalized, but there are a few instances where this is not so.

The atomic structure of crystalline porcine pancreatic elastase at 2.5 Å resolution: Comparisons with the structure of α-chymotrypsin

Three isomorphous heavy-atom derivatives have been used to calculate a 2.5 Å resolution electron density map of tosyl-elastase at pH 5.0, from which an accurate atomic model has been constructed. Atomic co-ordinates measured from this model have been refined using model building, real-space refinement and energy minimization programs. The three-dimensional conformation of the polypeptide chain is described in terms of conformational angles, hydrogen-bonding networks and the environment of different types of amino acid side-chain. Difference Fourier calculation of the high resolution structure of native elastase at pH 5.0 shows it to be virtually identical to that of the tosyl derivative, except near the tosyl group. The conformation of the catalytically important residues in native elastase is very similar to that of native α-chymotrypsin, except for the orientation of the active centre serine oxygen. The significance of important structural similarities and differences between these two enzymes is discussed. Elastase contains 25 internal water molecules which play an important role in stabilizing the active conformation of the enzyme. Many of these water molecules are in identical positions to those found in the interior of α-chymotrypsin

Structure of actinidin: Details of the polypeptide chain conformation and active site from an electron density map at 2·8 Å resolution

A 2·8 Å resolution electron density map of the sulphydryl protease, actinidin, has been calculated. Two isomorphous heavy-atom derivatives, prepared with uranyl acetate and dichloroethylenediamineplatinum(II), were used to calculate phases by the method of isomorphous replacement, giving an overall figure of merit of 0·81. The polypeptide chain is well-defined in the present map and many side-chains can be identified from their appearance. The molecule consists of a single chain of 220 residues, the last two of which appear disordered in the map and contains at least two, and probably three, disulphide bridges. The conformation of the polypeptide chain is remarkably similar to that of papain. It is folded into two domains, domain I consisting of residues 19–115 and 214–218, and domain II residues 1–18 and 116–213. There are three significant stretches of α-helix, involving residues 25–42, 69–81 and 120–129, together with several shorter pieces, while much of domain II consists of a twisted β-sheet structure. When compared with papain, actinidin has two additional residues inserted between 59 and 60, one inserted between 78 and 79, and four between 168 and 169 (papain numbering) while one residue (194) has been deleted from the papain structure. All these changes are in external parts of the molecule and have little effect on the conformation. The positions and orientations of the catalytically-important side-chains in the active site are virtually identical with those in papain, but some of the side-chains lining the non-polar binding pocket are clearly different.

Method for obtaining a high resolution protein map starting from a low resolution map.

A method is described for estimating the phases of high resolution single-crystal diffraction data from proteins, by using as a starting point a set of low resolution phases (about 3 A) derived by multiple isomorphous replacement (or other) methods. The method consists in refining by least-squares the positions and thermal parameters of a set of dummy atoms placed in the initial low resolution electron density map, so as to minimize the discrepancy between the calculated scattering intensities and the scattering intensities observed in the high resolution data set. Phases calculated from these refined atomic positions are used to extend the resolution and to improve the quality of the electron density map. The success of the method depends on a new least-squares algorithm that has a radius of convergence of about 0.75 A. This large radius of convergence, together with the severe restrictions placed on the initial positions of the dummy atoms by the requirement that they lie within limited regions of the isomorphous electron density map, and the constraint imposed by the polymeric nature of a polypeptide chain account for the success of the method. The method has been successfully used to phase the structure factors of 2-zinc insulin at a resolution of 2 A and 1.5 A, starting from a set of isomorphous phases at 3-A resolution.

Structure of 6-phosphogluconate dehydrogenase from sheep liver at 6 Å resolution

A 6 A resolution electron density map has been calculated for 6-phosphogluconate dehydrogenase from sheep liver. The dimeric enzyme crystallizes with a single subunit in the asymmetric unit. The structure was determined by the method of isomorphous replacement. The two heavy-atom derivatives used were obtained by soaking crystals in potassium dicyanoaurate, which bound at two sites, and potassium tetracyanoplatinate, which bound at three different sites in the monomer. The resulting electron density map shows the molecular boundary and indicates which of the crystallographic 2-fold axes is the molecular 2-fold. The subunit is ellipsoidal with a large indentation distant from the subunit boundary. Several columns of density of the correct dimensions for α-helices can be seen. Some of these are a part of the subunit contact area and two are more than 40 A units long.

Atomic coordinates for the chlorophyll core of a bacteriochlorophyll [formula omitted]-protein from green photosynthetic bacteria

Preliminary atomic coordinates are presented for the seven bacteriochlorophylls constituting the core of one subunit of the trimeric water soluble bacteriochlorophyll a -protein from the green photosynthetic bacterium Prosthecochloris aestuarii , strain 2K, formerly identified as Chlorobium limicola , strain 2K. The coordinates were derived from a 2.8 Å resolution electron density map based on four isomorphous heavy-atom derivatives, and adjusted to have stereochemically acceptable bond lengths and angles.

Atomic coordinates for T4 phage lysozyme

Atomic coordinates are presented for the lysozyme from T4 bacteriophage. The coordinates were derived from a 2.4 Å resolution electron density map based on two isomorphous heavy-atom derivatives, interpreted in terms of the known amino acid sequence, and adjusted to have stereochemically acceptable bond lengths and angles.

Structure of glycogen phosphorylase a at 3.0 A resolution and its ligand binding sites at 6 A.

A model of the polypeptide backbone of the dimer of glycogen phosphorylase a (EC 2.4.1.1) was built from a 3 A resolution electron density map derived from x-ray diffraction analysis of native tetragonal crystals and two heavy atom isomorphous replacement derivatives. Each identical subunit of the dimer has a compact shape with overall dimensions of 85 X 75 X 55 A and is tightly associated with its 2-fold symmetry related subunit. There are three major excursions of the polypeptide chain of one monomer across the 2-fold axis to make extensive contacts with the other subunit. The active site, of which there are two per dimer, is shared between the two subunits at their interface and comprises a pocket-like region within a "V"-shaped framework of two alpha helices. Within this region are found the binding sites for the substrates, glucose-1-P and arsenate, a competitive inhibitor, UDP-glucose, and the allosteric effector, AMP. The site of metabolic control, Ser-14 phosphate, is hydrogen-bonded to a side chain on the outside of one of the alpha helices forming the active site and is 15 A from the AMP binding site. Maltoheptaose, a glycogen analogue and substrate for these enzymatically active crystals, binds in a second region of interest. Even at concentrations above its Km, when binding is sufficiently tight that all seven glucose moieties may be discerned, the closest of these is 25 A from the glucose-1-P binding site. We suggest that this polysaccharide binding site may represent a storage site whereby phosphorylase is bound to the glycogen particle in the muscle cell. The polypeptide chain in a third region has the same topological structure as has been observed for the nucleotide binding domains in the dehydrogenases. Adenine or adenosine (but not AMP) bind here in a position similar to the adenine ring of NAD in the dehydrogenases while glucose binds 17 A away in an interior crevice near the center of the monomer.