Subunit analysis of bovine cytochrome c oxidase by reverse phase high performance liquid chromatography.

Bovine cytochrome c oxidase subunits were separated by reverse phase high performance liquid chromatography using a C4 column eluted with water and an acetonitrile gradient, both containing 0.1% trifluoroacetic acid. Subunits I and III precipitated in this solvent and could not be analyzed; the remaining eleven subunits were dissociated, denatured, soluble and could be resolved by elution from the column. The protein subunit eluting in each chromatographic peak was identified by a combination of polyacrylamide gel electrophoresis in sodium dodecyl sulfate, NH2-terminal amino acid sequencing, and amino acid analysis. Each subunit produced a single elution peak with the exception of subunit VIc (nomenclature of Kadenbach et al., 1983, Anal. Biochem. 129, 517-521), which eluted from the column as two well-resolved peaks. Sequence analysis showed that the two subunit VIc elution peaks resulted from partial chemical blockage of the alpha-amino serine residue of subunit VIc. The C4 reverse phase HPLC was used to document specific subunit removal from bovine cytochrome c oxidase either by tryptic digestion or by dodecyl maltoside extraction. The described HPLC method for separating cytochrome c oxidase subunits should be applicable for the analysis of other multisubunit proteins, especially other multisubunit membrane protein complexes.

Synthesis of cardiolipin derivatives with protection of the free hydroxyl: its application to the study of cardiolipin stimulation of cytochrome c oxidase.

Cardiolipin derivatives retaining the free hydroxyl on the polar head group were synthesized. With the use of a tetrahydropyranyl ether to protect this hydroxyl, fatty acyl substitutions were made at both of the 2-positions of cardiolipin (CL). The disubstituted derivatives were obtained in high yields. The stimulation of delipidated cytochrome c oxidase activity shows a hyperbolic dependence on the concentration of these CL derivatives. Both activation parameters, the apparent dissociation constant (Kd,app) and the maximum change in molecular activity (delta Actmax), depend on the chain length of the tails, with less dependence on the degree of saturation. Natural CL (92% C18:2, 8% C18:1) and CL disubstituted with oleic acid (47% C18:2, 52% C18:1) were equally effective at stimulating cytochrome c oxidase activity, with an apparent dissociation constant of approximately 1 microM when incubated in 0.3% Triton X-100 and assayed in lauryl maltoside. CL disubstituted with hexanoic acid, however, is a poorer activator, with an apparent dissociation constant of 6.8 microM and a delta Actmax that is 50% of that achieved with natural CL. Dilysocardiolipin, with complete removal of two of the fatty acid tails, shows negligible stimulation of cytochrome c oxidase activity.

Analysis of positional isotope exchange in ATP by cleavage of the beta P-O gamma P bond. Demonstration of negligible positional isotope exchange by myosin.

A method for analysis of positional isotope exchange (PIX) during ATP in equilibrium with HOH oxygen exchange is presented that uses a two-step degradation of ATP resulting in cleavage of the beta P-O gamma P bond. This cleavage yields Pi derived from the gamma-phosphoryl of ATP that contains all four of the gamma oxygens. Both PIX between the beta,gamma-bridge and beta-nonbridge positions and washout of the gamma-nonbridge oxygens can be simultaneously followed by using ATP labeled with 17O at the beta-nonbridge positions and 18O at the beta,gamma-bridge and gamma-nonbridge positions. Application of this method to ATP in equilibrium with HOH exchange during single turnovers of myosin indicates that the bulk of the ATP undergoes rapid washout of gamma-nonbridge oxygens in the virtual absence of PIX. At 25 degrees C with subfragment 1 the scrambling rate is at the limit of detectability of approximately 0.001 s-1, which is 50-fold slower than the steady-state rate. This corresponds to a probability of scrambling for the beta-oxygens of bound ADP of 1 in 10,000 for each cycle of reversible hydrolysis of bound ATP. A fraction of the ATP, however, does not undergo rapid washout. With myosin and stoichiometric ATP at 0 degrees C, this fraction corresponds to 10% of the ATP remaining at 36 s, or 2% of the initial ATP, and an equivalent level of ATP is found that does not bind irreversibly to myosin in a cold chase experiment. A significant level of apparent PIX is observed with subfragment 1 in the fraction that resists washout, and this apparent PIX is shown to be due to contaminant adenylate kinase activity. This apparent PIX due to adenylate kinase provides a possible explanation for the PIX observed by Geeves et al. [Geeves, M. A., Webb, M. R., Midelfort, C. F., & Trentham, D. R. (1980) Biochemistry 19, 4748-4754] with subfragment 1.

