The 1.9 A crystal structure of the "as isolated" rubrerythrin from Desulfovibrio vulgaris: some surprising results.

Rubrerythrin is a non-heme iron dimeric protein isolated from the sulfate-reducing bacterium Desulfovibrio vulgaris. Each monomer has one mononuclear iron center similar to rubredoxin and one dinuclear metal center similar to hemerythrin or ribonucleotide reductase. The 1.88 A X-ray structure of the "as isolated" molecule and a uranyl heavy atom derivative have been solved by molecular replacement techniques. The resulting model of the native "as isolated" molecule, including 164 water molecules, has been refined giving a final R factor of 0.197 (R(free) = 0.255). The structure has the same general protein fold, domain structure, and dimeric interactions as previously found for rubrerythrin [1, 2], but it also has some interesting undetected differences at the metal centers. The refined model of the protein structure has a cis peptide between residues 78 and 79. The Fe-Cys4 center has a previously undetected strong seventh N-H...S hydrogen bond in addition to the six N-H...S bonds usually found in rubredoxin. The dinuclear metal center has a hexacoordinate Fe atom and a tetracoordinate Zn atom. Each metal is coordinated by a GluXXHis polypeptide chain segment. The Zn atom binds at a site distinctly different from that found in the structure of a diiron rubrerythrin. Difference electron density for the uranyl derivative shows an extremely large peak adjacent to and replacing the Zn atom, indicating that this particular site is capable of binding other atoms. This feature/ability may give rise to some of the confusing activities ascribed to this molecule.

The primary and three-dimensional structures of a nine-haem cytochrome c from Desulfovibrio desulfuricans ATCC 27774 reveal a new member of the Hmc family.

BACKGROUND: Haem-containing proteins are directly involved in electron transfer as well as in enzymatic functions. The nine-haem cytochrome c (9Hcc), previously described as having 12 haem groups, was isolated from cells of Desulfovibrio desulfuricans ATCC 27774, grown under both nitrate- and sulphate-respiring conditions. RESULTS: Models for the primary and three-dimensional structures of this cytochrome, containing 292 amino acid residues and nine haem groups, were derived using the multiple wavelength anomalous dispersion phasing method and refined using 1.8 A diffraction data to an R value of 17.0%. The nine haem groups are arranged into two tetrahaem clusters, with Fe-Fe distances and local protein fold similar to tetrahaem cytochromes c3, while the extra haem is located asymmetrically between the two clusters. CONCLUSIONS: This is the first known three-dimensional structure in which multiple copies of a tetrahaem cytochrome c3-like fold are present in the same polypeptide chain. Sequence homology was found between this cytochrome and the C-terminal region (residues 229-514) of the high molecular weight cytochrome c from Desulfovibrio vulgaris Hildenborough (DvH Hmc). A new haem arrangement in domains III and IV of DvH Hmc is proposed. Kinetic experiments showed that 9Hcc can be reduced by the [NiFe] hydrogenase from D. desulfuricans ATCC 27774, but that this reduction is faster in the presence of tetrahaem cytochrome c3. As Hmc has never been found in D. desulfuricans ATCC 27774, we propose that 9Hcc replaces it in this organism and is therefore probably involved in electron transfer across the membrane.

Refinement of triclinic hen egg-white lysozyme at atomic resolution.

X-ray diffraction data have been collected at both low (120 K) and room temperature from triclinic crystals of hen egg-white lysozyme to 0.925 and 0.950 A resolution, respectively, using synchrotron radiation. Data from one crystal were sufficient for the low-temperature study, whereas three crystals were required at room temperature. Refinement was carried out using the programs PROLSQ, ARP and SHELXL to give final conventional R factors of 8.98 and 10.48% for data with F > 4sigma(F) for the low- and room-temperature structures, respectively. The estimated r.m.s. coordinate error is 0.032 A for protein atoms, 0.050 A for all atoms in the low-temperature study, and 0.038 A for protein atoms and 0.049 A for all atoms in the room-temperature case, as estimated from inversion of the blocked least-squares matrix. The low-temperature study revealed that the side chains of 24 amino acids had multiple conformations. A total of 250 waters, six nitrate ions and three acetate ions, two of which were modelled with alternate orientations were located in the electron-density maps. Three sections of the main chain were modelled in alternate conformations. The room-temperature study produced a model with multiple conformations for eight side chains and a total of 139 water molecules, six nitrate but no acetate ions. The occupancies of the water molecules were refined in both structures and this step was shown to be meaningful when assessed by use of the free R factor. A detailed description and comparison of the structures is made with reference to the previously reported structure refined at 2.0 A resolution.

