Electrochemical and spectroscopic investigations of the cytochrome bc1 complex from Rhodobacter capsulatus.

The cytochrome bc(1) complex from Rhodobacter capsulatus was investigated by protein electrochemistry and visible/IR spectroscopy. Infrared difference spectra, which represent redox-induced conformational changes of cofactors and their protein environments, show signals of the hemes, the quinone Q(i), and small conformational changes of the protein backbone. Furthermore, band features were tentatively assigned to protonated aspartic or glutamic acids involved in the redox transition of each of the b hemes, a proline in that of the [2Fe-2S] protein, and an arginine in that of cytochrome b(H). The midpoint potential of the [2Fe-2S] protein was determined for the first time at physiological temperature to be +290 mV at pH 8.7. The reduced minus oxidized difference extinction coefficients of the alpha-bands of the cytochromes were calculated as 11.5, 19, and 6.7 mM(-1) cm(-1) for cytochromes c(1), b(H), and b(L), respectively. A novel method has been developed to quantify protonation reactions of the complex during the redox reactions of its cofactors by evaluation of the buffer signals in the midinfrared region. Values will be discussed in relation to the pH dependence of the midpoint potentials.

Crystal structure of the beta-glycosidase from the hyperthermophile Thermosphaera aggregans: insights into its activity and thermostability.

The glycosyl hydrolases are an important group of enzymes that are responsible for cleaving a range of biologically significant carbohydrate compounds. Structural information on these enzymes has provided useful information on their molecular basis for the functional variations, while the characterization of the structural features that account for the high thermostability of proteins is of great scientific and biotechnological interest. To these ends we have determined the crystal structure of the beta-glycosidase from a hyperthermophilic archeon Thermosphaera aggregans. The structure is a (beta/alpha)8 barrel (TIM-barrel), as seen in other glycosyl hydrolase family 1 members, and forms a tetramer. Inspection of the active site and the surrounding area reveals two catalytic glutamate residues consistent with the retaining mechanism and the surrounding polar and aromatic residues consistent with a monosaccharide binding site. Comparison of this structure with its mesophilic counterparts implicates a variety of structural features that could contribute to the thermostability. These include an increased number of surface ion pairs, an increased number of internal water molecules and a decreased surface area upon forming an oligomeric quaternary structure.

Chemical and enzymatic synthesis of glycoconjugates. 5: One-pot regioselective synthesis of bioactive galactobiosides using a CLONEZYME thermophilic glycosidase library.

Enzymatic synthesis of galactobiosides using a versatile CLONEZYME thermostable glycosidase library was studied. One-pot transglycosylation reactions were demonstrated to synthesize beta(1-->4), beta(1-->6), and alpha(1-->6) disaccharide sequences with high regioselectivity and moderate to high yields.

Structural and electronic properties of the heme cofactors in a multi-heme synthetic cytochrome.

