The three-dimensional structure of porin from Rhodobacter capsulatus at 3 A resolution

The crystal structure of porin from Rhodobacter capsulatus strain 37b4 has been solved at 3.0 A (1 A = 0.1 nm) resolution by multiple isomorphous replacement and solvent-flattening. The three pores of the trimer are well defined in the electron density map. Each pore consists of a 16-stranded beta-barrel which traverses the membrane as a tube. Near its center the tube is narrowed by chain segments protruding from the inner wall of the barrel that form an eye-let with an irregular cross-section of about 6 A by 10 A. The eye-let has an axial length of about 10 A; it defines the exclusion limit for diffusing particles.     

Crystallization and preliminary X-ray analysis of porin from Rhodobacter capsulatus

Porin monomers of the phototrophic bacterium Rhodobacter capsulatus were purified. Crystals were obtained from a solution of porin solubilized with the detergent octyltetraoxyethylene within 5 days using the vapor phase equilibration technique. The crystals were rhombohedral with an edge length of 0.4 mm. The space group was trigonal R3 with unit cell constants of a = b = 95 A, c = 147 A. Reflexions were observed to 3.2 A.     

Prediction of the general structure of OmpF and PhoE from the sequence and structure of porin from Rhodobacter capsulatus. Orientation of porin in the membrane.

By comparing the hydrophilicity profiles and sequences of porin from Rhodobacter capsulatus with those of OmpF and PhoE from Escherichia coli, a set of insertions and deletions for alignment of the sequences has been deduced. With this alignment a similar folding of OmpF and PhoE has been predicted as found by X-ray structure analysis of porin from Rhodobacter capsulates. Furthermore, the orientation of the porin trimer in the outer membrane was inferred from topological data on PhoE. According to this result a single channel of approx. 30 A diameter starts at the outer surface. Near the middle of the outer membrane bilayer this channel branches out into three separate channels, each running within a single porin monomer to the periplasmic surface.     

Structurally identical porin in lithoauto- and organoheterotrophically grown Rhodobacter capsulatus cells

Abstract The porin from lithoautotrophically grown Rhodobacter capsulatus 37b4 cells migrated identically to porin from organoheterotrophic cells on SDS-PAGE. Moreover, it behaved comparably in isoelectric focusing, and it was also EDTA-sensitive. Furthermore, the porins of the two growth conditions were essentially identical in amino acid composition and N-terminal amino acid sequence. A final proof for structural identity could be obtained by crystallization and structural analysis showing identity in all non-hydrogen atoms.     

X-ray crystallographic and mass spectrometric structure determination and functional characterization of succinylated porin from Rhodobacter capsulatus: implications for ion selectivity and single-channel conductance.

The role of charges near the pore mouth has been discussed in theoretical work about ion channels. To introduce new negative charges in a channel protein, amino groups of porin from Rhodobacter capsulatus 37b4 were succinylated with succinic anhydride, and the precise extent and sites of succinylations and structures of the succinylporins determined by mass spectrometry and X-ray crystallography. Molecular weight and peptide mapping analyses using matrix-assisted laser desorption-ionization mass spectrometry identified selective succinylation of three lysine-epsilon-amino groups (Lys-46, Lys-298, Lys-300) and the N-terminal alpha-amino group. The structure of a tetra-succinylated porin (TS-porin) was determined to 2.4 A and was generally found unchanged in comparison to native porin to form a trimeric complex. All succinylated amino groups found in a mono/di-succinylated porin (MS-porin) and a TS-porin are localized at the inner channel surface and are solvent-accessible: Lys-46 is located at the channel constriction site, whereas Lys-298, Lys-300, and the N-terminus are all near the periplasmic entrance of the channel. The Lys-46 residue at the central constriction loop was modeled as succinyl-lysine from the electron density data and shown to bend toward the periplasmic pore mouth. The electrical properties of the MS-and TS-porins were determined by reconstitution into black lipid membranes, and showed a negative charge effect on ion transport and an increased cation selectivity through the porin channel. The properties of a typical general diffusion porin changed to those of a channel that contains point charges near the pore mouth. The single-channel conductance was no longer a linear function of the bulk aqueous salt concentration. The substantially higher cation selectivity of the succinylated porins compared with the native protein is consistent with the increase of negatively charged groups introduced. These results show tertiary structure-selective modification of charged residues as an efficient approach in the structure-function evaluation of ion channels, and X-ray crystallography and mass spectrometry as complementary analytical tools for defining precisely the chemically modified structures.     