Atomic resolution (0.94 A) structure of Clostridium acidurici ferredoxin. Detailed geometry of [4Fe-4S] clusters in a protein.

The crystal structure of the 2[4Fe-4S] ferredoxin from Clostridium acidurici has been solved using X-ray diffraction data extending to atomic resolution, 0.94 A, recorded at 100 K. The model was refined with anisotropic representation of atomic displacement parameters for all non-hydrogen atoms and with hydrogens riding on their parent atoms. Stereochemical restraints were applied to the protein chain but not to the iron-sulfur clusters. The final R factor is 10.03 % for all data. Inversion of the final least-squares matrix allowed direct estimation of the errors of individual parameters. The estimated errors in positions for protein main chain atoms are below 0.02 A and about 0.003 A for the heavier [4Fe-4S] cluster atoms. Significant differences between the stereochemistry of the two clusters and distortion of both of them from ideal Td tetrahedral symmetry can be defined in detail at this level of accuracy. Regions of alternative conformations include not only protein side chains but also two regions of the main chain. One such region is the loop of residues 25-29, which was highly disordered in the room temperature structure.

Preliminary crystallographic analysis and further characterization of a dodecaheme cytochrome c from Desulfovibrio desulfuricans ATCC 27774.

Dodecaheme cytochrome c has been purified from Desulfovibrio (D.) desulfuricans ATCC 27774 cells grown under both nitrate and sulfate-respiring conditions. Therefore, it is likely to play a role in the electron-transfer system of both respiratory chains. Its molecular mass (37768 kDa) was determined by electrospray mass spectrometry. Its first 39 amino acids were sequenced and a motif was found between amino acids 32 and 37 that seems to exist in all the cytochromes of the c(3) type from sulfate-reducing bacteria sequenced at present. The midpoint redox potentials of this cytochrome were estimated to be -68, -120, -248 and -310 mV. Electron paramagnetic resonance spectroscopy of the oxidized cytochrome shows several low-spin components with a g(max) spreading from 3.254 to 2.983. Two crystalline forms were obtained by vapour diffusion from a solution containing 2% PEG 6000 and 0.25-0.75 M acetate buffer pH = 5.5. Both crystals belong to monoclinic space groups: one is P2(1), with a = 61.00, b = 106.19, c = 82.05 A, beta = 103.61 degrees, and the other is C2 with a = 152.17, b = 98.45, c = 89.24 A, beta = 119.18 degrees. Density measurements of the P2(1) crystals suggest that there are two independent molecules in the asymmetric unit. Self-rotation function calculations indicate, in both crystal forms, the presence of a non-crystallographic axis perpendicular to the crystallographic twofold axis. This result and the calculated values for the volume per unit molecular weight of the C2 crystals suggest the presence of two or four molecules in the asymmetric unit.

Crystal structure of the 2[4Fe-4S] ferredoxin from Chromatium vinosum: evolutionary and mechanistic inferences for [3/4Fe-4S] ferredoxins.

The crystal structure of the 2[4Fe-4S] ferredoxin from Chromatium vinosum has been solved by molecular replacement using data recorded with synchrotron radiation. The crystals were hexagonal prisms that showed a strong tendency to develop into long tubes. The hexagonal prisms diffracted to 2.1 A resolution at best, and a structural model for C. vinosum ferredoxin has been built with a final R of 19.2%. The N-terminal domain coordinates the two [4Fe-4S] clusters in a fold that is almost identical to that of other known ferredoxins. However, the structure has two unique features. One is a six-residue insertion between two ligands of one cluster forming a two-turn external loop; this short loop changes the conformation of the Cys 40 ligand compared to other ferredoxins and hampers the building of one NH...S H-bond to one of the inorganic sulfurs. The other remarkable structural element is a 3.5-turn alpha-helix at the C-terminus that covers one side of the same cluster and is linked to the cluster-binding domain by a six-residue external chain segment. The charge distribution is highly asymmetric over the molecule. The structure of C. vinosum ferredoxin strongly suggests divergent evolution for bacterial [3/4Fe-4S] ferredoxins from a common ancestral cluster-binding core. The unexpected slow intramolecular electron transfer rate between the clusters in C. vinosum ferredoxin, compared to other similar proteins, may be attributed to the unusual electronic properties of one of the clusters arising from localized changes in its vicinity rather than to a global structural rearrangement.

