[NiFe] hydrogenase from Desulfovibrio desulfuricans ATCC 27774: gene sequencing, three-dimensional structure determination and refinement at 1.8 Å and modelling studies of its interaction with the tetrahaem cytochrome c 3

The primary and three-dimensional structures of a [NiFe] hydrogenase isolated from D. desulfitricans ATCC 27774 were determined, by nucleotide analysis and single-crystal X-ray crystallography. The three-dimensional structural model was refined to R=0.167 and Rfree=0.223 using data to 1.8 A resolution. Two unique structural features are observed: the [4Fe-4S] cluster nearest the [NiFe] centre has been modified [4Fe-3S-3O] by loss of one sulfur atom and inclusion of three oxygen atoms; a three-fold disorder was observed for Cys536 which binds to the nickel atom in the [NiFe] centre. Also, the bridging sulfur atom that caps the active site was found to have partial occupancy, thus corresponding to a partly activated enzyme. These structural features may have biological relevance. In particular, the two less-populated rotamers of Cys536 may be involved in the activation process of the enzyme, as well as in the catalytic cycle. Molecular modelling studies were carried out on the interaction between this [NiFe] hydrogenase and its physiological partner, the tetrahaem cytochrome c3 from the same organism. The lowest energy docking solutions were found to correspond to an interaction between the haem IV region in tetrahaem cytochrome c3 with the distal [4Fe-4S] cluster in [NiFe] hydrogenase. This interaction should correspond to efficient electron transfer and be physiologically relevant, given the proximity of the two redox centres and the fact that electron transfer decay coupling calculations show high coupling values and a short electron transfer pathway. On the other hand, other docking solutions have been found that, despite showing low electron transfer efficiency, may give clues on possible proton transfer mechanisms between the two molecules.     

A preliminary analysis of the three-dimensional structure of dimeric di-haem split-Soret cytochrome c from Desulfovibrio desulfuricans ATCC 27774 at 2.5-Å resolution using the MAD phasing method: a novel cytochrome fold with a stacked-haem arrangement

Haem-containing proteins are directly involved in electron transfer as well as in enzymatic functions. The \"split-Soret\" cytochrome (SSC) was isolated from the sulfate- and nitrate-reducing bacterium Desulfovibrio desulfuricans ATCC 27774 and has no significant nitrate or nitrite reductase activity. The protein received its name due its unusual spectral properties. It is a dimer containing two identical subunits of 26.3 kDa, each with two haem-c groups. A preliminary model for the three-dimensional structure of this cytochrome was derived using the Multiple Wavelength Anomalous Dispersion (MAD) phasing method. This model shows that SSC is indeed a dimer containing four haems at one end of the molecule. In each monomer the two haems have their edges overlapped within van der Waals contacts with an iron-to-iron distance of 9?Å. The polypeptide chain of each monomer supplies the sixth axial ligand to the haems of the other monomer. This work shows that SSC constitutes a new class of cytochrome. The stacking of the two haems in the monomer within van der Waals distances of each other, and also the short (van der Waals) distances between the two monomers in the dimeric molecule are unprecedented in hemoproteins. This particular haem arrangement is an excellent model for the spectral study (undertaken several years ago) of haem-haem interaction using the aggregated haem undecapeptide derived from mammalian cytochrome c.     

Hybrid cluster proteins (HCPs) from Desulfovibrio desulfuricans ATCC 27774 and Desulfovibrio vulgaris (Hildenborough): X-ray structures at 1.25 Å resolution using synchrotron radiation

The structures of the hybrid cluster proteins (HCPs) from the sulfate-reducing bacteria Desulfovibrio desulfuricans (ATCC 27774) and Desulfovibrio vulgaris (Hildenborough) have been elucidated at a resolution of 1.25 A using X-ray synchrotron radiation techniques. In the case of the D. desulfuricans protein, protein isolation, purification, crystallization and X-ray data collection were carried out under strict anaerobic conditions, whereas for the D. vulgaris protein the conditions were aerobic. However, both structures are essentially the same, comprising three domains and two iron-sulfur centres. One of these centres situated near the exterior of the molecules in domain 1 is a cubane [4Fe-4S] cluster, whereas the other, located at the interface of the three domains, contains the unusual four-iron cluster initially found in the D. vulgaris protein. Details of the structures and the associated EPR spectroscopy of the D. desulfuricans protein are reported herein. These structures show that the nature of the hybrid cluster, containing both oxygen and sulfur bridges, is independent of the presence of oxygen in the isolation and crystallization procedure and also does not vary significantly with changes in the oxidation state. The structures and amino acid sequences of the HCP are compared with the recently elucidated structure of the catalytic subunit of a carbon monoxide dehydrogenase from Carboxydothermus hydrogenoformans and related dehydrogenases. Electronic supplementary material to this paper can be obtained by using the Springer Link server located at http://dx.doi.org/10.1007/s00775-001-0326-y.     