Beta-glucosidase: substrate, solvent, and viscosity variation as probes of the rate-limiting steps.

The second-order rate constants (kcat/Km) for the beta-glucosidase-catalyzed hydrolysis of aryl beta-D-glucopyranosides show a bell-shaped dependence of pH. The pKas that characterize this dependence are 4.4 (delta Hion approximately equal to 0) and 6.7 (delta Hion approximately equal to 0). In D2O these pKas are increased by 0.5 (+/- 0.1) unit, but there is no solvent isotope effect on the pH-independent second-order rate constant. Nath and Rydon [Nath, R. L., & Rydon, H. N. (1954) Biochem. J. 57, 1-10] examined the kinetics of the beta-glucosidase-catalyzed hydrolysis of a series of substituted phenyl glucosides. We have extended this study to include glucosides with phenol leaving groups of pKa less than 7. Brønsted plots for this extended series were nonlinear for both kcat/Km and kcat. Brønsted coefficients for those compounds with leaving groups of pKa greater than 7 (for kcat/Km) or pKa greater than 8.5 (for kcat) were nearly equal to -1.0, indicating substantial negative charge buildup on the leaving group in the transition state. The nonlinearity indicates an intermediate in the reaction. This was confirmed by partitioning experiments in the presence of methanol as a competing glucose acceptor. A constant product ratio, [methyl glucoside]/[glucose], was found with aryl glucoside substrates varying over 16,000-fold in reactivity (V/K), indicative of a common intermediate. Viscosity variation (in sucrose-containing buffers) was used to probe the extent to which the beta-glucosidase reactions are diffusion-controlled.(ABSTRACT TRUNCATED AT 250 WORDS)

Reversible inhibitors of beta-glucosidase.

A variety of reversible inhibitors of sweet almond beta-glucosidase were examined. These included simple sugars and sugar derivatives, amines and phenols. With respect to the sugar inhibitors and, indeed, the various glycoside substrates, the enzyme has what can be considered a "relaxed specificity". No single substituent on glucose, for example, is essential for binding. Replacement of a hydroxyl group with an anionic substituent reduces the affinity while substitution with a cationic (amine) substituent enhances the affinity. Amines, in general, are good inhibitors, binding more tightly than the corresponding alcohols: pKiRNH3+ = 0.645pKiROH + 1.77 (n = 9, r = 0.97). The affinity of a series of 10 primary amines was found to be strongly influenced by substituent hydrophobicity: pKi = 0.52 pi + 1.32 (r = 0.95). The major binding determinant of the glycoside substrates is the aglycon moiety. Thus, the Ki values of phenols are similar in magnitude to the Ks values of the corresponding aryl beta-glucoside. The pH dependence for the inhibition by various phenols indicates that it is the un-ionized phenol which binds to the enzyme when an enzymic group of pKa = 6.8 (+/- 0.1) is protonated. The affinity of the phenol inhibitor is dependent on its basicity with a Brønsted coefficient for binding of beta = -0.26 (n = 14, r = 0.98). The pH dependence of the binding of two particularly potent beta-glucosidase inhibitors was also examined. 1-Deoxynojirimycin (1,5-dideoxy-1,5-imino-D-glucitol) has a pH-corrected Ki = 6.5 microM, and D-glucono-1,5-lactam has a pH-corrected Ki = 29 microM.(ABSTRACT TRUNCATED AT 250 WORDS)