Refined structures of bobwhite quail lysozyme uncomplexed and complexed with the HyHEL-5 Fab fragment.

The HyHEL-5 antibody has more than a thousandfold lower affinity for bobwhite quail lysozyme (BWQL) than for hen egg-white lysozyme (HEL). Four sequence differences exist between BWQL and HEL, of which only one is involved in the interface with the Fab. The structure of bobwhite quail lysozyme has been determined in the uncomplexed state in two different crystal forms and in the complexed state with HyHEL-5, an antihen egg-white lysozyme Fab. Similar backbone conformations are observed in the three molecules of the two crystal forms of uncomplexed BWQL, although they show considerable variability in side-chain conformation. A relatively mobile segment in uncomplexed BWQL is observed to be part of the HyHEL-5 epitope. No major backbone conformational differences are observed in the lysozyme upon complex formation, but side-chain conformational differences are seen in surface residues that are involved in the interface with the antibody. The hydrogen bonding in the interface between BWQL and HyHEL-5 is similar to that in previously determined lysozyme-HyHEL-5 complexes.

Zinc- and iron-rubredoxins from Clostridium pasteurianum at atomic resolution: a high-precision model of a ZnS4 coordination unit in a protein.

The Zn(Scys)4 unit is present in numerous proteins, where it assumes structural, regulatory, or catalytic roles. The same coordination is found naturally around iron in rubredoxins, several structures of which have been refined at resolutions of, or near to, 1 A. The fold of the small protein rubredoxin around its metal ion is an excellent model for many zinc finger proteins. Zn-substituted rubredoxin and its Fe-containing counterpart were both obtained as the products of the expression in Escherichia coli of the rubredoxin-encoding gene from Clostridium pasteurianum. The structures of both proteins have been refined with an anisotropic model at atomic resolution (1.1 A, R = 8.3% for Fe-rubredoxin, and 1.2 A, R = 9.6% for Zn-rubredoxin) and are very similar. The most significant differences are increased lengths of the M-S bonds in Zn-rubredoxin (average length, 2.345 A) as compared with Fe-rubredoxin (average length, 2.262 A). An increase of the CA-CB-SG-M dihedral angles involving Cys-6 and Cys-39, the first cysteines of each of the Cys-Xaa-Xaa-Cys metal binding motifs, has been observed. Another consequence of the replacement of iron by zinc is that the region around residues 36-46 undergoes larger displacements than the remainder of the polypeptide chain. Despite these changes, the main features of the FeS4 site, namely a local 2-fold symmetry and the characteristic network of N-H...S hydrogen bonds, are conserved in the ZnS4 site. The Zn-substituted rubredoxin provides the first precise structure of a Zn(Scys)4 unit in a protein. The nearly identical fold of rubredoxin around iron or zinc suggests that at least in some of the sites where the metal has mainly a structural role-e.g., zinc fingers-the choice of the relevant metal may be directed by its cellular availability and mobilization processes rather than by its chemical nature.

Refined structures at 2 and 2.2 A resolution of two forms of the H-protein, a lipoamide-containing protein of the glycine decarboxylase complex.

