GIẢI THƯỞNG NOBEL TRONG HÓA HỌC VỀ CHUYỂN HÓA NĂNG LƯỢNG SINH HỌC Nghiên cứu của Mitchell đã được thực hiện trong một lĩnh vực hóa sinh thường được gọi là năng lượng sinh học trong những năm gần đây, là nghiên cứu về các quá trình hóa học chịu trách nhiệm cung cấp năng lượng cho các tế bào sống.
European J Biochem (1967) 317-326 An Evaluation of the Mitchell Hypothesis of Chemiosmotic Coupling in Oxidative and Photosynthetic Phosphorylation E C SLATER Laboratory of Biochemistry, B.C.P Jansen Institute, University of Amsterdam (Received February 17, 1967) The Mitchell hypothesis of chemiosmotic coupling in oxidative phosphorylation is examined in the light of experimental data on oxidative phosphorylation a t present available The following objections are brought against the theory : Data on the magnitude of the phosphate potential against which ATP can be synthesized by the respiratory chain require, on the basis of chemiosmotic coupling, very effective proton and/or cation extrusion and/or anion uptake by the mitochondria Experimental evidence for this is lacking Indeed it has been shown by Chance and Mela that mitochondria in the controlled state not extrude protons Although reversible ATP-driven proton (chloroplasts) or cation (mitochondria, erythrocytes) pumps have been demonstrated, this is insufficient evidence for the postulate that the establishment of a proton gradient is the primary energy-conserving event of the chloroplast or mitochondrial respiratory chain Indeed, it is general experience that cations are required for the extrusion of protons by mitochondria The stoicheiometry of proton extrusion by mitochondria and of :proton uptake by chloroplasts, and the kinetics of the latter process are also difficult to reconcile with the chemiosmotic hypothesis Further experimental evidence is presented confirming that, under the conditions of the oxygen-pulse experiments of Mitchell and Moyle, the extrusion of Hf is not associated with the oxidation of mitochondrial NADH The respiratory chain included in the chemiosmotic hypothesis is difficult to reconcile with our present knowledge of the chain It is concluded that the chemiosmotic theory, in its present form, is untenable Until recently, discussions on the mechanism of respiratory-chain phosphorylation centred on what has become known as the “chemical” or “-” hypothesis of respiratory-chain phosphorylation, first formulated in 1953 [l].According to this hypothesis, the initial conservation of the energy made available by the oxido-reduction reactions of the respiratory chain is by formation of a high-energy compound of a n electron or hydrogen carrier in the chain This energy is utilized for various energy-requiring reactions, including the synthesis of ATP This hypothesis has provided a useful framework on which t o hang a large amount of information on the mechanism of respiratory-chain phosphorylation, especially on the effects of various uncouplers and inhibitors and on the utilization of the conserved energy in the absence of inorganic phosphate However, the failure to isolate the hypothetical high-energy intermediates has stimulated consideration of other possible mechanisms I n particular, the Mitchell Non-standard abbreviations dimethylurea, DCMU N-3,4-dichlorophenyl-N’- hypothesis [2,3] of chemiosmotic coupling in oxidative phosphorylation has attracted much attention A detailed account of the Mitchell hypothesis is given in a privately distributed Research Report [3], a shortened version of which has appeared in the published literature [4] Since several important modifications of the :chemiosmotic hypothesis have been made since it was first proposed [2], the subsequent discussion will be confined to the version discussed in the Research Report [3], and in subsequent publications [5,6] I n the Introduction [3], Mitchell makes thegood point that “the belief has been widely accepted that the energy-rich coupling intermediates (C, I,, C, I,, C, I,, etc.) must exist because there is no feasible alternative means of coupling electron transport to phosphorylation.” His objects in proposing a working chemiosmotic hypothesis in 1961 were (a) to provide a simple rationale for the organization of the components of the oxidative phosphorylation system in the lipid membrane systems of mitochondria and chloroplasts; (b) to formulate a - - - 318 Mitchell Chemiosmotic Hypothesis - type of coupling that would require no intermediates of the type of C I, (‘so that future work need no longer be so dependent upon or so circumscribed by the belief in the C -1 intermediates”; and (c) “to acknowledge the elusive character of the C I intermediates by admitting that they might not exist ” I believe that Mitchell has succeeded in putting forward a rational alternative to the hypothesis that deserves the serious consideration that it has received and is receiving from other workers in the field This article has the limited purpose of examining the chemiosmotic hypothesis in the light of experimental data on oxidative phosphorylation at present available Other hypotheses in which, as in Mitchell’s, the production of protons by reduction of the respiratory chain plays a central role [7-91 wiU be left out of consideration - - MITCHELL IIYPOTHESIS The chemiosmotic system consists of four basic postulates : (a) a proton-translocating reversible ATPase system ; (b) a proton-translocating oxidoreduction chain ; (c) an exchange-diffusion system, coupling proton translocation to that of anions and cations ; (d) an ion-impermeable coupling membrane, in which the three systems reside tions this may be written (representing the internal phase by the suffix R) f NADR+ (sH2)R (NAD+H2), (NADfH,), Sum : 2HRf + + + + + Proton-translocatingOxido-reductionChain Mitchellrepresents the proton-translocating oxidoreduction chain as in Fig (essentially identical with Fig 11 of Mitchell [3]) In terms of chemical equa- + f + QL + + + + - +2 (O& + + + QB 2(C12+)L 2H,+ f + ( Z C , ~ + ) ~ H20, MEMBRANE OUTSIDE INSIDE I a? -S SHZ +2H’ f 2Hf 4-SHZ + H20 + ATP + 2HL++ Pi + ADP N + + + XH, + IOH, (1) where the suffixes R and L refer to the right-hand and left-hand sides, respectively, of a membrane XI, XL-, IOL-, X H and IOH are in the membrane, whereas ATP, ADP and Pi are in the right-hand phase, which in mitochondria corresponds to the matrix or internal cristae phase X I is considered to be a low-energy compound when it is in contact with phase L having a high electrochemical potential of H+, and a high-energy compound (X I) when it is in contact with phase R with a low electrochemical potential of H+ (NAD+H2)li f + IOH, + ATP + XI, + Pi + ADP + XI, + XL- + 10, + 2H,+ + XI‘’ + XL- + 10,- + f (NAD+H,), (Fe, SH), +- NAD,+ 2HL+ (Fe, SH) 2e-), NADL+ + NAD,+ + (Fe, SH, 2e-), (Fe, SH, 2e-), (Fe, SH, 2e-), FMN, + (Fe, SH), (FMNH& ~HR+ (Fe, SH), + (Fe, SH), (FMNH2)R + (FMNH2)L +- FMN, bL2+ 2HL+ (FMNHz)L bL3+ bL2+ f bR2+ FMN, +- FMN, h + Qx f ~ H R + b f~(QH2)li ~ ~ +- bL3+ bR3+ (QH2)R (QHz), +QL (ci2+)L 2HL+ (QH2)L 2(ci3+)~ Proton-translocatingReversible ATPase Xystern Mitchell [ref 3, p 521 represents the oligomycinsensitive proton-translocating ATPase system by the reactions XH, XI, H2O 2H,+ European J Biochem S f 2Hf 1/202t2HS H20 Fig Mitchell’s [3] proposal of the proton-translocating respiratory chain for oxidation of NAD-linked and FAD-linked substrates in mitochondria The chain is shown branching a t ubiquinone (Q) Drawn from Mitchell [3], Fig.11 If we add to this equation that representing the synthesis of ATP by the proton-translocating ATPase, written with the appropriate stoicheiometry 6HL+ + ( P i ) +~ ~ A D P R6H,+ + ~ H ~ +O 3KL”J?, L f E C SLATEB Vol 1, No 3, 1967 we obtain +2 (SHZ)R (O2)Z + 3(Pi)R + SR 319 THERMODYNAMICS O F OXIDATIVE PHOSPHORYLATION + ZADPl% + HZO, + 3ATPR + 3HzOL which correctly represents the stoicheiometry of oxidative phosphorylation Exchange Diffusion System Mitchell postulates that the diffusion of ions other than protons (or OH- ions) down the electrical gradient across the coupling membrane, and their accumulation in osmotically disruptive concentrations in the internal phase, must be counteracted by specific extrusion in exchange for protons or OHions The exchange diffusion of H+ against cations is represented as the C+/H+ antiport in Fig.2 which is drawn from Fig 14 of Mitchell [3] It is a truism to state that the thermodynamics of a chemical reaction depend only upon the initial and final states, and not on the chemical mechanism of the reaction A calculation of the chemical potential against which ATP can be synthesized is, however, useful since it tells us the magnitude of the AG that can be built up by respiration in State (absence of phosphate acceptor) This in turn gives us information on the magnitude of the high-energy "content" of the hypothetical compound of the chemical hypothesis or of the pH differential (or its equivalent) of the Mitchell hypothesis The chemical potential of the system N ATP + ADP a t 25" is given by AG' OUTSIDE MEMBRANE INSIDE + Pi = ACT; + 1.36 log, [ADPI [Pi1 [ATP] Cockrell et al [lo] have recently calculated, on the basis of new data on the equilibrium in State 4, that, in the absence of Mg2+a t pH 7.8 and 25", o h system A CT' = -9.6 hid system = ~~~ H+ A-lH'symporI A-permeation Fig.2 The cationlproton antiport and the anion-proton symport proposed by Mitchell [3] Drawn from Mitchell [3], Big 14 Ion-impermeable Coupling Membrane This is identified as the inner membrane of the mitochondria, the grana membrane of chloroplasts and the plasma membrane and chromatophore membrane of bacteria The Mitchell theory will now be examined in the light of (a) the thermodynamics of oxidative phosphorylation ; (b) experimental evidence relating to the existence and function of a proton-translocating reversible ATPase system ; (c) the nature of the respiratory chain x 10-3 + 1.36 loglO i x 10-4.2 kcal/mole x 10-3 - 15.6 kcallmole This calculated value for the phosphate potential is rather greater than that generally used in calculations of the thermodynamics of oxidative phosphorylation From the dependence of the redox state of cytochrome G on the phosphate potential, Klingenberg and Schollmeyer [ i l l calculated a value of 12.8 kcal/mole at pH 7.2 Mitchell [3]assumes a value of 9.6 kcal Since, however, there appears no reason to question the value calculated by Cockrell et al.[lO], this will be used as the basis for subsequent calculations Thus, any hypothesis of oxidative phosphorylation has to provide a mechanism for the synthesis of ATP against a chemicalpotential of the ATP ADP Pi system of 15.6 kcal/mole According to the Mitchell hypothesis, as represented by Equation ( l ) ,the synthesis is driven by a pH differential, the magnitude of which may be calculated from the equation + AG = 1.36 loglo ~ + - _ 15.' kcal/mole i.e log,, [H+IL- log,, [H+IE= pH, - pH, - 15.6 1.36 x = Since the left-hand side of the membrane, in the Mitchell hypothesis, is the space between the outer and inner mitochondria1membranes, and this space is thought to be readily permeable to ions, pH= must Mitchell Chemiosmotic Hypothesis 320 be about the same as the pH of the suspension medium, i.e p H 7.8 in the experiment of Cockrell et al [lo] Thus, the p H of the cristae space, according to Mitchell’s hypothesis, would be 13.5, which is incompatible with the operation of the intracristae enzymic systems I n fact, Mitchell recognized this difficulty, although not its