Portal de Periódicos da CAPES

Membrane Potential-controlled Inhibition of Cytochromec Oxidase by Zinc

2002; Elsevier BV; Volume: 277; Issue: 17 Linguagem: Inglês

10.1074/jbc.m111922200

1083-351X

Denise A. Mills, Bryan Schmidt, Carrie Hiser, Erica Westley, Shelagh Ferguson‐Miller,

Photoreceptor and optogenetics research

Like many voltage-sensitive ion pumps, cytochromec oxidase is inhibited by zinc. Binding of zinc to the outside surface of Rhodobacter sphaeroides cytochromec oxidase inhibits the enzyme with a K Iof ≤ 5 μm when the enzyme is reconstituted into phospholipid vesicles in the presence of a membrane potential. In the absence of a membrane potential and a pH gradient, millimolar concentrations of zinc are required to inhibit. This differential inhibition causes a dramatic increase in the respiratory control ratio from 6 to 40 for wild-type oxidase. The external zinc inhibition is removed by EDTA and is not competitive with cytochrome cbinding but is competitive with protons. Only Cd2+ of the many metals tested (Mg2+, Mn2+, Ca2+, Ba2+, Li2+, Cs2+, Hg2+, Ni2+, Co2+, Cu2+Tb3+, Tm3+) showed inhibitory effects similar to Zn2+. Proton pumping is slower and less efficient with zinc. The results suggest that zinc inhibits proton movement through a proton exit path, which can allow proton back-leak at high membrane potentials. The physiological and mechanistic significance of proton movement in the exit pathway and its blockage by zinc is discussed in terms of regulation of the efficiency of energy transduction. Like many voltage-sensitive ion pumps, cytochromec oxidase is inhibited by zinc. Binding of zinc to the outside surface of Rhodobacter sphaeroides cytochromec oxidase inhibits the enzyme with a K Iof ≤ 5 μm when the enzyme is reconstituted into phospholipid vesicles in the presence of a membrane potential. In the absence of a membrane potential and a pH gradient, millimolar concentrations of zinc are required to inhibit. This differential inhibition causes a dramatic increase in the respiratory control ratio from 6 to 40 for wild-type oxidase. The external zinc inhibition is removed by EDTA and is not competitive with cytochrome cbinding but is competitive with protons. Only Cd2+ of the many metals tested (Mg2+, Mn2+, Ca2+, Ba2+, Li2+, Cs2+, Hg2+, Ni2+, Co2+, Cu2+Tb3+, Tm3+) showed inhibitory effects similar to Zn2+. Proton pumping is slower and less efficient with zinc. The results suggest that zinc inhibits proton movement through a proton exit path, which can allow proton back-leak at high membrane potentials. The physiological and mechanistic significance of proton movement in the exit pathway and its blockage by zinc is discussed in terms of regulation of the efficiency of energy transduction. Zinc is observed to have an unusually strong and specific inhibitory effect on a number of proton and ion channels, but the physiological significance of this inhibition is the subject of much debate (1.Outten C. O'Halloran T. Science. 2001; 292: 2488-2492Crossref PubMed Scopus (1155) Google Scholar). The controversy is heightened by the difficulty in accurately assessing free zinc levels in tissues and subcellular compartments (1.Outten C. O'Halloran T. Science. 2001; 292: 2488-2492Crossref PubMed Scopus (1155) Google Scholar, 2.Ye B. Marte W. Vallee B. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 2317-2322Crossref PubMed Scopus (220) Google Scholar). In mitochondria, zinc is known to inhibit thebc 1 complex at submicromolar levels (3.Paddock M. Graige M. Feher G. Okamura M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 6183-6188Crossref PubMed Scopus (116) Google Scholar), and similar strong inhibition is observed in bacterial photosynthetic reaction centers (3.Paddock M. Graige M. Feher G. Okamura M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 6183-6188Crossref PubMed Scopus (116) Google Scholar, 4.Axelrod H. Abresch E. Paddock M. Okamura M. Feher G. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 1542-1547Crossref PubMed Scopus (120) Google Scholar). In each case, the zinc inhibition is reversible and blocks a proton pathway. In the reaction center, where addition of Zn2+ limits a proton uptake step, the site for zinc binding was determined by x-ray crystallography (4.Axelrod H. Abresch E. Paddock M. Okamura M. Feher G. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 1542-1547Crossref PubMed Scopus (120) Google Scholar) allowing the identification of the predominant route for proton uptake and two aspartates that participate in that proton transfer pathway. Cytochrome c oxidase (CcO) 1The abbreviations used are: CcOcytochrome c oxidaseCCCPcarbonylcyanide-m-chlorophenylhydrazoneCOVscytochrome oxidase-containing phospholipid vesiclesFCCPcarbonylcyanide-p-trifluoromethoxy-phenylhydrazonehtIICcOcytochrome c oxidase with a His-tag on subunit II C-terminalMES2-(N-morpholino)ethanesulfonic acidNi2+-NTAnickel nitrilotriacetic acidRCRrespiratory control ratioRsR. sphaeroidesTMPDN,N,N′,N′-tetramethyl-p-phenylenediamineRsCcORhodobacter sphaeroides cytochrome c oxidaseCHES2-(cyclohexylamino)ethanesulfonic acid1The abbreviations used are: CcOcytochrome c oxidaseCCCPcarbonylcyanide-m-chlorophenylhydrazoneCOVscytochrome oxidase-containing phospholipid vesiclesFCCPcarbonylcyanide-p-trifluoromethoxy-phenylhydrazonehtIICcOcytochrome c oxidase with a His-tag on subunit II C-terminalMES2-(N-morpholino)ethanesulfonic acidNi2+-NTAnickel nitrilotriacetic acidRCRrespiratory control ratioRsR. sphaeroidesTMPDN,N,N′,N′-tetramethyl-p-phenylenediamineRsCcORhodobacter sphaeroides cytochrome c oxidaseCHES2-(cyclohexylamino)ethanesulfonic acid requires protons as a substrate, using 4 protons for the reduction of O2 to form 2H2O and translocating 4 protons across the membrane for each O2 reduced. A total of 8 protons are therefore taken up from the interior per 4 electrons donated externally by cytochrome c, the electron donor. Crystal structures of cytochrome c oxidase have helped identify two proposed channels for proton uptake in subunit I (5.Iwata S. Ostermeier C. Ludwig B. Michel H. Nature. 1995; 376: 660-669Crossref PubMed Scopus (1980) Google Scholar, 6.Ostermeier C. Harrenga A. Ermler U. Michel H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 10547-10553Crossref PubMed Scopus (715) Google Scholar, 7.Tsukihara T. Aoyama H. Yamashita E. Tomizaki T. Yamaguchi H. Shinzawa-Itoh K. Nakashima R. Yaono R. Yoshikawa S. Science. 1996; 272: 1136-1144Crossref PubMed Scopus (1921) Google Scholar), referred to as the D and K channels, because an aspartate and a lysine, respectively, are important for their activity (Fig. 1). The roles of these channels are still debated in terms of the number and destination of the protons they conduct and how the individual proton uptake events are coupled to electron transfer. Even less well understood is the pathway for proton release, although several proposals have been made (8.Florens L. Hoganson C. McCracken J. Fetter J. Mills D.A. Babcock G.T. Ferguson-Miller S. Peschek G.e.a. The Phototropic Prokaryotes. Kluwer Acad., Plenum Publishers, New York, NY1999: 329-339Google Scholar, 9.Wikström M. Bogachev A. Finel M. Morgan J.E. Puustinen A. Raitio M. Verkhovskaya M.L. Verkhovsky M.I. Biochim. Biophys. Acta. 1994; 1187: 106-111Crossref PubMed Scopus (113) Google Scholar). A hydrogen-bonded network, the H channel, has been suggested as a potential proton pathway that spans the membrane in the bovine oxidase (7.Tsukihara T. Aoyama H. Yamashita E. Tomizaki T. Yamaguchi H. Shinzawa-Itoh K. Nakashima R. Yaono R. Yoshikawa S. Science. 