i LIIRARY Mlchlgan State ‘ . University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or betore date due. r—'———.——_—_———_———_= DATE DUE DATE DUE DATE DUE L; MSU Is An Affirmative Action/Equal Opportunity Institution cm”: THE USE OF MONOCLONAL ANTIBODIES AND A PROTEIN MODIFYING REAGENT TO STUDY THE INTERACTION OF CYTOCHROME C WITH CYTOCHROME C OXIDASE By Taha S.M. Taha A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1990 m ABSTRACT THE USE OF MONOCLONAL ANTIBODIES AND A PROTEIN MODIFYING REAGENT TO STUDY THE INTERACTION OF CYTOCHROME C WITH BEEF HEART CYTOCHROME C OXIDASE By TAHA S. M. TAHA In order to investigate the functional roles of the subunits of beef heart cytochrome oxidase, 48 monoclonal antibodies were generated against the native enzyme. Two subunit 11 specific antibodies inhibited 90% of the enzyme activity. Kinetic and binding studies suggested that the inhibition was due to blockage of cytochrome c access to the enzyme. The epitope was mapped to a 1.5 kDa peptide at the C—terminus of subunit II, close to the region proposed to contain the CuA center and the high affinity cytochrome c binding site. Treatment of the enzyme with the carboxyl modifying reagent 1— ethyl—3—(3,3’—trimethy1aminopropyl) carbodiimide (ETC) inhibited more than 90% of the activity and affected the electrophoretic mobility of subunit II. The inhibition was biphasic with 50% of the activity lost in the first 20 minutes, while it took 3 hours to achieve maximal inactivation. During the short, apparently first order, reaction that produced 50% inhibi was p: respor. indicat and the. I carboxy occurs a complex normal {GUEST l fining Daranu PhOSph (I) . Ingle Obsery. prEVi 0 CYIOQ}1 inhibition, it appeared that the high affinity phase of the biphasic kinetics was preferentially inhibited. It was not possible to identify specific carboxyls responsible for this effect due to the variety of modified species produced, as indicated by the multiple enzyme forms eluted from anion exchange FPLC and the considerable variability in their enzymatic activity. The inhibitory effect of ETC modification results from conversion of carboxyl groups into bulky positively charged residues. When modification occurs at the active site, the reagent likely inactivates the enzyme completely. Modification at other sites may produce enzyme forms with normal activities and some with altered Km values. The various altered forms were not isolated or characterization. FPLC purification of the native enzyme reduced the cytochrome c binding stoichiometry without a significant effect on the kinetic parameters, suggesting that a non—functional binding site (possibly due to phospholipids) was removed. These results support the hypothesis that a single cytochrome c interaction may account for the biphasic kinetics observed between cytochrome c and cytochrome c oxidase. Although the minimal epitope did not include any of the carboxyls previously implicated in cytochrome c binding, both antibody and cytochrome c strongly protected subunit 11 against ETC modification suggesting that the native epitope does overlap with the cytochrome c binding site, and that both domains occupy a major portion of the exposed subunit II surface. Dedicated to the memory of my wife (Afaf) iv l Shelagl in mm oxidase Eileex their 1113ij I agre 2“eel .‘ ACKNOWLEDGEMENTS I would like to take this opportunity to thank my major professor Dr. Shelagh Ferguson-Miller for guidance and support. She has been patient in introducing me to the controversial field of bioenergetics and cytochrome oxidase, in particular. Her moral support and encouragement during the rough times is greatly acknowledged. My thanks are also due to all the members of my guidance committee; Dr. John Wilson, Dr. Clarence Suelter, Dr. William Smith and Dr. Shauna Sommerville for their excellent advice, and their encouragement during the whole project. It is with great pleasure that I thank all my labmates; (Wendy, Eileen, Jonathan, Claudia, Jian Li, and John) for their critical comments, their wonderful lab humor, and the great parties. Wendy has played a major role in introducing me to the American culture, and Eileen has been a great help whenever I needed it. I am also indebted to Dr. Joe Leykem for his help with the HPLC and Dr. John Wang for allowing me to use his tissue culture facility. Dr. Doug Gage has helped me with the mass spectrometric analysis and Dr. George Yefchak taught me everything I needed to know about computers; their help is greatly appreciated. Finally I would like to thank all my Sudanese friends for making me feel at home away from home and for the authentic Sudanese dinners on Sunday nights. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF ABBREVIATIONS INTRODUCTION The Respiratory Chain: An overview Structure of Cytochrome c Oxidase Subunit Composition Prosthetic Groups Arrangement of the Subunits in the Membrane Enzymatic Functions of Cytochrome c Oxidase Kinetics of Cytochrome c Oxidation Proton Translocation The Role of Monomers and Dimers The Role of Individual Subunits Subunit II: Structure Function Relationship Antibodies to the Oxidase and its Subunits CHAPTER 1. EFFECT OF SUBUNIT II ANTIBODY ON CYTOCHROME OXIDASE ACTIVITY Introduction Experimental Procedures Materials Enzyme and Cytochrome c Purification Production of Monoclonal Antibodies SDS-PAGE and Western Blotting Separation of Monomers and Dimers of Cytochrome Oxidase Immunoprecipitation Effect of the Antibody on Cytochrome c Binding Miscellaneous Results Antibody Production and Characterization viii ix xi H 8§EEGUSQmNMH $$8888 SSEESB 8 Inhfi Efibc DlSC‘u: CHAJY BEYDL hnrodx Expenr llater Subur quht Sepa: Carbt Char Nter 31355 Result Punf Trim Chen Dlges ISCU§ ClLAE’ CITFC) Introc ~‘per Blats Syn: Inhibition of Enzyme Activity Effect of Subunit II Antibody on Cytochrome c Binding Discussion CHAPTER 2. LOCALIZATION OF THE ANTIBODY BINDING EPITOPE Introduction Experimental Procedures Materials Subunit II Purification Tryptic and V3 Digestion Separation of theTryptic Peptides Carboxypeptidase Digestion Cleavage of Cysteine Residues N-terminal Sequencing Mass Spectrometric Analysis Results Purification of Subunit II Tryptic and V8 Digestion of Subunit 11 Chemical Cleavage of Subunit II Digestion of Subunit II with carboxypeptidase Discussion CHAPTER 3. EFFECT OF ETC MODIFICATION ON CYTOCHROME OXIDASE Introduction Experimental Procedures Materials Synthesis of ETC Reaction of ETC with Cytochrome Oxidase Fluorographic Analysis of the ETC Modified Peptides Effect of ETC Modification on Cytochrome c Binding Protection of the Enzyme from ETC Modification by Cytochrome c Protection of the Enzyme from ETC Modification by Subunit II Antibody Results Characterization of the ETC Reaction with Cytochrome Oxidase Effect of ETC on the Structure and Function of Cytochrome Oxidase Effect of ETC on Cytochrome 0 Binding to the Oxidase Protective Effect of Cytochrome c and Antibody Against ETC Modification Discussion 51 3 assessaasessssse E SE §8 s as a $833338 8 Bl LU. SUMMARY BIBLIOGRAPHY viii LIST OF TABLES Arrangement of the subunits of cytochrome oxidase in the membrane Effect of dilution on the affinity of the antibody to the oxidase Effect of subunit II antibody on cytochrome c binding Effect of ETC modification on the enzymatic activity and cytochrome 6 binding to cytochrome oxidase ix Page 10 43 55 110 1o 11- 13- 14— LIST OF FIGURES Subunit composition of beef heart cytochrome c oxidase Proposed folding pattern of subunit II Immunoblot analysis of the subunit specificity of the antibodies Kinetic analysis of the effect of subunit II antibody HPLC separation of monomers and dimers of beef heart cytochrome oxidase A standard curve for gel filtration HPLC analysis of beef heart cytochrome c oxidase Effect of subunit II antibody on the activity of monomeric cytochrome oxidase Purification of subunit II by reverse phase HPLC SDS-PAGE and immunoblot analysis of the tryptic digest of subunit II HPLC purification and characterization of the tryptic peptides of subunit II Amino acid sequence of subunit II of beef heart cytochrome c oxidase SDS-PAGE and immunoblot analysis of the cystine cleavage peptides of subunit II Separation of the cystine cleavage products by HPLC SDS-PAGE and immunoblot analysis of the carboxypeptidase digest of subunit II Base 5 8382‘ 71 76 78 lb- 16 17- 15- 16- 17- 18- 1e 21- Time couse of the inactivation of cytochrome oxidase by ETC Protection of cytochrome oxidase from ETC modification by cytochrome c Kinetic analysis of cytochrome oxidase after short term modification by ETC SDS-PAGE and fluorographic analysis of the effect of ETC on the subunits of cytochrome oxidase FPLC profiles of the native and ETC-modified enzyme Rechromatography of the native and ETC- modified cytochrome c oxidase Kinetic analysis of the effect of ETC modification on the FPLC purified cytochrome oxidase Kinetic analysis of the native and ETC-modified cytochrome oxidase in phosphate buffer Long term ETC modification in the presence and absence of cytochrome 0 Analysis of the effect of ETC on subunit II and its tryptic fragments Protection by antibody and cytochrome c against ETC modification of subunit II xi 101 103 106 108 114 116 119 INTRODUCTION n 1 ° 1 When Hogeboom et al. (1948) reported the first successful isolation of intact mitochondria, it had already been established that these organelles contained the enzymatic machinery for such important processes as the tricarboxylic acid cycle, B-oxidation of fatty acids and the electron transfer chain (reviewed by Tzagoloff, 1982). In all respiring cells the reducing power available from food materials is collected in the form of reduced nucleotides which donate their reducing equivalents to the electron transfer chain located in the inner mitochondrial membrane. The electrons and protons are initially transferred to complex I and complex II of the respiratory chain, also known as NADH-ubiquinone oxidoreductase and succinate dehydrogenase respectively. Electrons from these two complexes are used to reduce ubiquinone which transfers them to cytochrome bc1, (complex III). From there the electrons move via cytochrome c to cytochrome c oxidase (complex IV) where they are utilized in the reduction of molecular oxygen to water. An electrochemical gradient is generated from the vectorial transfer of protons associated with these electron transfer events. This in turn drives the synthesis of adenosine triphosphate (ATP) catalyzed by the enzyme ATP synthetase, according to the now widely accepted chemiosmotic mechanism of energy coupling (Mitchell, 1961). The respiratory chain complexes have been the focus of research for many years due to their central role in energy transduction, but further interest has been generated recently with the diagnosis of numerous mitochondrial myopathies resulting from deficiencies in these complexes (Kim et al., 1987; Ichiki et al., 1989; Muller-Hacker et al. ,1989; Palca 1990). The energy conserving electron transfer complexes (I, III, and IV) share several important features. First, they are composed of a large number of subunits; thus complex I has at least 26 polypeptides (Heron et al., 1979), complex III has eleven polypeptides (Schaager et al ., 1986), and cytochrome oxidase has 13 polypeptides (Kadenbach et al.,1983). Second, these complexes are coded for by both mitochondrial and nuclear genomes; i.e. some of the subunits are encoded by the mitochondrial DNA while others are made in the cytoplasm and are imported into the mitochondria posttranslationally (Schatz and Mason, 1974). Thirdly, in all cases, the mechanism of energy transduction remains to be elucidated. Stmctnmfiflflcchmmflxidase Subunit composition : Unlike the two or three subunit enzyme from prokaryote sources (reviewed by Ludwig, 1987), mammalian cytochrome oxidase is very complex in structure (Kadenbach, 1983). This structural complexity, together with technical difficulties involved in the purification and analysis of the enzyme, has generated much conflicting data in the field. Thus, using the technique of polyacrylamide gel electrophoresis in the presence of SDS, mammalian cytochrome oxidase was initially separated into two (Chuang and Crane, 1971), three (Capaldi and Hyashi, 1972), four (Kierns et al., 1971), five (Shakespeare and Mahler, 1971), and six (Briggs et al., 1975) polypeptides. Although this variability could be attributed in part to the purification procedure used, it is clear that the resolving power of the electrophoresis system is also a major factor. A good example of this is the result of Downer et al., (1976), who showed that with a highlv rather the res.