MECHANISM OF ENERGIZATION OF TRANSHYDROGENASE IN Escherichia coli MEMBRANES by DAVID YEUN BIN CHANG B.Sc, The University of British Columbia, 1988 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Biochemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1990 © David Yeun Bin Chang, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada Department DE-6 (2/88) ii ABSTRACT Low concentrations of the group IIA metals M g 2 + , C a 2 + , and B a 2 + stimulated energy-independent transhydrogenase activity. High concentrations of M g 2 + inhibited this activity. Transhydrogenase requires Mg 2+-complexed NADP(H) rather than free NADP(H) as its substrate. High concentrations of M g 2 + , however, may change the conformation of the enzyme to inhibit its enzymatic reaction by binding directly to the NADP(H) site. Upon transhydrogenation between NADPH and 3-acetylpyridine dinucleotide, E. coli pyridine nucleotide transhydrogenase can establish a proton gradient across the cell membrane. The primary component of the proton gradient for energization of transhydrogenase was found to be the pH gradient and not the membrane potential. A similar conclusion was drawn for the ATP-driven transhydrogenase reactions. ' In strains of E. coli that harbored plasmids to give the cells elevated levels of transhydrogenase, it was found that uncouplers stimulated the aerobic-driven transhydrogenase reaction. This is a chemiosmotic anomaly and is in contrast to the non-plasmid containing parent strains where uncouplers inhibited the activity. Further investigation revealed that the plasmid strains contained a much lower NADH oxidase activity than the non-plasmid strains and that neither KCN nor QNO can inhibit the aerobic-dependent activity in both types of strains even though they were effective in blocking the respiratory chain. These effects prompted us to inquire whether the anomaly was due to differences in the respiratory chain, but no differences were found between the NADH dehydrogenase activities, quinone and cytochrome contents of the plasmid and non-plasmid strains. The bacterial cells with amplified transhydrogenases induce extra intracellular tubular membrane structures to accomodate the extra proteins (Clark, D.M., Pyridine Nucleotide Transhydrogenase, PhD thesis, University of British Columbia, 1986). Separation of the E. coli membrane vesicles on a shallow sucrose gradient, iii however, did not reveal any differences between the vesicles of the plasmid and non-plasmid strains. Therefore, it seems unlikely that the anomaly is due to the plasmid strains performing a unique form of energization on these induced structures. Finally, it was established by S D S - P A G E and Western blot using anti-transhydrogenase antisera that the plasmid strains express a much higher level of transhydrogenase enzymes in their cell membranes than do the non-plasmid strains. iv TABLE OF CONTENTS Page Abstract ii Table of contents iv List of tables vi List of figures vii List of abbreviations ix Acknowledgement xi I. Introduction 1 A. Theories of Oxidative Phosphorylation 1. Early ideas of energy coupling - Chemical and Conformational 2 hypothesis 2. The Chemiosmotic theory 3 3. Localized Proton Theories 6 4. Collision hypothesis 9 B. The Electron Transport Chain 1. E. coli respiratory chain 10 2. Aerobic electron transport 11 C. Pyridine Nucleotide Transhydrogenase 14 D. Objective of this Thesis 15 II. Materials and Methods A. Materials 17 B. Growth of Bacterial Strains 17 C. Isolation of Cell Membranes 18 D. Determination of Protein Concentration 18 E. Enzymatic Assays 1. Energy-independent transhydrogenase assays 19 2. Energy-dependent transhydrogenase assays 19 3. Fluorescence assays 20 4. Ferricyanide and Qi reductase assays 20 5. NADH, d-NADH, and PMS/ascorbate oxidase assays 20 F. Measurement of Cytochrome Spectra 21 G. Quinone Extraction 22 H. S D S Polyacrylamide Gel Electrophoresis 22 V Page I. Immune-blots 23 J . Plasmid Isolation and Transformation of Competent cells 1. Mini plasmid preparation 24 2. Preparation of competent cells 25 3. Transformation 25 K. Membrane Fractionation 26 III. Results A. Effect of Various Metals on Energy-Independent Transhydrogenase 27 Activity B. Energiztion of Transhydrogenase and Energy-Dependent 30 Transhydrogenase Reactions 1. Energization of transhydrogenase 30 2. Energization of ATP-dependent transhydrogenase activity 35 3. Energization of aerobic-dependent transhydrogenase activity 36 C. The Components and activities of the respiratory chain from plasmid and non-plasmid containing strains of E. coli 1. NADH (d-NADH) oxidase activity and inhibitor effects on the 41 respiratory chain 2. Investigating the cause of low NADH oxidase activity of the 43 plasmid containing strains 3. Investigating the presence of a sodium respiratory chain 44 D. The high transhydrogenase content of the plasmid containing strains 1. Verifying the high transhydrogenase content of the plasmid 47 containing strains 2. Separation of membrane vesicles containing high levels of 50 transhydrogenase IV. Discussion A. Energization of Transhydrogenase 55 B. Energization of ATP-Dependent Activity 56 C. Energization of Aerobic-Dependent Activity 58 D. Heterogeneity among membrane vesicles of E. coli 61 V. Conclusion 61 VI. References 63 VI LIST OF TABLES Page 1. Oxidase activity with various substrates of the respiratory chain 42 2. Reductase and oxidase activity of the bacterial NADH dehydrogenase 42 complex 3. Quinone content of inverted bacterial membrane vesicles 42 LIST OF FIGURES P a g 1. Schemes of mitochondrial energy transduction 4 2. Bulk and localized proton circuits 8 3. E. coli respiratory chain 12 4. Effect of metal ions on energy-independent transhydrogenase activity 28 of strain JM83pDC21 5. Effect of high concentration of magnesium on energy-independent 29 transhydrogenase assays 6. Diagram of energy-independent and energy-dependent transhydrogenase 31 assays 7. Sample traces of quinacrine fluorescence and energy-dependent 32 transhydrogenase assays 8. Effect of uncouplers, ionophores, and respiratory chain inhibitors on 34 energy-independent transhydrogenase activity and ATP-dependent proton translocation in inverted membrane vesicles of strain JM83pDC21 9. Effect of uncouplers, ionophores, and respiratory chain inhibitors on 37 ATP-dependent transhydrogenase activity and ATP-dependent proton translocation in inverted membrane vesicles of strain JM83pDC21 10. Effect of uncouplers, ionophores, and respiratory chain inhibitors on 39 aerobic-dependent t?anshydrogenase activity and respiratory-dependent proton movement in inverted membrane vesicles of the plasmid containing strain JM83pDC21 11. Effect of uncouplers, ionophores, and respiratory chain inhibitors on 40 viii aerobic-dependent transhydrogenase activity and respiratory-dependent proton movement in inverted membrane vesicles of non-plasmid containing strain JM83 12. Comparison of the cytochrome content of the non-plasmid to the 45 plasmid containing strains of JM83 and JM83pDC21 13. Detail analysis of the cytochrome alpha band region in non-plasmid 46 and plasmid containing strains of JM83 and JM83pDC21 14. S D S polyacrylamide gel electrophoresis of membrane proteins from the 48 plasmid and non-plasmid containing strains 15. Western blot