Nucleotide transport in Rhodobacter capsulatus.

Cytoplasmic membrane vesicles isolated from the gram-negative photosynthetic bacterium Rhodobacter capsulatus catalyzed the transport of nucleotides. No transport occurred in the intact bacteria unless they were pretreated with EDTA. The transport rate was measured by incorporation of radioactive phosphate into externally added ADP or by incorporation of nonradioactive phosphate into added labeled ADP. The catalytic activities which utilized the added ADP were photosynthetic ATP synthesis, Pi-ADP exchange, and adenylate kinase. These activities were shown to occur on the cytoplasmic side of the internal membrane. The products were found in the outer medium. The rate of nucleotide transport across the membranes was comparable to the rate of photophosphorylation. These results indicated that nucleotides can be transported across the cytoplasmic membrane but not across the outer membrane of the native R. capsulatus cell. Therefore, by analogy to the mitochondrial ATP-ADP translocator, the exchange might function as an energy transfer system to the periplasm of these bacteria.     

Export of porin to the outer membrane of the phototrophic bacterium Rhodobacter capsulatus 37B4

Export of porin to the outer membrane of the phototrophic purple bacterium Rhodobacter capsulatus was studied with the use of the uncoupler of the electron transport chain, carbonylcyanide-m-chlorophenylhydrazone (CCCP). The agent reversibly blocked the transport of porin across the cytoplasmic membrane. By means of radioactive labeling and immunoprecipitation, porin was found to occur in two forms: (i) the exported form that was extractable from the outer membrane without disrupting the cells, and (ii) a pre-form with a slightly higher apparent molecular mass which accumulated in the cells during the block of the export process. Proteolysis studies revealed that the preform was highly sensitive to added proteases, whereas the exported form was resistant.Abbreviations CCCP carbonylcyanide-m-chlorophenylhydrazone - DMSO dimethylsulfoxide - EDTA ethylenediamine tetraacetic acid - OMP outer membrane porin; pre-OMP, form of outer membrane porin before export - PAGE polyacrylamide gel electrophoresis - PMSF phenylmethylsulfonyl fluoride - SDS sodium dodecyl sulfate     

Molecular structure of cytochrome c2 isolated from Rhodobacter capsulatus determined at 2.5 A resolution.

The molecular structure of the cytochrome c2, isolated from the purple photosynthetic bacterium Rhodobacter capsulatus, has been solved to a nominal resolution of 2.5 A and refined to a crystallographic R-factor of 16.8% for all observed X-ray data. Crystals used for this investigation belong to the space group R32 with two molecules in the asymmetric unit and unit cell dimensions of a = b = 100.03 A, c = 162.10 A as expressed in the hexagonal setting. An interpretable electron density map calculated at 2.5 A resolution was obtained by the combination of multiple isomorphous replacement with four heavy atom derivatives, molecular averaging and solvent flattening. At this stage of the structural analysis the electron densities corresponding to the side-chains are well ordered except for several surface lysine, glutamate and aspartate residues. Like other c-type cytochromes, the secondary structure of the protein consists of five alpha-helices forming a basket around the heme prosthetic group with one heme edge exposed to the solvent. The overall alpha-carbon trace of the molecule is very similar to that observed for the bacterial cytochrome c2, isolated from Rhodospirillum rubrum, with the exception of a loop, delineated by amino acid residues 21 to 32, that forms a two stranded beta-sheet-like motif in the Rb. capsulatus protein. As observed in the eukaryotic cytochrome c proteins, but not in the cytochrome c2 from Rsp. rubrum, there are two evolutionarily conserved solvent molecules buried within the heme binding pocket.     