- was (on: second, small pf. Conseqw ES com: highly cross-linked urea gel the enzyme could be separated into seven rather than six subunits. The idea of the "seven subunit" enzyme was largely accepted among the researchers in the field for two reasons: first, because the yeast enzyme was found also to have seven subunits (Poyton and Schatz, 1975a), and second, because mild proteolysis could effectively remove some additional small polypeptides without loss of enzyme activity (Ludwig et al., 1979). Consequently some workers considered the easily-removable polypeptides as contaminants (Ludwig et al.,1979; Saraste et al., 1981; Saraste, 1983). In 1981, the group of Kadenbach reported a high resolution urea gel system which separated the mammalian enzyme into twelve (Kadenbach and Merle, 1981) and then thirteen polypeptides (Kadenbach et al., 1983). A similar polypeptide pattern was seen in the enzyme purified from different mammalian species and tissues using various purification methods. The group also demonstrated that an antibody to subunit IV was able to immunoprecipitate all the subunits from Triton X-100 solubilized mitochondria (Merle et al., 1981). In addition, the polypeptides were found to exist in a 1:1 stoichiometric ratio, and to have different amino acid compositions (Merle and Kadenbach, 1980; Meincke and Buse, 1984; Buse et aL,1985) It is now generally accepted that the mammalian enzyme is composed of thirteen subunits (Kadenbach et al., 1983; Sinjorgo et al., 1987; Takarniya et al., 1987). The polypeptides have been purified and sequenced (see Buse et al., 1987 for a recent review), except for VIIb, but some confusion still remains regarding the nomenclature of the polypeptides. Figure 1 shows the subunit composition of our beef heart enzyme together with some of the most commonly used nomenclatures. Unless otherwise Kadm ‘etal Figure 1. Subunit composition of beef heart cytochrome c oxidase. The purified enzyme (30 ug/lane) was subjected to SDS- PAGE following the method of Kadenbach et a1 (1983). The right hand lane shows the true molecular masses of the different subunits as determined by amino acid sequencing. The three commonly used nomenclatures are are given. In each case the subunits are identified by roman numerals according to their electrophoretic mobility. The nomenclature of Capaldi et a1 (1987), uses the letters (Mt) or (C) depending on whether the subunit is encoded in the mitochondrial or the nuclear genome; For subunits that do not have equivalents in yeast the one letter code for the first three amino acids of the subunits are used. The nomenclature of Kadenbach et al. (1983) is used throughout this study. Kadenbach g, Buse et a1 Cfipaldi ~ Molecular -etal (1933) f (1987) et (1987) mass (Dal) I -b - I Mtl — 56993 n _ _ n MtII— ‘ ‘ —26049 111 — :55— m \MtIII— —29918 _ Iv CIV— . — 17153 a V V c b VI VI— 3 7 -10670 8 VII AED — —9419 VI ., vm ASA— —10068 c 1x STA- —8480 vn a — ' *‘ —x C — .3 —6244 b,c — .—xr,xn IHQ.Cvm— —sooo/5541 _ VIII _ ....... —xm CIx— —4962 Figure 1 Prosthe misdir- The en recent) Caug? stated the nomenclature of Kadenbach et al (1983) will be used throughout this study. Prosthetic groups: Earlier studies by Keilin and Hartee (1939) showed that cytochrome c oxidase contained two spectrally distinct hemes (a and a3). The enzyme was also reported to have two copper atoms (Beinert, 1966), and recently Zn (Einarsdottir and Caughey, 1984) and Mg (Einarsdottir and Caughey, 1985) were reported as integral components of the enzyme. More complete information regarding the metals associated with cytochrome oxidase has come from studies using inductively-coupled plasma atomic emission spectroscopy (ICP—AES). This technique allows the determination of a variety of metals on the same sample, allowing accurate ratios to be determined. However, the error for any particular metal can be up to 20% (Yewey and Caughey, 1988). Einarsdottir and Caughey (1985) applied the technique of ICP-AES and obtained a metal ratio of 2.5 Cu/ 2 Fe/ 1 Zn/ 1 Mg per beef heart oxidase monomer. These results were consistently seen with several different enzyme preparations including a crystalline form of the beef heart enzyme (Y oshikawa et al.,1988). The group of Caughey interpreted the data to mean that the enzyme exists as a dimer with each monomer containing two hemes and two coppers, while the additional copper is located at the junction between the two monomers (Yewey and Caughey, 1988). In contrast, the results of Steffens et al. (1987) and Bombelka et al. (1986) indicate that the monomer contains a metal ratio of 3 Cu:2 Fezl Zn:1 Mg. In their analysis the metal content is determined with reference to the protein content not assuming an exact iron content of two as is done by redn:cti renew y zinc in: With p result: bindir hasle assoc Afro Cry»;- Yewey and Caughey (1988). This difference may account for the disagreement in their data. The role of the two hemes and two copper atoms as oxidation reduction centers is well documented (see Wikstrom et al., 1981 for a review), but the function of the other metals is unknown. Some evidence for zinc involvement with proton translocation has been obtained from studies with p-hydroxy mercuribenzoate which causes displacement of the Zn and results in loss of proton pumping (Nilsson et al., 1988). Evidence for a binding site for ATP on cytochrome oxidase (Hiither and Kadenbach, 1986), has led to the proposal that the magnesium in the enzyme may be associated with that site. Arrangement of the subunits in the membrane: Based on X-ray crystallographic data, it is concluded that mitochondrial cytochrome c oxidase is asymmetrically inserted in the membrane. The enzyme has a "Y" shaped structure, with the tail of the "Y" extending in the cytoplasmic side and two arms penetrating the membrane and accessible to the matrix side (Capaldi et al., 1987). Recent studies on two dimensional vesicle crystals of the enzyme were conducted by Valpuesta et al. (1990). In this higher resolution analysis the enzyme appears to be packed as a dimer in the membrane with a larger portion of the protein (tail of the "Y") extending into the lumen of the vesicles (equivalent to the intermembrane space) and a less obvious division between the two arms of the "Y". The arrangement of the individual subunits with respect to each other and with respect to the membrane has been investigated using three different approaches. In one approach lipid soluble probes with photoaf‘i contact ' Gutweni I, ll, lll. submi‘ mean t‘r. has ex 9 Similar 1 sub uni Tra cy, photoaffinity crosslinking capacity were used to identify the subunits in contact with the membrane (Bisson et al.,1979; Prochaska et al.,1980; Gutweniger et al., 1981; Bisson et al., 1982). These studies showed subunits I, II, III, and VII to be the main site of labeling with these probes while subunits IV, V, and VI showed very low or no labeling at all. This could mean that these subunits are not deeply inserted in the membrane. However, it is also possible that they are shielded from labelling by other subunits. A second approach was aimed at studying the accessibility of the subunits to the two sides of the membrane (Eytan et al., 1975; Chan and Tracy, 1978; Ludwig et al., 1979; Zhang et al., 1984; J arausch and Kadenbach, 1985b; Zhang et al.,1988). Vesicles with right-side out orientation such as mitoplasts and in-side out orientation such as submitochondrial particles were used as the enzyme source. The vesicle preparation was treated with a membrane impermeable covalent modifying reagent, and the effect of the reagent on the individual subunits was investigated by sodium dodecyl sulfate polyacrylamide gel electrophoresis. The results from these studies agreed remarkably well on the orientation of subunits I-III, (see Table 1). It is difficult to draw conclusions for the smaller subunits (IV-VIII) because in most of these studies the electrophoresis system used to analyze the results was not powerful enough to resolve all the smaller polypeptides. The only information regarding the orientation of these small subunits (Table 1) comes from the studies by J arausch and Kadenbach (1985b), and Zhang et al. (1988), who used a highly resolving gel system to identify all thirteen subunits (Kadenbach et aL,1983) Table 1. Arrangement of the oxidase subunits in the membrane. The topography of the different subunits in membranous cytochrome c oxidase was investigated using chemical reagents that specifically label hydrophobic or hydrophylic domains. The accessability of the subunits to chemical modification, antibody binding, or proteolytic cleavage was determined using vesicles with known sidedness. The relationship between the different subunits was also studied using cross-linking reagents and two dimensional SDS-PAGE. The sources of the information are listed below but the letter (i) is not used to avoid confusion with (i) a = Prochaska et al. (1980) b = Zhang et al. (1988) c = Bisson et al. (1979) d = Ludwig et al. (1979) e = Gutweniger et al. (1981) f = J arausch and Kadenbach (1985b) g = Eytan et al. (1975) h = Zhang et al. (1984) j = Kornblatt and Lake ( 1980) k = Briggs and Capaldi (1977) 1 = J arausch and kadenbach (1985a) 10 .6qu u. >~3=no3aon ea Go 8898." «5:53. .o :8 5059.25. 55. §§§o8 38838 85 :5 83.30. om so 6885 8688 8 5a 0.8% +8 9. a. <3. <5. 55.. 5. 588583888 38.88 +c +~ <1. AVE. <9? <5. 55— 8688 8 En Bmfix man on So E +- +~ +G 4.. [I l B .5? +o éé n+c +~ .1. in K 9:5. <3. <5. 55— QA:

90% immunoprecipitation. 46 400 m 0 2 €1on72 7.392» 100 150 50 nus") Figure 4 47 Figure 5: HPLC separation of monomers and dimers of beef heart cytochrome oxidase. The separation was done using a combination of a Zorbax G-250 column (25 X .94 cm) connected in series with a G-450 column of the same dimensions. The buffer was composed of 100 mM Tris- H01, 300 mM NaCl, 1 mM EDTA and 2 mM lauryl maltoside pH 7.5. In A and B 20 ILL containing 0.39 nmoles of oxidase were injected to the system. In C, the dimer peak from B was reinjected. The flow rate was 0.8 ml per minute. A420 .04 .02 48 JU‘L l l I? 25 I? 25 c .008 - .004 0 I? 25 RETENTION TIME (Minutes) Figure 5 Figure 6. A standard curve for gel filtration HPLC analysis of beef heart cytochrome c oxidase. Gel filtration chromatography was performed as described in the methods section, using 20 uL samples of thyroglobulin (669 X 103), ferritin (440 X 103), catalase (232 X 103), aldolase (158 X 103), ovalbumin (44 X 103), myoglobin (17.2 X 103), and cytochrome c (12.4 X 103) all masses are given in Daltons. The open squares represent the monomer and dimer forms of cytochrome c oxidase with apparent molecular masses of (310 X 103), and (550 X 103)Da1tons respectively. 50 1000 1111 500 100 M,.(x10-3) l I ll 1 50 1 10111111 I 1 1111 1 18 ' . retention time Figure 6 25 (min.) 32 51 In our hands, however, rechromatography of the dimer one hour after its isolation resulted in a 50:50 mixture of monomers and dimers ( Figure 5C). Hakvoort et al. (1985) also showed that conversion of monomer to dimer could be achieved by incubation of the enzyme with cytochrome c and ascorbate. This result was reproduced in the present study. However, it took 42 hours to obtain 50% dimer starting with an enzyme that was already 10- 25% in the dimer form (not shown). As the hypothesis would predict, a completely monomeric enzyme should be completely inhibitable by the antibody. The HPLC profile of one enzyme preparation that was predominantly in the monomer form is shown in Figure 5A. The subunit H-specific antibody inhibited 90% of the activity of that enzyme (Figure 7), supporting the postulate that incomplete inhibition is related to the dimer content of the enzyme. Moreover, purification of this enzyme by anion exchange FPLC resulted in an enzyme that was more than 95% inhibitable by the antibody. This could mean that the remaining 10% of activity in the original enzyme was due to presence of higher aggregate forms. Effect of subunit H antibody on cytochrome c binding. Since subunit II contains a high affinity binding site for cytochrome c (Erecinska 1977; Bisson et al., 1980, 1982; Bisson and Montecucco, 1982; Kadenbach and Stroh, 1984), it appeared likely that blockage of substrate binding was the cause of inhibition of cytochrome oxidase activity. The effect of subunit II antibody on cytochrome c binding was therefore investigated using gel filtration chromatography as described by Ferguson-Miller et al. (1976). The enzyme/antibody ratio was chosen so that 95% of the enzyme was bound to the antibody as determined by immunoprecipitation. As shown in Table 3 Figure 7. Effect of subunit II antibody on the activity of monomeric cytochrome oxidase. The enzyme (0.02 nmoles aa3).was incubated with the indicated amounts of subunit II antibody or control (antibody against rat brain hexokinase) for one hour. The total volume was 30 uL made up with 25 mM Tris-HCl pH 7.5 containing 300 mM NaCl, 1 mM EDTA and 2 mM lauryl maltoside. The remaining enzyme activity was assayed polarographically as described in the Methods section. For each time point the remaining activity with the subunit 11 antibody was determined as a percent of the activity obtained with the hexokinase antibody at the same time. 120 100 O) O l % activity remaining 100 200 Incubation time (minutes) Figure 7 Table 3: Effect of subunit II antibody on cytochrome c binding. The enzyme (10 uM aa 3) was incubated with the subunit II antibody or control antibody at a 1:2 molar ratio. This was followed by the addition of cytochrome c (30 M) for 10 minutes. 50 uL of the mixture were chromatographed on a Sephadex G-75 as described in the methods section. A negative control with no enzyme added was done in parallel and was used to determine the non-specific binding of cytochrome c to the antibody. Table 3. Effect of subunit II antibodies on cytochrome c binding Incubation Conditions Cytochrome c / Oxidase 1 ratio observed corrected Enzyme + cytochrome c 1.3 1.3 Enzyme + control antibody + cytochrome c 2.5 1.1 Enzyme + subunit 11 antibody + 1.4 0.0 cytochrome c Control antibody + cytochrome c 1 4 0 0 the subunit H specific antibody blocked a significant portion of cytochrome c binding, suggesting that the antibody is binding at or close to the cytochrome c binding domain. No such effect was seen with the subunit IV antibody. 