of membrane proteins from the plasmid and non-plasmid 49 containing strains 16. Separation of E. coli membrane vesicles by sucrose density 51 centrifugation 17. Transhydrogenase and NADH-ferricyanide activities of sucrose density 52 gradient fractions 18. Growth curves of plasmid and non-plasmid containing strains 54 ix LIST OF ABBREVIATIONS AcNAD(H) 3-Acetylpyridine adenine dinucleotide ADP Adenosine-5'-diphosphate Asc Ascorbate ATP Adenosine-5'-triphosphate CCCP Carbonyl cyanide m-chlorophenylhydrazone DCCD N, N'-dicyclohexylcarbodiimide dh Dehydrogenase DNA Deoxyribonucleic acid DNase Deoxyribonuclease DTT Dithiothreitol d-NADH Deamino nicotinamide adenine dinucleotide EDTA (Ethylenedinitrilo)-tetraacetic acid E. coli Escherichia coli FAD Flavin adenine dinucleotide FMN Flavin mononucleotide F-i Enzymatic unit of ATPase F 0 Proton channel unit of ATPase Hepes N-2-hydroxyethylpiperazine NAD(H) Nicotinamide adenine dinucleotide NADP(H) Nicotinamide adenine dinucleotide phosphate Nig. Nigericin PAB Penassay broth PAGE Polyacrylamide gel electrophoresis Pi inorganic phosphate PMS Phenazine Methosulfate Q Quinone QNO 2-n-Heptyl-4-hydroxyquinoline N-oxide Qi Ubiquinone 1 SDS Sodium dodecyl sulfate TBTC Tri-n-butyltin chloride TCS 3, 3', 4', 5-Tetrachlorosalicylanilide TEMED N.N.N'.N'-tetraacetic acid TMAO Trimethylamine-N-oxide TPTC Triphenyltin chloride Tris Tris (hydroxymethyl)-aminoethane Vai. Valinomycin v/v Volume to volume ratio w/v Weight to volume ratio w/w Weight to weight ratio x g Times gravity AE Extinction coefficient Ap Proton motive force ApH Difference in pH across the membrane Proton electrochemical potential Ay Electrical potential xi ACKNOWLEDGEMENTS I am grateful to my teacher and supervisor, Dr. P.D. Bragg, for providing me with the opportunity and means to work in his lab. His patience and encouragement was helpful during these years. To Dr. Molday and Dr. Brownsey, I am thankful for their suggestions and understanding in my getting the writing of this thesis started. My sincere thanks to the people of our lab; Cynthia for her technical assistance and patience in putting up with me and all my friends who frequented the lab; Dr. Ted Sedgwick for his advice and witty remarks that fried the intellect; Natalie for her everpresent friendly smile that made our lab a much happier place to work in; and finally Ray for his late night companionship and off-earth sense of humour which helped us endure 'till "the death of day". I also wish to acknowledge the financial support from the University Graduate Fellowship. For with much wisdom, comes much sorrow; the more knowledge, the more grief. Ecclesiastes 1:18 1 I INTRODUCTION Over the past 40 years, much work has been done to elucidate how energy made available by the oxidation of substrates or by the absorption of light is used to synthesize ATP or to actively transport ions across the membrane. The "energy transducing" membranes which perform these functions vary from the plasma membrane of bacteria or blue green algae, to the thylakoid membrane of chloroplasts and the inner membrane of mitochondria (55). Yet in spite of this large diversity, sufficient similarities exist between these systems to allow uniform study of oxidative phosphorylation. In a simplified form, most energy transducing membranes can be viewed as consisting of two main protein assemblies. The first is the ATP synthase. Its function is to catalyze the energy requiring synthesis of ATP from ADP and Pj or visa versa. It is ubiquitous to most energy transducing membranes, though some minor alterations in structure may exist between organisms (89). The nature of the second assembly, the electron transport chain, is more complex and varies a great deal from system to system. One example is the mitochondrion which has a respiratory chain that can be operationally defined to four distinct complexes plus ubiquinone and cytochrome c. It oxidizes NADH as its primary energy source and reduces oxygen as its terminal electron acceptor. Most mitochondria have this pathway only (90). In contrast, E coli can synthesize several different respiratory chains depending on the carbon source, terminal electron acceptor, and growth conditions. Often, more than one respiratory chain is synthesized at a time (29). The chloroplast and photosynthetic bacteria also have complex and varied respiratory chains but they use light as their primary energy source (55). In all, the final product of the second assembly is to provide energy which can be used to drive "uphill" reactions such as ATP synthesis for the first assembly. 2 The mystery which remains to be solved is the nature of the "energy-transducing intermediate" linking these pairs of protein assemblies. Many hypotheses have been proposed but none have yet provided a satisfactory explanation. Nevertheless, from time to time during the development of ideas about oxidative phosphorylation, one or two concepts have dominated the thinking of the workers in the field. For example, hypotheses involving chemical coupling, conformational coupling, chemiosmosis, and localized protons have been proposed and they will be described in greater detail below. A. Theories of Oxidative Phosphorylation 1. Early ideas of energy coupling - Chemical and Conformational Hypothesis In 1939, Warburg and Christian (1) put forward the idea that ATP is synthesized by high energy phosphorylated intermediates during glycolysis. In 1941, Lipman established that ATP is the unit of "energy currency" in the cell. With the success of these two concepts, the general opinion was that a system of high energy compounds was the only convertible and transportable form of energy in the living cell. Therefore, when researchers were faced with the problem of oxidative phosphorylation, they naturally sought a high energy intermediate. The first results indicated that a high energy non-phosphorylated intermediate may be involved in oxidative phosphorylation. In experiments with mitochondria, it was demonstrated that substrates can be oxidized in the complete absence of ADP and phosphate. These observations, in direct contrast with glycolysis, prompted Slater (3) in 1953 to suggest an unknown high energy compound (x~y), possibly a thioester, to be the ATP precursor in oxidative phosphorylation. This hypothesis seemed for a time to receive support from experimental evidence. Lardy et al. (4) hypothesized that oligomycin prevented energy transfer between x*y and ATP. Furthermore, x-y produced by the second and third energy coupling sites of the 3 respiratory chain was proposed to drive reverse electron transport. Uncouplers were postulated to act by hydrolyzing the x~y compound. Some researchers even claimed to have isolated this high energy intermediate (26). Most of the cases were later proved to be based on faulty interpretation. Eventually, this "chemical coupling" hypothesis had to be abandoned. But the idea of a high energy intermediate did not fade away and in 1965 Boyer (5) suggested that the conformations of proteins may change during oxidation and reduction and that similar conformational changes may occur in the ATP synthase. He proposed that conformational changes in the respiratory chain components during oxidation and reduction were transmitted directly to a closely located ATP synthase and that the resultant "strain" induced in the latter provided energy for ATP synthesis. The evidence supporting this hypothesis has been at best circumstantial. But despite the lack of supporting evidence, the Concept of protein conformational changes has gained some support more recently as a possible mechanism for proton translocation in the chemiosmotic theory. 