Resonance Raman, absorption, and electron paramagnetic resonance spectra are reported for a water soluble, synthetic cytochrome. The protein is a variant of the cytochrome beta maquette described by Robertson et al. [Robertson, D. E., et al. (1995) Nature 368, 425-432] and is composed of 62 amino acid residues arranged in a di-alpha-helical unit which dimerizes in solution to form a four-helix bundle. Each di-alpha-helical unit contains histidine residues at the 10,10' positions which serve as ligands to the hemes. When protoheme IX is incorporated, both hemes in the dimer are bis-ligated and low spin. The two hemes are inequivalent with respect to both binding affinity and redox properties. To investigate the properties of the heme cofactors, spectroscopic studies were conducted on peptides reconstituted with protoheme IX (PHa) and several related variants. These hemes include 2-vinyldeuteroheme (2-VDH), 4-vinyldeuteroheme (4-VDH), protoheme III (PHs), and 1-methyl-2-oxomesoheme XIII (2-OMH). Collectively, the spectroscopic studies reveal the following: (1) 2-VDH, 4-VDH, and 2-OMH bind to the protein and form bis-ligated low-spin complexes similar to PHa. The structures of the two hemes in the dimers are identical as are the immediate protein environments around the bound cofactors. These results indicate that the redox inequivalence of the two hemes is due to heme-heme electronic interactions rather than structural and/or environmental differences between the two cofactors. (2) The two hemes in the dimer are arranged in a edge-to-edge arrangement wherein the oxo group (2-OMH) or the vinyl group(s) are in the hydrophobic interface between the two units which comprise the dimer. The propionic acid tails point outward toward the hydrophilic region and extend into the solvent. (3) The PHs protein differs from the other synthetic proteins in that it contains one pentacoordinate, high-spin and one hexacoordinate, low-spin heme rather than two hexacoordinate low-spin cofactors. The open coordination site on the high-spin heme is inaccessible to exogenous imidazole but readily bind cyanide, suggesting that the alpha-helix containing the unbound histidine is nearby and partially shields the coordination site. The high-spin heme converts to low-spin at low-temperature, presumably via binding of the histidine residue on this nearby alpha-helix. It is suggested that the different behavior observed for the PHs protein is due to the fact that this heme is symmetric with respect to rotation about the alpha,gamma-axis of the macrocycle which bisects the meso-carbons between the vinyl groups and propionic acid residues. This symmetry precludes rotational isomerism about the alpha,gamma-axis to establish an unhindered fit. In contrast, all the other hemes examined contain at least one substituent smaller than a vinyl group which together with the fact that two different alpha,gamma-rotational isomers are possible for each heme in the dimer could allow these hemes to avoid the like-substituent--like-substituent heme--heme interactions of PHs. The propensity to avoid such interactions could explain the inequivalent binding properties of the two hemes in the dimer. For the PHs protein wherein these these interactions cannot be mitigated by rotation of the heme, other rearrangements of the protein must occur. These rearrangements could force the second-bound heme to assume a high-spin configuration.

Ubiquinone pair in the Qo site central to the primary energy conversion reactions of cytochrome bc1 complex.

The mechanistic heart of the ubihydroquinone-cytochrome c oxidoreductase (cyt bc1 complex) is the catalytic oxidation of ubihydroquinone (QH2) at the Qo site. QH2 oxidation is initiated by ferri-cyt c, mediated by the cyt c1 and [2Fe-2S] cluster of the cytochrome bc1 complex. QH2 oxidation in turn drives transmembrane electronic charge separation through two b-type hemes to another ubiquinone (Q) at the Qi site. In earlier studies, residues F144 and G158 of the b-heme containing polypeptide of the Rhodobacter capsulatus cyt bc1 complex were shown to be influential in Qo site function. In the present study, F144 and G158 have each been singly substituted by neutral residues and the dissociation constants measured for both Q and QH2 at each of the strong and weak binding Qo site domains (Qos and Qow). Various substitutions at F144 or G158 were found to weaken the affinities for Q and QH2 at both the Qos and Qow domains variably from zero to beyond 10(3)-fold. This produced a family of Qo sites with Qos and Qow domain occupancies ranging from nearly full to nearly empty at the prevailing approximately 3 x 10(-2) M concentration of the membrane ubiquinone pool (Qpool). In each mutant, the affinity of the Qos domain remained typically 10-20-fold higher than that of the Qow domain, as is found for wild type, thereby indicating that the single mutations caused comparable extents of the weakening at each domain. Moreover, the substitutions were found to cause similar decreases of the affinities of both Q and QH2 in each domain, thereby maintaining the Q/QH2 redox midpoint potentials (Em7) of the Qo site at values similar to that of the wild type. Measurement of the yield and rate of QH2 oxidation generated by single turnover flashes in the family of mutants suggests that the Qos and Qow domains serve different roles for the catalytic process. The yield of the QH2 oxidation correlates linearly with Qos domain occupancy (QH2 or Q), suggesting that the Qos domain exchanges Q or QH2 with the Qpool at a rate which is much slower than the time scale of turnover.(ABSTRACT TRUNCATED AT 400 WORDS)

Association of flavin adenine dinucleotide with the Arabidopsis blue light receptor CRY1.