57 DISCUSSION In order to explain the incomplete inhibition obtained with subunit II antibodies, we postulated that dimeric forms of the oxidase may bind only one antibody molecule resulting in complete immunoprecipitation but only 50% inhibition of the enzyme activity (Taha and Ferguson-Miller, 1987). This possibility was of interest because the remaining activity was biphasic in character, and if it were due to the activity of one monomer within the dimer, it would rule out the idea of cooperativity between monomer sites as the source of the biphasic kinetic behavior (Nalecz et al., 1985). Support for the postulate comes from the fact that the dimer content of the enzyme used in these studies (> 25% as determined by HPLC) could account for the incomplete inhibition, considering that some percentage (10-15%) of the enzyme-antibody complex appears to dissociate in the dilute assay conditions. Moreover, monomeric cytochrome oxidase can be 90% inhibited with the antibody, while 65% inhibition was obtained with the partially dimeric enzyme. More definitive support for the idea would be obtainable if the enzyme could be converted more completely to the dimer form. Attempts to make a dimer enzyme were not successful because the conditions required to obtain a significant amount of dimer from our enzyme (incubation with ascorbate and cytochrome c for days at room temperature) are not ideal for enzyme stability. Furthermore, contrary to the results of Hakvoort et al. (1985) the isolated dimer tends to reequilibrate between the monomer and dimer forms. This phenomenon has also been recently reported by Michel and Bosshard (1989). The fact that a subunit II specific antibody can completely inhibit the electron transfer activity of cytochrome oxidase and also blocks the access of 58 cytochrome c to the enzyme, supports earlier reports that this subunit contains a binding site for cytochrome c (Bisson et al., 1980; 1982; Millett et al., 1983). Furthermore, the current results indicate that this site is active in electron transfer. In contrast, Miiller et al. (1988a,b) reported that subunit H of the Paracoccus enzyme could be completely removed by proteolytic degradation, with retention of the spectral characteristics as well as a native Km for cytochrome c. This latter finding, implying that subunit II is not essential, is inconsistent with our conclusion. Whether the binding of one (Anatalis and Palmer, 1982; Nalecz et al., 1985; Brzezinski and Malmstrom, 1986) or two (Ferguson-Miller et al., 1976; Errede and Kamen, 1978; Speck et al., 1984) cytochrome c molecules on the oxidase is necessary for producing the observed biphasic kinetics of cytochrome oxidase is an issue that remains to be clarified. The initial observation that the antibody completely occludes cytochrome c from the enzyme under conditions where a stoichiometry of 1.4 cytochrome c/oxidase was observed in the absence of antibody does not definitely answer this question , but does suggest that if more than one binding is important, both interactions occur on, or in close proximity to, subunit II. Our results provide significant evidence against the idea that biphasic kinetics are the result of negative cooperativity between monomers associated in a dimer. CHAPTER 2 LOCALIZATION OF THE ANTIBODY BINDING EPITOPE The use of antibodies to study the structure and function of enzymes was significantly improved after the development of the technique of somatic cell hybridization to produce monoclonal antibodies (Kohler and Milstein, 1975). Using this technique a stable hybrid cell line can be obtained that produces unlimited amounts of antibody to a specific epitope. Ifthe epitope happens to be at or close to the catalytic site of the enzyme, then the binding of the antibody may result in the loss of enzyme activity and by determining the location of the epitope within the enzyme structure, conclusions can be drawn regarding structure-function relationships (Wilson, 1987). The most common method for epitope mapping is called the direct method and is often used for mapping sequential epitopes. The method involves cleavage of the protein into small fragments using chemicals or proteolytic enzymes. Alternatively, pieces of the DNA that code for portions of the antigen can be expressed as fusion proteins, e.g. with B-galactosidase (Nunberg et al., 1984). The fragments are then physically separated by electrophoresis or HPLC and the reactivity of the fragments with the antibody can be examined by a suitable method. The most commonly used detection methods include immunoblotting (Polakis and Wilson, 1985; Ratman et al., 1986; ), ELISA (Djavadi-Ohaniance et al., 1984 ), or by the ability of the fragment to compete with the native enzyme for the antibody (Weldon et al., 1983). The location of an antibody epitope can be examined in the native form of the enzyme with reference to a certain structural feature within the enzyme molecule. This structural feature could be the epitope for a second antibody (Wilson and Smith, 1984; Hadikusumo et al., 1986). Thus the ability of one antibody to block the binding of an another antibody would indicate that the epitopes recognized by the two antibodies are within a 30 A X 20 A field, the area covered by the binding of an antibody (Amit et al., 1986). Since such studies utilize native enzymes, the actual sequences recognized by the two antibodies may or may not be as close. Alternatively, the ability of the antibody to protect certain residues in the protein from chemical modification can be investigated (Burnens et al., 1986). If a certain residue or group of residues are important for catalytic function, then the ability of the antibody to prevent loss of this function can be assayed. Moreover, if the residues in question are important for substrate binding then comparison of the protective effect of the antibody to that of the substrate can be used to compare the antibody epitope to the substrate binding site. Paterson et al. (1990) used hydrogen-deuterium exchange labelling and two dimensional nuclear magnetic resonance to map the binding area of a monoclonal antibody against cytochrome c. The epitope was found to be of the discontinuous type (conformational epitope) and the area covered by the antibody was about 750 square angstroms in good agreement with earlier results by Amit et al. (1986) The experiments described in this chapter are aimed at the localization of the binding domain for the subunit II antibody by peptide mapping using chemical and proteolytic cleavage methods. The ability of the antibody and cytochrome c to protect the enzyme against ETC modification was investigated in the next chapter in an attempt to define the antibody epitope with respect to the high affinity cytochrome c binding site. 61 EXPERIMENTAL PROCEDURES Materials. 5,5'-dithio-bis-(2-nitrobenzoic acid) (DTNB) and trypsin (TPCK treated) were from Sigma (St Louis, MO). Yeast carboxypeptidase (CPY), and endoproteinase Glu-C (V8 protease) were from Boeringer Mannheim. All HPLC solvents were from Burdick and Jackson (Muskegon, MI). Aminophenyl thioester (APT) blotting paper was purchased from Bio-Rad Laboratories (Richmond, CA.) and was activated with nitrous acid immediately before use. Subunit II purification Subunit II was isolated from beef heart oxidase using reverse phase HPLC. Beef heart cytochrome c oxidase (originally in cholate) was made 4 mM in lauryl maltoside and dialyzed for six hours at 4 °C against 20 mM phosphate buffer pH 7. The cholate free enzyme was diluted 1:10 in 0.1% trifluoroacetic acid (TFA). 2-3 mL of the enzyme solution (10 mg protein) were filtered through a 0.45 um filter and injected into a Vydac preparative butyl column (25 X1 cm). The peptides were eluted with a linear gradient from 0.1% TFA in water to 0.1% TFA in acetonitrile in 45 minutes. The flow rate was 5 mL per minute. For small scale isolation of subunit II, a (25 X 0.5 cm) column from the same manufacturer was used. The maximum protein load on this column was 1 mg and the flow rate was 0.8 ml per minute. Tryptic and V8 digestion The subunit II containing fraction was dried under a stream of nitrogen and the peptide (0.8-1 mg) was dissolved in 600 uL of the digestion buffer which composed of 100 mM ammonium bicarbonate, 1 mM calcium chloride, and 0.2% SDS, pH 8.2. To 500 uL of the 6‘2 subunit II solution, 5 uL (12.5 ug) of TPCK treated trypsin were added and the mixture was incubated at 37 °C for 1-3 hours in a rocking shaker. The digestion was stopped by incubating the digest for 5 minutes in a boiling water bath. Staphylococcal V3 protease digestion was done at an enzyme to subunit II ratio of 1:100. The digestion was done at 37 °C for 1-4 hours. Separation of the tryptic peptides This was done by reverse phase HPLC using a Water's Model 600 multi-solvent delivery system connected to a Model 712 Water's intelligent sample processor and a Model 738 UV monitor. The separation was done on a Vydac C3 column following the conditions of Millett et.al. (1983). Alternatively the digestion products were electroblotted to an Immobilon membrane. The membrane was stained with amidoblack and the band containing the peptide of interest was cut out and sent directly for sequencing. Carboxypeptidase digestion This was done essentially as described by Klemm (1984), with minor modifications. Purified subunit II (1 mg protein) was dissolved in 500 uL of 50 mM citrate buffer, pH 6.5, containing 0.1% SDS. Digestion was started by the addition of 20 ug of yeast carboxypeptidase. At the indicated times, 50 uL aliquots were withdrawn and immediately mixed with 10 uL of glacial acetic acid to stop the digestion. The samples were subjected to SDS-PAGE and immunoblotting to determine the reactivity of the resulting fragments with the antibody. Cleavage of cysteine residues Subunit II was chemically cleaved using the method of N efsky and Bretscher( 1989). The dried protein was dissolved in 100 mM Tris-HCI pH 8.0, containing 8 M urea and 5 mM dithiothreitol. The cysteine residues were modified by incubation with 15 mM DTNB for 15 minutes. The sample was exhaustively dialyzed against 50% acetic acid and then dried by blowing nitrogen gas. Cyanolation and cleavage of the 2- nitro-5-thiobenzoic acid was performed by dissolving the protein in 100 mM sodium borate buffer, pH 9.0, containing 8 M urea and 1 mM potassium cyanide. The cleavage reaction was done at 37 0C for 12 hours, and was stopped by the addition of excess dithiothreitol (100 mM final concentration). The cleavage products were diluted 20 fold with a 0.05% solution of SDS and concentrated using a Centricon—3 concentration device. This treatment removed most of the urea while the peptides were retained by the membrane. N-terminal sequencing This was done on a Model 477 pulsed-liquid protein peptide sequencer, with an on-line Model 120 PTH-amino acid analyzer (Applied Biosystems). Mass spectrometric analysis . The HPLC purified peptide (20-50 ug) was dried by blowing nitrogen gas and redissolved in 10 uL of of a mixture of water and isopropanol 20:80 respectively containing 0.1% trifluoroacetic acid. The peptide solution ( 2 uL) was carefully mixed with a matrix composed of dithiothreitol and dithioerythritol at a ratio of 5:1, on a probe tip prior to analysis by fast atom bombardment mass spectormetry (FAB MS). The analysis was done on a JEOL model HX-110 double focusing mass Spectrometer. Xenon gas was used in the FAB gun to produce the beam of accelerated atoms (energy 6 KeV). The accelerating potential of the instrument was 10 KV. The mass spectrometer was scanned at a rate of 2 minutes from 0-6000 Daltons, and the data was acquired in a single scan. RESULTS Purification of subunit II. Since the subunit II antibody inhibits the enzyme activity while preventing the access of cytochrome c to the enzyme, it was important to localize the the antibody epitope, in order to determine its relationship to cytochrome c binding domain that has been partially defined on subunit II. This required purification of subunit II, dissection by proteolytic and chemical methods and probing the resulting fragments with the antibody. Earlier attempts to purify subunit II included gel filtration chromatography in the presence of SDS (V erheul et al., 1981), and preparative polyacrylamide gel electrophoresis (Spiker and Isenberg, 1983). Both methods resulted in a pure protein. However, the high concentration of SDS used in the isolation (2-3%) made it difficult to do any proteolytic cleavage and attempts to reduce the level of detergent resulted in precipitation of the protein. Recently a reverse phase HPLC method was reported where most of the subunits of cytochrome oxidase can be isolated in a pure form, (Robinson et al., in press) As shown in Figure 8, subunit II is completely separated from the rest of the enzyme subunits. The purity of subunit II was determined by sodium dodecyl sulfate polyacrylamide gel electrophoresis (see below). This purification method provided a pure protein at higher yield than gel filtration chromatography. More importantly, since the protein is in acetonitrile it can be easily recovered by evaporating the solvent. Another added advantage of the HPLC purified subunit II is that it is readily soluble in 0.2% SDS, which makes the protein suitable for proteolytic digestion and chemical modification (see below). The protein was collected in tubes containing ammonium hydroxide to avoid Figure 8. Purification of subunit II by reverse phase HPLC. Lauryl maltoside dispersed cytochrome oxidase was diluted 1:10 in 0.1% trifluoroacetic acid, and loaded to a C4 column (25 X 0.5 cm). The enzyme subunits were eluted with a linear gradient going to 0.1% trifluoroacetic acid in acetonitrile in 60 minutes. Subunit II, indicated by (II) in the Figure is collected in a tube containing ammonium hydroxide to avoid acid hydrolysis. 