2. The Chemiosmotic Theory In spite of the popularity of the chemical coupling and conformational hypotheses, the theory that has dominated the bioenergetics field since its introduction in 1961, is the Chemiosmotic theory proposed by Peter Mitchell (6). This theory denounced the existence of any high energy compounds and suggested instead a protonmotive force (fig. 1). The essence of Mitchell's proposal as applied to mitochondria is that: (a) the respiratory chain is asymmetrically organized in the mitochondrial inner membrane such that protons are vectorially translocated from the inside to the outside of the mitochondria as electrons are passed from substrate to oxygen; (b) the inner mitochondrial membrane is essentially impermeable to most ions, including both OH" and H + . Consequently, the electrogenic pumping of protons 4 a Chemical hypothesis I II EH NADH : 0 2 Transhydrogenase uncouplers c a t i o n t r a n s l o c a t i o n o l i v o m y c i n AOP + Pi =^= ATP b. Chemiosmotic hypothesis i n TH NAOH *- 0 2 Transhydrogenase " _ A £ H + cation translocation uncouplers - oliqomycin AOP + Pi ATP Rq. 1: Schemes of mitochondrial energy transduction Mechanisms of membrane energization for mitochondria according to (a) chemical hypothesis and (b) chemiosmotic hypothesis. [~] represents an unknown 'high energy* intermediate. Figure reproduced from Nicholls, 1982 (55). 5 across the membrane creates an electrochemical gradient (AU,H) which consists of both a pH gradient (ApH) and an electrical potential (Ay): A|J.H = A y - Z ApH where Z=2.303RT/F and serves to convert ApH into electrical units; (c) the electrochemical potential produced can then be used to drive the synthesis of ATP, translocate ions, perform reverse electron transport, drive energy-linked transhydrogenase reactions, or other forms of active transport processes. An analogous explanation is postulated for oxidative phosphorylation in bacteria and photophosphorylation in chloroplasts and photosynthetic bacteria. At first, Mitchell's theory was not widely accepted, but a large flood of evidence soon poured in to support his hypothesis. It is briefly summarized here: 1. Respiration and ATP hydrolysis can generate an electrochemical proton gradient. Direct measurements of proton translocation by energy transduction systems were first reported in 1964 by Neumann and Jagendorf (7) for chloroplasts and by Mitchell in 1965 (8) with mitochondria. 2. An electrochemical gradient will lead to ATP synthesis. Jagendorf and Uribe (9) showed in 1966 that when an artificial A|XH is imposed on chloroplasts, ATP . can be synthesized. Furthermore, Racker and Stoeckenius (10) used Au. H generated by illuminating bacteriorhodopsin in liposomes containing ATP synthase and bacteriorhodopsin to drive ATP synthesis. 3. The electrochemical gradient was shown to be a kinetically competent intermediate in oxidative phosphorylation by Thayer and Hinkle (11). They reported that ATP synthesis by an artificially imposed A(XH in submitochondrial particles is as fast or faster than the rate of ATP synthesis by respiration. 4. Uncouplers may act by dissipating the proton gradient (12). 6 The chemiosmotic theory reached its height of popularity by the late 1970's. Since then, evidence has been accumulating which argues against Mitchell's theory as originally put forward. A few of the more impressive data will be presented here: 1. Uncoupler-resistant mutants exist. One such example is Bacillus megaterium which can synthesize ATP even when Ann is abolished by uncouplers (13). 2. Ap.H can be very low in alkalophilic (14) and halophilic (15) bacteria under conditions in which they can synthesize ATP. For instance_£?ac/7/us- — alkalophilus, when grown with an external pH of 11, has only a Ap of 15 mV and yet it is still able to synthesize ATP. 3. ATP synthesis can be abolished without affecting A^LH by a group of substances known as decouplers (16). 4. Open membrane fragments of rabbit skeletal muscle have been proposed to be capable of energy transduction (17), when they should be structurally incompetent according to the chemiosmotic theory which requires enclosed membrane vesicles. These discrepancies have dramatically lessened the attractiveness of the chemiosmotic theory. Many workers in the field have tried to search for an alternative explanation. Some have taken refuge in the belief that localized protons are the intermediates. 3. Localized Proton Theories At about the same time that Mitchell proposed his chemiosmotic hypothesis, Williams (18) made an independent proposal that also involved protons as the intermediate in oxidative phosphorylation. He suggested that protons produced as a result of electron transfer are delivered directly to the ATP synthase where they promote the condensation of inorganic phosphate and ADP to form ATP. In this formulation, the protons remained in the membrane phase (fig. 2). It differs from 7 Mitchell's hypothesis in that a transmembrane potential is not necessary for ATP synthesis. In fact, the appearance of a bulk proton gradient is envisioned as a "leak" that is counter-productive to the energization of the membrane. Williams has had to modify his theory several times as new facts emerged. One of his most recent modifications is that direct proton binding on ATP synthase acts to drive two conformational changes (19). One releases ATP from F i , the other opens the F 0 channel to the inside of the mitochondria, allowing downhill diffusion of protons. This modification was made in light of the finding that the energy-requiring reaction in ATP synthesis is the dissociation of ATP and not the esterification of ADP. Several other interesting "localized proton" theories have also been proposed by other researchers. Kell (20) suggested a 'protoneural' network of proteinaceous components that conducts protons between various protonmotive sources and proton accepting sinks by allowing rapid movement of protons along the surfaces of energy-transducing membranes. Along the same line of thought, Prats et al. (21) postulated the polar head groups of phospholipids at the surface of membranes to act as pathways for rapid proton conduction. This might occur by 'proton hopping' via a Grotthus mechanism (22). Westerhoff and colleagues (23) put forward the model of "mosaic chemiosmosis". They envisioned a mitochondrial membrane with different compartments for protons, all in communication with the bulk phase, but through a barrier of significant resistance. The protons are still translocated by a chemiosmotic mechanism but in small independently operating coupling units consisting of several respiratory chain assemblies associated with a limited number of ATP synthase molecules. Using a completely different approach to this problem, Malpress (24) proposed a "coulombic hypothesis" where the primary mediating force in ATP synthesis is the electrostatic interaction generated between protons and fixed negative charges in selected, localized areas on the surface of the energized membrane. This is similar 8 outside or periplasm membrane cytoplasm (a) redox system \ ^ ADP + Pi i -ATP + H30 (b) redox system H" ADP + Pi ATP + H20 Figure 2: Bulk and localized proton circuits Schematic diagram of the proton ciruit according to (a) the chemiosmotic hypothesis, and (b) the localized proton hypothesis. F0F^ symbolizes the ATPase. The dotted line in (b) represents a proton 'leak' to the bulk gradient. Figure modified from Jones, 1982 (57). 