Studies on cytochrome oxidase. Interactions of the cytochrome oxidase protein with phospholipids and cytochrome c

The crystallization of mitochondrial cytochrome oxidase-cytochrome c complex

Conditions are described under which crystals are formed with equimolar complex of mitochondrial cytochrome oxidase and cytochrome c. Characteristic absorption bands of the solubilized crystals could be attributed to the cytochrome oxidase-cytochrome c complex with heme a:c ratio of 2:1. Activity of crystals shows more close heme-heme interaction between two cytochromes than that of the mixture.

Behavior of beef-heart cytochrome c oxidase in reconstituted proteovesicles. A systemtic evaluation of the influence of phospholipid polar headgroup and fatty-acyl side chains

1. Phospholipid-depleted cytochrome c oxidase is incorporated in vesicles, built up of phospholipids of known polar headgroup and fatty-acyl side chains. 2. Maximal reactivation is obtained only when the fatty-acyl side chains provide a fluid environment. 3. Fluid zwitterionic phospholipids are found to be more efficient reactivators than fluid anionic ones. 4. Irrespective of the polar headgroup type, two narrow ranges of activation energies for the enzymatic reaction are calculated from the Arrhenius plots: 81--92 kJ/mol in solid and 51--61 kJ/mol in fluid conditions. 5. Cytochrome c oxidase is also incorporated in a series of vesicles, each built up of an equimolar amount of two phospholipids which differ in their polar headgroup type and/or their fatty-acyl side chain characteristics. From the localization of the enzyme activity profiles, obtained with these mixtures, tentative deductions are made about the preference of cytochrome c oxidase for different phospholipid molecules.

Relationships between bilayer lipid, motional freedom of oxidoreductase components, and electron transfer in the mitochondrial inner membrane

The relationships between bilayer lipid, diffusional and conformational activities of oxidoreduction components, and electron transfer activity in the mitochondrial inner membrane are considered. Using a new, low pH method to fuse liposome phospholipid (asolectin) with the isolated mitochondrial inner membrane, the membrane bilayer is enriched up to 700% with exogenous phospholipid. During such enrichment, ultrastructural analysis reveals that integral proteins diffuse freely and randomly into the expanding bilayer. Kinetic analysis reveals that a diffusion limited step occurs between succinate- and NADH dehydrogenase and cytochromes bc1, and that the dehydrogenases, ubiquinone, and cytochromes bc1 are free to diffuse independently of one another in the membrane plane. Whether cytochromes bc1 and cytochrome c oxidase codiffuse in the membrane plane, or diffuse independently of one another remains unclear. The specific activities of succinate- and NADH-dehydrogenase as well as cytochrome c oxidase are affected by bilayer enrichment. This most likely occurs through the direct modulation by the newly incorporated phospholipid on conformational activity required in the oxidoreductases for electron transfer.

Protein-associated lipid of Bacillus stearothermophilus

The composition and patterns of metabolism of phospholipids isolated as part of a lipid-depleted membrane fragment (LDM fragment) and associated with the membrane adenosine triphosphatase complex have been compared with those of the bulk membrane phospholipid. The bulk lipid was extracted from washed membranes with sodium cholate. The LDM fragments, which contained a portion of the electron transport system and the membrane adenosine triphosphatase complex, were purified by chromatography with Sepharose 6B. The LDM fragment preparations contained 0.10 +/- 0.02 mumol of lipid phosphorus per mg of protein, compared with 0.54 +/- 0.05 mumol of lipid phosphorus per mg of protein for washed membranes. The phospholipid associated with the LDM fragments consisted of 78 +/- 4% cardiolipin, 7 +/- 1% phosphatidylglycerol, and 15 +/- 3% phosphatidylethanolamine. Changes in the total membrane lipid composition (produced by culture conditions) did not alter the phospholipid composition of the LDM fragments. The adenosine triphosphate complex was separated from the other components of the LDM fragments by suspension of the fragments in 1% Triton X-100 and precipitation with antibody specific for the F(1) component of the adenosine triphosphatase complex. The phospholipid isolated with the adenosine triphosphatase complex consisted of 86% cardiolipin, 8% phosphatidylglycerol, and 6% phosphatidylethanolamine. In pulse-chase experiments with (32)P and [2-(3)H]glycerol, the labeling patterns of the phosphatididylglycerol and phosphatidylethanolamine associated with the LDM fragments were different from those of the bulk membrane phosphatidylglycerol and phosphatidylethanolamine. It was concluded that at least a portion of the phospholipid isolated with the LDM fragments was part of a native lipid-protein complex.