The arabidopsis thaliana HY4 gene encodes CRY1, a 75-kilodalton flavoprotein mediating blue light-dependent regulation of seedling development. CRY1 is demonstrated here to noncovalently bind stoichiometric amounts of flavin adenine dinucleotide (FAD). The redox properties of FAD bound by CRY1 include an unexpected stability of the neutral radical flavosemiquinone (FADH.). The absorption properties of this flavosemiquinone provide a likely explanation for the additional sensitivity exhibited by CRY1-mediated responses in the green region of the visible spectrum. Despite the sequence homology to microbial DNA photolyases, CRY1 was found to have no detectable photolyase activity.

Characterization of flavocytochrome C552 from the thermophilic photosynthetic bacterium Chromatium tepidum.

A M(r) 68 kDa flavocytochrome c552 has been isolated from the thermophilic photosynthetic purple sulfur bacterium Chromatium tepidum and shown to consist of a M(r) 25 kDa subunit that contains two covalently bound heme c and a M(r) 43 kDa subunit that probably contains a single FAD. The prosthetic group content, absorbance spectra, and subunit composition of the C. tepidum flavocytochrome are quite similar to those previously reported for the flavocytochrome c552 isolated from a mesophilic Chromatium species, Chromatium vinosum. The oxidation-reduction properties of the hemes present in the C. tepidum flavocytochrome have been characterized by titrations, the effect of temperature on the catalytic activity of the protein has been investigated, and the heme environment has been characterized using resonance Raman spectroscopy.

Design and synthesis of multi-haem proteins.

A water-soluble, 62-residue, di-alpha-helical peptide has been synthesized which accommodates two bis-histidyl haem groups. The peptide assembles into a four-helix dimer with 2-fold symmetry and four parallel haems that closely resemble native haems in their spectral and electrochemical properties, including haem-haem redox interaction. This protein is an essential intermediate in the synthesis of molecular 'maquettes', a novel class of simplified versions of the metalloproteins involved in redox catalysis and in energy conversion in respiratory and photosynthetic electron transfer.

Electrochemical and spectral analysis of the long-range interactions between the Qo and Qi sites and the heme prosthetic groups in ubiquinol-cytochrome c oxidoreductase.

The results are presented of an electrochemical and high-resolution spectral analysis of the heme prosthetic groups in the bc1 complex from mouse cells. To study the long-range interactions between the Qo and Qi quinone redox sites and the b heme groups, we analyzed the effects on the proximal and distal b heme groups, and the c1 heme, of inhibitors that tightly and specifically bind to the Qi or Qo redox site. A number of results emerged from these studies. (1) There is inhomogeneous broadening of the b heme alpha band absorption spectra. Furthermore, contrary to the conclusion from low-resolution spectral analysis, the higher energy transition in the split-alpha band spectrum of the bL heme is more intense than the lower energy transition. (2) Inhibitors that bind at the Qi site have significant effects upon the electronic environment of the distal bL heme. Conversely, Qo site inhibitors induced changes in the electronic environment of the distal bH heme. (3) In contrast, inhibitor binding at either site has little effect upon the midpoint potential of the distal heme. (4) Experiments in which both a Qi and a Qo inhibitor are bound at the redox sites indicate that the long-range effects of one inhibitor are not blocked by the second inhibitor; enhanced effects are often observed. (5) In the double-inhibitor titrations involving the Qo inhibitor myxothiazol, there is evidence for two electrochemically and spectrally distinct species of the bL heme group, a phenomenon not observed previously. (6) The high-resolution deconvolutions of alpha band absorption spectra allow an interpretation of these inhibitor-induced changes in terms of homogeneous broadening, inhomogeneous broadening, and changes in x-y degeneracy. The general conclusion from these experiments is that when an inhibitor binds to a quinone redox site of the cytochrome b protein, it produces local conformational changes that, in turn, are transmitted to distal regions of the protein. The ligation of the bH and bL hemes between two parallel transmembrane helices provides a mechanism by which long-distance interactions can be propagated. The lack of long-range effects upon the midpoint potentials of the heme groups suggests, however, that protein conformational changes are unlikely to be a major control mechanism for the transmembrane electron- and proton-transfer steps of the Q cycle.

The interaction between cytochrome c2 and the cytochrome bc1 complex in the photosynthetic purple bacteria Rhodobacter capsulatus and Rhodopseudomonas viridis.