9 to the "local electric field" hypothesis put forward earlier by Skulachev (25). He considered the concept of local electric fields arising due to charge separation in a molecule of Au.H-generating enzyme in affecting the adjacent Au.H-utilizing enzyme such as ATP synthase. 4. Collision Hypothesis In 1985, Slater and colleagues (26) postulated what they termed as the "collision hypothesis". They suggested that collision between freely diffusable energized redox enzymes and the ATP synthase is the manner by which energy is transferred between these two systems. This proposal may be written as: Ared + B o x + E f ^ 1 Ao X+Bred + E r ~ h + 5 = 4 E r + A|I H (1) E r~h+ + EATP i '* E r + E ATp~ h + i E A T P + Au.H (2) E ATP~h+ + ADP + Pi 5=*E A TP + ATP (3) Reaction 1 describes a redox reaction, the end product of which is the 'high energy' conformation of the redox enzyme (E fh + ) . This energy can be utilized either to create an electrochemical gradient (A;IH) or can be transferred to the ATP synthase to form an energized form of the synthase (EATPTI +) as shown by reaction 2. Subsequently, the energized synthase can use its energy to synthesize ATP or create AJJ.H (reaction 3). So similar to chemiosmosis, the redox enzymes and ATP synthase can still behave as proton pumps that deliver protons to the bulk phase. But the major route of energization is the transfer of energy via direct collision between the two molecules. A hypothesis almost identical to this one was independently proposed by Boyer (27) at about the same time. 10 Both their theories implicated the action of protonophoric uncouplers to be the direct interference between the redox enzymes and ATP synthase. Herweijer et al. (28) showed that the concentration of uncouplers needed is related only to the concentration of the interacting enzymes concerned with energy transduction. Evidence for this is found in the case of the uncoupler S13. Herweijer and colleagues demonstrated that low concentrations of S13 affected the direct energy transfer whereas high concentrations exhibited the classical mode of chemiosmotic uncoupling. That is, S13 uncoupled by behaving as a proton shuttling system. B. The Electron Transport Chain 1. E. coli Respiratory Chain To fully appreciate how the mechanism of energy coupling might operate, one must have some understanding of the components which make up the energy transducing membrane. One of the major assemblies in this membrane is the respiratory chain. The respiratory chains of E. coli and mitochondria are very similar in their overall arrangement. In both, substrates are oxidized by dehydrogenases which are often flavoproteins associated with iron-sulfur centers. The dehydrogenases reduce quinones which in turn transfer electrons to cytochromes. The cytochromes or complexes containing cytochromes then transfer electrons to the terminal electron acceptor such as oxygen (29-30). In contrast to the one major respiratory pathway of mitochondria, E. coli can synthesize a variety of respiratory chains depending on the conditions. Thus E. coli has the ability to induce several different dehydrogenases, each oxidizing specific substrates which may be NADH, formate, lactate, succinate, or a-glycerophosphate. The reduced dehydrogenases may subsequently reduce one or both species of ubiquinone-8 and menaquinone-8 (33). The reduced quinones in turn can terminate 11 in two different pathways under aerobic conditions. Under anaerobic conditions, the terminal electron acceptor may be nitrate, fumarate, or trimethylamine-N-oxide (TMAO) (29-32). Though these pathways are widely branched, there appears to be some favoured interactions between certain electron donors and specific electron acceptors. For instance, formate tends to donate its electrons to nitrate reductase while NADH is the preferred donor of electron to oxygen. Since the present study involves only the aerobic respiratory chain of E. coli, this will be described in more detail. 2. Aerobic Electron Transport There are two aerobic respiratory pathways in E. coli; one terminating with cytochrome o and the other with cytochrome d. Both may, and often do, exist simultaneously under aerobic growth conditions. They share the same NADH dehydrogenases and both involve ubiquinone-8. They differ from each other at the level of the cytochromes (29). Two immunologically distinct NADH dehydrogenases are found in the cytoplasmic membrane of E. coli (34). Matsushita et al. (35) differentiated the two enzymes by demonstrating that one enzyme system oxidizes both NADH and deamino NADH (dNADH), involves [4 Fe-4 S] and [2 Fe-2 S] type iron sulfur clusters, and is coupled to the generation of an electrochemical proton gradient. This is called NADH dh I. The other enzyme system, named NADH dh II, oxidizes NADH exclusively, has no iron-sulfur clusters, and does not generate a AU.H- Hayashi et al. (36) further characterized the systems by showing that NADH dh I contains flavin mononucleotide (FMN) as a cofactor and catalyzes the reduction of ferricyanide. On the other hand, NADH dh II contains flavin adenine dinucleotide (FAD) as a cofactor and catalyzes the reduction of ubiquinone-1 (Q-i). Their ability to selectively reduce 12 SH2 Dehydrogenases Flavoprotein Fe-S protein Ubiquinone (Menaquinone) 1/2 0 2 + 2 H 1/2 02 + 2 H Cytochrome o Complex Cytochrome d Complex Figure 3: E. coli Respiratory Chain Proposed arrangement of components in the respiratory chain of E. coli. No specific order is implied for the components listed in the boxes for dehydrogenase, cytochrome o and d complex. Fe-S, iron sulfur centre; cyt, cytochrome; S , Substrates. 13 either ferricyanide or Qi has formed the basis for enzymatic assays to distinguish between them. E. coli can synthesize both ubiquinone-8 and menaquinone-8, although cells grown aerobically generally have a higher concentration of ubiquinone-8 than menaquinone-8. The converse is true under anaerobic growth conditions (37). However when the activity of the aerobic respiratory chain is impaired such as in heme deficient mutants, menaquinone is synthesized in high concentrations (38). It is not understood how the synthesis of the two quinones is" regulated, but it is known that under some circumstances, they can substitute for each other to maintain electron transport activity (39). When E. coli is grown under vigorous aeration, a set of redox carriers that involve ubiquinone-8 and the cytochrome o complex are induced. The latter consists of the cytochromes bs5 4 (555). bs64 (563) and probably bss7.5 (91). Which one of these cytochromes contains the oxygen reacting heme for the complex (cytochrome 0) is not yet known. However, we know this heme reacts with carbon monoxide and is sensitive to cyanide. Kita and Anraku (93) have proposed this heme to be the one responsible for the peak at 555 nm whereas Withers and Bragg (91) argue that it is the heme which exhibits an absorbance peak at 564 nm. The cytochrome d pathway is induced under conditions of low aeration. Cytochrome d differs from cytochrome 0 in that it has different spectral properties and is much less sensitive to inhibition by cyanide. Moreover, cytochrome bsss and bsgs are associated with cytochrome d. Cytochrome bsss has been suggested to reduce cytochrome d but the role of cytochrome bsgs, however, still remains a mystery (29). In addition to induction under low oxygen tension, the cytochrome d pathway has also been observed in a variety of apparently unrelated growth conditions such as aerobic growth in the presence of cyanide (41) or with limiting concentrations of sulfate (42). 