Gaseous hydrocarbons associated with black layer induced by the interaction of cyanobacteria and Desulfovibrio desulfuricans†

Black layer is a condition of high-sand-content golf greens that results in a subsurface blackened layer in the sand produced by sulfate-reducing bacteria. Black layer can be the product of an interaction of cyanobacteria and sulfate-reducing bacteria and may or may not be toxic to the grass growing on the sand. The organic byproducts of the cyanobacteria coat and plug the sand thereby creating an anoxic environment for development of the sulfate- reducing bacteria. The present study was initiated to determine the range of gaseous hydrocarbons evolved from black layered sand produced by the interaction of two genera of cyanobacteria, Nostoc and Oscillatoria, and Desulfovibrio desulfuricans. The gaseous hydrocarbons measured included methane, ethane, ethylene, and propylene. In nonblackened sand, Nostoc evolved the highest levels of these gases, Oscillatoria evolved relatively low levels except for propylene, and D. desulfuricans evolved the smallest quantities of the gases. When the cyanobacteria and D. desulfuricans were combined to develop black layered sand some changes occurred in the evolution of the gases. Evolution of the gases from Nostoc + D. desulfuricans decreased or remained the same relative to Nostoc alone, and increased relative to D. desulfuricans alone. Except for propylene evolution, gases from Oscillatoria + D. desulfuricans increased relative to Oscillatoria or D. desulfuricans alone. Propylene evolution from Oscillatoria + D. desulfuricans remained unchanged relative to Oscillatoria alone, but increased relative to D. desulfuricans alone. The gases measured are discussed relative to the organisms observed and the conditions of the study.     

M?ssbauer characterization of the tetraheme cytochrome c3 from Desulfovibrio baculatus (DSM 1743). Spectral deconvolution of the heme components.

M?ssbauer spectroscopy was used to study the tetraheme cytochrome c3 from Desulfovibrio baculatus (DSM 1743). Samples with different degrees of reduction were prepared using a redoxtitration technique. In the reduced cytochrome c3, all four hemes are reduced and exhibit diamagnetic M?ssbauer spectra typical for low-spin ferrous hemes (S = 0). In the oxidized protein, the hemes are low-spin ferric (S = 1/2) and exhibit overlapping magnetic M?ssbauer spectra. A method of differential spectroscopy was applied to deconvolute the four overlapping heme spectra and a crystal-field model was used for data analysis. Characteristic M?ssbauer spectral components for each heme group are obtained. Hyperfine and crystal-field parameters for all four hemes are determined from these deconvoluted spectra.     

The anaerobic interaction of ferrocytochrome c with the “ferric” and “oxygenated” forms of purified cytochrome c oxidase. XVI. Cytochrome oxidase and its derivatives