magnitude, from the beginning, and proposed the exchange-diffusion system t o overcome it [Z] It was suggested that, by exchange of cations for protons, the p H differential would be replaced by a membrane potential However, since the first version of the Mitchell theory [Z], the “sidedness” of the mitochondrion has been reversed Now 131, the inside of the mitochondrion is phase R, the outside is phase L Thus, it is the low concentration of H+ in phase R, rather than the high concentration of H+ in phase L, that is supposed to drive the ATPase towards ATP synthesis I n the most recent version of the chemiosmotic hypothesis, Mitchell has proposed two exchangediffusion systems, a cation/proton antiport and an anionlproton symport Thus Equation (1) could be supplemented by the equation of the C+/H+ antiport, namely 2H,+ + 2CR+s 2HR+f 2C,+ (3) The sum of Equations (1)and (3) is H,O + ATP + 2CR+ + Pi + ADP + 2CL+ Thus, the ATP synthesis is now driven by a concentration gradient of cations across the inner membrane, the concentration being greater outside than inside The magnitude of the concentration gradient required is given by log1,- [CL+I = 5.7 i.e [cR+1 Alternatively, Equation (1) could be supplemented by the equation of the A-/H+ symport, e.g 2HL+ + AL2-+ 2HR++ AR2- (4) assuming a dicarboxylic acid as an anion The sum of Equations (1)and (4)is H20 + ATP + AL2-+ Pi + ADP + A,’- Thus, the ATP synthesis is now driven by a concentration gradient of anions across the inner membrane, the concentration between greater inside than outside The magnitude of the concentration gradient required is given by 1.36 loglO-= 15.6, [AL’-] [-k2-] i.e log,, - = 11.5 [AL I European J Biochem The total energy available for the synthesis of ATP is the sum of the contributions made by the proton, cation and anion gradients Thermodynamic requirements for ATP synthesis are met if d p H - 0.5 ApA2- + ApC+ = 5.7 where A refers to the difference, R (inside) minus L (outside), pA2- = log, [A2-] and pC+ = -log, [C+] Various possible combinations yielding the required phosphate potential of 15.6 kcal are listed in Table I Table Various possible combinations of proton, cation and anion gradients yielding a phosphate potential of 15.6 kcal A PH 5.7 2.0 2.0 2.0 2.0 2.0 0 0 -A pAZ- 7.5 5.5 3.5 1.5 11.5 7.5 3.5 A pC+ 0 1.o 2.0 3.0 3.7 0’ 2.0 4.0 5.7 Even if a pH differential of 2.0 is allowed (the p H of the cristae space in the experiments of Cockrell et al [lo] would then be 9.8), very high ionic gradients are required Since oxidative phosphorylation proceeds actively with 0.6 M succinate in the medium, it is highly probable that -ApA2- would be much less than This would require a ApC+ of 3.5, i.e the external concentration of cation would have to be IO3e5 times that of the internal concentration Since there is no evidence that the contents of mitochondria become highly alkaline during steady-state oxidative phosphorylation (see below), the requirements for cation extrusion and/or anion uptake by the mitochondria are even more stringent I n the absence of any experimental evidence of the extrusion of cations and uptake of anions of the magnitude required by the chemiosmotic hypothesis, the thermodynamic difficulties become overwhelming I n chloroplasts, the direction of proton movements during electron transport [I21 is the opposite from that found with mitochondria The significance of this will be discussed below EXISTENCE AND FUNCTION O F A PROTON-TRANSLOCATING REVERSIBLE ATPASE SYSTEM Existence As first emphasized by SARIS[13], the energydependent accumulation of cations by mitochondria is accompanied by the liberation of a n equivalent amount of protons The energy may be provided either by the operation of the respiratory chain, in E C SLATER Vol.1, N0.3, 1967 which case the accumulation of cations is insensitive t o oligomycin, or by the hydrolysis of ATP in an oligomycin-sensitive reaction, thus 321 greater than that predicted purely from reversal of the equilibrium ATP3- + OH- + ADP2- + Pi2- After I sec a t pH 4.2-4.3 in the experiments of respiratory chain Reid et al [5], [ADP] [PI] - (5 x - [ATP] aligornycin (cation uptake,proton extrusion) where E, represents energy conserved during the operation of the chain, without specifying the nature of the energy conservation Thus, the existence of a proton-translocating ATPase, under certain conditions, is well-established The question a t issue is whether these conditions include all those in which oxidative phosphorylation occurs The reversibility of the proton-translocating ATPase has also been demonstrated, first by Jagendorf and Uribe [14] with chloroplasts, in which the direction of proton translocation is in the reverse direction from that found with mitochondria After chloroplasts had been left for a short time a t 0” and p H 3.8 in the presence of a n organic acid (such as succinic acid) and DCMU (to inhibit the light reaction), the suspension was brought to p H 8.0 in the presence of 32Piand ADP After a few seconds, the reaction was stopped with trichloroacetic acid and the ATP synthesis calculated from the uptake of 32Pi into the organic phosphate fraction or the net ATP synthesis measured by the luciferase assay This so-called “acid-bath phosphorylation”, which can amount t o about 40 pmoles ATP/g protein, has been confirmed by Dr K G Rienits in our laboratory [I51 and by McCarty and Racker 1161 It seems likely that this phosphorylation is driven by a difference of p H between the outside and inside of the chloroplasts, which, a t the moment of bringing the p H of the suspension medium t o 8.0, would contain a high concentration of organic acid that had diffused into the chloroplasts as uncharged acid during the acid bath If the