14 Both of these pathways are usually present in cells grown aerobically. Figure 3 gives a general sequential order of electron transfer in both these pathways. C. Pyridine Nucleotide Transhydrogenase Pyridine nucleotide transhydrogenase is an integral enzyme found in the cytoplasmic membrane of E. coli and in the inner membrane of mitochondria. It catalyzes a reversible and direct transfer of a hydride ion equivalent between oxidized and reduced forms of NAD(H) and NADP(H) (44, 45). The transhydrogenase reaction is coupled to proton translocation across native mitochondrial membranes according to the following reaction (46): NADPH + NAD+ + n H m a t r j x s ** NADP+ + NADH + n Hc yto8oi where n is the number of protons translocated per hydride transferred. Values ranging from 0.2 - 2 have been determined for n, though no definitive value has yet been agreed upon (47). Kinetic data indicates the enzyme follows a random ternary-complex mechanism (95). The transhydrogenase reaction can be driven by other energy transfer systems in the membrane such as the respiratory chain and the ATPase. These 'energy-dependent' transhydrogenase reactions produce a 5-10 fold increase in the rate (48) and a 500 fold increase in the equilibrium constant for the reduction of NADP+ by NADH (49). Bovine heart mitochondrial transhydrogenase has been purified to homogeneity employing methods that involved affinity chromatography on immobilized NAD+ (96) or NADP+ (97) and fast protein liquid chromatography (98). The E. coli transhydrogenase has also been purified by pre-extraction of the cytoplasmic membrane with sodium cholate and Triton X-100 followed by 15 solubilization of the enzyme with sodium deoxycholate in the presence of 1 M potassium chloride and centrifugation through a discontinuous sucrose gradient (99). Reconstitution studies of both mitochondrial and E. coli enzymes have confirmed that it is a proton pump, using 9-aminoacrine as a pH probe (99, 100). Structurally, transhydrogenase is a relatively simple enzyme. Its amino acid sequences from bovine mitochondria (50) and £ coli (51) have been deduced via gene clonmg and sequencing. The mitochondrial transhydrogenase was found to be a homodimer whose monomers have a Mr=110,000. In contrast, the E. coli enzyme is composed of two types subunits a (Mr=54,000) and (3 (Mr=48,700), and exists as an C12P2 tetramer (94). The NAD(H) and NADP(H) binding sites are believed to be near the N and C termini respectively of the bovine mitochondrial enzyme (52). These domains correspond to the N terminal region of the a and the C terminal region of the p subunit of E. coli transhydrogenase respectively. Hydropathy plots predict as many as 14 transmembrane regions for the mitochondrial transhydrogenase (50) and 12 for the E. coli transhydrogenase (72). A number of other facts are also known about the enzyme. For example, cysteine residues are believed to be involved in its catalytic activity (53). The enzyme can undergo large, energy-induced conformational changes during catalysis (54). D C C D inactivates both the proton translocating and catalytic activity of the enzyme (101). Magnesium inhibits the enzyme (102). Yet despite all these data, the physiological significance of this enzyme is still unclear. D. Objective of this thesis The structural simplicity of the transhydrogenase, the existence of convenient real time assays, together with its unique relationship to other energy transducing systems has made this enzyme an attractive system by which to study the mechanism of membrane energization. Indeed, Slater (3) had interpreted the energy-dependent 16 transhydrogenase reaction in terms of using a non-phosphorylated high energy intermediate during the era of the "chemical" hypothesis. This was reinterpreted a decade later in favour of the chemiosmotic theory by Mitchell (55) as shown in figure 1. The work from this thesis shall present some evidence for the "collision hypothesis". Prior to the commencement of this project, Clarke and Bragg (56) had cloned the E. coli transhydrogenase gene into a multicopy vector pUC 13 after screening for clones in the Clarke and Carbon colony bank (56) that showed elevated levels of transhydrogenase activity. The recombinant plasmid was termed pDC21. The plasmid was then transformed into the E. coli JM83 host to create the transformant JM83pDC21F*. This transformant showed a 70 fold increase in transhydrogenase activity in comparison to the host strain. This thesis will describe the pattern of energization of the transhydrogenase enzyme by itself and under conditions whereby it is energized by other systems (i.e. electron transport chain or ATPase). The strategy pursued was one where the enzymatic and proton translocating activities were measured with a wide variety of uncouplers and ionophores. Both the parent and transformed strains of JM83 and W6 were compared for their mode of energization. Furthermore, the respiratory chains of these strains were investigated spectrometrically in an attempt to detect any differences between the parent and transformed strains. Finally, a partial separation of the transhydrogenase containing vesicles from the vesicles containing respiratory chain components was demonstrated for JM83pDC21 F". 17 II MATERIALS AND METHODS A. Materials All chemicals were obtained from commercial sources and were either of reagent or analytical grade. Special chemicals were purchased from the following suppliers: DIFCO -Bactotryptone, agar, and yeast extract SIGMA -NAD(H), NADP(H), 3Ac NAD(H), yeast alcohol dehydrogenase, ATP, QNO, DCCD, deamino NADH, ampicillin, CCCP, valinomycin, nigericin, PMS BIORAD -All polyacrylamide gel electrophoresis reagents and molecular weight standards, alkaline phosphatase conjugated goat anti-rabbit IgG, nitrocellulose membrane Nutri. Biochem. Co. -Quinacrine (Atebrin) BDH FISHER Promega Biotec Eastman J.T. Baker Chem. Alfa products -Hydrogen peroxide, sodium dithionite, ascorbate -Potassium ferricyanide, potassium cyanide -(BCIP/NBT) 5-bromo-4-chloro-3-indoyl phosphate/nitro blue tetrazolium alkaline phosphatase development system -TCS -Thiamine -TBTC, TPTC B. Growth of Bacterial Strains The E. coli strains of JM83pDC21 F - ara A lac pro strA thi or N a + and assayed activity of transhydrogenase in a medium containing Na+ or K> respectively. Addition of valinomycin gave high rates of NADPH to NAD+ and NADH to NADP+ transhydrogenation