1. The anaerobic interaction of ferrocytochrome c with “ferric” cytochrome c oxidase (EC 1.9.3.1) in the absence of ascorbate or ferro/ferricyanide cannot be interpreted as the reaction of a uniform ferric oxidase. The increase of the α-band at 605 mμ precedes that of the γ-band at 444 mμ and the γ/α ratio increases over a period of several minutes.2. The biphasic form of the reaction is far less noticeable in the reaction of “oxygenated” oxidase with ferrocytochrome c. A similar difference has been previously found in the reduction by dithionite.3. A spectrophotometric method is developed which allows calculations of the concentrations of ferrous and ferric cytochromes a, a₃ and c at various times after mixing oxidase and ferrocytochrome c. This is based on the extinction coefficients of cytochrome c and those of ferro- and ferricytochromes a and a₃ (refs. 1 and 2) at 605, 550 and 444 mμ.4. With ferric oxidase, rapid reduction of ferricytochrome a is followed by a far slower reduction of ferricytochrome a₃ by ferrocytochromes c and a. No rapid equilibrium is established. Two equivalents of ferrocytochrome c are required for the reduction of one ferricytochrome a and two for that of ferricytochrome a₃ (4 per 2 moles haem a). The additional electron acceptor is probably the intrinsic copper of the oxidase.5. With oxygenated oxidase, initial rapid conversion of the ferryl form of cytochrome a₃ is suggested by the fact that initially more than two equivalents of ferrocytochrome c are oxidized per mole of haem a. The ferricytochrome thus formed (a_3x³⁺, ref. 2) is spectroscopically similar to or identical with ferricytochrome a₃³⁺, but differs from it by its far more rapid reduction by ferrocytochromes c and a.6. Ferricytochrome a_3x³⁺ probably plays a larger role in the enzymic reaction of the oxidase than does the ferricytochrome known as a₃³⁺ present in the ferric oxidase preparation.     

Three-dimensional structure of the complex between pancreatic secretory trypsin inhibitor (Kazal type) and trypsinogen at 1.8 Å resolution: Structure solution,crystallographic refinement and preliminary structural interpretation

The three-dimensional structure of the proteic complex formed by bovine trypsinogen and the porcine pancreatic secretory trypsin inhibitor (Kazal type) has been solved by means of Patterson search techniques, using a predicted model of the trypsin-ovomucoid complex (Papamokos et al., 1982). The structure of the complex, including 162 solvent molecules, has been refined at 1.8 Å resolution (26,341 unique reflections) to a conventional crystallographic R factor of 0.195. The inhibitor molecule binds to trypsinogen via hydrogen bonds and/or apolar interactions at sites P9, P7, P6, P5, P3, P1, P1′, P2′ and P3′ of the contact area. The structure of the inhibitor itself resembles closely that of the third domain of Japanese quail ovomucoid inhibitor, recently reported by Weber et al. (1981). The trypsinogen part of the complex resembles trypsin, as is the case in the trypsinogen-basic pancreatic trypsin inhibitor complex, but two segments of the activation domain adopt a different conformation. Most notably in the N-terminal region the Ile16-Gly19 loop, which is disordered in free trypsinogen and in the trypsinogen-basic pancreatic trypsin inhibitor complex (Huber & Bode, 1978), assumes a regular structure and the polypeptide chain can be traced as far as residue Asp14. This new and fixed structure allows the formation of a buried salt link between the side-chains of Lys156 and Asp194. Conformations differing from those of trypsin are also found for residues 20 to 28 and residues 141 to 155. Some structural perturbation is observed in other parts of the molecule, including the calcium loop.     

Structure of cytochrome c551 from Pseudomonas aeruginosa refined at 1.6 Å resolution and comparison of the two redox forms

The molecular structures of ferri- and ferrocytochrome c₅₅₁ from Pseudomonas aeruginosa have been refined at a resolution of 1.6 Å, to an R factor of 19.5% for the oxidized molecule and 18.7% for the reduced. Reduction of oxidized crystals with ascorbate produced little change in cell dimensions, a 10% mean change in F_obs, and no damage to the crystals. The heme iron is not significantly displaced from the porphyrin plane. Bond lengths from axial ligands to the heme iron are as expected in a low-spin iron compound. A total of 67 solvent molecules were incorporated in the oxidized structure, and 73 in the reduced, of which four are found inside the protein molecule. The oxidized and reduced forms have virtually identical tertiary structures with 2 ° root-mean-square differences in main-chain torsion angles φ and ψ, but with larger differences along the two edges of the heme crevice. The difference map and pyrrole ring tilt suggest that a partially buried water molecule (no. 23) in the heme crevice moves upon change of oxidation state.Pseudomonas cytochrome c₅₅₁ differs from tuna cytochrome c in having: (1) a water molecule (no. 23) at the upper left of the heme crevice; that is, between Pro62 and the heme pyrrol 3 ring on the sixth ligand Met61 side, where tuna cytochrome c has an evolutionary invariant Phe82 ring; (2) a string of hydrophobic side-chains along the left side of the heme crevice, and fewer positively charged lysines in the vicinity; and (3) a more exposed and presumably more easily ionizable heme propionate group at the bottom of the molecule. A network of hydrogen bonds in the heme crevice is reminiscent of that inside the heme crevice of tuna cytochrome c. As in tuna, a slight motion of the water molecule toward the heme is observed in the oxidized state, helping to give the heme a more polar microenvironment. The continuity of solvent environment between the heme crevice and the outer medium could explain the greater dependence of redox potential on pH in cytochrome c₅₅₁ than in cytochrome c.     