The Rhodobacter capsulatus genome

The genome of Rhodobacter capsulatus has been completely sequenced. It consists of a single chromosome containing 3.5 Mb and a circular plasmid of 134 kb. This effort, started in 1992, began with a fine-structure restriction map of an overlapping set of cosmids that covered the genome. Cosmid sequencing led to a gapped genome that was filled by primer walking on the chromosome and by using lambda clones. Methods had to be developed to handle strong stops in the high GC (68%) inserts. Annotation was done with the ERGO system at Integrated Genomics, as was the reconstruction of the cell's metabolism. It was possible to recognize 3709 orfs of which functional assignments could be made with high confidence to 2392 (65%). Unusual features include the presence of numerous cryptic phage genomes embedded in the chromosome. This revised version was published online in June 2006 with corrections to the Cover Date.     

Physiological functions of hydroperoxidases in Rhodobacter capsulatus.

Rhodobacter capsulatus J1 has two hydroperoxidases: a catalase-peroxidase and a peroxidase. A mutant strain, AH18, that had no catalase-peroxidase was isolated. The growth rate under aerobic and photosynthetic conditions, respiration, superoxide dismutase and peroxidase activities, and pigment content of the mutant were similar to those of the wild type. AH18 was more susceptible to killing and to inhibition of nitrogenase by H2O2 but not by molecular oxygen. The incidences of spontaneous mutations were similar in both strains. Viable counts in aerobic but not anaerobic cultures of AH18 started to decline as soon as the cultures reached the stationary phase, and the rate of cell death was much higher in AH18 than in the wild type. It is inferred that the peroxidase provides protection against H2O2 in log-phase cells and that the catalase-peroxidase provides protection under the oxidative conditions that prevail in aging cultures. This protective function might be related to the dual activity of the latter as a catalase and a peroxidase or to its capacity to oxidize NADH, NADPH, and cytochrome c.     

Cloning of the Rhodobacter capsulatus hemA gene.

Portions of the Rhodobacter capsulatus hemA gene have been cloned from a hemA::Tn5 insertion strain into the lambda bacteriophage derivative EMBL3. A cosmid containing the wild-type R. capsulatus hemA gene was isolated by complementation of the hemA::Tn5 mutant. The cosmid contains a 1.4-kilobase EcoRI fragment that spans the hemA::Tn5 insertion site. The entire hemA gene is contained in this fragment and the adjacent 0.6-kilobase EcoRI fragment.     

Cytochrome c2 mutants of Rhodobacter capsulatus.

Although structurally related to other members of the class I c-type cytochromes, the cytochromes c2 have little amino acid sequence homology to the eukaryotic cytochromes c. Moreover, the cytochromes c2 exhibit distinct properties such as redox potential and an isoelectric point. In an effort to understand the differences between the cytochromes c2 and the other class I c-type cytochromes, we have developed a genetic system to study Rhodobacter capsulatus cytochrome c2 by site-directed mutagenesis. We describe here overproduction of R. capsulatus wild-type cytochrome c2 in cytochrome c2-minus strains of R. capsulatus and Rhodobacter sphaeroides. We demonstrate that R. capsulatus wild-type cytochrome c2 can transcomplement for photosynthetic growth in R. sphaeroides. Further, we describe the generation, expression, and in vivo functionality properties of nine R. capsulatus site-directed mutants. We show that mutants K12D, K14E, K32E, K14E/K32E, P35A, W67Y, and Y75F are overproduced and functional in vivo. In contrast, mutants Y75C and Y75S are expressed at low levels and exhibit poor functionality in vivo. These findings establish an effective system for the production of R. capsulatus site-directed mutants and demonstrate that interspecies complementation can be used to detect defective cytochrome c2 mutants.     

