Analysis of Molecular Mechanisms of ATP Synthesis from the Standpoint of the Principle of Electrical Neutrality

Biophysical Chemistry 224 (2017) 49–58

Contents lists available at ScienceDirect

Biophysical Chemistry

journal homepage: http://www.elsevier.com/locate/biophyschem

Review Analysis of molecular mechanisms of ATP synthesis from the standpoint of the principle of electrical neutrality

Sunil Nath

Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology, Delhi, Hauz Khas, New Delhi 110016, India

HIGHLIGHTS GRAPHICAL ABSTRACT

• Theories of biological energy coupling and mechanisms of ATP synthesis are reviewed. • Current ATP theories are evaluated based on the principle of electrical neutrality. • Mitchell's chemiosmotic theory is shown to violate the electroneutrality principle. • A dynamically electrogenic but overall electroneutral mode of ion transport is proposed. • Nath's torsional mechanism satisﬁes electroneutrality and is a more complete theory.

article info abstract

Article history: Theories of biological energy coupling in oxidative phosphorylation (OX PHOS) and photophosphorylation Received 13 February 2017 (PHOTO PHOS) are reviewed and applied to ATP synthesis by an experimental system containing purified ATP Received in revised form 6 March 2017 synthase reconstituted into liposomes. The theories are critically evaluated from the standpoint of the principle Accepted 6 March 2017 of electrical neutrality. It is shown that the obligatory requirement to maintain overall electroneutrality of bulk Available online 8 March 2017 aqueous phases imposes strong constraints on possible theories of energy coupling and molecular mechanisms of ATP synthesis. Mitchell's chemiosmotic theory is found to violate the electroneutrality of bulk aqueous phases Keywords: Principle of electrical neutrality and is shown to be untenable on these grounds. Purely electroneutral mechanisms or mechanisms where the Electroneutrality anion/countercation gradient is dissipated or simply flows through the lipid bilayer are also shown to be inade- ATP synthesis quate. A dynamically electrogenic but overall electroneutral mode of ion transport postulated by Nath's torsional Bioenergetics mechanism of energy transduction and ATP synthesis is shown to be consistent both with the experimental find- Oxidative phosphorylation ings and the principle of electrical neutrality. It is concluded that the ATP synthase functions as a proton-dicar- Biological energy coupling boxylic acid anion cotransporter in OX PHOS or PHOTO PHOS. A logical chemical explanation for the selection Nonequilibrium thermodynamics of dicarboxylic acids as intermediates in OX PHOS and PHOTO PHOS is suggested based on the pioneering classi- Mitchell's chemiosmotic theory cal thermodynamic work of Christensen, Izatt, and Hansen. The nonequilibrium thermodynamic consequences Nath's torsional mechanism of energy transduction and ATP synthesis for theories in which the protons originate from water vis-a-vis weak organic acids are compared and contrasted, and several new mechanistic and thermodynamic insights into biological energy transduction by ATP synthase F1FO Ion transport are offered. These considerations make the new theory of energy coupling more complete, and lead to a deeper understanding of the molecular mechanism of ATP synthesis. © 2017 Elsevier B.V. All rights reserved.

E-mail addresses: [email protected], [email protected].