0.30 66 100 I I I/ 0: 1’ '— 2 O 1’ E I 402 1” I- 2 M II 0 0: LL] - a l 1 o 30 60 ELUTION TIME (minutes) Figure 8 acid hydrolysis by the trifluoroacetic acid in the HPLC solvents. Tryptic and V8 digestion of subunit H. Tryptic digestion of subunit II resulted in a number of peptides ranging in molecular masses from 20 kDa to less than 5 kDa as determined by SDS-PAGE (Figure 9 lane 1). Two peptides with molecular masses of 15 kDa and 10 kDa reacted with the subunit II antibody (lane 3). Purification of the two peptides was done by reverse phase HPLC using a Vydac C3 column as shown in Figure 10 (top panel). The purity of the two peptides was determined by SDS-PAGE and immunoblotting (Figure 10 lower panel). Each of the two peptides eluted in more than one fraction; however, fraction 2 provided a pure preparation of the 15 kDa peptide while fraction 3 contained the 10 kDa peptide slightly contaminated with the 15 kDa peptide as shown in the immunoblot. Nevertheless, the purity of the two peptides was suffecient to allow for six cycles of N-terminal sequencing up to . The sequence of both peptides was confirmed by subjecting the digest to SDS-PAGE followed by Western blotting to an Immobilon membrane. The bands corresponding to the two peptides were sequenced directly from the Immobilon membrane. The N- terminal sequence of the 15 kDa peptide was found to be Thr—Met—Gly—His confirming cleavage at Lys 98 and that of the 10 kDa peptide was Leu—Leu— Glu-Val confirming cleavage at Arg 134 (Figure 11). Initial attempts to digest subunit II with V3 protease (specific for Asp and Glu residues) resulted in a large number of poorly separated peptides. However, when the enzyme was treated with a carboxy] modifying reagent, the purified subunit II became resistant to proteolysis at the modified sites and fewer proteolytic products were obtained . The smallest peptide that Figure 9. SDS-PAGE and immunoblot analysis of the tryptic digest of subunit II. Subunit II was subjected to tryptic digestion as described in the Methods section. The tryptic products were dried by blowing nitrogen and redissolved in sample buffer containing 5% v/v B- mercaptoethanol. Lanes 1 and 2 show the silver stained gel of the tryptic products and the undigested subunit II respectively, while lanes 3 and 4 show the immunoblot of the digest and the intact subunit II probed with subunit II antibody. The molecular weight standards used were the cyanogen bromide fragments of myoglobin with molecular masses of 17.2, 14.6, 10.8, 8.2, and 2.6 kDa respectively. 17.2 — — 15K 14.6 — I 10.8 — . hub - — 10K 8.2 — 2.6 .— Figure 9 THIS PHOTOGRAPH WAS PRODUCED BY MSU I lNSTRUCTlONAL MEDIA CENTER (517,) 353-3960 MSU rsan Affirmative A530” Equal 0000mm“)! 1nstmmon 70 Figure 10. HPLC purification and characterization of the tryptic peptides of subunit II. The purified subunit II (800 ug protein) was dissolved in 0.6 mL of 200 mM ammonium bicarbonate buffer pH 8.2, containing 0.2% SDS. The protein was incubated with trypsin at a 20:1 ratio in the presence of 1 mM CaC12. After a 3 hour incubation at 37 0C, the digestion was stopped by incubation in a boiling water bath for 5 minutes. 200 uL of the products were chromatographed on a C8 column (top panel). The peaks were analyzed by SDS-PAGE followed by silver staining (not shown), or immunoblotting with the subunit II antibody (bottom panel). The lane numbers 1-4 in the immunoblot correspond to the peaks in the HPLC profile, while lane (0) denotes the undigested subunit II. 71 0.032 - 0.016 -— 0.0 — 0 60 1 20 Elution Time (minutes) Figure 10 Figure 11. Amino acid sequence of subunit II of beef heart cytochrome c oxidase. The cleavage sites by trypsin (1), V3 protease (11) and cysteine specific chemical reaction (V) that resulted in immunogenic peptides are indicated by the symbols. The shortest peptide obtained from enzymatic digestion that reacted with the antibody is underlined, while the minimum immunogenic peptide that resulted from chemical cleavage is double underlined. Also shown in bold letters are the carboxyls that may be involved in cytochrome c binding. The starred residues have been suggested to be ligands for the CuA center. Peptides 99-227 and 135-227 resulted from tryptic cleavage. Peptide 138-227 resulted from V3 protease digestion while peptides 196-227 and 200-227 were produced by chemical cleavage D 33* Q D A I V F L T H A I I I N N Y T D P G E I R M L K T G L Y V L E E L 20 r. y 40 Q s so p s so T M 100 D s 120 LN 140 v L 160 L N 180 7* r g 200 F E 220 M L G T L T T I L M D s Y E L E M s L s R * M P L 227 Figure 11 74 reacted with the antibody had a molecular mass of 10 kDa. The N-terminal sequence was Val—Asp—Asn—Arg resulting from cleavage at Glu 137. Chemical cleavage of subunit II. The proteolytic cleavage results showed the antibody epitope to be in the C-terminal part of subunit II. However, no information could be obtained as to whether the identified peptides contain the entire C-terminal of the protein. Attempts to determine the total mass of the peptides by mass spectrometry were not successful; hence an alternative cleavage reaction was attempted where subunit II was dissected with the cysteine specific reagent DTNB in the presence of cyanide and high pH (see Methods section for details of the procedure). Figure 12 shows the resulting peptides as resolved by SDS-PAGE. Aside from the uncleaved subunit II, two peptides at the range of 22-23 kDa, and a small peptide with a molecular mass of less than 5 kDa were obtained. Only the small molecular weight peptide was recognized by the antibody as shown in the immunoblot (lane b). This peptide was purified by reverse phase HPLC as Shown in Figure 13. The molecular mass of this peptide was determined by mass spectrometry to be 3.2 kDa corresponding to residues 200-227. Digestion of subunit II with carboxypeptidase. The importance of the C- terminal end of the protein for antibody binding was further explored by subjecting subunit II to carboxypeptidase digestion. Figure 14 shows the Coomassie blue staining pattern (panel A), and the reactivity with the SUbunit II antibody (panel B) of the carboxypeptidase digests taken at the indicated times. The appearance of a band below subunit II was indicative 0f C-terminal cleavage. Moreover, this new band was not detectable by the antibody indicating that removal of a peptide considerably less than 3.2 kDa (Compare lanes 0 and 20 in panel A) from the C-terminal of the protein had 75 Figure 12. SDS-PAGE and immunoblot analysis of the cysteine cleavage products of subunit II. Purified subunit II was dissolved in 600 uL of modification buffer (100 mM Tris, 4 M urea, 0.1% SDS, pH 8.0) containing 5 mM dithiothreitol. Modification of the cysteine residues was done for 15 minutes at room temperature, by the addition of 15 mM 5,5'-dithiobis-2-nitrobenzoic acid. The excess reagent was removed by dialysis against 50% acetic acid, and the protein was recovered by lyophilization. The cleavage reaction was done for 12 hours at 37 0C, in glycyl glycine/borate buffer pH 9.0 and was initiated by the addition of potassium cyanide to a final concentration of 1 mM. lane a shows the Coomassie blue-stained gel of the cleavage product while the immunoblot with the subunit II antibody is shown in lane b. 76 22.8 — 3.2 _ ’ Figure 12 Figure 13. Separation of the cysteine cleavage products by HPLC. Subunit II (500-600 ug) was subjected to chemical cleavage as described in the methods section. 250 uL of the digest were injected to a C4 column equilibrated with 0.1% TFA. The peptides were eluted with a linear gradient from 100% A (0.1% TFA in water) to 100% B (0.1% TFA in a 50:50 mixture of isopropanol and acetonitrile) in 120 minutes, at a flow rate of 250 uL per minute. The first 20 minutes of the profile were omitted because of the high absorbance due to the unreacted 2- nitro-5-thiobenzoic acid. The arrow indicates the position of the immunogenic peptide (p). 78 A2l4 ..N 0.0 . _ . Figure 13 no .3 mrcjoz .225 $12.33 _No Figure 14. SDS-PAGE and immunoblot analysis of the carboxypeptidase digest of subunit II. The purified subunit II was dissolved in 50 mM sodium citrate buffer pH 5.7 containing 0.1% SDS. Digestion with yeast carboxypeptidase was carried out as described in the Methods section. The digest products were subjected to electrophoresis (panel A) and immunoblotting with subunit II antibody (panel B) for the indicated times (all in minutes). Lane c represent the cysteine cleavage products of subunit II. The major peptide produced by carboxypeptidase digestion is labelled (cpy). c302010521 302010521 25 _ epv ——-> 59d 23 _ Ln.’ lie bufer St 11 the acted to all] lmes 1.2-1 'age ced by Figure 14 81 caused loss of immunogenicity. This result indicated that the C-terminus of subunit II is necessary for the antibody binding. Furthermore, it indicated that the two tryptic peptides that reacted with the antibody likely extend to the end of subunit II as shown in Figure 11. In an attempt to correlate between the loss of immunogenicity and the type of amino acids released by carboxypeptidase, the digestion products at different times were subjected to amino acid analysis. This experiment was not successful possibly because of incomplete derivatization of the free amino acid, or more likely that there is an internal cleavage that resulted in the removal of the peptide as one piece. Nevertheless, it can be estimated from the SDS-PAGE analysis that removal of about 1.5 KDa from the C-terminal of subunit II is accompanied by loss of immunogenicity implicating less than 14 residues of the C-terminal in the minimal epitope. Another piece of information regarding residues important in the epitope comes from comparing the subunit H sequences from different species. The amino acid sequence of the 10 KDa tryptic peptide from beef heart subunit II was compared with the corresponding sequences from rat liver, mouse liver, yeast, and maize. Of the eleven mismatches between beef heart and rat liver peptides there was one case where a charged residue became neutral or oppositely charged. Lys 221 in the beef heart was replaced by asparagine in the rat liver, tryptophan in yeast and aspartate in the maize sequence. All these enzymes showed weak reactivity with the antibody to beef heart subunit II. This suggests that lysine 221 may be important for the antibody binding. It should be noted that a complete antibody binding domain is undoubtedly much larger than a minimum epitope (Amit et al., 1986; Paterson et al., 1990); however, the binding affinity can be significantly altered if the charge of one residue is neutralized or reversed as is the case here. DISCUSSION Subunit II is asymmetrically inserted in the inner mitochondrial membrane with the amino terminal half of the protein spanning the membrane while the carboxy terminal half extends into the intermembrane space (Bisson et al., 1982; Millett et al., 1983; Wikstrom et al., 1985; Holm et al., 1986)). An important outcome of this structural feature is that in the purified enzyme the carboxy terminal half of subunit II is more accessible to the surrounding environment; therefore, using the whole enzyme as an antigen, this region of subunit II will be expected to elicit a stronger immunogenic response than the N-terminal region (Berzofsky, 1985). This is found to be the case as both of the subunit II antibodies have epitopes in the carboxy terminal half of the peptide. Moreover, since the carboxy terminal half is accessible to the solvent the antibodies raised against that area will have easier access to their epitopes. Thus in the study by Mariottini et al. (1986) an antibody against a synthetic peptide corresponding to the last eleven amino acids from the C-terminal of subunit II immunoprecipitated the whole enzyme while an antibody against the N—terminal decapeptide failed to do so. The carboxy terminal half of subunit II also contains the cytochrome c binding site (Millett et al., 1983; Luntz and Margoliash, 1987), and the CuA center (Steffens and Buse, 1979; Holm et al., 1986; Covello and Gray, 1990). Hence, the complete inhibition of the enzyme activity by the subunit II antibody could be due to interruption of the electron transfer process by blockage of cytochrome c access, as the binding data suggests (see Table 3 in chapter 1). Moreover, since the antibody showed a direct effect on cytochrome c binding, it is unlikely that the inhibitory effect is a result of a 84 conformational change caused by the binding of the antibody to the enzyme. Such a phenomenon was suggested by Nicholls et al. (1988b) to explain the effect of a subunit V antibody on cytochrome oxidase activity. These studies showed that the minimal antibody epitope may be located in a 1.5 KDa peptide at the C-terminus of subunit II, a region that does not include the residues defined by Millett et al. (1983) or Luntz and Margoliash (1987) to be a part of the cytochrome c binding domain. However, the ability of the antibody to block cytochrome c binding indicates that the antibody binding domain and the cytochrome c binding site do overlap, presumably because the antibody blocks a much larger surface area than indicated by the minimal epitope. In order to clarify this question, the approach of chemical protection (Burnens et al., 1986) was employed to define the antibody epitope in a broader sense. CHAPTER 3 EFFECT OF ETC MODIFICATION ON CYTOCHROME OXIDASE Chemically modifying enzymes is a powerful tool for studying their functional properties (reviewed by Eyzaguirre,1987). In most cases, the choice of a particular chemical reagent depends on which amino acid residues could be expected to play a key role in substrate binding and/or catalysis, but, even more importantly, which reagent would modify the residue(s) of interest without causing unwanted side reactions. This latter criterion is often difficult to meet; hence, it is usually necessary to identify the modified residue(s) before a final conclusion is made regarding the results of such studies. An alternative to chemical modification is now possible using the technique of site-directed mutagenesis. This approach is very specific compared to chemical modification, so that changes in functional properties of the enzyme can be directly related to specific residue changes, although there are still problems determining whether overall changes in structure have resulted from the single residue change. In addition, it is possible to get secondary effects from a mutation despite a specific change in the DNA sequence, such as altered degree of posttranscriptional modification (Luntz and Margoliash, 1988). Moreover, some proteins such as cytochrome oxidase contain subunits that are mitochondrially encoded and require the difficult task of changing the genes into the universal codon forms before they can be expressed in available expression systems (Cao et al., in press). The interaction between cytochrome c and cytochrome c oxidase is largely electrostatic in nature involving lysine residues in cytochrome c and carboxyl side chains on the enzyme. Converting of some of the lysines in cytochrome c to monocarboxydinitrophenyl residues affects the binding of cytochrome c to the enzyme (Ferguson-Miller et al., 1978). Similarly the modification of the carboxyls on cytochrome oxidase using the water soluble carbodiimide, EDC, or its quaternary amine analog, ETC, resulted in loss of enzyme activity (Millett et al., 1982). Most of the modification is in subunit II, although some of the smaller subunits were also found to be affected (Bisson and Montecucco, 1982; Kadenbach and Stroh, 1984). Moreover, the addition of an equimolar concentration of cytochrome c prior to the reagent protects over 70% of the enzyme activity (Millett et al., 1982). The experiments described in this chapter were designed to define the optimal conditions of this reaction, to examine the effect of the modification on various aspects of the interaction between cytochrome c and cytochrome c oxidase, and to use the technique of protection against modification to further define the relationship of the antibody epitope to the cytochrome c binding domain on subunit II. EXPERIMENTAL PROCEDURES Materials. 