1996; 272: 1136-1144Crossref PubMed Scopus (1921) Google Scholar), but its presence in bacterial oxidases has not been confirmed (10.Lee H.-M. Das T. Rousseau D. Mills D. Ferguson-Miller S. Gennis R. Biochemistry. 2000; 39: 2989-2996Crossref PubMed Scopus (99) Google Scholar). cytochrome c oxidase carbonylcyanide-m-chlorophenylhydrazone cytochrome oxidase-containing phospholipid vesicles carbonylcyanide-p-trifluoromethoxy-phenylhydrazone cytochrome c oxidase with a His-tag on subunit II C-terminal 2-(N-morpholino)ethanesulfonic acid nickel nitrilotriacetic acid respiratory control ratio R. sphaeroides N,N,N′,N′-tetramethyl-p-phenylenediamine Rhodobacter sphaeroides cytochrome c oxidase 2-(cyclohexylamino)ethanesulfonic acid cytochrome c oxidase carbonylcyanide-m-chlorophenylhydrazone cytochrome oxidase-containing phospholipid vesicles carbonylcyanide-p-trifluoromethoxy-phenylhydrazone cytochrome c oxidase with a His-tag on subunit II C-terminal 2-(N-morpholino)ethanesulfonic acid nickel nitrilotriacetic acid respiratory control ratio R. sphaeroides N,N,N′,N′-tetramethyl-p-phenylenediamine Rhodobacter sphaeroides cytochrome c oxidase 2-(cyclohexylamino)ethanesulfonic acid A reversible metal inhibitor that selectively binds to one of the proton paths would be an excellent tool for elucidating the function of the different proton transfer channels in oxidase. Although studies in 1988 (11.Nicholls P. Singh A.P. Life Sci. Adv. (Agra, India). 1988; 7: 321-326Google Scholar) showed zinc inhibition of bovine heart oxidase reconstituted into proteoliposomes, the estimated inhibition constant of ∼20–40 μm was quite high. However, in the reconstitutedEscherichia coli bo 3 oxidase, 10 μm ZnSO4 gave significant inhibition (12.Kita K. Kasahara M. Anraku Y. J. Biol. Chem. 1982; 257: 7933-7935Abstract Full Text PDF PubMed Google Scholar). Recent studies on purified RsCcO show two different inhibitory effects on steps in the reaction cycle (13.Aagaard A. Brzezinski P. FEBS Lett. 2001; 494: 157-160Crossref PubMed Scopus (53) Google Scholar) with high and low affinities for Zn2+. Similarly,Paracoccus denitrificans oxidase shows evidence of Zn2+ sites that affect proton uptake steps in the reaction cycle similar to those identified in the purified Rs oxidase (14.Kannt A. Ostermann T. Muller H. Ruitenberg M. FEBS Lett. 2001; 503: 142-146Crossref PubMed Scopus (34) Google Scholar). The studies reported here demonstrate that there is a zinc binding site on the external side (P side) of the oxidase that is strongly inhibitory in the presence of a membrane potential (ΔΨ, negative inside), but whose inhibitory effect is much diminished when the electrical potential is removed. The zinc inhibition ofRsCcO in lipid vesicles is readily reversed by a water-soluble chelator, demonstrating that the zinc site is external and solvent accessible. We suggest that the potential-sensitive inhibitory effect involves blockage of a proton exit path. Cytochrome coxidase was purified from Rhodobacter sphaeroides as previously reported (15.Zhen Y. Qian J. Follmann K. Hosler J. Hayward T. Nilsson T. Ferguson-Miller S. Protein Expr. Purif. 1998; 13: 326-336Crossref PubMed Scopus (72) Google Scholar) using an Ni2+-NTA-agarose (Qiagen) affinity step followed by a further fast protein liquid chromatography (Amersham Biosciences, Inc., AKTA-519) purification procedure with tandem DEAE-5PW columns (Toso-Haas) (16.Hiser C. Mills D.A. Schall M. Ferguson-Miller S. Biochemistry. 2001; 40: 1606-1615Crossref PubMed Scopus (40) Google Scholar). Bovine heart oxidase was purified by the Yoshikawa method (17.Yoshikawa S. Tera T. Takahashi Y. Tsukihara T. Caughey W.S. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 1354-1358Crossref PubMed Scopus (64) Google Scholar). Site-directed mutants were made previously (M263LII (18.Zhen Y. Wang K. Kang U.G. Millett F. Ferguson-Miller S. Biophys. J. 1999; 76: A237Google Scholar); D132AI (19.Fetter J.R. Qian J. Shapleigh J. Thomas J.W. Garcı́a-Horsman J.A. Schmidt E. Hosler J. Babcock G.T. Gennis R.B. Ferguson-Miller S. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 1604-1608Crossref PubMed Scopus (216) Google Scholar); D407AI and R481KI (20.Qian J. Shi W. Pressler M. Hoganson C. Mills D. Babcock G.T. Ferguson-Miller S. Biochemistry. 1997; 36: 2539-2543Crossref PubMed Scopus (35) Google Scholar); H277AI(21.Hosler J.P. Ferguson-Miller S. Calhoun M.W. Thomas J.W. Hill J. Lemieux L. Ma J. Georgiou C. Fetter J. Shapleigh J.P. Tecklenburg M.M.J. Babcock G.T. Gennis R.B. J. Bioenerg. Biomembr. 1993; 25: 121-136Crossref PubMed Scopus (242) Google Scholar); R234H/CII, D412AI, and H411QII (22.Florens L. McCracken J. Ferguson-Miller S. Biophys. J. 1999; 76: A238Google Scholar, 23.Mills D.A. Florens L. Hiser C. Qian J. Ferguson-Miller S. Biochim. Biophys. Acta. 2000; 1458: 180-187Crossref PubMed Scopus (62) Google Scholar); and H93C/NI 2C. Hiser and S. Ferguson-Miller, unpublished. and subunit III-less oxidase (24.Bratton M. Pressler M. Hosler J. Biochemistry. 1999; 38: 16236-16245Crossref PubMed Scopus (68) Google Scholar)) and purified in the same way as wild-type. Reconstitution of cytochrome c oxidase vesicles (COVs) was performed using a dialysis method, giving a final concentration of 20 mg/ml asolectin (recrystallized from Associated Concentrates using the procedure of Sone et al. (25.Smith H.T. Staudenmayer N. Millett F. Biochemistry. 1977; 16: 4971-4974Crossref PubMed Scopus (127) Google Scholar)) and 2 μmoxidase in 50 μm HEPES-KOH, pH 7.4, + 44 mmKCl + 38 mm sucrose. Alternatively, purified COVs were made using an oxidase construct with a His-tag on subunit II (htIICcO) that is able to bind to an Ni2+-NTA affinity chromatography column for isolation and concentration of COVs with correctly oriented oxidase, as described in Hiser et al. (16.Hiser C. Mills D.A. Schall M. Ferguson-Miller S. Biochemistry. 2001; 40: 1606-1615Crossref PubMed Scopus (40) Google Scholar). Basically, detergent was removed by a Sephadex G-25 column equilibrated with 75 mm HEPES-KOH, pH 7.4, +14 mm KCl. The COVs were collected and diluted into 20 mm HEPES-KOH, pH 8.0, 27 mm KCl and 38 mm sucrose, and the pH was adjusted to 8.0 with KOH before addition to Ni2+-NTA-agarose. Vesicles containing correctly oriented oxidase bound to the column and were eluted with 20 mm HEPES-KOH, pH 7.4, 100 mm histidine, 38 mm sucrose. Histidine and buffer were removed by concentration of the COVs on a Centriplus 100 centrifugal spin filter (Amicon) and dialyzed against 2 × 1000 volumes for 6 h each of 50 μm HEPES-KOH, pH 7.4, 45 mm KCl, and 44 mm sucrose. The concentration of oxidase in the purified vesicles was calculated from the dithionite-reduced spectrum at 605 nm using an extinction coefficient of 33.35 mm−1cm−1 after background baseline subtraction (26.Vanneste W.H. Biochemistry. 1966; 5: 838-848Crossref PubMed Scopus (190) Google Scholar). ZnSO4 was from Columbus Chemical Industries, Inc., and all other metals used were of ACS grade. Na-EDTA (Invitrogen) at pH 7.4 was used as a chelator. Oxygen consumption was measured with a Gilson oxygraph at 25 °C using ascorbate, TMPD (Kodak), and horse heart cytochrome c (Sigma), which further was purified using carboxymethylcellulose ion-exchange chromatography (27.Brautigan D.L. Ferguson-Miller S. Margoliash E. Methods Enzymol. 