The rates of electron transfer from a ubiquinol analogue to cytochrome c2 catalyzed by the cytochrome bc1 complexes of Rhodobacter capsulatus and Rhodopseudomonas viridis were measured as a function of ionic strength. The effects of ionic strength on the kinetic parameters for the reactions are consistent with a role for electrostatic complex formation between cytochrome c2 and the cytochrome bc1 complex in the electron-transfer pathways in both photosynthetic purple non-sulfur bacteria. Additional support for a docking model in which positively charged lysines on cytochrome c2 interact with negatively charged groups on the Rb. capsulatus cytochrome bc1 complex was obtained from kinetic experiments using Rb. capsulatus cytochrome c2 and equine cytochrome c in which specific lysine residues were altered by site-directed mutagenesis and chemical modification, respectively. Equine cytochrome c, which is a poor electron donor to the reaction center of Rps. viridis, is an effective electron acceptor for the Rps. viridis cytochrome bc1 complex. Chemical modification of lysine residues on Rps. viridis cytochrome c2 has a substantially greater effect on the reduction of the Rps. viridis reaction center by ferrocytochrome c2 than on the oxidation of the Rps. viridis cytochrome bc1 complex by ferricytochrome c2. These data suggest that the docking site for Rps. viridis cytochrome c2 on the Rps. viridis reaction center tetraheme subunit differs in structure from the docking site for the cytochrome on the Rps. viridis cytochrome bc1 complex to a significant extent. In this respect, Rps. viridis differs from photosynthetic purple non-sulfur bacteria in which the reaction center does not contain a tetraheme subunit, where the binding sites for cytochrome c2 on the reaction center and the cytochrome bc1 complex appear to be quite similar.

Hydroubiquinone-cytochrome c2 oxidoreductase from Rhodobacter capsulatus: definition of a minimal, functional isolated preparation.

The hydroubiquinone-cytochrome c2 oxidoreductase (cyt bc1) from Rhodobacter capsulatus has been solubilized according to the dodecyl maltoside method and isolated, and its minimal functional composition has been characterized. We find the complex to be composed of three protein subunits corresponding to polypeptides of cyt b (44 kDa), cyt c1 (33 kDa), and 2Fe2S cluster (24 kDa). A fourth band sometimes discernable at 22 kDa appears to be an artifact of the polyacrylamide gel electrophoresis procedure. Its appearance is shown to be derived from the 2Fe2S cluster subunit by the similarity of the binding of subunit-specific monoclonal antibodies and the identical N-terminal sequence of the 24- and 22-kDa bands. The cofactors of cyt bc1, namely, cyt bH, cyt bL, cyt c1, and the 2Fe2S center, the Qos and Qow domains of the Qo site, and the Qi site appear intact as indicated by their optical and EPR spectral signatures, redox properties, and inhibitor binding. The electron paramagnetic resonance spectrum of the cyt bH heme is altered by antimycin, consistent with a change in the dihedral angle between the ligating histidine imidazoles, while the spectrum of the cyt bL heme is broadened by stigmatellin. The ubiquinone-10 content is variable, ranging from 0.8 to 3 molecules/cyt bc1. Activity studies define this three-subunit cyt bc1 complex as a minimal structure, equipped as the enzyme in the native state and capable of full catalytic activity.

Arginine 54 in the active site of Escherichia coli aspartate transcarbamoylase is critical for catalysis: a site-specific mutagenesis, NMR, and X-ray crystallographic study.