1-ethyl-3-(3,3-dimethyl aminopropyl) carbodiimide (EDC) was purchased from Sigma (St. Louis, Mo.) as the hydrochloride salt. 14C glycine ethyl ester hydrochloride (46.5 mCi/mmol) was from New England Nuclear research Products (Mount Prospect, 11). Synthesis of ETC. 300 mgs of EDC-HCL were dissolved in 3 mL of 40% potassium carbonate. The resulting solution was immediately transferred to a separatory funnel and mixed with 3 mL of anhydrous diethyl ether. The ether layer which contains the dissolved EDC, was mixed with dry magnesium sulfate to remove any traces of moisture. The EDC solution was centrifuged to remove magnesium sulfate and transferred to a stoppered round bottom flask to which 250 1.1L of methyl iodide were added. The reaction was left to proceed overnight with stirring. The resulting ETC precipitate was dried by blowing nitrogen and kept in a sealed vial at -20 0C until used. Reaction of ETC with cytochrome oxidase. The enzyme (10 11M M3) and ETC (2 mM) were incubated at room temperature for 20 minutes (short term modification) or 3 hours (long term modification) in 10 mM sodium phosphate buffer pH 7, containing 2 mM lauryl maltoside. Radioactive labeling of the enzyme subunits was done in the presence of 0.5 mM 14C glycine ethyl ester as described by Kadenbach and Stroh (1984). The reaction was stopped by the addition of 1 M ammonium acetate (0.1 M final concentration) Fluorographic Analysis of the ETC modified peptides. Following electrophoresis, the gels containing the peptides modified with ETC in the presence of 14 C glycine ethyl ester were subjected to fluorography as described by Bonner and Laskey (1974). The dried gel was exposed to a Kodak X-OMAT AR film at -70 °C. Effect of ETC modification on cytochrome c binding. Following the modification reaction, the enzyme solution (0.5 mL of 10 uM aa3) was diluted 4-5 fold with 25 mM Tris-acetate buffer, pH 7.9, and concentrated on a Centricon-30 to give a final concentration of at least 50 uM aa3. The enzyme was then incubated with cytochrome c for 10 minutes and the mixture was subjected to gel filtration chromatography as described in Chapter 1. Protection of oxidase from ETC modification by cytochrome c. The enzyme (10 uM aa3) was incubated with 20 uM ferri-cytochrome c for 10 minutes on ice prior to the addition of the reagent. Purified cytochrome c was diluted with 25 mM Tris-acetate buffer, pH 7.9, and concentrated on a Centricon 10 to remove the salts. Protection of the enzyme fi'om ETC modification by subunit H antibody. Cytochrome c oxidase (2.5 uM aa3) was incubated with subunit II antibody, subunit 1V antibody, or hexokinase antibody to achieve > 90% binding of the subunit II and subunit IV antibodies as judged by immunoprecipitation. After a one hour incubation at room temperature each enzyme-antibody solution (200 uL) was transferred to a tube containing 1‘er glycine ethyl ester to achieve a final concentration of 0.5 mM of the reagent. The tubes were kept in a rocking shaker to dissolve the dried glycine ethyl ester and the reaction was then started by adding ETC to a final concentration of 4 mM. After a two hour incubation at room temperature the excess reagent was inactivated by the addition of ammonium acetate to a final concentration of 100 mM. 100 uL of the resulting mixture (50 pg enzyme) were prepared for electrophoresis by TCA precipitation, and redissolved in sample buffer containing 4% SDS. No B-mercaptoethanol was added to avoid dissociation of the antibodies. The samples were subjected to SDS-PAGE and fluorography as described above. Alternatively, the Affi-Gel coupled antibody was incubated with the enzyme (2.5 uM aa3) for one hour, and the bound enzyme was recovered by centrifugation. The enzyme-antibody complex was cleared of excess enzyme by repeated centrifugation and resuspension in 10 mM phosphate buffer pH 7. The modification reaction was started by the addition of ETC (4 mM) in the presence of 14C labelled glycine ethyl ester (0.5 mM), and was allowed to proceed for two hours. RESULTS Characterization of the ETC reaction with cytochrome oxidase. As an additional approach to defining the antibody epitope, we decided to use the principle of protection from chemical modification to compare the interaction domains of the subunit II antibody and cytochrome c. ETC appeared to be an appropriate reagent for such studies since it had been shown to react with carboxyl groups on subunit II and some of these could be protected by cytochrome c binding (Bisson and Montecucco, 1982; Millett et al., 1983; Kadenbach and Stroh, 1984). A careful examination of the modification conditions showed that the reaction occurred in two phases including a fast phase that resulted in 50% inhibition of the enzyme activity in 20 minutes, and a slower phase that took nearly 150 minutes to achieve maximal inactivation of the enzyme (Figure 15). The three hour modification conditions showed similar results to those observed after a 12 hour incubation by Millett et al. (1982) despite the higher enzyme and reagent concentrations used by those investigators (100 M aa3, 4 mM ETC). Figure 16 shows the effect of ETC on the electrophoretic mobility of the subunits (top) and the electron transfer activity of the enzyme (bottom) in the presence and absence of cytochrome c. The two fast migrating derivatives of subunit II reported by Millett et a1 (1982) are labelled as 11’ and II” in the t0p panel. The use of BSA in the controls indicated that the protective effect of cytochrome c was a result of a specific binding to the enzyme, rather than to the presence of excess protein in the reaction tube. Effect of ETC on the structure and function of cytochrome oxidase. Analysis of the kinetics of the partially inhibited enzyme (Figure 17) indicated that the high affinity phase of the biphasic kinetic plot was more strongly 91 Figure 15. Time course of the inactivation of cytochrome oxidase by ETC. Beef heart cytochrome c oxidase (10 uM aa3) was incubated with 2 mM ETC in 10 mM sodium phosphate buffer pH 7.0 containing 2 mM lauryl maltoside. At the indicated times aliquotes of 45 1.1L were withdrawn and the reaction was immediately stopped by the addition of ammonium acetate to a final concentration of 100 mM. Enzyme activity was assayed polarographically in 50 mM potassium phosphate buffer pH 6.5. Other conditions of the assay are as described in the methods section. The open circles represent the enzyme that has been subjected to anion exchange FPLC prior to ETC modification while the solid circles represent the enzyme that did not receive this treatment. log. % activity remaining 2.2 2.0 1.8“ 1.6- 1.4- 1.2- 1.0 a . - °— 0 100 200 incubation time (minutes) Figure 15 Figure 16 Protection of cytochrome oxidase from ETC modification by cytochrome c. The enzyme (10 uM aa3) was incubated with 0.8 mM or 2 mM ETC for 3 hours with no addition, 20 uM cytochome c, or 20 uM BSA. In the top panel the the protective effect of cytochrome c (lanes 2,4,7) and BSA (lanes 1,3,5,6) was analyzed by examining the electrophoretic mobility of the modified subunits. Lanes 1 and 2 show the native enzyme that has not recieved ETC. The lower panel depicts the ability of cytochrome c to protect against loss of enzyme activity due to ETC modification. The open bars represent the native enzyme activity. The solid bars are for the enzyme modified without protection, while the dotted and striped bars represent the protection by cytochrome c and BSA respectively. The enzyme activity was assayed polarographically in 50 mM potassium phosphate buffer, pH 6.5. II III VII 100 - ~—~ ‘— % Activity Remaining 94 _. BSA -—Cytc — 2.0 mM ETC Figure 16 Figure 17. Kinetic analysis of cytochrome oxidase after short term modification with ETC. The enzyme activity was assayed polarographically in 10 mM HEPES, 40 mM KCl buffer pH 7.4, with a range of cytochrome c concentration from 0.04-22 uM. The dotted lines were drawn to best represent the data points (control enzyme,O; modified enzyme,0) while the solid lines were obtained from a computer analysis which separates the contribution of the two kinetic phases. Lines C1 and Cg represent the high and low affinity phases respectively for the control enzyme, while lines M1 and M2 are the high and low affinity phases for the modified enzyme. m m m €3-23 5:2... 100 ,8 Figure 17 affected (80% inhibited) than the low affinity phase (20% inhibited). This observation was of particular interest since the significance of the two kinetic phases and how they relate to cytochrome c binding is a subject of much controversy (see Cooper 1990 for a recent review). The short term ETC inhibition characteristics suggested that modification of a few reactive carboxyls might preferentially affects the high affinity reaction of cytochrome c. If those residues could be identified, it would provide evidence for the location of a physically distinct high affinity site. However, analysis of the short-term modified enzyme, using 1‘er glycine ethyl ester to tag the sites of ETC reaction, revealed that a number of different modified species were produced even during the short labelling time as indicated by the forms of subunit II with different electrophoretic mobility seen on SDS gels (Figure 18). This was further confirmed by subjecting the modified enzyme to ion exchange FPLC as shown in Figure 19. In addition to the inhomogeneity originally present in the enzyme (panel A), the partially modified enzyme (panel B) and the completely modified enzyme (panel C) gave rise to a variety of modified forms in terms of their affinities for the anion exchange resin and their enzymatic activity. Thus, contrary to the implication of a first order time course, it is clear that even at the early stages of reaction we are producing a wide range of modified products whose activities give an average of 50% inhibition. Attempts to eliminate the inhomogeneity by purifying the native enzyme by FPLC and taking the first major peak for ETC modification were not successful. As shown by the profiles in Figure 20, both the control (A) and the partially modified enzyme (B) again showed multiple peaks. This suggests that some of the inhomogeneity of the native enzyme is the result of equilibration between different oligomeric forms or phospholipid binding states. The time course Figure 18. SDS-PAGE and fluorographic analysis of the effect of ETC modification on the subunits of cytochrome oxidase. The enzyme (10 uM aa3) was incubated with (lanes 2 and 4) or without (lanes 1 and 3) an equimolar amount of cytochrome c for 10 minutes in 10 mM sodium phosphate, pH 7.0 containing 2 mM lauryl maltoside. ETC (2 mM) and 14C glycine ethyl ester (0.5 mM) were added and the mixture was incubated for 20 minutes (lanes 1 and 2) or 3 hours (lanes 3 and 4). The reaction mixture was made .1 M in ammonium acetate to stop the reaction. 50 uL of each sample were treated with 80% acetone and the precipitated enzyme was subjected to SDS-PAGE (panel A) and fluorography (panel B). Lane c in panel A denotes the control enzyme that has not received ETC A 2 4 I _. II—T win-wt» IV "" ...-.. Figure 18 cram — II" THIS PHOTOGRAPH WAS PRODUCED BY MSU / INSTRUCTIONAL MEDIA CENTER (517) 353-3960 MSU is an Affirmative Action .' Equal Opportunity Institution. 100 Figure 19. FPLC profiles of native and ETC modified enzyme. Beef heart cytochrome c oxidase (10 uM aa3) was unmodified (panel A), treated with 2 mM ETC for 20 minutes (profile B), or 3 hours (profile C). In each case 400 uL of the enzyme solution were equilibrated with 20 mM sodium phosphate buffer pH 7.0 containing 2 mM lauryl maltoside using a Centricon-30 concentration device. The enzyme was then loaded on a mono-Q 5\5 column previously equilibrated with the same buffer. The collected fractions were analyzed spectrally for oxidase content and polarographically for enzyme activity. The numbers under the arrows denote the turnover number of the enzyme from the fractions eluted at the indicated times. 8200 101 ELUTION TIME (Minutes) Figure 19 . 4 - I 300 .07 - 150 0 O . 4 - | 300 .07 - / - 150 o / I l / IZO 240 ’4” r 1 r 1 r l r r 1 r r r l 1 O C "4 ' ’,/’/ - 300 .07 — ,,/” T ’$/”1 - 150 o ”’ //56 so 45 "I 1 1 r l r r l l 1 I 1 1 r l o 0 IO 20 30 40 NoCI (mM) 102 Figure 20. Rechromatography of native and ETC modified cytochrome c oxidase. The FPLC purified enzyme was reequilibrated with 10 mM sodium phosphate buffer pH 7.0 containing 2 mM lauryl maltoside. The enzyme was incubated with 2 mM ETC as described in the Methods section. The modified enzyme was then applied to mono- Q 5/5 column previously equilibrated with 20 mM phosphate buffer pH 7.2 containing 2 mM lauryl maltoside. The enzyme was eluted with a sodium chloride gradient (dashed lines). Panel A shows the control untreated enzyme while panel B shows the enzyme modified with ETC for 20 minutes. 