Methylammonium uptake by Rhodobacter capsulatus

The ammonium uptake system of Rhodobacter capsulatus B100 was examined using the ammonium analog methylammonium. This analog was not transported when cells were grown aerobically on ammonium. When cultured on glutamate as a nitrogen source, or when nitrogen-starved, cells would take up methylammonium. Therefore, in cells grown under nitrogen-limiting conditions, a second system of ammonium uptake (or a modified form of the first) is present which is distinguished by its capacity for transporting the analog in addition to ammonium. The methylammonium uptake system exhibited saturation kinetics with a K _m of 22 M and a V _max of about 3 nmol per min · mg protein. Ammonium completely inhibited analog transport with a K _i in the range of 1 M. Once inside the cell methylammonium was rapidly converted to -N-methylglutamine; however, a small concentration gradient of methylammonium could still be observed. Kinetic parameters reflect the effects of assimilation.The methylammonium uptake system was temperature and pH dependent, and inhibition studies indicated that energy was required for the system to be operative. A glutamine auxotroph (G29) lacking the structural gene for glutanime synthetase did not accumulate the analog, even when nitrogen starved. The Nif^- mutant J61, which is unable to express nitrogenase structural genes, also did not transport methylammonium, regardless of the nitrogen source for growth. However, the mutant exhibited wild-type ammonium uptake and glutamine synthetase activity. These data suggest that transport of ammonium is required for growth on limited nitrogen and is under the control of the Ntr system in R. capsulatus.Abbreviations CCCP carbonyl cyanide-m-chlorophenyl hydrazone - CHES cyclohexylaminoethanesulfonic acid - DMSO dimethyl sulfoxide - GMAD -N-methylglutamine - GS glutamine synthetase - MES 2-(N-morpholino) ethanesulfonic acid - MSX methionine-Dl-sulfoximine - pCMB p-chloromercuribenzoate - Tricine N-tris(hydroxymethyl)methylglycine     

Mediated electrochemistry of dimethyl sulfoxide reductase from Rhodobacter capsulatus

Electrochemically driven catalysis of the bacterial enzyme dimethyl sulfoxide (DMSO) reductase (Rhodobacter capsulatus) has been studied using the macrocyclic complex (trans-6,13-dimethyl-1,4,8,11-tetraazacyclotetradecane-6,13-diamine)cobalt(III) as a mediator. In the presence of both DMSO and DMSO reductase, the normal transient Co^III/II voltammetric response of the complex is transformed into an amplified and sigmoidal (steady-state) waveform characteristic of a catalytic EC′ mechanism. At low concentrations of DMSO (approximately K _M) or high mediator concentrations (more than the concentration of DMSO reductase), the steady-state character of the voltammetric response disappears and is replaced by more complicated waveforms that are a convolution of transient and steady-state behavior as different steps within the catalytic cycle become rate limiting. Through digital simulation of cyclic voltammetry performed under conditions where the sweep rate, DMSO concentration, DMSO reductase concentration and mediator concentration were varied systematically, we were able to model all voltammograms with a single set of rate and equilibrium constants which provide new insights into the kinetics of the DMSO reductase catalytic mechanism that have hitherto been inaccessible from steady state or stopped flow kinetic studies.

Characterization of a tungsten-substituted nitrogenase isolated from Rhodobacter capsulatus