ADP and phosphate are added only sec after adjusting the p H to 8.0, the yield of ATP is decreased by 83O/,, 1141 Analogous experiments with rat-liver mitochondria, carried out by Dr K G Rienits and Dr M Koivusalo in our laboratory, have failed to show any synthesis of ATP, either when the p H was raised from 5.0 t o 8.4, or lowered from 10.3 to 7.0 (see Table for a typical experiment) Reid et al 151, however, have reported the synthesis of a small amount of ATP when the p H was lowered from 8.8-9.0 to 4.2-4.3, in the presence of K+ and valinomycin, which is known t o increase the permeability of the mitochondria1 membrane to K+ Although the amount of ATP found is small, it is x (1 x 10-3) 12.5 x 10-6 M = 0.4 At pH’s between and 6, [ATPI a t equilibrium equals about lo5M Although there are some features of this experiment that require further examination, we shall assume in what follows that the existence of a protontranslocating reversible ATPase system in both chloroplasts and mitochondria has been experimentally demonstrated Table Failure of alkali bath to effect ATP synthesis by rat-liver mitochondria ml of mitochondrial suspension (containing 72 mg protein) were mixed with 400 pmoles potassium phosphate buffer (pH 7.0) containing 3zP~,150 pmoles MgCl,, 50 pmoles EDTA and 2.5 mmoles sucrose in a total volume of 24 ml After at 25”, 0.24 ml of 0.1 M KCN (pH 7.0) was added (final concentration, mM), and later ml of a solution containing 108 mmoles ethanolamine chloride (pH l l O ) , 24 pmoles MgCl,, pmoles EDTA, 60 pmoles ADP, pmoles KCN and 54 pmoles KOH This addition brought the p H to 10.5 After 30sec the p H was brought back t o 7.0 by the addition of ml of a solution containing 60 pmoles ADP, pmoles KCN, 24 pmoles MgCl,, pmoles EDTA, 790 pmoles HC1 and 720 pmoles Tris-HC1 buffer (pH 7.0) Samples were withdrawn at various times and the amount of 3zPincorporated into esterified phosphate was measured Control experiments showed that mitochondria could be kept at pHs between 5.0 and 10.3 for 30 sec at 25O without any loss of respiratory control or effect on the P: ratio Time PH STincorporated ~ nmoleslmg firotein see 260 270 280 290 305 310 320 325 335 345 355 370 7.0 7.0 7.0 7.0 10.5 10.5 10.5 10.5 O 7.0 7.0 7.0 58 62 59 59 60 59 59 55 59 54 53 66 Function The relationship between energy conservation, cation uptake and proton extrusion can be expressed, quite generally, in three possible ways Scheme A represents the chemiosmotic hypothesis in which the primary energy conservation is the form of translocated Hf (indicated by t H+ to 322 Mitchell Chemiosmotic Hypothesis SCHEME A : respiratory chain H+ -c II11 SCHEME B: respiratory choin -& ATP oligomycin *li C”+ - E, - ATP oligomycin Hf II + c”+ SCHEME C: E, respiratory chain 11 +C & ATP aligomycin n+ 41 I t HS show the translocation of H+ from right to left) Cation accumulation (+ Cn+) takes place from left t o right, as result of the membrane potential set up by the C+/H+ antiport and the H+/anion symport of Fig.2 According to Scheme B, the conserved energy (E,) may be utilized t o make ATP, in an oligomycinsensitive reaction, or t o drive a proton pump, that may be linked, perhaps via a n exchange-diffusion carrier, with cation accumulation According to Scheme C, the conserved energy may operate a cation pump, that is linked with the movement of H+ in the opposite direction Scheme C covers the more specific mechanisms proposed by Cockrell et al [lo] and Chance and Mela [17] All reactions in the three schemes are written reversibly It has been amply demonstrated that the direction of that part of the respiratory chain lying between NAD and cytochrome c can be reversed by ATP [is] Reversal of the respiratory chain by a p H or cation concentration differential has not yet been demonstrated in either mitochondria or chloroplasts However, the recently reported luminescence of chlorophyll in spinach chloroplasts [19] induced by the acid-bath treatment of Jagendorf and Uribe [14] points in this direction The function of the proton-translocating ATPase differs fundamentally in the three schemes According t o SchemeA (the chemiosmotic hypothesis), it is on European J Biochem the main path of ATP synthesis According to Schemes B and C it is on a side path that can act as a n energy store (cf Jagendorf and Uribe [20]) or may be utilized for cation uptake I n Scheme B, E, drives a proton pump, in Scheme C a cation pump Since all three schemes provide an explanation for the synthesis of ATP linked to a p H differential, I cannot agree with Reid et al [5] that their findings “would thus seem to help t o promote the status of the chemiosmotic hypothesis towards that of a theory Conversely, the chemical hypothesis is weakened to the extent that a further ad hoc assumption is required t o account for the new observations.” ATP-driven cation pumps have for a long time featured in explanations ofthe Na+ and K+fluxesthrough the cell membrane A simple allosteric model has recently been proposed [21] Garrahan and Glynn [22] have recently reported the synthesis of ATP by reversal of the (Na+ K+)-stimulated ATPase of the erythrocyte membrane A critical test between the three schemes above is whether or not permeant cation is necessary for the extrusion of H+ by mitochondria or the uptake of H+ by chloroplasts, driven by respiration or by ATP According to Schemes A and B, permeant cations need not be necessary; according to Scheme C, they must be The careful studies of Chance and Mela [17] provide strong evidence that, although a proton concentration gradient can be clearly demonstrated when respiration is linked t o the uptake of Ca2+ in the absence of a permeant anion (cf Lehninger et al [as]), no gradient is set up in the absence