The replacement of Arg-54 by Ala in the active site of Escherichia coli aspartate transcarbamoylase causes a 17,000-fold loss of activity but does not significantly influence the binding of substrates or substrate analogs (Stebbins, J.W., Xu, W., & Kantrowitz, E.R., 1989, Biochemistry 28, 2592-2600). In the X-ray structure of the wild-type enzyme, Arg-54 interacts with both the anhydride oxygen and a phosphate oxygen of carbamoyl phosphate (CP) (Gouaux, J.E. & Lipscomb, W.N., 1988, Proc. Natl. Acad. Sci. USA 85, 4205-4208). The Arg-54-->Ala enzyme was crystallized in the presence of the transition state analog N-phosphonacetyl-L-aspartate (PALA), data were collected to a resolution limit of 2.8 A, and the structure was solved by molecular replacement. The analysis of the refined structure (R factor = 0.18) indicates that the substitution did not cause any significant alterations to the active site, except that the side chain of the arginine was replaced by two water molecules. 31P-NMR studies indicate that the binding of CP to the wild-type catalytic subunit produces an upfield chemical shift that cannot reflect a significant change in the ionization state of the CP but rather indicates that there are perturbations in the electronic environment around the phosphate moiety when CP binds to the enzyme. The pH dependence of this upfield shift for bound CP indicates that the catalytic subunit undergoes a conformational change with a pKa approximately 7.7 upon CP binding. Furthermore, the linewidth of the 31P signal of CP bound to the Arg-54-->Ala enzyme is significantly narrower than that of CP bound to the wild-type catalytic subunit at any pH, although the change in chemical shift for the CP bound to the mutant enzyme is unaltered. 31P-NMR studies of PALA complexed to the wild-type catalytic subunit indicate that the phosphonate group of the bound PALA exists as the dianion at pH 7.0 and 8.8, whereas in the Arg-54-->Ala catalytic subunit the phosphonate group of the bound PALA exists as the monoanion at pH 7.0 and 8.8. Thus, the side chain of Arg-54 is essential for the proper ionization of the phosphonate group of PALA and by analogy the phosphate group in the transition state. These data support the previously proposed proton transfer mechanism, in which a fully ionized phosphate group in the transition state accepts a proton during catalysis.

Composition, Variation, and Dynamics of Major Osmotic Solutes in Methanohalophilus Strain FDF1.

Methanohalophilus strain FDF1, a member of the halophilic genus of methanogens, can grow over a range of external NaCl concentrations from 1.2 to 2.9 M and utilize methanol, trimethylamine, and dimethyl sulfide as substrates for methanogenesis. It produces the osmolytes glycine betaine, beta-glutamine, and N-acetyl-beta-lysine with increasing external NaCl, but the relative ratio of these zwitterions depends primarily on the methanogenic substrate and less on the external osmolarity. When the cells are grown on methanol in defined medium, accumulation of glycine betaine predominates over the other zwitterionic solutes. The cells also synthesized a carbohydrate which was not detected in cells grown on trimethylamine. This negatively charged compound, identified as alpha-glucosylglycerate from the C and H chemical shifts, does not act as an osmoregulatory solute in the salt range 1.4 to 2.7 M in this methanogen as evidenced by its invariant intracellular concentration. CH(3)OH-pulse/CH(3)OH-chase experiments were used to determine half-lifes for these organic solute pools in the cells. l-alpha-Glutamate showed a rapid loss of heavy isotope, indicating that l-alpha-glutamate functions as a biosynthetic intermediate in these cells. Measurable turnover rates for both beta-glutamine, which acts as an osmolyte, and alpha-glucosylglycerate suggest that they function as metabolic intermediates as well. Molecules which function solely as osmolytes (glycine betaine and N-acetyl-beta-lysine) showed a slower turnover consistent with their roles as osmotic solutes in Methanohalophilus strain FDF1.

Free amino acid dynamics in marine methanogens. beta-Amino acids as compatible solutes.