H-protein, a 14 kDa lipoic acid-containing protein is a component of the glycine decarboxylase complex. This complex which consists of four protein components (P-, H-, T- and L-protein) catalyzes the oxidative decarboxylation of glycine. The mechanistic heart of the complex is provided by the lipoic acid attached to a lysine residue of the H-protein. It undergoes a cycle of transformations, i.e. reductive methylamination, methylamine transfer, and electron transfer. We present details of the crystal structures of the H-protein, in its two forms, H-Pro(Ox) with oxidized lipoamide and H-Pro(Met) with methylamine-loaded lipoamide. X-ray diffraction data were collected from crystals of H-Pro(Ox) to 2 and H-Pro(Met) to 2.2 A resolution. The final R-factor value for the H-Pro(Ox) is 18.5% for data with F > 2sigma. in the range of 8.0-2.0 A resolution. The refinement confirmed our previous model, refined to 2.6 A, of a beta-fold sandwich structure with two beta-sheets. The lipoamide arm attached to Lys63, located in the loop of a hairpin conformation, is clearly visible at the surface of the protein. The H-Pro(Met) has been crystallized in orthorhombic and monoclinic forms and the structures were solved by molecular replacement, starting from the H-Pro(Ox) model. The orthorhombic structure has been refined with a final R-factor value of 18.5% for data with F > 2sigma in the range of 8.0-2.2 A resolution. The structure of the monoclinic form has been refined with a final R-factor value of 17.5% for data with F > 2sigma in the range of 15.0-3.0 A. In these two structures which have similar packing, the protein conformation is identical to the conformation found in the H-Pro(Ox). The main change lies in the position of the lipoamide group which has moved significantly when loaded with methylamine. In this case the methylamine-lipoamide group is tucked into a cleft at the surface of the protein where it is stabilized by hydrogen bonds and hydrophobic contacts. Thus, it is totally protected and not free to move in aqueous solvent. In addition, the H-protein presents some sequence and structural analogies with other lipoate- and biotin-containing proteins and also with proteins of the phosphoenolpyruvate:sugar phosphotransferase system.

Crystal structure of desulforedoxin from Desulfovibrio gigas determined at 1.8 A resolution: a novel non-heme iron protein structure.

The crystal structure of desulforedoxin from Desulfovibrio gigas, a new homo-dimeric (2 x 36 amino acids) non-heme iron protein, has been solved by the SIRAS method using the indium-substituted protein as the single derivative. The structure was refined to a crystallographic R-factor of 16.9% at 1.8 A resolution. Native desulforedoxin crystals were grown from either PEG 4K or lithium sulfate, with cell constants a = b = 42.18 A, c = 72.22 A (for crystals grown from PEG 4K), and they belong to space group P3(2)21. The indium-substituted protein crystallized isomorphously under the same conditions. The 2-fold symmetric dimer is firmly hydrogen bonded and folds as an incomplete beta-barrel with the two iron centers placed on opposite poles of the molecule. Each iron atom is coordinated to four cysteinyl residues in a distorted tetrahedral arrangement. Both iron atoms are 16 A apart but connected across the 2-fold axis by 14 covalent bonds along the polypeptide chain plus two hydrogen bonds. Desulforedoxin and rubredoxin share some structural features but show significant differences in terms of metal environment and water structure, which account for the known spectroscopic differences between rubredoxin and desulforedoxin.

Refined crystal structure of the 2[4Fe-4S] ferredoxin from Clostridium acidurici at 1.84 A resolution.

The crystal structure of the 2[4Fe-4S] ferredoxin from Clostridium acidurici has been determined at a resolution of 1.84 A and refined to an R-factor of 0.169. Crystals belong to space group P4(3)2(1)2 with unit cell dimensions a = b = 34.44 A and c = 74.78 A. The structure was determined by molecular replacement using the previously published model of an homologous ferredoxin and refined by molecular dynamics techniques. The model contains the protein and 46 water molecules. Only two amino acid residues, Asp27 and Asp28, are poorly defined in the electron density maps. The molecule has an overall chain fold similar to that of other [4Fe-4S] bacterial ferredoxins of known structure. The two [4Fe-4S] clusters display similar bond distances and angles. In both of them the co-ordination of one iron atom (bound to Cys11 and Cys40) is slightly distorted as compared with that of the other iron atoms. A core of hydrophobic residues and a few water molecules contribute to the stability of the structure. The [4Fe-4S] clusters interact with the polypeptide chain through eight hydrogen bonds each, in addition to the covalent Fe-Scys bonds. The ferredoxin from Clostridium acidurici is the most typical clostridial ferredoxin crystallized so far and the biological implications of the newly determined structure are discussed.

Sequential 1H NMR assignment of the complex of aponeocarzinostatin with ethidium bromide and investigation of protein-drug interactions in the chromophore binding site.