of a permeant cation By using bromthymol blue as a n indicator of membrane pH and bromcresol purple as an indicator of extramitochondrial pH, Chance and Mela were able directly t o demonstrate the production of H+ outside the mitochondria and the alkalinization of the membrane on the addition of amounts of Ca2+ exceeding 30 pmoles per g protein t o a mitochondria1 suspension in State (absence of ADP) Additions of Ca2+ below this limit caused no alkalinization of the membrane, indicating the existence of a buffer capacity, tentatively identified with phospholipid Thus, State itself, the condition of maximum conservation of energy, is not associated with the establishment of a p H gradient Moreover, this buffer capacity would have to be neutralized before a proton gradient could be built up sufficient to drive ATP synthesis by the chemiosmotic mechanism Uncoupling agents, which abolished the alkalinization of the membrane by Ca2+, caused no change in p H in the absence of Ca2+,further supporting the conclusion that there is no detectable difference in the pH gradient in the fully activated State and in the uncoupled state Taking into account the sensitivit y of the technique used, Chance and Mela concluded that no p H gradient exceeding 0.02 pH exists + E C SLATER Vol.1, No.3,1967 across the mitochondrial membrane in the energized State With rat-liver mitochondria especially prepared to minimize cation accumulation, Chance and Mela found no detectable alkalinization of the mitochondrial membrane or acidification of the medium (less than 0.36 pmole H+ per g mitochondrial protein) on aerating an anaerobic suspension I n contrast to this conclusion, Mitchell and Moyle [24] claimed to have demonstrated the primary production of protons They pre-incubated a suspension of rat-liver mitochondria in a medium containing KC1, glycylglycine buffer and /?-hydroxybutyrate or succinate for 20min at 25” in the absence of oxygen When small amounts of oxygen were introduced, H+was rapidly liberated into the medium, followed by a slower decline in H+ concentration The H+:0 ratio was with /?-hydroxybutyrate and with succinate When ATP was added instead of oxygen to an anaerobic mitochondrial suspension in the absence of substrate, moles H+ were produced for each mole of ATP hydrolysed, over and above the 0.8 mole H+ expected for the hydrolysis of ATP t o ADP and Pi a t this pH Mitchell and Moyle [24] concluded that the H+ production measured was that predicted by the chemiosmotic hypothesis Tager et al [15] have, however, objected t o this interpretation We pointed out that, according to Scheme A, when oxygen is added to mitochondria in the presence of substrate and the absence of Pi and ADP, it may be expected that the respiratory chain will function until a sufficient p H differential (or its equivalent) is built up t o prevent the flow of reducing equivalents along the respiratory chain from substrate to oxygen I n the words of Mitchell and Moyle [25]: “under the conditions of this kind of experiment there is a ‘backlash’ before the translocation of protons across the coupling membrane builds up a protonmotive force sufficient to retard oxido-reduction.” Thus, one must expect that, immediately after adding the oxygen, the respiration would be uncontrolled even though phosphate and phosphate acceptor are absent With sufficient oxygen, one would proceed through the following sequences of states [26] : State f State + State + State The amount of oxygen consumed (and substrate oxidized) during the uncontrolled State would depend upon the amount of H+ required to build up the required protonmotive force With smaller amounts of oxygen, State would never be reached and the sequence would be : State + State + State The ‘backlash’ limit, i.e the amount of H+ liberated by oxygen-pulsed mitochondria before the 22 European J Bioehem., Vol.1 323 steady-state oxygen consumption (State 4) sets in, was found by Mitchell and Moyle [24,25] to be about 12 pmoles per g mitochondrial protein, equivalent t o 6pmoles ATP per g mitochondrial protein This ‘backlash’ limit is equivalent to the amount of energy (E,) that can be conserved in the respiratory chain in the presence of substrate and oxygen, but absence of ADP and Pi Van Dam [27] has shown that this is less than 0.4pmole ATP per g mitochondrial protein Thus, the amount of H+ that can be liberated in an oxygen-pulse experiment is more than an order of magnitude greater than the amount of E, that can be conserved in an ADP-pulse experiment [27] We [15] concluded then that the H+ liberated in the oxygen-pulse experiments does not represent the primary conservation of energy in the respiratory chain, as envisaged by Scheme A (the chemiosmotic hypothesis) Mitchell and Moyle [3] calculate that “the translocation of only about pequiv proteins per g protons in rat-liver mitochondria should bring the membrane potential t o its presumed respiratory control value of some 250 mV.” They [25] explain the unexpectedly large ‘backlash’ found in their experiments by postulating the presence of “a, considerable quantity of mobile charges or dipoles.” The membrane potential, which they postulate “is the main cause of respiratory control and of the reentry of the outwardly translocated protons, would thus tend to be collapsed while the mobile charges or dipoles were redistributing in the electric field; and the major part of the backlash of some 12 pg ion H+, or 12 p electron equivalents, per g mitochondrial protein would be equated t o the quantity of the displaceable charge.” They suggest that endogenous calcium ions may be partly responsible for the ‘backlash’ Since, according to Mitchell and Moyle, ‘mobile charges or dipoles’ are responsible for more than 9001, of the protons liberated by mitochondria in oxygen-pulse experiments, these experiments cannot be brought forward in support of Scheme A, which implies that a pH differential could be built up in the absence of ‘movable ionic constituents.’ Indeed, we brought forward experimental evidence that cast doubt on the conclusion that the burst of H+ obtained by Mitchell and Moyle [25] in their oxygen-pulse experiments is associated with uncontrolled respiration A sensitive method of detecting a transition from controlled to uncontrolled respiration is to follow the oxidation of intramitochondrial nicotinamide nucleotide [26] Making use of the data of Van Dam 11271 on the kinetics of the oxidation of NADH on the transition from State to State 3, we calculated that, if the chemiosmotic hypothesis and the interpretation of the oxygenpulse experiments given by Mitchell and Moyle [24] were correct, a t least 0.9pmole NADH should be 324 Mitchell Chemiosmotic Hypothesis oxidized on the addition of oxygen [15] However, no oxidation was found unless phosphate was added simultaneously with the oxygen I n these experiments, NAD+ was determined enzyrnically in extracts obtained by stopping the reaction with perchloric acid We pointed out that erroneous conclusions would be drawn if the direct spectrophotometric assay (based on changes in A340-374 m,) for reducednicotinamide nucleotides were used, since the transition from anaerobic t o aerobic conditions is accompanied by a large oxidation of cytochrome which interferes with the assay for the reduced nicotinarnide nucleotides This was demonstrated by carrying out oxygen-pulse experiments in I TIME Fig.3 The effect of an oxygen pulse on the p H of a mitochondria1 suspension and on the amount of intramitoclumdrial N AD+ The upper figure shows traces of a recording pH meter (upward trace represents a fall in pH) In this experiment, rat-liver mitochondria (16.8 mg) were preincubated in the absence of air in 3.5ml of a medium containing 3.3mM glycylglycine buffer and 0.15 M KCI The p H before addition of the mitochondria was 7.1 Nitrogen was bubbled through all solutions to free them of oxygen Anaerobiosis was maintained by layering oxygen-free paraffin on top of the reaction mixture The effectiveness of the procedure was tested in control experiments with an oxygen polarograph After 20 min, 50 pl 0.15 M KCl were added, either saturated with C0,-free air (upper trace) or with N, (lower trace) I n the experiment shown in the lower figure, rat-liver mitochondria (5.6 mg protein) were preincubated anaerobically in I ml of the same medium as above After 20 min, 0.1 ml 0.15 M KCI saturated with C0,-free air (Curve B) or 0.1 ml 0.15M KCl saturated with C0,-free air and containing 0.5pmole ADP and 5pmoles phosphate (Curve A) was added At the times indicated, the reaction was stopped with HClO,, and NAD+ determined in the acid extracts as already described [15] Each point represents a separate incubation European J Biochem the presence of rotenone which inhibits the oxidation of NADH Mitchell and Moyle [6], while admitting the seriousness of this criticism, have very recently challenged the experimental results Although they agree with our conclusion [I51 that the direct spectrophotometric method cannot be applied without correction, they claim t o have demonstrated by applying appropriate corrections to the spectrophotometric traces that NADH is oxidized No direct chemical assays for NAD+ were made Fig.3 shows the results of a re-examination of this question carried out by Dr R D VeldsemaCurrie I n order to make conditions more favourable for NADH oxidation, no substrate was added According t o Scheme A, the introduction of oxygen would be expected t o lead to the oxidation of NADH a t an initial rate as great as that obtained in the presence of ADP and Pi, and to the extent of more than 1pmole NAD+ per g mitochondrial protein The results clearly show the expected State f State transition in the presence of ADP and Pi, but in the absence of ADP and Pi, no oxidation of NADH, characteristic of the State + State transition, is observed The upper trace shows the transient H+ production reported by Mitchell and Moyle [24] Thus, protons are produced under these conditions without the expenditure of a detectable amount of the energy conserved by the respiratory chain in State Whether or not Mitchell and Moyle [S] have demonstrated oxidation of NADH in their experimentsl is relatively unimportant The fact remains that in our preparations of rat-liver mitochondria which are capable of rapid phosphorylation when ADP and Pi are added together with the oxygen, NADH is not oxidized by oxygen in the absence of Pi and ADP The Mitchell chemiosmotic hypothesis and the interpretation of the experiments of Mitchell and Moyle [24] given by these authors require that the oxidation of NADH after second in the experiment shown in Fig.3 would be as great in the absence as in the presence of ADP and Pi We [I51 showed that, under the conditions of the preincubation used by Mitchell and Moyle [24], a large part of the Mg2+leaked out of the mitochondria We suggested that the proton production observed by Mitchell and Moyle 1241 was due to the reaccumulation of this Mg2+and possibly K+ (cf [lo]) into the mitochondria The accumulation of Mg2+ by ratliver mitochondria requires a low rate of expenditure of energy I n summary, Chance and Mela [I71 have shown that, with some preparations of rat-liver mitochonIt remains possible t h a t slowly reacting components of the chain, such as cytochrome or ubiquinone, cause errors in the calculation by Mitchell and Moyle [6] of t h a t part of the absorbancy change a t 34Omp-374mp that they attribute t o NADH E C SLATER Vol 