Methanogenic archaebacteria respond to osmotic stress by accumulating a series of organic molecules which function as compatible solutes. In two strains of marine methanogenic archaebacteria, Methanogenium cariaci and Methanococcus thermolithotrophicus, four key organic solutes are observed: L-alpha-glutamate, beta-glutamate, N epsilon-acetyl-beta-lysine, and betaine. The first three of these are synthesized de novo; betaine is transported into the Mg. cariaci cells from the medium. Mesophilic Mg. cariaci will preferentially transport betaine from the extracellular medium if it is present to counterbalance the external NaCl. In its absence it synthesizes N epsilon-acetyl-beta-lysine as the dominant osmolyte. This zwitterionic compound occurs at levels in Mg. cariaci which are considerably greater (based on mumol/mg of protein) than in Mc. thermolithotrophicus grown in media of the same ionic strength. Intracellular potassium ion concentrations, determined by 39K NMR spectroscopy and atomic absorption, differ significantly in the two cells. In Mc. thermolithotrophicus, intracellular K+ is balanced by the total concentration of anionic amino acid species, glutamate, and beta-glutamate. Turnover of the organic solutes has been monitored using 13C-pulse/12C-chase, and 15N-pulse/14N-chase experiments. Both beta-amino acids exhibit slower turnover rates when compared to L-alpha-glutamate or aspartate, consistent with their roles as compatible solutes. Biosynthetic information for the beta-amino acids is also provided by 13C-labeling experiments. beta-Glutamate shows a lag in 13C uptake from 13CO2, indicative of its biosynthesis from a precursor (probably a macromolecule) not in equilibrium with the soluble L-alpha-glutamate pool. Confirmation of a novel route for beta-glutamate synthesis and the production of the beta-lysine moiety from the diaminopimelate pathway is deduced from [13C2]acetate labeling patterns.

Oxidation-reduction properties of the ferredoxin-linked glutamate synthase from spinach leaf.

Oxidation-reduction titrations have been conducted to determine the midpoint potential (Em) values of the three electron-carrying prosthetic groups of the ferredoxin-linked glutamate synthase isolated from spinach leaves. Titrations using electron paramagnetic resonance (EPR) signals to monitor the oxidation state of the [3Fe-4S]+,0 cluster found in the enzyme, indicated the presence of a single n = 1 component with Em = -170 mV at pH 7.7. Titrations using absorbance changes in the visible region to monitor the oxidation states of the FAD and FMN groups present in the enzyme could be fit to a single n = 2 Nernst curve with Em = -180 mV at pH 7.7. The magnitude of the absorbance change observed during this titration accounts for all of the FMN and FAD found in the enzyme, indicating that the two flavins are either isopotential or differ in Em by less than about 30 mV. Neither optical nor EPR titrations gave any evidence for the presence of stable flavin free radicals. These results represent the first characterization of the redox properties of the prosthetic groups of a ferredoxin-dependent glutamate synthase.

Enzymatic degradation of cyclic 2,3-diphosphoglycerate to 2,3-diphosphoglycerate in Methanobacterium thermoautotrophicum.

2,3-Diphosphoglycerate (2,3-DPG) has been found to be the product of the enzymatic degradation of cyclic 2,3-diphosphoglycerate (cDPG) in the archaebacterium Methanobacterium thermoautotrophicum delta H. Although 2,3-DPG has not previously been detected as a major soluble component of M. thermoautotrophicum, large pools accumulated at an incubation temperature of 50 degrees C (below the optimum growth temperature of 62 degrees C). Under these conditions, cellular activity was significantly decreased; a return of the culture to the optimum growth temperature restored the 2,3-DPG pool back to original low levels and caused steady-state cDPG levels to increase again. While 13CO2-pulse/12CO2-chase experiments at 50 degrees C showed that the cDPG turned over, the appearance of 2,3-DPG at NMR-visible concentrations required at least 10 h. Production of 2,3-DPG in vivo was prevented by exposure of the cells to O2. The enzyme responsible for this hydrolysis of cDPG was purified by affinity chromatography and appears to be a 33-kDa protein. Activity was detected in the presence of oxygen and was enhanced by a solution of 1 M KCl, 25 mM MgCl2, and dithiothreitol. Both Km and Vmax have been determined at 37 degrees C; kinetics also indicate that in vitro the product, 2,3-DPG, is an inhibitor of cDPG hydrolysis. These findings are discussed in view of a proposed role for cDPG in methanogens.

Cytochrome bc1 complex [2Fe-2S] cluster and its interaction with ubiquinone and ubihydroquinone at the Qo site: a double-occupancy Qo site model.