In the phototrophic non-sulfur bacterium Rhodobacter capsulatus, the biosynthesis of the conventional Mo-nitrogenase is strictly Mo-regulated. Significant amounts of both dinitrogenase and dinitrogenase reductase were only formed when the growth medium was supplemented with molybdate (1 microM). During cell growth under Mo-deficient conditions, tungstate, at high concentrations (1 mM), was capable of partially (approximately 25%) substituting for molybdate in the induction of nitrogenase synthesis. On the basis of such conditions, a tungsten-substituted nitrogenase was isolated from R. capsulatus with the aid of anfA (Fe-only nitrogenase defective) mutant cells and partially purified by Q-sepharose chromatography. Metal analyses revealed the protein to contain an average of 1 W-, 16 Fe-, and less than 0.01 Mo atoms per alpha(2)beta(2)-tetramer. The tungsten-substituted (WFe) protein was inactive in reducing N(2) and marginally active in acetylene reduction, but it was found to show considerable activity with respect to the generation of H(2) from protons. The EPR spectrum of the WFe protein, recorded at 4 K, exhibited three distinct signals: (i) an S = 3/2 signal, which dominates the low-field region of the spectrum (g = 4.19, 3.93) and is indicative of a tungsten-substituted cofactor (termed FeWco), (ii) a marginal S = 3/2 signal (g = 4.29, 3.67) that can be attributed to residual amounts of FeMoco present in the protein, and (iii) a broad S = 1/2 signal (g = 2.09, 1.95, 1.86) arising from at least two paramagnetic species. Redox titrational analysis of the WFe protein revealed the midpoint potential of the FeWco (E(m) < -200 mV) to be shifted to distinctly lower potentials as compared to that of the FeMoco (E(m) approximately -50 mV) present in the native enzyme. The P clusters of both the WFe and the MoFe protein appear indistinguishable with respect to their midpoint potentials. EPR spectra recorded with the WFe protein under turnover conditions exhibited a 20% decrease in the intensity of the FeWco signal, indicating that the cofactor can be enzymatically reduced only to a small extent. The data presented in the current study demonstrate the pivotal role of molybdenum in optimal N(2) fixation and provides direct evidence that the inability of a tungsten-substituted nitrogenase to reduce N(2) is due to the difficulty to effectively reduce the FeW cofactor beyond its semi-reduced state.     

X-ray absorption spectroscopy of dimethylsulfoxide reductase from Rhodobacter capsulatus

Mo K-edge X-ray absorption spectroscopy (XAS) has been used to probe the environment of Mo in dimethylsulfoxide (DMSO) reductase from Rhodobacter capsulatus in concert with protein crystallographic studies. The oxidised (Mo^VI) protein has been investigated in solution at 77?K; the Mo K-edge position (20006.4?eV) is consistent with the presence of Mo^VI and, in agreement with the protein crystallographic results, the extended X-ray absorption fine structure (EXAFS) is also consistent with a seven-coordinate site. The site is composed of one oxo-group (Mo=O 1.71?Å), four S atoms (considered to arise from the dithiolene groups of the two molybdopterins, two at 2.32?Å and two at 2.47?Å, and two O atoms, one at 1.92?Å (considered to be H-bonded to Trp 116) and one at 2.27?Å (considered to arise from Ser 147). The Mo K-edge XAS recorded for single crystals of oxidised (Mo^VI) DMSO reductase at 77?K showed a close correspondence to the data for the frozen solution but had an inferior signal:noise ratio. The dithionite-reduced form of the enzyme and a unique form of the enzyme produced by the addition of dimethylsulfide (DMS) to the oxidised (Mo^VI) enzyme have essentially identical energies for the Mo K-edge, at 20004.4?eV and 20004.5?eV, respectively; these values, together with the lack of a significant presence of Mo^V in the samples as monitored by EPR spectroscopy, are taken to indicate the presence of Mo^IV. For the dithionite-reduced sample, the Mo K-edge EXAFS indicates a coordination environment for Mo of two O atoms, one at 2.05?Å and one at 2.51?Å, and four S atoms at 2.36?Å. The coordination environment of the Mo in the DMS-reduced form of the enzyme involves three O atoms, one at 1.69?Å, one at 1.91?Å and one at 2.11?Å, plus four S atoms, two at 2.28?Å and two at 2.37?Å. The EXAFS and the protein crystallographic results for the DMS-reduced form of the enzyme are consistent with the formation of the substrate, DMSO, bound to Mo^IV with an Mo-O bond of length 1.92?Å.