1, No 3,1967 dria, aerobiosis causes no Hf production, whereas we [I51 have shown that H+ production can be obtained in the absence of NADH oxidation It must be concluded from these experiments that the evidence that the proton-translocating reversible ATPase system is directly concerned in oxidative phosphorylation, as indicated by Scheme A, is very weak Indeed, the requirement of a cation for proton translocation [I71 speaks strongly in favour of Scheme C over Scheme A or B A second method of deciding between Scheme A on one hand and Schemes B and C on the other is t o measure the stoicheiometry between t H+, +Cn+ and ATP According to the chemiosmotic theory, ATP: t H+:+Cn+ = 1:2:2/n According t o Schemes B and C, there need be no strict stoicheiometry between the ATP (or - P equivalent) and the proton or cation translocations, although one would expect stoicheiometry between the proton and cation movements Cockrell et al [lo] have found K+:- P ratios of (representing a thermodynamic efficiency of nearly 800/,), which as they emphasize are inconsistent with the chemiosmotic hypothesis Carafoli et al [28] and Azzone et al 11291 have also obtained Ca2+: P ratios for calcium uptake greater than under certain circumstances As also pointed out by Cockrell et al [lo], the yield of ATP in the acid-bath experiment of Jagendorf and Uribe [14] is uncomfortably high t o be consistent with Equation (1) According t o the calculations of Jagendorf and Uribe [20], the AG for ATP formation under the conditions of their experiments (pH 8.4, ” ) presence of Mg2+)is 10.95 kcal/ mole On the basis of the mechanism given in Equa10.95 tion (l),a pH difference of 1.25 = 4.4 would be - required t o yield this amount of ATP, compared with an experimental pH jump of 4.2 I n the experiments of Rienits [30], 15.7 @ ATP, 63 @ ADP and 0.83 mM Pi were found after a jump of pH from 3.8 t o 7.4 (i.e 3.6 units) Assuming a value for AG,,’ of 7.1 kcal/mole under these conditions, AG becomes 10.2 kcal/mole which would require a p H jump of 4.1, in considerable excess of the experimental 3.6 Moreover, as pointed out by Cockrell et al [lo], the initial pH differential would be expected t o be rapidly dissipated These authors conclude that, if these experiments represent the harnessing of a p H gradient to form ATP, the H+: P ratio is probably considerably greater than A third method of distinguishing between Scheme A on the one hand and Schemes B and C on the other is to compare the rate of the overall phosphorylation with that of the formation of the pH differential The chloroplast system of Jagendorf and Uribe [20] provides the opportunity of testing this When grana are irradiated in the presence of a redox dye but in the absence of ADP or Pi, a considerable amount of N 22 325 - energy (up to 20 pmoles P equivalent per g protein) can be conserved in a form that may be converted t o ATP on the addition of ADP and Pi in the dark There is a good correlation between the conservation of energy in this way (X,) and the uptake of H+ by the grana Jagendorf and Uribe [20] have shown that the rate of accumulation of X , is much slower (less than one-fifth) than the rate of ATP synthesis in a complete phosphorylating system This is most easily explained by supposing that the proton translocation is on a side-path from the main path of ATP synthesis, as in Schemes B and C Mitchell and Moyle 11251 bring forward an additional argument in favour of the primary origin of the proton translocation during respiratory activity, based upon the fact that oxidation of ferrocyanide does not cause proton translocation, but results in the uptake of two protons in the inner mitochondria1 compartment per oxygen atom reduced by cytochrome oxidase They state: “The important point is that if the proton translocation during normal succinate or ,B-hydroxybutyrate oxidation were due t o the utilization of an intermediate high-energy compound by a specific proton pump, or by a cation pump that incidentally translocated protons, the same energy-rich compound should have resulted in proton-translocation during ferrocyanide oxidation.” There is a n obvious alternative trivial explanation of this experiment The oxidation of ferrocyanide by oxygen Fe(CN),4- + H+ + 0, + Fe(CN)63- + H20 is a sufficient explanation of the uptake of protons per oxygen atom consumed I n the absence of uncouplers, there was no net production or consumption of protons during the first 30 sec after addition of ferrocyanide If ferrocyanide was oxidized during this period, protons must have been consumed and must, therefore, have been compensated by the production of protons, possibly by the utilization of a high-energy compound for a specific proton or cation pump THE NATURE OF THE RESPIRATORY CHAIN According to Mitchell (ref [3], pp 90-91), “The view ofthe respiratory chain, according t o the chemical coupling hypothesis, would require us to believe that the chemical complexity of the respiratory chain is considerably greater than it seems, and that the known physical complexity has some functional significance that is, as yet, a matter of conjecture On the other hand, the view of the respiratory chain, according to the chemiosmotic coupling hypothesis, does not require us t o bias, one way or the other, the chemical and physical facts as far as they are known a t present.” 