Tyrosine triad at the interface between the Rieske iron–sulfur protein,cytochrome c1 and cytochrome c2 in the bc1 complex of Rhodobacter capsulatus

A triad of tyrosine residues (Y152–154) in the cytochrome c₁ subunit (C1) of the Rhodobacter capsulatus cytochrome bc₁ complex (BC1) is ideally positioned to interact with cytochrome c₂ (C2). Mutational analysis of these three tyrosines showed that, of the three, Y154 is the most important, since its mutation to alanine resulted in significantly reduced levels, destabilization, and inactivation of BC1. A second-site revertant of this mutant that regained photosynthetic capacity was found to have acquired two further mutations—A181T and A200V. The Y152Q mutation did not change the spectral or electrochemical properties of C1, and showed wild-type enzymatic C2 reduction rates, indicating that this mutation did not introduce major structural changes in C1 nor affect overall activity. Mutations Y153Q and Y153A, on the other hand, clearly affect the redox properties of C1 (e.g. by lowering the midpoint potential as much as 117 mV in Y153Q) and the activity by 90% and 50%, respectively. A more conservative Y153F mutant on the other hand, behaves similarly to wild-type. This underscores the importance of an aromatic residue at position Y153, presumably to maintain close packing with P184, which modeling indicates is likely to stabilize the sixth heme ligand conformation.     

1-(3-Deoxy-3-fluoro-β-d-glucopyranosyl) pyrimidine derivatives as inhibitors of glycogen phosphorylase b: Kinetic,crystallographic and modelling studies

Design of inhibitors of glycogen phosphorylase (GP) with pharmaceutical applications in improving glycaemic control in type 2 diabetes is a promising therapeutic strategy. The catalytic site of muscle glycogen phosphorylase b (GPb) has been probed with five deoxy-fluro-glucose derivatives. These inhibitors had fluorine instead of hydroxyl at the 3′ position of the glucose moiety and a variety of pyrimidine derivatives at the 1′ position. The best of this carbohydrate-based family of five inhibitors displays a K_i value of 46 μM. To elucidate the mechanism of inhibition for these compounds, the crystal structures of GPb in complex with each ligand were determined and refined to high resolution. The structures demonstrated that the inhibitors bind preferentially at the catalytic site and promote the less active T state conformation of the enzyme by making several favorable contacts with residues of the 280s loop. Fluorine is engaged in hydrogen bond interactions but does not improve glucose potency. The pyrimidine groups are located between residues 284–286 of the 280s loop, Ala383 of the 380s loop, and His341 of the β-pocket. These interactions appear important in stabilizing the inactive quaternary T state of the enzyme. As a follow up to recent computations performed on β-d-glucose pyrimidine derivatives, tautomeric forms of ligands 1–5 were considered as potential binding states. Using Glide-XP docking and QM/MM calculations, the ligands 2 and 5 are predicted to bind in different tautomeric states in their respective GPb complexes. Also, using α-d-glucose as a benchmark model, a series of substitutions for glucose –OH at the 3′ (equatorial) position were investigated for their potential to improve the binding affinity of glucose-based GPb catalytic site inhibitors. Glide-XP and quantum mechanics polarized ligand (QPLD-SP/XP) docking calculations revealed favorable binding at this position to be dominated by hydrogen bond contributions; none of the substitutions (including fluorine) out-performed the native –OH substituent which can act both as hydrogen bond donor and acceptor. The structural analyses of these compounds can be exploited towards the development of better inhibitors.     