1978; 53: 128-164Crossref PubMed Scopus (334) Google Scholar). Zinc inhibition constants (K I) were calculated from the fits of the data using non-linear least-squares fitting in Microcal Origin to 1 or 2 inhibitory binding sites using the equations,v=Vmax/(1+[zinc]/KI)Equation 1 for uncontrolled COV measurements, and,v=Vmax1/(1+[zinc]/KI1)+Vmax2/(1+[zinc]/KI2)Equation 2 for controlled COV and purified enzyme. The data from the steady-state kinetics of cytochromec reaction with controlled COVs was fitted in Microcal Origin to Michaelis-Menten plots using the equation for two cytochromec binding sites (28.Zhen Y. Hoganson C.W. Babcock G.T. Ferguson-Miller S. J. Biol. Chem. 1999; 274: 38032-38041Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar),v={Va([S]/Ka)+Vb([S] 2/KaKb)}/{1Equation 3 +[S]/Ka+[S]/Kb=[S] 2/KaKb} Measurements of cytochromec oxidation were made in an Olis-rsm stopped-flow apparatus with COVs. Scans were collected (1000/s), and from these the rates of cytochrome c oxidation (550 nm) were fit either using Global analysis or by single-exponential fitting to the kinetic traces using Microcal Origin. The electron donor was pre-reduced horse heart cytochrome c made by sodium dithionite reduction followed by desalting and depolymerizing through a Sephadex G-75 (Amersham Biosciences, Inc.) gel filtration column into 0.5 mm HEPES-KOH, pH 7.4, + 45 mm KCl + 1 mm Na-EDTA. Proton pumping measurements were made as described previously (16.Hiser C. Mills D.A. Schall M. Ferguson-Miller S. Biochemistry. 2001; 40: 1606-1615Crossref PubMed Scopus (40) Google Scholar). Scans were collected, and kinetic traces for cytochrome c oxidation at 550 nm or phenol red changes at 557 nm (isosbestic point of cytochromec) were extracted after averaging at least three data sets and creating difference spectra (reduced minus oxidized). A small mixing artifact at 0–200 ms was subtracted from the phenol red changes (16.Hiser C. Mills D.A. Schall M. Ferguson-Miller S. Biochemistry. 2001; 40: 1606-1615Crossref PubMed Scopus (40) Google Scholar). Rates of proton uptake or release were measured by fitting the kinetic traces for phenol red at 557 nm to one exponential with Microcal Origin. The cytochrome c oxidation rates were obtained from Global fitting analysis by the Olis software to a single-component exponential and were similar to those obtained from the kinetic traces at 550 nm, but the latter were more perturbed by the influence of the phenol red absorbance. The inhibition with zinc was examined with cytochrome c oxidase reconstituted in lipid vesicles to physically separate the outside of the enzyme (P side) from the inside (N side). Measurements of cytochrome c oxidation with reconstituted cytochrome-c oxidase (COVs) were made: (a) in the controlled condition (no ionophores added; i.e. with ΔpH and ΔΨ), (b) with valinomycin (no ΔΨ), and (c) in the uncontrolled state with uncoupler (no ΔpH or ΔΨ) (Fig. 2). It was immediately obvious that zinc was more effective in inhibiting the controlled state, in stopped-flow measurements with few or many turnovers. This was also true in steady-state measurements of oxygen consumption, where increasing ZnSO4 concentrations revealed aK I of ∼5 μm, reaching ∼75% inhibition in the controlled state at 300 μmZnSO4 (Fig. 3 A).Figure 3Differential inhibition by nickel and zinc of purified enzyme compared with COVs. Steady-state measurements of ZnSO4 (solid lines) or NiSO4(dashed lines) inhibition were made as described under "Experimental Procedures" with increasing metal added immediately prior to the addition of TMPD, ascorbate, enzyme, and 30 μm horse heart cytochrome c. A, controlled (no ionophores); B, uncontrolled states (+1 μm valinomycin + 6 μm FCCP), COVs (14 nm aa 3), 5.7 mmascorbate, and 0.28 mm TMPD were added. C, purified enzyme (2 nm aa 