Two-dimensional 1H NMR spectroscopy has been used to investigate the binding site, binding interactions, and the conformation of a 1:1 complex of aponeocarzinostatin (apo-NCS) with ethidium bromide in solution. The protein component in the complex has been sequence-specifically assigned using information derived from coherence transfer and two-dimensional homonuclear 1H NOESY experiments. The conformation of the protein in the complex has been found to be similar to the free form of the apoprotein, and intermolecular NOEs between the residues of the protein to protons on the ethidium bromide suggest that the ethidium bromide is bound to the protein in the same cleft in which the neocarzinostatin chromophore binds. Protons on ethidium show NOE interactions to the following protein residues: Trp-39, Leu-45, Cys-47, and Gln-94 which interact with the phenanthridine ring system of ethidium, Gly-102 and Asn-103 which interact with the alkyl chain of ethidium, and Phe-52 which interacts with the phenyl ring of ethidium. The orientation of ethidium in the cleft of apo-NCS is compared and contrasted to orientation of the chromophore as determined by high-resolution NMR and X-ray diffraction studies.

26-10 Fab-digoxin complex: affinity and specificity due to surface complementarity.

We have determined the three-dimensional structures of the antigen-binding fragment of the anti-digoxin monoclonal antibody 26-10 in the uncomplexed state at 2.7 A resolution and as a complex with digoxin at 2.5 A resolution. Neither the antibody nor digoxin undergoes any significant conformational changes upon forming the complex. Digoxin interacts primarily with the antibody heavy chain and is oriented such that the carbohydrate groups are exposed to solvent and the lactone ring is buried in a deep pocket at the bottom of the combining site. Despite extensive interactions between antibody and antigen, no hydrogen bonds or salt links are formed between 26-10 and digoxin. Thus the 26-10-digoxin complex is unique among the known three-dimensional structures of antibody-antigen complexes in that specificity and high affinity arise primarily from shape complementarity.

Structure at pH 6.5 of ferredoxin I from Azotobacter vinelandii at 2.3 A resolution.

Ferredoxin I from Azotobacter vinelandii (AvFdI) is an iron-sulfur protein composed of 106 amino acids, seven Fe atoms and eight inorganic S* atoms. A crystallographic redetermination of its structure showed the originally reported structure to be incorrect. We report here the crystal structure of AvFdI at pH 6.5. Extensive refinement has led to a final R value of 0.170 for all 6986 non-extinct reflections in the range 10-2.3 A using a solvent model which includes 98 discrete solvent atoms with occupancies between 0.3 and 1.0 and an average B value of 22.5 A(2). The first half of the peptide chain closely resembles that of the 55-residue ferredoxin from Peptococcus aerogenes (PaFd), while the remainder consists of three turns of helix and a series of loops which form a cap over part of the molecular core. Despite the similarities in structure and surroundings, the corresponding 4Fe4S* clusters in PaFd and AvFdI have strikingly different redox potentials; a possible explanation has been sought in the differing hydration models for the two molecules.

The application of direct methods and Patterson interpretation to high-resolution native protein data.

Conventional small-molecule methods of solving the phase problem from native data alone, without the use of heavy-atom derivatives, known fragment geometries or anomalous dispersion, have been tested on 0.9 A resolution data for two small proteins: rubredoxin, from Desulfovibrio vulgaris, and crambin. The presence of three disulfide bridges in crambin and an FeS(4) unit in rubredoxin enabled automated Patterson interpretation as well as direct methods to be tried. Although both structures were already well established, the known structures were not used in the phasing attempts, except for identifying successful solutions. Direct methods were not successful for crambin, although the correct phases were stable to phase refinement and gave figures of merit clearly superior to any obtained in the ca 500 000 random starting phase sets that were refined. It appears that the presence of an iron atom in rubredoxin reduces the scale of the search problem by many orders of magnitude, but at the cost of producing 'over-consistent' phase sets that overemphasize the iron atom and involve partial loss of enantiomorph information. However, about 1% of direct-methods trials were successful for rubredoxin, giving mean phase errors of about 56 degrees (for all E > 1.2) that could be reduced to about 20 degrees by standard E-Fourier recycling methods. Limiting the resolution of the data degraded the quality of the solutions and suggested that the limiting resolution for routine direct-methods solution of rubredoxin is about 1.2 A. With the 0.9 A data, automated Patterson interpretation convincingly finds the three disulfide bridges in crambin and the FeS(4) unit in rubredoxin, and in both cases E-Fourier recycling starting from these 'heavier' atoms yields almost the complete structure. Whereas crambin could only be solved in this way at very high resolution, rubredoxin could be solved by Patterson interpretation down to 1.6 A. These results emphasize the benefits of collecting protein data to the highest possible resolution, and indicate that when a few 'heavier' atoms are present, it may prove possible in favorable cases to solve the phase problem from a single native data set collected to 'atomic resolution'.