when K + movement occured in a direction opposite to that proposed for transhydrogenase-coupled proton movements. Furthermore, they claimed that pH gradients had very little effect in transhydrogenation. In contrast to the above experiments which were done with artificial proteoliposomes and purified transhydrogenase, the work done in this thesis has concerned the energetics of everted, native membrane vesicles. Fig. 8(a) showed that uncouplers stimulated transhydrogenase activity suggesting that respiratory control may exist in these vesicles from a balancing of the electrochemical hydrogen ion gradient across the membrane against the substrate oxidation-reduction potential. Dissipating the proton gradient, as evidenced by the relieving of the fluorescence quenching with uncouplers (fig. 8b), removes this control and consequently transhydrogenation is stimulated. These data establish that a electrochemical 5 6 gradient is involved in transhydrogenation but it does not reveal whether ApH or A y or both are important in this process. The importance of ApH and A y is unveiled by the results of experiments with the ionophores valinomycin and nigericin (added separately). These results imply that ApH is the most important component of the electrochemical gradient which drives transhydrogenase and not A y . TBTC and TPTC's effects on the energy independent assays do not agree with the above explanation. This may be rationalized by arguing that these agents do not function as OH7CI- exchangers in the assays as previously mentioned. Singh and Bragg (69) have proposed an alternative mechanism of action for TBTC on transhydrogenase energetics. They suggest that TBTC may be a sulfydryl reacting agent that acts directly on the transhydrogenase. Since critical sulfydryl groups exist on the transhydrogenase (79), TBTC's action may well inhibit both the enzymatic and proton-translocating ability of the enzyme. TPTC should have a similar effect because it differs from TBTC only in having phenyl groups instead of butyl groups in its structure making it more membrane soluble than its counterpart. B. Energization of ATP-dependent activity ATP dependent transhydrogenation was found to be inhibited by uncoupler (80) and specifically by inhibitors of phosphorylation such as oligomycin and DCCD (81). Because of these properties, ATP was hypothesized to drive the transhydrogenase reaction by creating a AU.H as predicted by Mitchell. However, in submitochondrial particles, Ernster et al. (82) have demonstrated a direct relationship between inhibition of the ATPase activity with increasing amounts of the Pullman-Monroy inhibitor and inhibition of the ATP-linked transhydrogenase reaction. This relationship exists in spite of the large excess of ATPase over transhydrogenase. The classical chemiosmotic theory would not have predicted this because inhibiting the excess ATPase should not adversely affect the AU.H which energizes the ATP-57 dependent transhydrogenation. These findings were interpreted as a localized interaction between ATPase and transhydrogenase with each ATPase being able to "energize" transhydrogenase in a limited domain within the membrane. Furthermore, Kay and Bragg (83) have isolated a mutant strain of Salmonella typhymurium HfrA that cannot perform ATP-dependent transhydrogenation but does have a functional transhydrogenase and is capable of carrying out all its other energy-dependent processes including ATP-dependent quenching of atebrin fluorescence. The properties of this mutant are also difficult to explain in a chemiosmotic manner. Therefore, some controversy exists about the mechanism of ATP-dependent transhydrogenation. The work in this thesis investigated the ATP-dependent energization of everted, E. coli membrane vesicles. Fig. 9 indicated that the addition of uncouplers inhibited this reaction and dissipated fluorescence quenching for both the plasmid and non-plasmid containing strains to a similar extent. These results argue for a bulk proton gradient to be the energy intermediate between ATPase and transhydrogenase. Furthermore, the effects of valinomycin and nigericin (added separately) suggest once more that ApH is the important component of the electrochemical gradient in transhydrogenase. The ability of both TBTC and TPTC to inhibit both the enzymatic and fluorescence quenching activities support the previous notion that these compounds are sulfydryl agents which act directly on the transhydrogenase. If they behaved as electroneutral OH7CI- exchangers, they would have stimulated the ATP-dependent reaction and maintained or stimulated fluorescence quenching by increasing the concentration of protons inside the membrane vesicle. Lastly, inhibitors of electron transport were found not to significantly affect the ATP-driven transhydrogenase reaction in agreement with Ernster and Danielson's work (84) with submitochondrial particles. 58 C. Energization of aerobic-dependent activity Comparatively little work has been done on the aerobic-driven transhydrogenase reaction. The reason is that no simple system has yet been developed to study this process in a systematic way. The work described in this thesis implicates the existence of a localized pathway between the respiratory chain components and transhydrogenase in E. coli membrane vesicles. R g . 10 and 11 show that addition of uncouplers stimulated the aerobic-dependent activities of transhydrogenase in the plasmid containing strains but inhibited the same activities in the non-plasmid containing strains. The chemiosmotic theory would have predicted an inhibition of the enzymatic activities for both cases since the uncouplers should dissipate the aerobically generated internal protons used to energize the transhydrogenase reaction (see fig. 6). One explanation may be i that the uncouplers do not fully dissipate the proton gradients in the plasmid strains. i Although the fluorescent quenching data (fig. 10) show that indeed the proton gradient is relieved upon uncoupler addition, nevertheless, one must be careful with the interpretation of these data. For example, fluorescent probes may not probe the internal pH of the vesicle exclusively, but may also interact with the surface or the interior of the membrane (85). Quinacrine has been proposed to act by decreasing its fluorescence in response to binding to the "energized" membrane via its diprotonated form (86). The other explanation is that a localized interaction exists between the respiratory components and transhydrogenase in the plasmid containing strain due to higher levels of transhydrogenase protein. This interaction may not exist or may occur to only a very limited extent in the non-plasmid containing strains. The results of experiments with the ionophores agree with the interpretation that the ineffectiveness of valinomycin indicates that A y is not significant in this process. The observation that the presence of nigericin did not affect the aerobic-dependent transhydrogenase activity of the plasmid containing strains but inhibited 59 the activity of the non-plasmid containing strains, confirms the involvement of ApH in this "localized interaction". The inability of nigericin to dramatically stimulate the enzymatic activity as did the uncouplers (fig. 10a) may be due to