3), 0.05% lauryl maltoside, 1.4 mm ascorbate, and 1.1 mmTMPD were added. The inset to B shows the full range of metal concentrations tested with the uncontrolled COVs. The data from at least two separate measurements is fitted to 1 or 2 binding sites as under "Experimental Procedures."View Large Image Figure ViewerDownload Hi-res image Download (PPT) Zinc also inhibited after the addition of nigericin, which removes the ΔpH leaving a ΔΨ, both in steady-state and stopped-flow measurements where activity was decreased ∼60% with 100 μm zinc. Zinc was much less inhibitory when valinomycin or uncoupler was added (K I ∼ 1 mm) (Fig. 3 B). In the purified, unreconstituted enzyme, the activity profile with increasing zinc (Fig. 3 C) fit to two inhibition constants (K I ∼ 9 μmand ∼ 400 μm). The effect of zinc on bovine heart-controlled COVs (data not shown) was less potent, giving inhibition constants of ∼ 25 μm and in the millimolar range, similar to the earlier report (11.Nicholls P. Singh A.P. Life Sci. Adv. (Agra, India). 1988; 7: 321-326Google Scholar). Inhibition by Zn2+ was reversed by addition of Na-EDTA under all conditions and within the dead-time of mixing in the stopped-flow apparatus, indicating that zinc binds on the outside of the enzyme surface and is solvent-accessible. In the controlled state, slight additional inhibition at very high concentrations of zinc (K I ∼ 1–2 mm) was discerned from biphasic fitting of the data of Fig. 3 A, but this was likely due to vesicle aggregation, which can be monitored by light-scattering changes. Examination for these scattering effects showed substantial vesicle aggregation with 5 μmpoly-l-lysine but no measurable effect up to 100 μm Zn2+, indicating that vesicle aggregation is not a factor in the zinc inhibition at the lower concentrations. The question of whether Zn2+ inhibition is mediated by binding to the phospholipid membrane rather than the oxidase was addressed by altering the phospholipid composition of the vesicles. When the asolectin vesicles were supplemented with 25% phosphatidylcholine (neutral charge) or 25% phosphatidylserine (negative charge), the resulting COVs gave an unchanged inhibition constant with Zn2+ (K I ≤ 5 μm), arguing against a lipid-mediated effect. Further evidence that Zn2+ inhibition involves direct interaction with the RsCcO rather than the membrane comes from comparison of normal and htIICcO-purified vesicles, where the lipid-to-CcO ratio is decreased 10-fold. The decrease in lipids did not affect the K I for Zn2+. Specific zinc inhibition of the controlled state was observed whether the histidine-tag, used for purification of the overexpressed enzyme, was on subunit I and therefore on the inside of the COVs or on subunit II at the C-terminal end (htIICcO) on the outside of the COVs (16.Hiser C. Mills D.A. Schall M. Ferguson-Miller S. Biochemistry. 2001; 40: 1606-1615Crossref PubMed Scopus (40) Google Scholar), implying that the His-tag is not also involved in zinc inhibition. A comparison of nickel and zinc in the controlled state (Fig. 3 A) showed that, although nickel appears to bind with high affinity to the outside of oxidase its inhibition is limited, only inhibiting to 20% even at high metal concentrations. Unlike the Ni2+ binding on the outside of COVs, Ni2+ strongly inhibits with medium affinity (∼40 μm) the purified enzyme (Fig. 3 C) but lacks the low affinity binding site seen with Zn2+. Other divalent cations were found to be either non-inhibitory beyond ionic strength effects (Mg2+, Mn2+, Ca2+, Ba2+, Li2+, Cs2+, Tb3+, Tm3+), or slightly inhibitory (Hg2+>Ni2+>Co2+), with only Cd2+ being as effective as Zn2+ (data not shown with the exception of Co2+, Ca2+, and Zn2+ in Table I). These results indicate a