Contents

1. Introduction...... 50 2. ClassicaltheoriesofenergycouplinginOXPHOS...... 50 2.1. TheongoingdebateonOXPHOSsincethe1960s...... 50 3. NewconceptualframeworkforenergycouplinginOXPHOSandPHOTOPHOS...... 50 4. TheissueofelectricalneutralityanditsbiologicalimplicationsforenergycouplingandATPsynthesis...... 51 4.1. Theprincipleofelectricalneutrality...... 51 4.2. Experimentalsystem...... 51 4.3. ViolationoftheprincipleofelectricalneutralitybyMitchell'schemiosmotictheory...... 51 4.4. Otherimportantconsequencesforthechemiosmotictheory...... 53 4.4.1. Protonmotive force, ΔpisnotanessentialintermediateinATPsynthesis...... 53 4.4.2. Delocalized electrical potential, Δφ isnotanessentialintermediateinATPsynthesis...... 53 4.5. Adynamicallyelectrogenicbutoverallelectroneutralmodeofiontransportasanewparadigminbioenergetics...... 53 4.6. Sitesofentryandexitofthedicarboxylicacidanion...... 54 5. PossiblechemicallogicfortheselectionofdicarboxylicacidsasintermediatesofenergycouplinginOXPHOSandPHOTOPHOS...... 54 6. Furtherdetails...... 55 6.1. Estimation of electrogenic H+ chargetransferrequiredbychemiosmosisinmitochondrialOXPHOS...... 55 6.2. EnergytransductionbyaMitchellianfuelcellversusaNatheanbiologicalmolecularmachine...... 55 6.3. The issue of H+/2e− and H+/ATPstoichiometries...... 55 6.4. Usage of the term “electrogenic” and stationary vs. transient electrical ﬁelds...... 56 6.5. RelationshiptotheNernstequation...... 56 6.6. Transient departure from electrical neutrality as an ordering principle for transport and reaction steps and thestringencyofelectricalneutrality...... 57 7. Conclusions...... 57 Acknowledgements...... 58 References...... 58

1. Introduction coupling oxidation and phosphorylation [3,4]. According to the theory, the membrane itself plays no active role in the energy transduction, serv- The synthesis of adenosine-5′-triphosphate (ATP), the universal bio- ing only as “insulation material” for separating two bulk phases, and logical energy currency, catalyzed by the enzyme F1FO-ATP synthase is Mitchell repeatedly emphasized this point and the key role of such “ener- the fundamental means of cellular energy production in animals, plants gized bulk aqueous media” [4]. The chemiosmotic theory is an equilibrium and microorganisms. It is among the most important and frequently oc- theory lying in the realm of classical equilibrium thermodynamics, and the curring enzyme reactions in biology. ATP synthesis is achieved by cou- Δp is considered by the theory to equilibrate both with the respiratory pling of the reactions of oxidation and phosphorylation in the process of chain and the high-energy squiggle (the ~) in ATP [2–4]. oxidative phosphorylation (OX PHOS) or light reactions and phosphorylation (PHOTO PHOS) in photosynthesis. The prevailing theory of energy 2.1. The ongoing debate on OX PHOS since the 1960s coupling serves as a guiding compass for several sectors of biochemical and biophysical research. Hence, the mechanism underlying the coupling The chemiosmotic theory itself had a contentious history, with part of chemical reactions on the redox/photo and ATP sides is among the of the accumulated body of experimental evidence supporting the the- most important questions of physical and biophysical chemistry. In this ory, and part of it in conflict with it [summarized in refs. 5–7]. The the- fl invited article, the major theories of biological energy coupling are brie y ory was severely criticized by many stalwarts of 20th century delineated (Sections 2 and 3). In Section 4, these theories of energy cou- biochemistry, including E. C. Slater [8],R.J.P.Williams[9], Albert pling and molecular mechanisms of ATP synthesis are critically examined Lehninger [10],GregorioWeber[11], and David Green [12] on the from the standpoint of the principle of electrical neutrality. It is shown grounds of it either being “physically unsound,” involving “unrealistic that the principle imposes strong constraints on possible mechanisms of assumptions,” or being “unsupported by experiments” and having “no energy transduction, coupling, and ATP synthesis. basis in fact” [8–12], and specifically of violating “the necessity to ob- serve charge neutrality in chemical reactions” [12]. However, despite 2. Classical theories of energy coupling in OX PHOS long and heated debates, among the most acrimonious in modern science, and commonly known as “the OX PHOS wars” [13], the controver- fi The rsttheoryofenergycouplinginOXPHOSproposedbySlater[1] sies were never resolved [5–7]. Even the scientific ingenuity of the envisaged a chemical high-energy intermediate as the link between oxida- above stalwarts of biochemistry and biophysics and the efforts of a tion and phosphorylation. The theory was abandoned after a massive and large number of researchers in the field of bioenergetics in the 20th lengthy search (lasting ~20 years) for the elusive chemical intermediate century could not formulate a theory of energy coupling that might be proved futile, and was eventually replaced by Mitchell's chemiosmotic the- put in place of chemiosmosis. Moreover, the design of experiments oryofenergycoupling[2,3]. The chemiosmotic theory postulated that en- and the interpretation of data failed to offer any guidance either in “ ” Δ ergy coupling was achieved in OX PHOS by a protonmotive force p[Eq. devising a rational substitute. The chemiosmotic theory was incorporat- Δ (1),where p is in units of mV] obtained by addition of the bulk-to-bulk ed into the textbooks more by erosion of the opposition than by any Δ Δφ pH and a delocalized electrical potential, between bulk aqueous decisive experimentation or theoretical analysis. phases created by translocation of uncompensated protons across the membrane by the redox complexes in the respiratory chain. 3. New conceptual framework for energy coupling in OX PHOS and Δp ¼ Δφ−60ΔpH ð1Þ PHOTO PHOS