103 woo .uo Azso woo .uo mrcjoz ._._z_m A3352... NoCl(mM) Figure 20 104 of the ETC modification reaction using FPLC purified oxidase was similar to that found for the enzyme, prior to FPLC purification, and the kinetics of short and long term modified species displayed similar biphasicity. However, the apparently greater inhibition of the high affinity phase during the short term modification was no longer evident (Figure 21). A comparison of the kinetic parameters of the modified enzyme under different buffer conditions revealed some other important features. When assayed in low ionic strength buffer that favors tight binding of cytochrome c, the activities of the modified forms of the enzyme were higher than when assayed at higher ionic strength (Figure 22). The long term modified enzyme was less than 10% active in the phosphate buffer normally used to follow loss of activity, but was 40% active in the low ionic strength buffer; the short term modified enzyme was 50% active in phosphate buffer but 65% active at low ionic strength. In no case did the modified species show a dramatic alteration in Km values, as might be expected from previous studies on cytochrome c modified at lysine residues (Ferguson- Miller et al., 1978). These results suggest that the predominant effect of modification is on electron transfer rather than simply on binding, as observed for cytochrome c. It also appears that ETC modification is affecting a step that contributes more significantly to the rate limitation of the overall reaction under high ionic strength conditions than under low ionic strength conditions. Further elaboration on this point will be offered in the discussion. Effect of ETC on cytochrome c binding to the oxidase. Analysis of cytochrome c binding stoichiometry (Table 4) indicated that the partially modified enzyme and the completely modified enzyme bound cytochrome c 105 Figure 21. Kinetic analysis of the effect of ETC modification on the FPLC purified cytochrome oxidase. The enzyme was subjected to ion exchange FPLC as described in the methods section. The purified enzyme (10 uM aa3) was treated with ETC (2 mM) for 20 minutes (0), or three hours (A). A control enzyme from the same FPLC fraction was incubated under the same experimental conditions except that ETC was omitted (0). In each case the rate of oxygen consumption was measured using 0.03 nmoles aa3, and a range of cytochrome c concentration from 0.04-22 uM. The assay buffer contained 25 mM Tris acetate pH 7.9. 1 mM Lauryl maltoside, 2.5 mM sodium ascorbate, and 0.5 mM TMPD were also added to the assay chamber. The maximal velocities from the intercept on the x-axis were 240 8'1 for the native enzyme, 145 S'1 for the partially modified enzyme and 105 5'1 for the highly modified enzyme. 106 400 300‘ 200‘ TN/[S] 100‘ 0 100 200 Figure 21 107 Figure 22. Kinetic analysis of native and ETC modified cytochrome c oxidase in phosphate buffer. The enzyme (10 uM aa3) was incubated with 2 mM ETC for 20 minutes (0) or 3 hours (CI). The native enzyme was treated similarly except that no ETC was added (0). In each case the enzyme activity was assayed in 50 mM potassium phosphate buffer pH 6.5. Other conditions of the experiment were as in the legend to figure 21. TN/(S) 300 108 100 200 300 TN Figure 22 109 Table 4. Effect of ETC modification on the enzymatic activity and cytochrome c binding to beef heart cytochrome c oxidase. The enzyme (10 uM aa3) was incubated with 2 mM ETC in the presence (+ cyt. c) or absence of 20 uM cytochrome c for the indicated times. The reaction was stopped by the addition of ammonium acetate to a final concentration of 0.1 M, and the enzyme solution was concentrated and equilibrated with the binding buffer (25 mM Tris-acetate pH 7.9) using a Centricon-30 concentration device. For cytochrome c binding studies the enzyme (10 11M aa3) was incubated with 30 uM cytochrome c for 10 minutes in the Tris acetate buffer containing 2 mM lauryl maltoside. Cytochrome c binding was determined by gel filtration chromatography as described in the Methods section. The enzyme activity is expressed as turnover number (TN 8'1), and is meaured polarographically in the Tris-acetate buffer (T-A), and 50 mM potassium phosphate buffer (Kpi). 110 Table 4. Effect of ETC modification on the enzymatic activity and cytochrome c binding to cytochrome oxidase rEgerimental conditions Cyt .c Activity in Activity in binding TEA K i ratio % TN % TN % (8'1) (3'1) control 2.0 1(X) 220 100 445 100 ETC (20 minutes) 1.2 so 145 65 196 44 ETC (3 hours) 0.8 40 1(1) 45 45 10 ETC (20 minutes) + cyt. c 1.8 90 ND ND 365 82 111 at ratios of 1.2:1 and 0.8:1 respectively, under conditions where 2:1 binding was observed with the native enzyme. This result indicated that some of the modified carboxyls are essential for cytochrome c binding, and that modified enzyme activity is lost in proportion to the loss of binding under low ionic strength conditions. Purification of the enzyme by FPLC resulted in a lowered binding ratio in some cases approaching one cytochrome c per oxidase for the native enzyme. Two possible explanations for the change in binding are that 1) two binding sites are no longer saturated at the 25 uM cytochrome c concentration used; or 2) a second binding of cytochrome c is not involved in electron transfer, since biphasic kinetics are still observed. Related to the first explanation, an increased apparent Km value for the low affinity phase of the kinetics was observed after FPLC treatment, but an absolute correlation between Km and KD values cannot be made due to the complexity of the oxidase kinetics. Since anion exchange FPLC in the presence of lauryl maltoside has been found to cause delipidation of the enzyme (Gregory, 1988), the lower level of cytochrome c binding may relate to loss of phospholipids that could directly or indirectly affect cytochrome c binding. Protective effect of cytochrome c and antibody against ETC modification. Due to the multiple forms produced by even short term ETC modification, it was not feasible to attempt to identify specific carboxyls that might be related to the high or low affinity reactions of cytochrome c. It was still possible that a difference in the pattern of labelling of the whole enzyme might be observed under short and long modification conditions, that might reveal differential involvement of subunits in the cytochrome c interaction. 112 Furthermore, examination of the labelling pattern of subunit II and the two immunogenic tryptic fragments under short and long term modification conditions in the presence and absence of cytochrome c, could provide some information regarding the location of the carboxyls that are readily labelled and strongly protected by cytochrome c. Such information could be used for defining the antibody epitope in relation to the high affinity cytochrome c binding site. Fluorographic analysis of the pattern of labelling among the different subunits of the enzyme was carried out for short and long term modification in the presence and absence of cytochrome c. As shown in Figure 18 the same set of subunits was labelled in the short and long term cases. Thus in addition to subunit II, subunits I, Vb, and VIIc were significantly labelled, but only subunits Vb, and VIIc were protected by cytochrome c, suggesting that they might be involved in or close to the cytochrome c binding domain (protection of subunit Vb is not obvious in the short term reaction because of the lower protein load in lane 1 of Figure 18). Further analysis of the labelling patterns after long term modification are presented in Figure 23. The reagent labelled most strongly subunits I, II, Vab, and VIIbc. There was protection by cytochrome c to all of these subunits except subunit I. This lack of protection of subunit I is in agreement with the results of Bisson and Montecucco (1982). Analysis of the labelling patterns of subunit II and its tryptic fi'agments is shown in Figure 24. An interesting observation from this figure is that despite a further loss of activity during the long term modification, cytochrome c completely protects against additional modification of subunit II. This could be attributed to modification of subunit I which is not protected by cytochrome c and constitutes a major part of the catalytic core of the enzyme (Muller et 113 Figure 23. ETC labelling pattern of cytochrome oxidase subunits under long term conditions. ETC modification was carried out for 3 hours in the presence of 0.5 mM 14C glycine ethyl ester. The modification was done in the presence (dashed lines) or absence (solid lines) of cytochrome c. The y-axis indicates the integrated intensity of the fluorograms obtained by exposing the radioactive gel at —80 °C for 5 days. siiaieuiimui 99 osi .0 o 114 INTEGRATED INTENSITY 9 .- 0| 0 Figure 23 115 Figure 24. Analysis of the effect of ETC on subunit II and its tryptic fragments. Cytochrome oxidase was modified with ETC in the presence of 14C glycine ethyl ester. The labelling was done under short and long conditions in the presence or absence of cytochrome c. In each case subunit II was purified by HPLC and subjected to tryptic digestion as described in the Methods section. Following SDS-PAGE the gel was subjected to fluorography and the fluorograph was scanned using a Visage 110 scanning analyzer. The scan pattern from short term modification is shown by the dashed lines while the solid lines represent the long term modification. )IQI >101 116 INTEGRATED INTENSITY 'o o ,0 O f 0| \‘ I x/ ”xx .‘ .r‘" I’ll ’ 0 l ‘\ ~‘\\... + — . ‘\’ ' Q“ ‘ . ‘\ a - o - IO ,' I0 ’I I O I” 0” I, 0“.“ l’ ,1" I ’0 ’O I I I I r I r | \ \~ ' ~ : > I ” I I u r r a i i i ,I I’ ”’ r” r .i' Figure 24 117 al., 1988a). Indeed labelling experiments showed substantially increased level of modification of subunit I under long term conditions. (The result shown in Figure 18 is misleading regarding the labelling of subunit I due to the lower amount of protein in lane 4). This type of experiment was not done in the presence of the antibody because of the need for large amounts of enzyme and therefore antibody to yield enough purified subunit II. However, the subunit 11 antibody afforded a strong protective effect against ETC modification of subunit II when compared to a subunit IV antibody or one against hexokinase, as shown in Figure 25. It is interesting to note that both the antibody and cytochrome c protected subunit II from labelling, but the nature of the protection was somewhat different (see discussion). 118 Figure 25. Protection by antibody and cytochrome c against ETC modification of subunit II. In lanes 1-3 the enzyme (2.5 uM aa3) was incubated with antibody to hexokinase (1), subunit IV (2) or subunit II (3). In each case the enzyme-antibody complex was treated with 4 mM ETC in the presence of 0.5 mM 14C glycine ethyl ester. The proteins were precipitated by the addition of trichloroacetic acid and subjected to SDS-PAGE followed by fluorography. In lanes 4 and 5, the enzyme (2 uM aa3) was incubated with Affi-Gel-coupled antibody to hexokinase in the presence (4 )or absence (5) of cytochrome c for 2 hours. ETC (4 mM) was added for 2 hours in the presence of 0.5 mM 140 glycine ethyl ester. The enzyme was recovered by centrifuging out the cells and precipitated by the addition of an equal volume of 10% trichloroacetic acid and redissolved in electrophoresis sample buffer. 119 Figure 25 THIS PHOTOGRAPH WAS PRODUCED BY MSU / INSTRUCTIONAL MEDIA CENTER (Si 7) 353-3960 MSU is an Affirmative Action ,- Equal Cmponunmj Institution~ DISCUSSION The significance of the biphasic kinetics observed between cytochrome c and the oxidase is still an unsettled question. Based on cytochrome c binding data, Ferguson-Miller et al. (1976) proposed that the enzyme has two interaction sites, both active in electron transfer but with different affinities for cytochrome c. Alternatively, the biphasic kinetics were explained on the basis of a single productive site with cytochrome c binding at a second regulatory site that altered the binding at the productive site (Speck et al., 1984). It has also been argued that biphasic kinetics can result from a single cytochrome c interaction if the sequential internal electron transfer is slow enough to contribute to limitation of the overall rate (Anatalis and Palmer, 1982), or if there is negative cooperativity between monomers in a dimer (Nalecz et al., 1985), or if the enzyme exists in two different conformations with different affinities for cytochrome c (Brzezinski and Malmstrom, 1986). Our data seem to suggest that a single cytochrome c interaction may be sufficient for the biphasic kinetics, since reduction in the ratio of cytochrome c binding to the oxidase was observed after treating the enzyme by anion exchange FPLC, while there was little effect on the biphasic kinetics. This result could mean that the chromatographic procedure is eliminating a non-specific cytochrome c interaction that is detectable by the binding assay but is not involved in the enzyme function. A complete binding curve for both forms of the oxidase would be required to substantiate this possibility. If partial delipidation is the major effect of anion exchange FPLC on the enzyme (Gregory, 1988), the current data would argue against a role of phospholipids as modulators of the kinetics (Vik et al., 1981) and in support of the findings of Marsh and 121 Powell (1988) who showed that replacement of the acidic phospholipids in cytochrome c oxidase by dimyristoyl phosphatidylcholine, a neutral phospholipid, had no effect on the biphasic kinetics. Cytochrome c binding stoichiometry was determined by gel filtration chromatography. This method has the advantage of detecting both strong and weak binding, unlike the spectral assay introduced by Michel and Bosshard (1984). We have observed some variability in our results with the native enzyme (without FPLC purification) giving 1.4-2.0 cytochrome c bound per oxidase under conditions which might be expected from previous data to fully saturate two sites. This variability appeared to be mainly due to the use of a low enzyme concentration (10 uM aa3) for the binding study, resulting in less than 1 uM enzyme after gel filtration, with a background of 25 uM cytochrome c. The low concentration was dictated by practical considerations, to