The ubiquinone complement of Rhodobacter capsulatus chromatophore membranes has been characterized by its isooctane solvent extractability and electrochemistry; we find that the main ubiquinone pool (Qpool) amounts to about 80% of the total ubiquinone and has an Em7 value close to 90 mV. To investigate the interactions of ubiquinone with the cyt bc1 complex, we have examined the distinctive EPR line shapes of the [2Fe-2S] cluster of the cyt bc1 complex when the Qpool-cyt bc1 complex interactions are modulated by changing the numbers of Q or QH2 present (by solvent extraction and reconstitution), by the exposure of the [2Fe-2S] to the Qpool in different redox states, by the presence of inhibitors specific for the Qo site (myxothiazol and stigmatellin) and Qi site (antimycin), and by site-specific mutations of side chains of the cyt b polypeptide (mutants F144L and F144G) previously identified as important for Qo site structure. Evidence suggests that the Qo site can accommodate two ubiquinone molecules. One (designated Qos) is bound relatively strongly and is second only to the ubiquinone of the QA site of the reaction center in its resistance to solvent extraction. In this strong interaction, the Qo site binds Q and QH2 with approximately equal affinities. Their bound states are distinguished by their effects on the [2Fe-2S] cluster spectral feature at gx at 1.783 (Q) and gx at 1.777 (QH2); titration of the line-shape change reveals an Em7 value of approximately 95 mV. The other molecule (Qow) is bound more weakly, in the same range as the ubiquinone of the QB site of the reaction center. Again, the affinities of the Q form (gx at 1.800) and QH2 form (gx at 1.777) are nearly equal, and the Em7 value measured is approximately 80 mV. These results are discussed in terms of earlier EPR analyses of the cyt bc1 complexes of other systems. A Qo site double-occupancy model is considered that builds on the previous model based on Qo site mutants [Robertson, D. E., Daldal, F.,& Dutton, P. L. (1990) Biochemistry 29, 11249-11260] and includes the recent suggestion that two of the [2F3-2S] cluster ligands of the R. capsulatus cyt bc1 complex are histidines [Gurbiel, R. J. Ohnishi, T., Robertson, D. E. Daldal, F., & Hoffman, B. M. (1991) Biochemistry 30, 11579-11584]. We speculate that the cyt bc1 complex complexes a full enzymatic turnover without necessary exchange of ubiquinone with the Qpool.

Q-band ENDOR spectra of the Rieske protein from Rhodobactor capsulatus ubiquinol-cytochrome c oxidoreductase show two histidines coordinated to the [2Fe-2S] cluster.

Electron nuclear double resonance (ENDOR) experiments were performed on 14N (natural abundance) and 15N-enriched iron-sulfur Rieske protein in the ubiquinol-cytochrome c2 oxidoreductase from Rhodobactor capsulatus. The experiments proved that two distinct nitrogenous ligands, histidines, are undoubtedly ligated to the Rieske [2Fe-2S] center. The calculations of hyperfine tensors give values similar but not identical to those of the Rieske-type cluster in phthalate dioxygenase of Pseudomonas cepacia and suggest a slightly different geometry of the iron-sulfur cluster in the two proteins.

Distribution of compatible solutes in the halophilic methanogenic archaebacteria.

Accumulation of compatible solutes, by uptake or de novo synthesis, enables bacteria to reduce the difference between osmotic potentials of the cell cytoplasm and the extracellular environment. To examine this process in the halophilic and halotolerant methanogenic archaebacteria, 14 strains were tested for the accumulation of compatible solutes in response to growth in various extracellular concentrations of NaCl. In external NaCl concentrations of 0.7 to 3.4 M, the halophilic methanogens accumulated K+ ion and low-molecular-weight organic compounds. beta-Glutamate was detected in two halotolerant strains that grew below 1.5 M NaCl. Two unusual beta-amino acids, N epsilon-acetyl-beta-lysine and beta-glutamine (3-aminoglutaramic acid), as well as L-alpha-glutamate were compatible solutes among all of these strains. De novo synthesis of glycine betaine was also detected in several strains of moderately and extremely halophilic methanogens. The zwitterionic compounds (beta-glutamine, N epsilon-acetyl-beta-lysine, and glycine betaine) and potassium were the predominant compatible solutes among the moderately and extremely halophilic methanogens. This is the first report of beta-glutamine as a compatible solute and de novo biosynthesis of glycine betaine in the methanogenic archaebacteria.