a difference in mechanism between these two classes of compounds. The effects of TBTC and TPTC can probably be explained as for the case of ATP-dependent activities. An interesting fact is that neither KCN nor QNO inhibited the aerobic dependent activities of both the plasmid and non-plasmid containing strains (fig. 10 and 11) even though the concentrations used were effective in blocking the NADH oxidase activity (table I). This indicates that the aerobic dependent energization process does not require a fully functional respiratory chain. Another anomaly is that the plasmid strains contain a much lower NADH oxidase activity than do the non-plasmid strains (table I and II). This effect is not unique. Weiner et al. (75) have also found that cells containing amplified levels of fumarate reductase have a complete absence of NADH oxidase activity in the cell envelope. The low oxidase activities may originate from a segment of the respiratory chain before the cytochromes, as hinted at by the results with PMS/ascorbate (table I), but no direct evidence was obtained to support this. These anomalous effects together with the great flexibility of E. coli to induce different respiratory chains prompted us to inquire whether the effects were due to the induction of a different respiratory pathway in the plasmid strains. The results from tables II and III and figures 12 and 13, indicated that there were no differences between the NADH dehydrogenases, quinone, or cytochrome content of the plasmid strains and the non-plasmid strains. Furthermore, the anomaly could not be explained by the presence of a sodium-translocating respiratory chain. It appears then that both the plasmid and non-plasmid strains have the same respiratory chain. This is not too surprising considering that different respiratory chains are usually induced only under different growth conditions including the 60 presence of different terminal electron acceptors and carbon sources. Overcrowding the membrane with transhydrogenase probably does not pose a sufficient challenge to the proper respiratory function of the cell to warrant the induction of a different respiratory chain. The other explanation of the chemiosmotic anomaly is the presence of the high transhydrogenase content in the membranes of plasmid containing strains as demonstrated by the S D S - P A G E gel, Western blot, and transformation data (fig.14 and 15). It may be that energy transfer occurs by collision between complexes as suggested by Slater and colleagues (26) in their "collision hypothesis". High transhydrogenase levels in the plasmid strain would promote this localized transport because its presence would increase the frequency of collisions. Uncouplers might act in some yet undefined manner in assisting this mode of energization thus stimulating the aerobic dependent activities. Uncouplers probably do not influence the respiratory chain components directly as allosteric compounds because of the wide variation in structure and mechanism of uncouplers that show this chemiosmotic anomaly. Respiratory chain inhibitors are ineffective in this model due to the ability of the individual high energy respiratory components to directly act with transhydrogenase in energy transduction. Furthermore, NADH oxidase activity would be impaired in the plasmid strains because redox pairs would collide less frequently with each other in the presence of the high transhydrogenase content. This model, however, does not exclude the possibility of a bulk proton gradient in energization. It merely hypothesizes this gradient to be a secondary effect and the collision of energized complexes to be the primary driving force (see introduction). In all, a collision model would provide an explanation, albeit unproven at this stage, to a great number of the observations described in this thesis. 61 D. Heterogeneity among membrane vesicles of E. coli Fig. 16 revealed that membrane proteins are associated with different subsets of membrane vesicles that can be separated by a shallow sucrose gradient. Fig. 17 showed that the transhydrogenase enzymes tend to be localized to one type of membrane vesicle. NADH dh I, and possibly also the other respiratory chain components, tend to gather together in another type. The existence of these associations among vesicles and proteins suggest that there is a situational organization of at least some proteins within the inner membrane of E. coli, possibly by segregating proteins into distinct membrane regions or domains. In other words, proteins do not flow freely in the membrane surface as predicted by the fluid mosaic model. This concept is feasible because the bacterial cell must somehow perform site-specific events such as cell division, elongation or maintainance of shape by organizing its proteins in localized domains (87). The aforementioned separation of different membrane vesicle types was similar for both the plasmid and non-plasmid strains. There was no detectable, specific separation of the induced tubular structures in the plasmid containing strains which would be expected if they contained components that help them perform a distinctly different sort of energetics. However, this separation technique may not give sufficiently resolution to show the difference. Young et al. (88) have used other methods to fractionate bacterial membrane vesicles such as electrophoresis through dilute agarose and sizing chromatography through Sephacryl S-1000. These might yield different results to those described here. V. CONCLUSION The work from this thesis demonstrated that strains of E. coli containing high levels of transhydrogenase have aerobic-dependent transhydrogenase activity that are insensitive to uncouplers. This effect cannot be explained by the classical 62 Mitchell chemiosmotic hypothesis. It is concluded that some form of localized interaction between respiratory chain components and the transhydrogenase possibly involving collision between energized complexes, as suggested by Slater et al. (26), may account for the observed results. 63 VI. REFERENCES 1. Warburg, O. and W. Christian (1939), Biochem. J. 303, 40-68 2. Lipmann, F. (1941), Adv. Enzymol. 1, 99-162 3. Slater, E.C. (1953), Nature 172, 975-978 4. Lardy, H.A., D. Johnson and W.C. McMurray (1958), Arch. Biochem. Biophys. 78, 587-597 5. Boyer, P.D. (1965), Oxidases and related redox systems (King, T.E., H.S. Mason and M. Morrison Ed.) vol 2. 994-1008, Wiley Press, New York 6. Mitchell, P. (1961), Nature 191, 144-148 7. Neumann, J . and A T . Jagendorf (1964), Arch. Biochem. Biophys. 107. 109-110 8. Mitchell, P. and J . Moyle (1965), Nature (London) 208,. 147-151 9. Jagendorf, A T . and E. Uribe (1966), Proc. Natl. Acad. Sci. USA 55, 170-177 10. Racker, E. and W. Stoeckenius (1974), J. Biol. Chem. 249, 662-663 11. Thayer, W.C. and P.C. Hinkle (1975), J. Biol. Chem. 250, 5336-5342 12 : Montal, M., B. Lee and C P . Chance (1970), J. Memb. Biol. 2, 201-234 13. Decker, S . J . and D.R. Lang (1978), J. Biol. Chem. 253, 6738-6743 14. Krulvich, T.A. and A.A. Guffanti (1983), Adv. Microbiol. Physiol. 