Refinement of rubredoxin from Desulfovibrio vulgaris at 1.0 A with and without restraints.

X-ray data have been recorded from crystals of rubredoxin derived from the bacterium Desulfovibrio vulgaris to a resolution of 1.0 A using in part synchrotron radiation and in part X-rays from a sealed-tube Mo K alpha source. In both cases an imaging-plate scanner was used as detector. The space group of the crystals is P2(1) with cell dimensions a = 19.97, b = 41.45, c = 24.41 A and beta = 108.3 degrees. The overall merging R(I) factor between symmetry-related reflections was 5.8%. The model was refined by least-squares minimization initially with stereochemical restraints to an R factor of 16.4%. Only atomic positional parameters and isotropic temperature factors for non-H atoms were used in the refinement. There were 18,532 independent X-ray observations for a total of 1916 atomic parameters. A round of unrestrained refinement gave an R factor of 16.0%, acceptable geometry for more than 90% of the protein atoms, but emphasized the disorder inherent in eight of the residues. A final round of restrained refinement gave an R factor of 14.7%. Three of the 389 protein atoms in the molecule, in the side chain of Lys2, have been assigned zero occupancy in the model. A total of eight atoms in three side chains have been assigned two conformations, giving 393 protein atomic sites in the model. In addition there is one Fe atom, a sulfate ion and 102 water sites. 339 H atoms were included at their calculated positions, which were not refined. There is clear evidence for anisotropic thermal motion. This has not been incorporated in the present model.

Refined crystal structure of ferredoxin II from Desulfovibrio gigas at 1.7 A.

The crystal structure of ferredoxin II from Desulfovibrio gigas has been determined using phasing from anomalous scattering data at a resolution of 1.7 A and refined to an R-factor of 0.157. The molecule has an overall chain fold similar to that of the other bacterial ferredoxins of known structure. The molecule contains a single 3Fe-4S cluster with geometry indistinguishable from the 4Fe-4S clusters, and a disulfide bond near the site corresponding to the position of the second cluster of two-cluster ferredoxins. The cluster is bound by cysteine residues 8, 14 and 50. The side-chain of cysteine 11 extends away from the cluster, but could rotate to become the fourth cysteine ligand in the four-iron form of the molecule given a local adjustment of the polypeptide chain. This residue is modified, however, by what appears to be a methanethiol group. There are a total of eight NH . . . S bonds to the inorganic and cysteine sulfur atoms of the Fe-S cluster. There is an additional residue found that is not reported for the chemical sequence: according to the electron density a valine residue should be inserted after residue 55.

Structures of deoxy and oxy hemerythrin at 2.0 A resolution.

The crystallographic structure analyses of deoxy and oxy hemerythrin have been carried out at 2.0 A resolution to extend the low resolution views of the physiological forms of this oxygen-binding protein. Restrained least-squares refinement has produced molecular models giving R-values of 16.8% for deoxy (41,064 reflections from 10 A to 2.0 A) and 17.3% for oxy hemerythrin (40,413 reflections from 10.0 A to 2.0 A). The protein structure in each derivative is very similar to that of myohemerythrin and the various met forms of hemerythrin. The binuclear complex in each derivative retains an oxygen atom bridging the two iron atoms, but the bond lengths found in deoxy hemerythrin support the idea that, in that form, the bridge is protonated, i.e. the bridging group is a hydroxyl. Dioxygen binds to the pentaco-ordinate iron atom in deoxy hemerythrin in the conversion to oxy hemerythrin. The interatomic distances are consistent with the proposed mechanism where the proton from the bridging group is transferred to the bound dioxygen, stabilizing it in the peroxo oxidation state by forming a hydrogen bond between the peroxy group and the bridging oxygen atom.