326 E C SLATER:Mitchell Chemiosmotic Hypothesis I n fact, the chemiosmotic hypothesis takes considerable liberties with our present knowledge of the respiratory chain : a) It would have us accept that a proton would remain attached t o NADH as it moved from the right-hand side (inside) of the membrane t o the left b) According t o the respiratory chain proposed by Mitchell 131, the reduction of flavin in NADH and succinate dehydrogenases proceeds to the level of oxidation of the fully reduced leucoflavin There is, in fact, much evidence in favour of the view that the transition is from the oxidized flavin t o the semiquinone [311 c) There is no experimental evidence that the iron-sulphur system of NADH or succinate dehydrogenase is reduced before the ffavin as required by the chemiosmotic hypothesis d) The chemiosmotic hypothesis requires alternating hydrogen and electron carriers Accordingly, ubiquinone is placed between cytochromes b and c,, instead of between fiavin and cytochrome b where it is usually placed [32] The sensitivity of &H2 oxidation to antimycin, and the close association of antimycin inhibition with cytochrome b, supports the conventional respiratory chain Indeed, it is very difficult, although not perhaps impossible [33], t o reconcile the chemiosmotic hypothesis with what is known about the respiratory chain (cf [17,33]) CONCLUSIONS Of the four basic postulates of the chemiosmotic system, experimental evidence exists for only two, the proton-translocating reversible ATPase system and the ion-impermeable membrane There is no evidence for the existence of a proton-translocating oxidoreduction chain or for the exchange-diffusion system Indeed, recent experimental data make it very doubtful if the respiratory chain is capable of proton translocation, in the absence of cations Furthermore, the thermodynamic difficulties of the chemiosmotic theory are formidable It must be concluded that the chemiosmotic theory, in its present form, is untenable Our knowledge of the mechanism of oxidative phosphorylation and of cation movements is best summarized by Scheme C above The unravelling of the nature of E, remains a challenge The possibility that protons, produced by the respiratory chain, raise the effective acidity of the membrane, in such a way as to favour ATP synthesis, as suggested by Williams [8,9], is not contradicted by the arguments brought forward in this article and deserves further consideration The experimental work reported in this paper was supported in part by grants from the Life Insurance Medical Research Fund and the U.S Public Health Service European J Biochem REFERENCES Slater, E C., Nature, 172 (1953) 975 Mitchell, P., Nature, 191 (1961) 144 Mitchell, P., Chemiosmotic coupling in oxidative and photosynthetic phosphorylation, Glynn Research Ltd., Bodwin 1966 Mitchell, P., Biol Rev 41 (1966) 445 Reid, R A., Moyle, J., and Mitchell, P., Nature, 212 (1966) 257 Mitchell, P., and Moyle, J., Nature, 213 (1967) 137 Robertson, R N., Biol Rev 35 (1960) 231 Williams, R J P., J Theoret Biol (1961) I Williams, R J P., J Theoret Biol (1962) 209 10 Cockrell, R S., Harris, E J., and Pressman, B C., Biochemistry, (1966) 2326 11 Klingenberg, M., and Schollmeyer, P., in Intracellular Respiration: Phosphorylating and Non-Phosphorylating Oxidution Reactions, Proceedings of the Fifth International Congress of Biochemistry, Moscow 1961, Pergamon Press, Oxford 1963, Vol 5, p 46 12 Jagendorf, A T., and Hind, G., in Photosynthesis mechanisms in green plants (edited by B IZok and A T Jagendorf ), National Academy of Sciences, Washington 1963, p 599 13 Saris, N C., Dissertation, University of Helsinki, 1963 14 Jagendorf A T and Uribe., E , Proc Natl Acad Sci 5.8 55 (1966) 170 15 Taeer J M Veldsema-Currie, R D., and Slater, E C., %uture, 212 (1966) 376 16 McCarty, R.E., and Racker,E., Federation Proc 25 (1966) 226 17 Chance, B., and Mela, L., Nature, 212 (1966) 369, 372 18 Chance, B., and Hollunger, G., J Biol Chem 236 (1961) 1524 19 Mayne, B C., and Clayton, R K., Proc Natl Acad Sci U S 55 (1966) 494 20 Jagendorf, A T., and Uribe, E., in Brookhaven Symp Biol 19: BNL 989 (C-48), 1967, in the press 21 Jardetzky, O., Nature, 211 (1966) 969 22 Garrahan, P J., and Glynn, I M., Nature, 211 (1966) 1414 23 Lehninger, A L., Rossi, C S., Bielawski, J., and Gear, A., in Mitochondrondrial Structure and Compartmentation (edited by S Papa, J M Tager, E Quagliariello, and E C Slater), Adriatica Editrice, Bari, 1967, in the press 24 Mitchell, P., and Moyle, J., Nature, 208 (1965) 147 25 Mitchell, P., and Moyle, J., in Biochemistry of Mitochondria (edited by E C Slater, Z Kaniuga, and L Wojtczak), Academic Press and P.W.N., London and Warsaw, 1967, p 53 26 Chance, B., and Williams, G R., Advan Enzymol 17 (1956) 65 27 Dam, K van, Biochim Biophys Acta, 128 (1966) 337 28 Carafoli, E., Gamble, R L., and Lehninger, A L., Biochem Biophys Res Commun 21 (1965) 215 29 Rossi, C., and Azzone, G F., Biochim Biophys Acta, 110 (1965) 434 30 Rienits, K G., unpublished observations 31 Slater, E C (Ed.), Flavins and Flavoproteins (BBA Library, Vol 8), Elsevier, Amsterdam, 1966 32 Slater, E C., Colpa-Boonstra,J P., and Links, J., in Ciba Foundation Symp on Quinones in Electron Transport, Churchill, London, 1961, p 161 33 Slater, E C., in Biochemistry of Mitochondria (edited by E C Slater, Z Kaniuga, and L Wojtczak), Academic Press and P.W.N., London and Warsaw, 1967, p E C Slater Laboratorium voor Biochemie, B C.P Jansen Institute Plantage Muidergracht 12, Amsterdam-C, The Netherlands ... Nature, 172 (1953) 975 Mitchell, P., Nature, 191 (1961) 144 Mitchell, P., Chemiosmotic coupling in oxidative and photosynthetic phosphorylation, Glynn Research Ltd., Bodwin 1966 Mitchell, P., Biol... Proton-translocatingOxido-reductionChain Mitchellrepresents the proton-translocating oxidoreduction chain as in Fig (essentially identical with Fig 11 of Mitchell [3]) In terms of chemical equa-... ATPase Xystern Mitchell [ref 3, p 521 represents the oligomycinsensitive proton-translocating ATPase system by the reactions XH, XI, H2O 2H,+ European J Biochem S f 2Hf 1/202t2HS H20 Fig Mitchell? ??s