24, 173-214 15. Helgerson, S.L., C. Requadt and W. Stoeckenius (1983), Biochemistry 22, 5746-5753 16. Terada, H., O. Shima, K. Yoshida, Y. Shinohara (1990), J. Biol. Chem. 265, 7837-7842 • * '17. Story, B.T., D.M. Scott, and C P . Lee (1980), J. Biol. Chem. 255, 5224-5229 18. Williams, R.J.P. (1961), J. Theor. Biol.V, 1-13 19. Williams, R.J.P. (1985), The enzymology of biological membranes (Martonosi, A.N., ed.) vol 4. 71-110, Plenum Press, New York 64 20. Kell, D.B., D.J. Clarke and J.G. Morris (1981), FEMS Microbiol. Letters H , 1-11 21. Prats, M., J . Teissie and J.F. Tocanne (1986), Nature (London) 322, 756-758 22. Nagel, J.F. and H.S. Morowritz (1978), Proc. Natl. Acad. Sci. USA 75, 298-302 23. Westerhoff, H.V., B.A. Melandri, G. Venturoli, G.F. Azzone, D.B. Kell (1984), Biochim. Biophys. Acta 768, 257-292 24. Malpress, F.H. (1984), Biochem. Soc. Trans. 12, 399-401 25. Skulachev, V.P. (1982), FEBS letters 146, 1-4 26. Slater, E . C , J.A. Berden and H.A. Herweijer (1985), Biochim. Biophys. Acta 811. 217-231 27. Boyer, P.D. (1984), H+-ATP synthase. Structure, function, regulation (Papa, S., ed.) 329-338, ICSU Press and Adriatica Editrice, Bari 28. Herweijer, M.A., J.A. Berden and E.C. Slater (1986), Biochim. Biophys. Acta 849. 276-287 29. Haddock, B.A. and C W . Jones (1977), Bacteriol. Rev. 41, 47-99 30. Bragg, P.D. (1980), Diversity of bacterial respiratory systems (In Knowles, C.J. , ed.) 115-136, CRC Press, Florida 31. Kroger, A. (1980), Diversity of bacterial respiratory systems (In Knowles, C.J. , ed.) 1-18, C R C Press, Florida 32. Stouthamer, A.H., J . van't Riet and L F . Ottman (1980), Diversity of bacterial respiratory systems (In Knowles, C.J. , ed.), CRC Press Florida 33. Wallace, B.J., I.G. Young (1977), Biochim. Biophy. Acta 461_, 84-100 34. Owen, P. and H.R. Kaback (1979), Biochemistry 18, 1413-1422 35. Matsushita K., T. Ohnishi and H.R. Kaback (1987), Biochemistry 26, 7732-7737 36. Hayashi, M., T. Miyoshi, S . Takashina, T. Unemoto (1989), Biochim. Biophys. Acta 977, 62-69 37. E. Hackimi, Z., O. Samuel and R. Azerad (1974), Biochimie 56, 1239-1247 65 38. Haddock, B.A. and H.U. Schairer (1973), Eur. J. Biochem. 35, 34-35 39. Newton, N.A., G.B. Cox and F. Gibson (1971), Biochim. Biophys. Acta 244. 155-166 40. Kita, K. and Y. Anraku (1981), Biochem. Int. 2, 105-112 41. Ashcroft, J.R. and B.A. Haddock (1975), Biochem. J. 148, 349-352 42. Poole, R.K. and B.A. Haddock (1975), Biochem. J. 152, 537-546 43. Downie, J.A. and G.B. Cox (1978), J. Bacteriol. 133, 477-484 44. Rydstrom, J . (1977), Biochim. Biophys. Acta 463, 155-184 45. Fisher, R.R. and S.R. Earle (1982), The pyridine nucleotide coenzymes, 279-324 46. Rydstrom, J . (1979), J. Biol. Chem. 254, 8611-8619 47. Earle, S.R. and R.R. Fisher (1980), J. Biol. Chem. 255, 827-830 48. Danielson, L. and L. Ernster (1963), Energy-Linked Functions of Mitochondria (Chance B. ed.) 157-175, Academic Press, New York 49. Rydstrom, J . , A. Teixeira da Cruz and L. Ernster (1970), Eur. J. Biochem. 17, 56-62 50. Yamaguchi, M., Y. Hatefi, K. Trach and J.A. Hock (1988), J. Biol. Chem. 263. 2761-2767 51. Clarke, D.M., T.W. Loo, S . Gillam and P.D. Bragg (1986), Eur. J. Biochem. 158. 647-653 52. Wakabayashi, S . and Y. Hatefi (1987), Biochem. Int. 15, 915-924 53. O' Neal, S .G . and R.R. Fischer (1977), J. Biol. Chem. 252, 4552-4556 54. Hatefi, Y., D.C. Phelps and Y.M. Galante (1980), J. Biol. Chem. 255, 9526-9529 55. Nicholls, D.G. (1982), Bioenergetics: An introduction to chemiosmotic theory (United States ed.) 1-22, Academic Press, London 56. Clarke, D.M. and P.D. Bragg (1985), J. Bacteriol. 162, 367-373 66 57. Jones, C W . (1982), Bacterial Respiration and Photosynthesis (Nelson ed.) 6-10, Thompson Press, New Delhi 58. Lowry, O.H., N.J. Rosegrough, a.L. Farr and A.J. Randall (1951), J. Biol. Chem. 193. 265-275 59. Kaplan, N.O. (1967), Methods Enzymol.10, 317-322 60. Fisher, R.J. and D.R. Sanadai (1971), Biochim. Biophys. Acta 245, 34-41 61. Kaback, H.R., T. Ohnishi and K. Matsuchita (1987), Biochemistry 26, 7732-7737 62. Kroger, A. (1987), Methods Enzymol. U N , 579-591 63. Laemmli, U.K. (1970), Nature (London) 227, 680-685 64. Towbin, H., T. Staehelin and J . Gordon (1979), Proc. Natl. Acad. Sci. USA 76., 4350-4354 65. Harlow, E. and D. Lane (1988), Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory, New York 66. Maniatis, J . , E.F. Fritsch and J . Sambrook (1982), Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York 67. Hanstein, W.G. (1976), TIBS Mar.. 65-67 68. Pressman, B.C. (1976) Ann. Rev. Biochem. 45, 501-529 69. Singh, A.P. and P.D. Bragg (1979), Can J. Biochem. 57, 1384-1391 70. Kita, K., K. Konishi and Y. Anraku (1984), J. Biol. Chem. 259, 3368-3374 71. Huang, C .S . , S . J . Kopacz and C P . Lee (1977), Biochim. Biophys. Acta 459. 241-249 72. Clarke, D.M. (1986), Pyridine Nucleotide Transhydrogenase (PhD thesis, University of British Columbia), 77-89 73. Lee, C P . and L. Ernster (1966), Biochim. Biophys. Acta 7, 218-234 74. Danielson, L. and L. Ernster (1963), Biochem. Biophys. Res. Commun.TO, 91-96 67 75. Weiner, J.H., B.D. Lemire, M.L. Elmes, R.D. Bradley and D.G. Scraba (1984), J. of Bacteriol. 1_58, 590-596 76. Jacoby, G.H. and K.D. Young (1990), Analyt. Biochem. 184, 48-54 77. von Meyenburg, K., B.B. Jorgensen and B. van Deurs (1984), EMBO J. 3, 1791-1797 78. Kaback, H.R. (1977), Biochem. Membr. Transp.76, 598-625 79. Yamaguchi, M. and Y. Hatefi (1989), Biochemistry 28, 6050-6056 80. Montal, M., B. Chance, C P . Lee and A. Azzie (1969), Biochem. Biophys. Res. Commun. 34. 104-110 81. Keister, D.L. and N.J. Yike (1967), Biochemistry 6, 3847-3857 82. Ernster, L , K. Juntti and K. Asami (1972), J. Bioenerg. 4, 149-159 83. Kay, W.W. and P.D. Bragg (1975), Biochem. J. 150, 21-29 84. Danielson, L. and L. Ernster (1963), Biochem. J. 338, 118-205 85. Fiolet, J.W.T., E.P. Bakker and K. van Dam (1974), Biochim. Biophys. Acta 368. 432-445 86. Huang, C .S . , S . J . Kopacz and C P . Lee (1977), Biochim. Biophys. Acta 459, 241-249 87. Young, K.D. and G.H. Jacoby (1990), Analyt. Biochem. 184, 48-54 88. Young, K.D., M.J. Leidenix, G.H. Jacoby, T.A. Henderson (1989), J. of Bacteriol. 171. 5680-5686 89. Simoni, R.D. and P.W. Postma (1975), Annu. Rev. Biochem. 44, 523-554 90. Hatefi, Y. (1985), Annu. Rev. Biochem. 54, 1015-1067 91. Withers, H.K. and P.D. Bragg (1989), Biochem. Cell Biol. 68, 83-90 92. Kita, K., K. Konishi and Y. Anraku (1984), J. Biol. Chem. 259, 3368-3374 93. Kita, K. and Y. Anraku (1981), Biochem. Int.2, 105-112 68 94. Hou, C , M. Potier and P.D. Bragg (1990), Biochim. Biophy. Acta (In press) 95. Enander, K. and J . Rydstrom (1982), J. Biol. Chem. 256. 14760-14766 96. Wu, L.N.Y., R.M. Pennington, T.D. Everett and R.R. Fisher (1982), J. Biol. Chem. 257. 4052-4055 97. Carlenor, E., H.L. Tang and J . Rydstrom (1985), Analyt. Biochem. 148. 518-523 98. Persson, B., K. Enander, H.L. Tang and J . Rydstrom (1984), J. Biol. Chem. 259, 8626-8632 99. Clarke, D.M. and P.D. Bragg (1985), Eur. J. Biochem. 149. 517-523 100. Rydstrom, J . , B. Persson and H.L. Tang (1984), Bioenergetics (Ernster ed.) chp 7, 207-219, Elsevier Science Publisher, B.V. 101. Pennington, R.M. and R.R. Fisher (1981), J. Biol. Chem. 256. 8963-8969 102. O'Neal, S.G. , S.R. Earle and R.R. Fisher (1980), Biochim. Biophy. Acta 589. 217-230 103. Skulachev, V P . (1989), J. Bioenerg. Biomemb.2±, 635-647 104. Lee, C P . and L. Ernster (1964), Biochim. Biophy. Acta 8, 187-190 105. Rydstrom, J . , A. Teixeira da Cruz and L. Ernster (1970), Eur. J. Biochem. 17, • 56-66