The surface organization of diacylglycerol pyrophosphate and its interaction with phosphatidic acid at the air–water interface

Diacylglycerol pyrophosphate (DGPP), a phosphorylated form of phosphatidic acid (PA), gained attention recently due to its role as signaling lipid. However, little is known about its surface organization and potential impact on membrane-mediated function. In this work we investigated the interfacial behavior of Langmuir monolayers formed with pure DGPP and of its mixtures with PA. We found that changes of the subphase pH affect the surface behavior of DGPP. At pH 8, DGPP forms liquid expanded monolayers with a compressibility modulus of about 60 mN m^?1 at collapse. On acidic solutions, the compressibility modulus increases to 90 mN m^?1 and the average molecular area is smaller. At pH 8, DGPP and its precursor PA form thermodynamically favored topographically homogeneous non-ideal mixtures. The interaction among these lipids leads to a non-ideal diminution of the mean molecular area and consequently, to an increase of the compressibility modulus, with variations of the surface electrostatics. The favorable interaction of PA and DGPP, leading to changes of the film packing suggest that DGPP may act as a structural signal transducer in membrane-mediated cellular processes.     

Computer simulation of the interaction of α3hclix of λ Cro protein with DNA and origin of sequence specific recognition

The conformation of the α3 helix of Cro protein (residues 27–36) of bacteriophageλ is optimised by the damped least square minimization technique, with the steric constraint that Cα atom positions should match the crystallographic data available to date. On the basis of minimization of total interaction and conformation energy, models for complexes of this peptide sequence with heptanucleotide duplexes from native and altered O_R3 operator are obtained in the major groove of B DNA. Analysis of the energetics for 3 sequences of the DNA show that binding strength is derived mainly from the interaction of side chains of the peptide with DNA. Sequence specificity (maximum difference in binding energy for different DNA sequences) is due to hydrogen bonding interaction. A small amount of sequence specificity is derived from non-bonded interaction also. Stereochemical aspects of peptide DNA interaction and their role in DNA recognition are discussed in this paper.     

Properties of the monomeric form of human 14-3-3ζ protein and its interaction with tau and HspB6

Dimers formed by seven isoforms of the human 14-3-3 protein participate in multiple cellular processes. The dimeric form has been extensively characterized; however, little is known about the structure and properties of the monomeric form of 14-3-3. The monomeric form is involved in the assembly of homo- and heterodimers, which could partially dissociate back into monomers in response to phosphorylation at Ser58. To obtain monomeric forms of human 14-3-3ζ, we produced four protein constructs with different combinations of mutated (M) or wild-type (W) segments E(5), (12)LAE(14), and (82)YREKIE(87). Under a wide range of expression conditions in Escherichia coli, the MMM and WMM mutants were insoluble, whereas WMW and MMW mutants were soluble, highly expressed, and purified to homogeneity. WMW and MMW mutants remained monomeric over a wide range of concentrations while retaining the α-helical structure characteristic of wild-type 14-3-3. However, WMW and MMW mutants were highly susceptible to proteolysis and had much lower thermal stabilities than the wild-type protein. Using WMW and MMW mutants, we show that the monomeric form interacts with the tau protein and with the HspB6 protein, in both cases forming complexes with a 1:1 stoichiometry, in contrast to the 2:1 and/or 2:2 complexes formed by wild-type 14-3-3. Significantly, this interaction requires phosphorylation of tau protein and HspB6. Because of minimal changes in structure, MMW and especially WMW mutant proteins are promising candidates for analyzing the effect of monomerization on the physiologically important properties of 14-3-3ζ.     

M?ssbauer study of beef heart cytochrome oxidase. Comparative study of the bovine enzyme and cytochrome c1aa3 from Thermus thermophilus