specific interaction and not, for example, an effect of ionic strength. This was verified by the measurement of steady-state turnover of the oxidase with high concentrations of Ca2+, which did not remove the inhibition by zinc (data not shown). Neither was it an effect on the interaction of cytochrome c with the oxidase, as shown by examining the kinetics of interaction of cytochrome c in the absence or presence of zinc (Fig. 4). This result clearly shows that addition of zinc in the micromolar range to controlled COVs reduces the V max from 900 to 600 e−/s/aa 3. However, from the negative inverse of both of the slopes in the Eadie-Scatchard plot and from the fitting of the Michaelis-Menten plot (as under "Experimental Procedures," not shown), there was no significant effect on theK m values for cytochrome c(K a = 0.04 μm andK b = 0.8 μm) with or without zinc. These apparent Michaelis constants are similar to those published for free enzyme (28.Zhen Y. Hoganson C.W. Babcock G.T. Ferguson-Miller S. J. Biol. Chem. 1999; 274: 38032-38041Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar).Table IZinc inhibits the controlled condition and increases the respiratory control ratioCOVsActivityRCR1-aRCR = respiratory control ratio (uncontrolled activity/controlled activity).Controlled+ValinomycinUncontrollede− /s/aa3Wild-type99 ± 30460 ± 40620 ± 706.2+Ca2+130 ± 30410 ± 30600 ± 704.5+Zn2+15 ± 1440 ± 20650 ± 4044D132AI15 ± 0.47.6 ± 0.19.2 ± 0.30.6+Co2+12 ± 0.25.3 ± 0.19.2 ± 0.30.6+Zn2+4.0 ± 0.14.8 ± 0.18.3 ± 0.12.0M263LII9.1 ± 0.222 ± 124 ± 52.7+Zn2+5.3 ± 0.230 ± 0.239 ± 27.3R481KI11 ± 127 ± 3250 ± 1323+Zn2+3.7 ± 0.225 ± 2200 ± 3541-a RCR = respiratory control ratio (uncontrolled activity/controlled activity). Open table in a new tab The purified oxidized enzyme or COVs after turnover with reduced cytochrome c were examined by EPR (electron paramagnetic resonance): the CuA and heme a spectra revealed no changes with zinc addition up to millimolar concentrations (data not shown). To investigate the nature of zinc inhibition, mutants with already low activity due to blockage at varying sites were examined for additive Zn2+effects. A mutant of the D channel path (D132AI), has very low activity, normally 5% of wild-type activity when comparing the rates of the purified enzymes, and does not pump protons (19.Fetter J.R. Qian J. Shapleigh J. Thomas J.W. Garcı́a-Horsman J.A. Schmidt E. Hosler J. Babcock G.T. Gennis R.B. Ferguson-Miller S. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 1604-1608Crossref PubMed Scopus (216) Google Scholar, 29.Fetter J.R. Sharpe M. Qian J. Mills D. Ferguson-Miller S. Nicholls P. FEBS Lett. 1996; 393: 155-160Crossref PubMed Scopus (41) Google Scholar, 30.Zhen Y. Mills D. Hoganson C.W. Lucas R.L. Shi W. Babcock G. Ferguson-Miller S. Papa S. Guerrieri F. Tager J.M. Frontiers of Cellular Bioenergetics: Molecular Biology, Biochemistry and Physiopathology. Plenum Press, New York1999: 157-178Crossref Google Scholar). In vesicles under controlled conditions, D132A (15 e−/s/aa 3) has 15% of the wild-type rate. D132A is inhibited by zinc to an even lower rate (4 e−/s/aa 3) with a similarK I as wild-type (Table I). The D132AI mutant is rate-limited by blockage of proton uptake from the inside via thed-path, but there is substantial evidence that protons can be supplied from the outside at a slow rate that is stimulated by the membrane potential (19.Fetter J.R. Qian J. Shapleigh J. Thomas J.W. Garcı́a-Horsman J.A. Schmidt E. Hosler J. Babcock G.T. Gennis R.B. Ferguson-Miller S. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 1604-1608Crossref PubMed Scopus (216) Google Scholar, 30.Zhen Y. Mills D. Hoganson C.W. Lucas R.L. Shi W. Babcock G. Ferguson-Miller S. Papa S. Guerrieri F. Tager J.M. Frontiers of Cellular Bioenergetics: Molecular Biology, Biochemistry and Physiopathology. Plenum Press, New York1999: 157-178Crossref Google Scholar). The rate of proton uptake from the outside for D132AI COVs, observed as an alkalinization of the vesicle exterior in the controlled condition, is the same as the electron transfer rate (Table II) whether or not Zn2+ is added. In the wild-type COVs, this apparent proton leak rate, under controlled conditions, is much slower than the electron transfer rate, because proton uptake from the inside is supporting activity and pumping. But the addition of Zn2+slows the electron transfer rate to that of the alkalinization rate, similar to the D132AI mutant. A mutation of the CuA ligand, M263LII, has very low activity due to a 100-mV increase in the CuA redox potential (31.Karpefors M. Adelroth P. Zhen Y. Ferguson-Miller S. Brzezinski P. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 13606-13611Crossref PubMed Scopus (75) Google Scholar, 32.Zhen Y. Wang K. Mills D. Ferguson-Miller S. Millett F. Biophys. J. 1997; 72: A93Google Scholar). Zn2+ also inhibited this mutant to a similar extent (to 5 e−/s/aa 3 with 250 μmZn2+) as D132AI (Table I). Also, Zn2+ inhibits a mutant, R481K, which already has high respiratory control.Table IIMeasurements of cytochrome c oxidation and phenol red changes with controlled COVs in the stopped-flow Olis-rsmCOVsRatesCyt.c(e−)2-aCyt.c, cytochromec.H+alkalinizations−1Wild-type206 ± 1443 ± 1+Ca2+182 ± 530 ± 0.4+Zn2+8 ± 0.39 ± 0.8D132A29 ± 0.630 ± 0.3+Ca2+22 ± 0.525 ± 0.4+Zn2+4 ± 0.36 ± 1.72-a Cyt.c, cytochromec. Open table in a new tab In an effort to identify the zinc binding site, various mutants of outside sites were examined for their ability to be inhibited by zinc in the controlled state after reconstitution. The preferred ligands for zinc are cysteines, histidines, carboxyls, carbonyls, other charged groups, and H2O (33.Cowan J.A. Inorganic Biochemistry: An Introduction. 2nd Ed. Wiley-VCH, Inc., New York, NY1997: 1-63Google Scholar). There are only a few conserved charged residues on the outside of oxidase, excluding the cytochrome cbinding site on subunit II. The subunit I mutants H277LI, H411QI, D412AI, and D407AI retained inhibition by zinc (Fig. 1). Mutations of a surface histidine (H93C/NI) were made, above heme a close to the aspartate 51 in the bovine oxidase. This region has been identified as having an altered conformation in the reduced versusoxidized crystal structures of the bovine enzyme and was suggested to be the proton exit site (34.Yoshikawa S. Shinzawa-Itoh K. Nakashima R. Yaono R. Yamashita E. Inoue N. Yao M. Fei M.J. Libeu C.P. Mizushima T. Yamaguchi H. Tomizaki T. Tsukihara T. Science. 1998; 280: 1723-1729Crossref PubMed Scopus (973) Google Scholar). However, the reconstituted H93 mutants were still inhibited by zinc. Additionally, a mutant that lacked subunit III (Cox III(−)) (24.Bratton M. Pressler M. Hosler J. Biochemistry. 1999; 38: 16236-16245Crossref PubMed Scopus (68) Google Scholar) was inhibited by zinc, eliminating subunit III as a candidate for the external zinc binding site. In stopped-flow measurements of cytochrome c oxidation, zinc inhibition was decreased from ∼85% at 240 μm Zn2+ to 5% after the addition of valinomycin, which equilibrates potassium across the membrane to dissipate the ΔΨ, to wild-type COVs (Table I). Upon further removal of both ΔΨ and ΔpH with valinomycin and the uncoupler FCCP (uncontrolled condition), little inhibition was seen. This differential inhibition of the controlled state results in a very high respiratory control ratio (RCR = 44) for the wild-type COVs with Zn2+ added (Table I) whereas the normal RCR for wild-type is expected to be from 6 to 10.