allow studies with the modified enzyme and with the subunit II antibody. The significance of the observed difference in the maximal activity of the ETC modified enzyme under different ionic strength conditions is not completely clear. Since the interaction between cytochrome c and the oxidase is electrostatic in nature, involving negative charges on subunit II and positive charges on cytochrome c (Luntz and Margoliash 1987), then ETC modification would be expected to completely prevent cytochrome c binding or lower its affinity for the enzyme by converting the carboxyls into positively charged groups, provided that these carboxyls are essential for binding. It should be noted that due to the abundance of carboxyls and the randomness of the reaction there is likely to be a mixture of modified species produced some at the active site resulting in complete inactivation of the enzyme and others at the periphery or even more distant in which case the functional properties of the enzyme may be only partially affected or not at all. A certain proportion of these latter species will have intact active sites but altered binding properties. These forms will be influenced in their activity by the ionic strength conditions such that at low ionic strength, they will bind cytochrome c sufficiently well for electron transfer to occur, while at high ionic strength they will be inactive. The fact that kinetic analysis of the ETC modified did not reveal major changes in Km values suggests that the majority of the modified carboxyls are directly involved in the active site domain so that their conversion to bulky positively charged derivatives results in complete inactivation of the enzyme. The fact that native and modified cytochrome oxidase gave multiple peaks when chromatographed by the FPLC is an indication of inhomogeneity in the enzyme preparation as well as the randomness of modification of the large number of carboxyls. Inhomogeneity of the purified enzyme is not unexpected since it is isolated as a multi-peptide complex with varying amounts of bound phospholipids and detergent. The enzyme can also acquire a number of aggregation states that add to the inhomogeneity (Hartzell et al., 1988). The unavoidable consequence of the inhomogeneity is that the different enzyme forms may react differently when treated with various chemical and biological probes used to study structure-function relationships. This fact, combined with the numerous carboxyls available to react, is likely to give a variety of products even in the short term. A multitude of products were also observed when lysines of cytochrome c were derivatized (Brautigan et al., 1978). In the case of cytochrome oxidase, both short term and long term modification conditions resulted in a number of different species that were partially resolved by FPLC. They not only varied in their elution properties but also in their enzymatic activity, despite the apparent first order time course for the short term modification. The fact that we must deal with both inhomgenous enzyme and numerous modified forms makes it difficult to foresee that singly modified species will be purified from which more conclusive structure-function relationships could be drawn. These results indicate that in complex membrane proteins such as cytochrome c oxidase the term “homogeneous enzyme” should be used with caution, regardless of the purification scheme (Powers et al., 1987; Hartzell et al., 1988), since the multiple forms of the enzyme appear to be in dynamic equilibrium. This is evident from the fast reequilibration of the isolated dimer (Figure 5C) and the appearance of multiple forms after rechromatography of the native enzyme (Figure 19A). The ability of cytochrome c to protect subunit II from ETC modification was taken as an evidence that this subunit contains a binding site for cytochrome c (Millett et al., 1982; 1983; Bisson and Montecucco, 1982; Kadenbach and Stroh, 1984). Under the conditions used in this study similar cytochrome c protection was observed. Moreover, the extent of labelling of subunit II and the two tryptic peptides in the presence of cytochrome c is the same in short and long term modification, suggesting that the further inactivation observed in the long term modification (which is not protected by cytochrome c) must be occurring at a different subunit. Indeed examination of the extent of labelling under the two conditions revealed that subunit I was more modified under long term conditions. The cytochrome c interaction domain on subunit II is proposed to be centered around residues 114-137 (Luntz and Margoliash 1987); however, Millett et al., (1983) reported that Asp-112, Glu-114, and Glu-198 are the only three carboxyls protected by cytochrome c. In this study we have observed that the 15 kDa peptide (residues 99-227), and the 10 kDa peptide (residues 135-227) were modified and protected in a similar way as was the intact subunit II suggesting that modification occurred in the C-terminal half of the protein and was relatively random. One goal of this study was to determine the location of the antibody epitope in relation to the cytochrome c binding domain. The fact that cytochrome c protected the 10 kDa immunogenic peptide against ETC modification suggested that the antibody epitope and the cytochrome c binding domain may be overlapping. However, examination of the electrophoretic mobility of subunit H after ETC modification in the presence of cytochrome c, as compared to that observed in the presence of the antibody, indicated that a small amount of faster migrating, highly modified species was produced during antibody protection, while a small amount of slower migrating less labelled form was produced in the presence of cytochrome c. One explanation for this phenomenon could be that different sets of carboxyls were protected in the two cases, but it is more likely that a small portion of the enzyme was heavily labelled in the presence of antibody because it was not associated with the antibody, while the majority of the enzyme molecules were completely protected by essentially irreversible binding to the antibody. In the case of cytochrome c, there would be protection to all the enzyme molecules, but since there is a dissociation/association reaction going on, the reagent would get access to all enzyme molecules but only to a limited extent. In any case, the results show strong protection of subunit II modification by the antibody as well as cytochrome c, indicating that both molecules occupy a significant proportion of the water accessible surface of subunit II. SUMMARY The functional roles of the individual subunits of mammalian cytochrome c oxidase are not well established. Previous studies using cross- linking and chemical modification techniques indicate that subunit 11 contains a binding site for cytohrome c. This study is in agreement with these reports since a monoclonal antibody against subunit II inhibited the enzyme activity and blocked cytochrome c binding to the enzyme. Moreover, the addition of cytochrome c prior to modification of the enzyme by ETC protected against chemical modification of subunit II and against inactivation of the enzyme. Thus contrary to a recent report indicating that removal of subunit II from a homologous bacterial oxidase had no effect on the Km for cytochrome c, subunit H appears to have an essential role mediating the interaction of cytochrome c with cytochrome oxidase. The biphasic Eadie-Hoffstee plots obtained when the enzyme is titrated with increasing cytochrome c concentration have been interpreted as resulting from two cytochrome c interactions sites with different affinities for the substrate, based on binding studies revealing two dissociation constants similar to the two Km values calculated from the kinetics. Alternative models that explain the biphasic kinetics on the basis of a single cytochrome c interaction have also been proposed. Some of the results presented here favor the idea that a single interaction may be sufficient for biphasic kinetics; in particular, anion exchange FPLC lowered the binding ratio of cytochrome c to the enzyme to about 1:1 without a significant effect on activity or on the biphasicity of the kinetics. The current binding studies were conducted using a single concentration of cytochrome c (25 uM) to equilibrate the column because it was usually found to be sufficient to saturate two binding sites on the standard (non- FPLC treated) enzyme. It will be necessary to establish a binding curve for the FPLC purified enzyme to determine whether one binding site has been removed or whether the binding parameters of one or both sites have been altered. In any case, the results indicate that a direct correlation between the Km and the K1) values does not pertain under all conditions. The binding of cytochrome c to the oxidase is governed by electrostatic interactions. Modification of the carboxyls in subunit II by ETC reverses the charge on these residues and therefore would be expected to weaken or prevent the interaction with cytochrome c. A strong inhibition of activity (V max) was observed, an effect that was intensified at high ionic strength and diminished at low ionic strength. A definitive analysis of the effect of ETC on the kinetic parameters of the enzyme would only be possible if singly modified species were isolated. However, FPLC separation of the modified enzyme revealed a large number of enzyme species that were only partially resolved by anion exchange resin but had a wide range of enzyme activities. Thus isolating pure monoderivatized forms would be a formidable task, especially since the native enzyme also contains forms that have different affinities for the anion exchange resin, possibly reflecting variable phospholipid contents and states of association. The minimum binding epitope of the antibody was mapped to a 1.5 kDa peptide at the C-terminal of subunit II, a region which does not contain the carboxyls that have so far been implicated in the cytochrome c binding domain. However, both antibody and cytochrome c strongly protected subunit II from ETC modification, suggesting a significant overlap in their binding domains. Further clarification on this issue could be achieved if the carboxyls protected by the antibody and cytochrome c could be identified. Efforts to identify the modified and protected residues in a 3.2 kDa C- terminal peptide using mass spectrometry have so far been unsuccessful. 128 BIBLIOGRAPHY Amit, A.G., Mariuzza, R.A., Phillips, S.E.V., and Poljak, R.J.(1986) Science, 233, 747-753. Anatalis, T.M., and Palmer, G. (1982) J. Biol. Chem. 251, 6194—6206. Beinert, H. (1966) In “The biochemistry of copper” J. Peisarch P. Aisen, and W. Blumberg eds. Acad. Press, New York pp 213-234. Berzofsky, J. (1985) Science W9, Bickar, D., Lehninger, A., Brunori, M., Bonaventura, J ., and Bonaventura, C. (1985) J. Inorg. Biochem. 23, 365-372. Bisson, R. (1990) In "Bioelectrochemistry III: Charge Separation across Biomembranes". Milazzo, G., and Blank, M. eds. Plenum Press, N.Y. London Bisson, R., Gutweniger, H., Montecucco, C., Colonna, R., Zanotti, A., and Azzi, A. (1977) FEBS Lett. 81, 147-150 Bisson, R., Azzi, A., Gutweniger, H., Colonna, R., Montecucco, C., and Zanotti, A. (1978) J. Biol. Chem. 253, 1874-1880. Bisson, R., Montecucco, C., Gutweniger, H., and Azzi, A(1979). J. Biol. Chem._25_4, 9962-9965. Bisson, R., Jacobs, B., and Capaldi, R.A. (1980).Biochemistry, 19,4173-4178 Bisson, R., and Montecucco, C. (1982a) FEBS Lett. 150, 49-53. Bisson, R., Steffens, G., Capaldi, R.A. , and Buse, G.(1982b) FEBS Lett. 1514, 359-363. Bisson,R., Steffens, G.M., and Buse, G. (1982). J. Biol.Chem. 251, 6716-6720. Bisson, R., Schiavo, G., and Montecucco, C. (1987) J. Biol. Chem. 262, 5992-5998. Bolli, R., Nalecz, K.A., and Azzi, A. (1985a) Arch. Biochem. Biophys. 240, 102-116. 129 Bolli, R., Nalecz, K.A., and Azzi, A. (1985b) Biochimie 61, 119-128. Bolli, R., Nalecz, K.A., and Azzi, A. (1986) J. Bioenerg. Biomembranes 18, 277-284. Bombelka, E., Richter, F.W., Stroh, A., and Kadenbach, B.(1986) Biochem. Biophys. Res. Comm. JAQ. 1007-1014. Bonen, L., Boer, P.H., and Gray, M.W. (1984) EMBO J. 3, 2531-2536. Bonner, W.M., and Laskey, RA. (1974) Eur. J. Biochem. 46, 83-88. Brautigan, D., Ferguson-Miller, M., and Margoliash, E (1978) Methods Enzymol. 53, 128-164. Briggs, M., and Capaldi, RA. (1977) Biochemistry 16, 73-77. Briggs, M., and Capaldi, RA. (1978) Biochem. Biophys. Res. Comm. 8_Q, 553-559. Briggs, M., Kamp, P.F., Robinson, NC, and Capaldi, R.A. (197 5) Biochemistry 14. 5123-5128. Brzezinski, P., and Malmstrom, B.G. (1986) Pro. Natl. Acad. Sci. 83, 4282-4286. Burnens, A., Demotz, S., Corradin, G., Binz, H., and Bosshard, H. (1986) Science 235. 780-783. Buse, G., Meinecke, L., and Bruch, B. (1985) J. Inorg.Biochem. 23, 149-153. Buse, G., Steffens, G.C.M., Biewald, R., Bruch, B., and Hensel, S. (1987) in "Cytochrome systems. Molecular Biology and Bioenergetics" (S. Papa, B. Chance, and L, Ernester eds) Plenum Press, New York, pp 791-793. Cao, J ., Revzin, A., and Ferguson-Miller, S. Biochemistry, in press. Capaldi, R.A., and Hayashi, H. (1972) FEBS Lett. 26, 261-263. Capaldi, R.A., Bell, R.L., and Branchek, T. (1977) Biochem. Biophys. Res. Comm. 14, 425-433. Capaldi, R.A., Halphen, D.G., Zhang, Y., and Yanamura, W. (1988) J. Bioenerg. Biomembr. 20, 291-311. Capaldi, R.A., Takamiya, S., Zhang, Y., Halphen, D.G., and Yanamura, W. (1987) Curr. Topics Bioenerg. 15, 91-112 Casey, R.P., Thelen, M., and Azzi, A. (1979) Biochem. Biophys. Res. Comm. 81, 1044-1051. Casey, R.P., Thelen, M., and Azzi, A. (1980) J. Biol. Chem. 255, 3994-4000. Casey, RR, and Azzi, A. (1983) FEBS Lett. 154, 237-242. Chan, SR, and Tracy, RP. (1978) Eur. J. Biochem. 