Labelling of cardiolipin in vitro

A method for labelling the polar head groups of cardiolipin is described. Labelling was carried out on sonicated cardiolipin/water suspensions. The free hydroxyl group of cardiolipin was oxidised with an excess of p-(diazonium) benzenesulfonic acid (DABS) and then reduced with NaB3H4. Isopropanol was oxidised in the presence of DABS to test the reactivity of the diazonium salts, and the reaction product was analysed by means of gas-chromatography. Labelled cardiolipin, identified by thin-layer chromatography (TLC), was chromatographically pure and identical to untreated cardiolipin. The hydrolysis of cardiolipin confirmed that the labelling was at the level of polar head groups.

Azethoxyl nitroxide spin labels. ESR studies involving thiourea crystals, model membrane systems and chromatophores, and chemical reduction with ascorbate and dithiothreitol

Trans- and cis-azethoxyl nitroxides 1, 2, 3 and 4 can be trapped in the cavities of thiourea crystals. The presence of a single gauche conformation on either side of the pyrrolidine ring within the crystals was indicated by the ESR spectra. Rotation about the long molecular axis then corresponds approximately to y-axis motion of the nitroxide moiety. Proxyl nitroxides in which the nitroxide group is located on the penultimate carbon of long chain lipids can also be trapped and were shown to adopt the azethoxyl conformation in the thiourea crystals. The measured deltaA values (A parallel to - A perpendicular) of oriented egg lecithin multilayers containing trans- and cis-azethoxyl nitroxides 1 and 2 were quite small, consistent with the unique orientation of the nitroxide principal axes with respect to the long axis of the molecule. The deltaA values for a series of lipids bearing a label near the terminus of the chain were very similar to that of 1, showing that the azethoxyl conformation is likely the predominant one in these labels in orienting systems. Computer simulations of the ESR spectra of 1 and 2 in egg lecithin vesicles provided values for molecular orientation and motion parameters consistent with those expected from a consideration of molecular models in the extended (all trans) conformation. Azethoxyl nitroxides have also proven useful in the investigation of motion restricted (boundary) lipid in a lipid-protein system. Thus, the values (69 +/- 10%) for the amount of boundary lipid in the chromatophore membranes from Rhodopseudomonas sphaeroides as determined using trans- 2 and cis- 2 are in good agreement with values using 16-doxylstearic acid (64 +/- 3%). The fact that all three labels show about the same fraction of boundary lipid in this system indicates that the lipid binding sites are relatively insensitive to the geometry of the lipid chain. Also, both 1 and 2 appear to be able to detect a third lipid environment not seen with the doxyl fatty acid. The apparent fluidity of this component lies between that of boundary and bilayer lipid. The unique orientation of the nitroxide principal axes with respect to the long molecular axis in azethoxyl nitroxides 1 and 2 allows detection of hindrance to rotation about the long molecular axis, in contrast to the analogous doxyl and proxyl fatty acids. Comparative reduction studies using ascorbate and dithiothreitol indicate that azethoxyl nitroxides are slightly more resistant toward reduction than proxyl nitroxides and much more resistant than doxyl nitroxides.

Electron nuclear double resonance of cytochrome oxidase: nitrogen and proton ENDOR from the 'copper' EPR signal

Proton ENDOR resonances have been found from at least two different protons with fairly large and isotropic couplings of about 12 and 19 MHz. It is possible that such protons are attached to carbons that are one bond removed from the point of ligation to copper. A number of weakly coupled protons with anisotropic couplings have also been seen. None of the protons, either weakly or strongly coupled, appears to exchange with 2H2O. We have obtained nitrogen ENDOR from at least one nitrogen with a hyperfine coupling large enough for the nitrogen to be a ligand of copper. We have not yet demonstrated experimentally ENDOR characteristic of the copper nucleus itself.

Enzyme behaviour and molecular environment. The effects of ionic strength, detergents, linear polyanions and phospholipids on the pH profile of soluble cytochrome oxidase

The activity vs. pH profile for the oxidation of ferrocytochrome c by purified cytochrome oxidase (ferrocytochrome c:oxygen oxidoreductase, EC 1.9.3.1) was investigated as a function of ionic strength (from 10 to 200 mM) in the absence and in the presence of various perturbants: Tween 20, linear polyanions (RNA, heparin, polyglutamic acid) and phospholipids (asolectin, phosphatidylcholine, phosphatidic acid and cardiolipin). The activation induced by Tween 20 and "zero net charge" phospholipid liposomes was not pH dependent. On the other hand, linear polyanions and polyanionic liposomes strongly perturbed the pH profile, mostly at low ionic strength, by shifting the pH optimum about 1.7 pH units towards alkaline pH values. This effect was reversed by increasing ionic strength. These observations are interpreted in the light of polyelectrolyte theory. Since these results show striking with membrane-bound enzyme, it is concluded that in vivo cytochrome oxidase is located within polyanionic sites of the micochondrial membrane. The activation broght about by phospholipids may result from two posible processes: creation of a hydrophobic environment by the non-polar tails, preventing autoaggregation; and creation of a suitable polyelectrolytic environment by the polar heads (of non zero net charge), increasing the intrinsic reaction rate.