We have studied beef heart cytochrome c oxidase at 4.2 K with M?ssbauer spectroscopy using the 57Fe present in natural abundance. The spectra observed are very similar to those of the a- and a3-sites of cytochrome c1aa3 from Thermus thermophilus. Thus, many conclusions derived from studies of the bacterial oxidase (available with enriched 57Fe) also apply to the mammalian enzyme. In the resting (as isolated) state, cytochrome a3 of the mammalian enzyme exhibits a doublet with quadrupole splitting, delta EQ = 1.0 mm/s and isomer shift, delta = 0.48 mm/s. These parameters suggest a high spin ferric heme and rule out an Fe(IV) assignment. The absence of magnetic features in the 4.2 K spectrum is consistent with earlier proposals that cytochrome a3 is spin-coupled to a cupric ion. The absorption lines are rather broad, suggesting that the a3-site is heterogeneous in the resting enzyme. Reduced cytochrome a3 has delta EQ = 1.85 mm/s and delta = 0.93 mm/s, demonstrating that the heme iron is high spin ferrous. The observed value for delta EQ is smaller than those of hemoglobin (2.4 mm/s), myoglobin (2.2 mm/s), and cytochrome a3 from T. thermophilus (2.06 mm/s). The M?ssbauer spectra of oxidized cytochrome a3-CN show that the heme iron is low spin ferric and that the ground state has integer spin S greater than or equal to 1, which plausibly results from ferromagnetic coupling of the S = 1/2 heme to an S = 1/2 cupric ion. Reduced cytochrome a is low spin ferrous, with parameters similar to those of cytochrome b5 and cytochrome c.     

Purification of (1→3)-β-glucan endohydrolase isoenzyme II from germinated barley and determination of its primary structure from a cDNA clone

A (13)--D-glucan 3-glucanonydrolase (EC 3.2.1.39) of apparent M _r 32 000, designated GII, has been purified from germinated barley grain and characterized. The isoenzyme is resolved from a previously purified isoenzyme (GI) on the basis of differences in their isoelectric points; (13)--glucanases GI and GII have pI values of 8.6 and 10.0, respectively. Comparison of the sequences of their 40 NH₂-terminal amino acids reveals 68% positional identity. A 1265 nucleotide pair cDNA encoding (13)--glucanase isoenzyme GII has been isolated from a library prepared with mRNA of 2-day germinated barley scutella. Nucleotide sequence analysis of the cDNA has enabled the complete primary structure of the 306 amino acid (13)--glucanase to be deduced, together with that of a putative NH₂-terminal signal peptide of 28 amino acid residues. The (13)--glucanase cDNA is characterized by a high (G+C) content, which reflects a strong bias for the use of G or C in the wobble base position of codons. The amino acid sequence of the (13)--glucanase shows highly conserved internal domains and 52% overall positional identity with barley (13, 14)--glucanase isoenzyme EII, an enzyme of related but quite distinct substrate specificity. Thus, the (13)--glucanases, which may provide a degree of protection against microbial invasion of germinated barley grain through their ability to degrade fungal cell wall polysaccharides, appear to share a common evolutionary origin with the (13, 14)--glucanases, which function to depolymerize endosperm cell walls in the germinated grain.     

Evidence for structural heterogeneities and a study of exchange coupling. M?ssbauer studies of cytochrome c1aa3 from Thermus thermophilus

We have studied samples of oxidized (as isolated) cytochrome c1aa3 from Thermus thermophilus in the pH range 5.7 to 9.3 with M?ssbauer spectroscopy. In this pH range, the spectra of cytochromes c1 and a are independent of pH, whereas the spectra of cytochrome a3 are not. Most importantly, spectra taken in applied fields up to 6.0 T reveal the presence of multiple ferric forms of cytochrome a3. At any given pH value, at least two high-spin ferric cytochrome a3 species can be distinguished; in addition, most samples contain a low-spin ferric cytochrome a3 species (less than 20% of cytochrome a3). The M?ssbauer spectra show clearly that all forms of cytochrome a3 are spin coupled (to CuB). We have analyzed the high field (H greater than or equal to 1.5 tesla) spectra of a sample at pH 6.5 in the framework of a model that considers isotropic exchange-coupling, JS1.S2, between a high-spin ferric (S1 = 5/2) cytochrome a3 and cupric CuB (S2 = 1/2). In strong applied fields, the spectra can be fitted with any value for J greater than or equal to 0.5 cm-1. In the strong coupling case (J/D1 approximately greater than 3), a zero field splitting parameter D1 approximately 2.5 cm-1 is required for cytochrome a3; this value is distinctly smaller than those observed for high-spin ferric hemes (4-20 cm-1). A model assuming weak coupling magnitude of J approximately 1 cm-1, yields D1 approximately 8 cm-1 and a parameter set for cytochrome a3 quite similar to that reported for metmyoglobin. A J-value of approximately 1 cm-1 does not demand the presence of a ligand bridging between cytochrome a3 and CuB.