88, 595 -605. Chan, S.H.P., and Freedman, J .A. (1983) FEBS Lett. 162, 344-348 Chan, S., and Gai, W. (1987) in "Advances inMembrane Biochemistry and Bioenergetics" (C.H. Kim, H.Tedesche, J .J . Diwan, and J .C. Salerno eds) Plenum Press, New York. pp 193-202. Chuang, T.F., and Crane, FL. (1971) Biochem. Biophys. Res. Comm. 42, 1076- Cooper, C. (1990) Biochim. Biophys. Acta 1611, 187-203. Costello, J ., Escaig, J ., Matsushita, K., Viitanen, P., Menick, D., and Kaback, H. (1987) J. Biol. Chem. 262, 17072-17082. Covello, P., and Gray, M. (1989) Nature 841, 662-666. Covello, P., and Gray, M. (1990) FEBS Lett. 268, 5-7. DeWitt, D.L., Day, J .S., Gauger, J .A. and Smith, W.L. (1982) Meth. Enzymol. 86, 229-240. Deathrage, J .F., Henderson, R., and Capaldi, RA. (1982) J. Mol. Biol. 158, 487-499. Djavadi-Ohaniance, L., Friguet, B., and Goldberg, M. (1984) Biochemistry 28, 97-104. Downer, N.W., Robinson, N.C., and Capaldi, RA. (1976) Biochemistry 15, 2930-2936. Einarsdottir, O., and Caughey, W. (1984) Biochem. Biophys. 131 Res. Comm. 124, 836-842. Einarsdottir, 0., and Caughey, W. (1985) Biochem. Biophys. Res. Comm. 128, 840-847. Erecinska, M. (1977) Biochem. Biophys. Res. Comm. 16, 495- 501 Errede, B. (197 6) Doctoral Thesis, University of California at San Diego. Errede, B., and Kamen, M. (1978) Biochemistry 11, 1015- 1027 By, P.L., Prowse, S.J., and Jenkin, GR. (1978) Immunochem. 15, 429-436. Eytan, G.D., Carroll, R.C., Schatz, G., and Racker, E. (1975) J. Biol. Chem. 250, 8598-8603. Eyzaguirre, J. (1987) in “Chemical modification of enzymes: Active site studies”. J. Eyzaguirre editor, Ellis Horwood Ltd. J. Wiley and Sons, N.Y, Chichester, Brisbane, Toronto. pp 9-22. Ferguson-Miller, S., Brautigan, D.L., and Margoliash, E. (1976). J. Biol. Chem. 251, 1104-1115 Ferguson-Miller, S., Brautigan, D.L., and Margoliash, E. (1978). J. Biol. Chem. 258, 149- 159. Finel, M. (1987) Biochim. Biophys. Acta 894, 174-179. Finel, M., and Wikstrom, M. (1986) Biochim. Biophys. Acta 851, 99-108. Finel, M., and Wikstrom, M. (1988) Eur. J. Biochem. 176, 125-129 Finney, K.,Messer, J .,DeWitt, D., and Wilson, J. (1984) J. Biol. Chem. 259, 8232-8237. Freedman, J .A., and Chan, S.H.P (1983) J. Biol. Chem. 258, 5885-5892. Freedman, J .A., Bratcher, R.L., and Chan, S.H.P (1984) Biochem. Biophys. Res. Comm. 120, 52-58. Gai, W., Sun, S., Ding, Y., Freedman, J .A., and Chan, S.H.P (1988) Arch. Biochem. Biophys. 266, 628-638. 132 Galfre, G., and Milstein, C. (1981) Meth. Enzymol. 13, 1-46. Garber, E.A.E., Luntz, T.L., and Margoliash, E. (1988) in "oxidases and related redox systems" King, T.E., Mason, HS, and Morrison, M. eds.Allan,R. Liss Inc. New York.pp 749-769 Garber, E., and Margoliash, E. (1990) Biochim. Biophys. Acta 1615, 279-287. Georgevich, G., Darly-Usmar, V.M., Malatesta, F., and Capaldi, RA. (1983) Biochemistry 22, Gregory, L. (1988) Doctoral Thesis, Michigan State University Gregory, L., and Ferguson-Miller, S. (1987) in "Advances in Membrane Biochemistry and Bioenergetics" (C.H. Kim, H. Tedesche, J .J . Diwan, and J .C. Salerno eds) Plenum Press, New York. pp 301-309. Gregory, L., and Ferguson-Miller, S. (1988) Ann. NY. Acad. Sci. 56, 260-268. Gutweniger, H., Bisson, R., and Montecucco, C. (1981) Biochim. Biophys. Acata 685, 187-193. Hadikusumo, R., Hertzog, P., and Marzuki, S. (1986) Biochim. Biophys. Acta 85Q, 33-40 Hackenbrock, GR, and Hammon, KM. (1975) J. Biol. Chem. 2m, 9185-9197. Hakvoort, T., Sinjorgo, K., Van Gelder, B., and Muijsers, A. (1985). J. Inorg. Biochem. 28, 381-388. Hakvoort, T.B.M., Moolenaar, K., Lankvelt, A.H.M., Sinjorgo, KM.C., Dekker, H.L., anf Muijsers, AD. (1987). Biochim. Biophys. Acta 8&4, 347-354. Hartzell, C., Beinert, H., Babcock, G., Chan, S., Palmer, G., and Scott, R. (1988) FEBS Lett. 23.6. 14. Henderson, R., Capaldi, R.A., and Leigh, J .S. (1977) J. Mol. Biol. 112. 631-648. Heron, C., Smith, S., and Ragan, C.I. (1979) Biochem. J. 181, 435-443. 133 Hogeboom, G., Schneider, W.C., and Palade, G. (1948) J. Biol. Chem. 121, 619- . Holm, L., Saraste, M., and Wikstrom, M. (1987) EMBO J. 6, 2819-2823. Hfither, F.J., and Kadenbach, B. (1986) FEBS Lett. 261, 89-94 Ichiki, T., Tanaka, M., Kobayashi, M., Sugiyama, N., Suzuki, H., Nishikimi, M., Ohnishi, T., Nonaka, I., Wada, Y., and Ozawa, T. (1989) Pediatr. Res. 25, 194-201. J arausch, J ., and Kadenbach, B. (1985a) Eur. J. Biochem. 146, 211-217. J arausch, J ., and Kadenbach, B. (1985b) Eur. J. Biochem. 146, 219-225. Kadenbach, B. (1983) Trends Biochem. Sci. 8, 398-400. Kadenbach, B., and Merle, P. (1981) FEBS Lett. 135, 1-11. Kadenbach, B., Jarausch, J ., Hartmann, R., and Merle,P. (1983). Anal. Biochem.129, 517-521. Kadenbach, B., and Stroh, A. (1984). FEBS Lett. 113, 374- 380. Keilin,D., and Hartee, F. (1939) Proc. Roy. Soc, B 121, 167-191. Kierns, J .J ., Yang, 0.8., and Gilmour, M.V. (1971) Biochem. Biophys. Res. Comm. 45, 835- Kim, S.J., Lee, K.O., Takamiya, S., and Capaldi, R. (1987) Biochim. Biophys. Acta 894, 27 0-276. Chem. 260, 8571-8574. Klemm, P (1984) In “Methods in Molecular Biology” vol. 1 Proteins J .M. Walker, editor. pp. 255-259. Humana Press Clifton, NJ. Kiihler, G., and Milstein, C. (197 5) Nature (London) 256, 495-497. Kornblatt, J ., and Lake, D. (1980) Can. J. Biochem. 58, 219- 224. Kuhn-Nentwig, L., and Kadenbach, B. (1985) Eur. J. Biochem. 149, 147-158. 134 Lehninger, A.L., Reynafarje, B., Davies, P., Alexandre, A., Villalobo, A., and Beavis, A. (1981) in "Mitochondria and Microsomes" (C.P. Lee, G. Schatz, and G. Dallner eds) Addison-Wesley Publishing Co. pp 459-479. Ludeen, M.C. (1986) Inorg. Chem. 25, 4852-4856. Ludwig, B. (1980) Biochim. Biophys. Acta 594, 177-189. Ludwig, B. (1987) FEMS Microbiol. Rev. 46, 41-56. Ludwig, B., Downer, N.W., and Capaldi, RA. (1979) Biochemistry , 1401-1407. Luntz, T.L., and Margoliash, E. (1987) in "Cytochrome Systems. Molecular Biology and Bioenergetics" (S. Papa, B. Chance and L. Ernester eds) Plenum Press, New York pp 271-279 Marsh, D., and Powell, G. (1988) Bioelectrochem. Bioenerg. 26, 73-82. Mariottini, P., Chomyn, A., Doolittle, RE, and Attardi, G. (1986) J. Biol. Chem. 261, 3355-3362. Meinecke, L., and Buse, G. (1984) Eur. Bioenerg. Conf. Report 8, 177-178 Merle, P., and Kadenbach, B. (1980) Eur. J. Biochem. 165, 499-507. Merle, P., and Kadenbach, B. (1982) Eur. J. Biochem. 125, 239-244. Merle, P., Jarausch, J ., Trapp, M., Scherka, R. and Kadenbach, B (1981) Biochim. Biophys. Acata 669, 222- 230. Michel, B., and Bosshard, H. (1984) J. Biol. Chem. 259, 10085-10091. Michel, B., and Bosshard, H. (1989) Biochemistry 28, 244-252. Millett, F., Darly-Usmar , V., and Capaldi, RA. (1982). Biochemistry 2.1. 3857-3862. Millett, F., De Jong, K., Paulson, L., and Capaldi, RA. (1983). Biochemistry 22, 546-552. Mitchell, P. (1961) Nature 191, 144-148. Mitchell, P., and Moyle, J. (1979) in "Cytochrome Oxidase" (T.E. King, Y. Orii, B. Chance, and K. Okunuki eds) Elsevier, North-Holland pp 361-372. Mitchell, P., and Moyle, J. (1983) FEBS Lett. 151, 167-178. Mochan, B.S., Lang, R.W., and Elliott, W.R. (1970a) Biochim. Biophys. Acta 216, 96-105. Mochan, B.S., Lang, R.W., and Elliott, W.R. (1970b) Biochim. Biophys. Acta 216, 106-111. Montecucco, C., Schiavo, G., and Bisson, R. (1986) Biochem. J. 284, 241-243. Moody, A.J., and Rich, PR. (1989). Biochim. Biophys. Acta 818, 29-34. Moroney, P., Scholes, T.A., and Hinkle, RC. (1984) Biochemistry 28, 4991-4997. Miiller-Hécker, J ., Droste, M., Kadenbach, B., Bongratz, D. and Hubner, G. (1989) Hum. Pathol. 26, 666-672. Muller, M., Schlapfer, B., and Azzi, A. (1988a) Biochemistry 21, 7546-7551. Miiller, M., Schlapfer, B., and Azzi, A. (1988b) Proc. Natl. Acad. Sci. (U.S.A) 85, 6647-6651. Myers, D., and Palmer, G. (1988) Anals of New York Acad. Sci. 556, 85-97. Nalecz, K.A., Bolli, R., and Azzi, A. (1983). Biochem. Biophys. Res. Comm. 114, 822-828. Nefsky, B., and Bretscher, A (1989) Proc. Natl. Acad. Sci. 86, 3549-3553. Nicholls, P. (1965) Cytochrome oxidase in situ : Concepts and questions. In "Oxidases and Related Redox Systems" T.E. King, H.S. Mason, and M. Morrison eds. pp 764-777 Wiley, New York. Nicholls, P., Hildebrandt, V., Nicholls, F., and Wrigglesworth, J. (1980) Can. J. Biochem. 58, 969-977. Nicholls, P., Cooper, C.E., , Freedman, J.A., and Leece, B. ( 1988a) Biochem. Cell Biol. 66, 1218-1224. Nicholls, P., Cooper, C.E., Leece, B., Freedman, J .A., and Chan, S.H.P (1988b) in "oxidases and related redox systems" King, T.E., Mason, HS, and Morrison, M. eds. Allan, R. Liss Inc. New York. Nilsson, T. -J., Gelles, P. M., and Chan, S. I. (1988) Biochemistry, 21, 296-301. Nunberg, J ., Rodgers, G., Gilbert, J ., and Snead, R. (1984) Proc. Natl. Acad. Sci. (USA) 81, 3675-3679. Palca, J. (1990) Science 249, 1104-1105. Paterson, Y., Englander, S., and Roder, H. (1990) Science 249, 755-759. Patterson, T.E., and Poyton, RD. (1986) J. Biol. Chem. 261, 17192-17197. Patterson, T.E., Trueblood, C.E., Wright, R.M., and Poyton, R.O. (1987) in "Cytochrome Systems. Molecular Biology and Bioenergetics" (S. Papa, B. Chance and L. Ernester eds) Plenum Press, New York pp 253-260 Penttila, T. (1983) Eur. J. Biochem. 188, 355-361. Polakis, P., and Wilson, J. (1985) Arch. Biochem. Biophys. 286, 328-337. Powers, L., Chance, B., Ching, Y., and Lee, C.-P. (1987) J. Biol. Chem. 262, 3160-3164. Poyton, R.O., and Schatz, G. (1975a) J. Biol. Chem. 256, 752- 761. Poyton, R.O., and Schatz, G. (1975b) J. Biol. Chem. 256, 762- 766. Prochaska, L.J., Bisson, R., and Capaldi, RA. (1980) Biochemistry 19, 3174-3179. Prochaska, L.J., Bisson, R., Capaldi, R.A., Steffens, G.C.M., and Buse, G. (1981) Biochim. Biophys. Acta 681, 360- 373. 137 Puetner, 1., Carafoli, E., and Malatesta, F. (1985) J. Biol. Chem. 266, 3719-3723. Puustinen, A., Finel, M., Virkki, M., and Wikstrom, M. (1989) FEBS Lett. 249, 163-167. Ratman, M., Sargent, P., Sarin, V., Fox, L.,Nguyen, D., Rivier, J ., Criado, M., and Lindstrom, J. (1986) Biochemistry 25, 2621-2632. Robinson, N., and Talbert, L. (1986) Biochemistry, 25, 2328 -2335. Robinson, N., Dale, M., and Talbert, L. Arch. Biochem. Biophys. in press. Saraste, M. (1983) Trends Biochem. Sci. 8, 139- 142. Saraste, M., Penttila, T., and Wikstrom, M. (1981) Eur. J. Biochem. 115, 261-268. Sarti, P., Jones, M., Antonini, G., Malatesta, F., Colosimo, A., Wilson, M.T., and Brunori, M. (1985) Proc. Natl. Acad. Sci. (U SA) 82. 4876-4880. Schagger, H., Link, T.A., Engle, W.D., and Von J agow, G. (1986) Meth. Enzymol. 126, 224-237. Schatz, G., and Mason, T. L. (1974) Ann. Rev. Biochem. 48, 51- . Shakespeare, P.G., and Mahler, HR. (1971) J. Biol. Chem. 246, 7649- Sinjorgo, K.M.C., Meijling, J .H., and Muijsers, AD. (1984) Biochim. Biophys. Acta 161, 48-56 Sinjorgo, K.M.C., Steinebach, O.M., Dekker, H.L., and Muijsers, A.O. (1986) Biochim. Biophys. Acta 856, 108- 115 Sinjorgo, K.M.C., Durak, I., Dekker, H.L., Edel, G.M., Hakvoort, T.B.M., Van Gelder, BR, and Muijsers, AD. (1987) Biochim. Biophys. Acta 898, 251-258. Smith, L., and Conrad, H. (1956) Arch. Biochem. Biophys. 68, 403-413. Smith, P.K.,Krohn, R.l., Hermanson, G.T., Mallia, A.K., 138 Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, M.N., Olson, B.J., and Klenk, D.G., (1985). Anal. Biochem. 156, 76-85. Solioz, M., Carafoli, E., and Ludwig, B. (1982) J. Biol. Chem. 251, 1579-1582. Speck, S.H., Dye, D., and Margoliash, E. (1984) Proc. Natl. Acad. Sci. (USA) 81, 347-351. Spiker, S., and Isenberg, I (1983) Methods Enzymol. 91, 214-226. Steffens, G.J., and Buse, G (1979) in "cytochrome oxidase" T.E. King et al eds. Steffens, G., Buse, G., Opplinger, W., and Ludwig, B. (1983) Biochem. Biophys. Res. Comm. 116, 335-340. Steffens, G.C.M., Biewald, R., and Base, G. (1987) Eur. J. Biochem.1fi, 295-300. Steinrucke, P., Steffens, G.C.M., Panskus, G., Buse, G., and Ludwig, B. (1987) Eur. J. Biochem. 161, 431-439. Stevens, T.H., Martin, G.T., Wang, H., Brudvig, G.W., Scholes, GR, and Chan, 8.1. (1982) J. Biol. Chem. 251, 12106-12113. Suarez, M., Revzin, A., Narlock, R., Kempner, E., Thompson, D.A., and Ferguson-Miller, S. (1984). J. Biol. Chem. 259, 13791-13799. Taha, T. , and Ferguson-Miller, S. (1987) Fed. Proc. 46, 1930. Takamiya, S., Lindorfer, M.A., and Capaldi, RA. (1987) FEBS Lett. 218, 277-282. Thelen, M., O'Shea, P.S., Pertone, G., and Azzi, A. (1985) J. Biol. Chem. 25.0. 3626-3631. Thompson, D.A., Suarez, M., and Ferguson-Miller, S. (1982) Biophys. J. 81, 285-292. Thompson, D.A., and Ferguson-Miller, S. (1983). Biochemistry 28, 3178-3187. Thompson, D.A., Gregory, L., and Ferguson-Miller, S. (1985) J. Inorg. Biochem. 28, 357-364. Tzagoloff, A. (1982) In "Mitochondria" Plenum Press, New York. Valpuesta, J ., Henderson, R., and Frey, T. (1990) J. Mol. Biol. Veerman, E., Wilms, J ., Casteleijn, G., and Van Gelder, B. (1980) Biochim, Biophys. Acta 596, 117-127. Verheul, F.E.A.M., Draijer, J.W., Dentener, I.K., and Muijsers, A.O. (1981) Eur. J. Biochem. 119, 401-408. Vik, S., Georgevich, G., and Capaldi, R. (1981) Porc. Natl. Acad. Sci. (USA) 18, 1456-1460. Weldon, 8., Mumby, M., Beavo, J ., and Taylor, S. (1983) J. Biol. Chem. 258, 1129-1135 West, I.C., Mitcell, R., Moody, J ., and Mitchell, P. (1986) Biochem. J. 286, 15-21. Wikstrom, M. (1977) Nature, 266, 271-273. Wikstrom, M. (1984) Nature, 868, 558-560. Wikstrom, M., and Saari, H. (1977) Biochim. Biophys. Acta 462, 347-361. Wikstrom, M., Krab, K., and Saraste, M. (1981). In "Cytochrome c oxidase: A synthesis" Academic press, London-New York. Wikstrom, M., Saraste, M., and Pentilla, T. (1985) in "The enzymes of biological membranes" Martonosi, A. editor. Plenum Publishing Corp. Wilson, J. (1987) in “Chemical modification of enzymes: Active site studies”. J. Eyzaguirre editor, Ellis Horwood Ltd. J. Wiley and Sons, N.Y, Chichester, Brisbane, Toronto. pp 171-181 Wilson, J ., and Smith, A. (1984) Anal. Biochem. 148, 179-187. Wilson, M.T., Peterson, J ., Antonini, E., Brunori, M., Colosimo, A., and Wayman, J. (1981) Pro. Natl. Acad. Sci. 18, 7115-7118. Winter, D.B., Bruyninckx, W.J., Foulke, F.G., Grinich, N.P., 140 and Mason, HS. (1980) J. Biol. Chem. 255, 11408-11414. Yewey, G.L., and Caughey, W.S. (1988) Ann. NY. Acad. Sci. 556, 22-32. Yoshikawa, S., Tera, T., Takahashi, Y., Tsukihara, T., and Caughey, W.S (1988) Por. Natl. Acad. Sci. 85, 1354-1358 Zhang, Y., Georgevich, G., and Capaldi, RA. (1984) Biochemistry 28, 5616-5621. Zhang, Y., Lindorfer, M.A., and Capaldi, RA (1988) Biochemistry 21, 1389-1394 "IIIIIIIIIIIIIIIIIIIIT