Nuclear magnetic resonance studies of lipid-protein interactions. A model of the dynamics and energetics of phosphatidylcholine bilayers that contain cytochrome c oxidase

Reconstituted membrane systems of synthetic phosphatidylcholines and the integral membrane enzyme cytochrome c oxidase were prepared in order to conduct nuclear magnetic resonance studies of lipid-protein interactions. These lipids, labeled with a geminate difluoro group on the 1-position hydrocarbon chain, were combined with the enzyme to give active lipid-protein particles with a well-defined ratio of lipid to protein. The fluorine magnetic resonance spectra of a series of preparations with different lipid/protein ratios suggest that the hydrocarbon chain mobility of the lipid is substantially reduced with increasing amounts of protein. The fluorine spectra of a single lipid-protein preparation show a dramatic increase in the number of the more mobile lipid chains with increasing temperature. The results suggest that the enzyme orders the lipid bilayer well beyond those lipids in direct contact with the protein surface, and that the amount of the lipid restricted by the enzyme is dependent upon temperature. The exchange of lipid between the restricted and the more mobile lipid environments most probably does not occur over the time scale measurable by the magnetic resonance techniques, about 10(-3) s.

Identification of subunits of bovine heart cytochrome oxidase

Purified lipid-depleted cytochrome oxidase, at purity of 12--14 nmol heme a per mg protein, has been shown to contain seven non-identical subunits in the ratio of unity. Their molucular weights on polyacrylamide gel are, in thousands, 40, 21, 14.8, 13.5, 11.6, 9.5, and 7.6 from gel electrophoresis after dissociation in sodium dodecyl sulfate and beta-mercaptoethanol. The molar ratio is determined by the amino acid composition of each subunit obtained from direct hydrolysis of the stained polyacrylamide gel slices. The amino acid composition of the isolated subunits I and II determined by regular hydrolysis method is found practically the same as that from direct hydrolysis of gel slices. The heme-associated polypeptides are identified with subunits of molecular weights of 40.10(3) and 11.6.10(3). One of the two coppers associated with the polypeptide of molecular weight of 21 000. The second copper may be associated with heme in the subunit of 40.10(3). Evidence of the existence of interpolypeptide disulfide linkages is presented.

Mitochondrial biogenesis in cultured mammalian cells. II. Mitochondrial protein and phospholipid synthesis in chloramphenicol-treated BHK-21 cells

The effect of growth of BHK-21 cells in chloramphenicol on the synthesis of cellular proteins and phospholipids has been examined. The incorporation of leucine into total cellular proteins, or into the proteins of specific subcellular fractions are not significantly reduced by cell culture in the presence of chloramphenicol. In cells treated with cycloheximide, a small amount of chloramphenicol-sensitive labelling of protein was detected within the first hour of exposure to the drug. Chloramphenicol inhibits the incorporation of delta-amino-levulinic acid into hemoproteins, only if it is present during both the 48-h culturing and 4-h labelling period. De novo synthesis of cellular lipids as measured by pulse labelling with 32Pi or [3H]glycerol, is decreased in chloramphenicol-treated cells. This decrease is observed in all sub-cellular fractions, although the mitochondrial fraction is most affected. All phospholipids are affected, with diphosphatidylglycerol labelling reduced to the greatest extent. Although fatty acid synthesis is inhibited, the labelling of diphosphatidylglycerol with fatty acids is stimulated on chloramphenicol treatment.

An EPR study of the lineshape of copper in cytochrome c oxidase

The EPR spectrum of copper in cytochrome c oxidase (EC 1.9.3.1) has been studied between 5 and 220 degreesK, and the spectral parameters have been determined for both forms of EPR-detectable copper by computer simulation methods. Numerical methods have been developed to separate the spectra of intrinsic copper and inactive copper. Evidence is presented to show that inactive copper is probably formed by denaturation. The EPR parameters for intrinsic copper were determined as gx = 1.99, gy = 2.03, gz = 2.185, / Ax(Cu) / = 0.0020 cm-1, / Ay(Cu) / = 0.0025 cm-1, / Az(Cu) / = 0.0030 cm-1. The principal values of the g tensor and the small value of /Az(Cu) / are interpreted in terms of mixing of 3d, 4s, and 4p metal orbitals. A flattened-tetrahedral stereochemistry about Cu2+ with an additional rhombic distrotion is in best agreement with all of the data. The peak-to-peak linewidth is found to be orientation dependent, and is described by a tensor with principal values deltaHx = 45G, deltaHy = 65 G, deltaHz = 85 G. A weak dipolar interaction with a low-spin ferric species stereochemistry for the copper ion is consistent with the electron transport function of the enzyme. Broad EPR signals with a very short spin-lattice relaxation time has been observed near g = 14 and g = 3 at 5 degrees K in oxidized cytochrome oxidase but not in the reduced or denatured enzyme. The possibility that these are due to the "EPR-undetectable" iron and copper is raised.