In the theory of chemiosmosis, the electrogenic Δp created across bulk Following a fresh molecular systems biology/engineering approach to aqueous phases is considered as the obligatory energy intermediate the problem, an alternative theory of energy coupling and molecular S. Nath / Biophysical Chemistry 224 (2017) 49–58 51 mechanism of ATP synthesis has been formulated and embellished by the author during the last two decades [5–7,14–21]. Nath's torsional mechanism of energy transduction and ATP synthesis has also been extensively discussed by other authors [22–28]. The molecular mechanism appealed to E. C. Slater and R. J. P. Williams (who, among the stalwarts mentioned above, were the only ones alive or scientifically active during the author's time), as evidenced by a long 15-year corre- spondence. In this new conceptual framework for energy coupling, the overall driving force for ATP synthesis is postulated to be the sum of the electrochemical potential difference of protons and anions/ countercations that translocate through the access channels at the interface of the a-c subunits, i.e., ΔμH̃ + ΔμA/C̃ ,orequivalently,ΔpH and ΔpA/ C (or because the driving force has to change form to act, ΔpH and ΔψA/ C,orΔψH and ΔψA/C, depending on the stage of the conformational cycle and where one draws the boundary surface). In other words, the driving force for ATP synthesis is created by two independent sources of energy, i.e., FO-permeable anions/countercations and protons respectively, in contrast to all other mechanisms, including chemiosmosis, that postulate a single energy source only (e.g., protons). Both molecular driving forces are thermodynamically intensive properties, related to the discrete, ordered and sequential elementary events of proton and anion/ countercation translocations (binding and unbinding) through the FO portion of the F1FO-ATP synthase at the a-c interface. New terminology of symsequenceport and antisequenceport was coined to enable a more perceptive understanding of the temporal sequence of events in the new paradigm in order to extract energy from anion/countercation Δψ translocationinadditiontoprotontranslocation.The in this mecha- Fig. 1. The experimental model system containing purified F1FO-ATP synthase nism is local and transient in the access half-channels of FO and has no re- reconstituted into a liposome. lationship with the delocalized potential, Δφ [Eq. (1)] postulated by chemiosmosis across bulk aqueous phases [6,17,18,20]. The mechanism is step-wise or dynamically electrogenic but overall electroneutral on great assistance in discriminating between various theories of energy both redox/photo and ATP sides, and hence does not violate overall coupling. It can be profitably applied to the routinely-used purified, electroneutrality of bulk media. These principles were postulated to be reconstituted F1FO-ATP synthase experimental system shown in Fig. 1. of a general and universal nature in biological systems, and applicable also to the related P-type, V-type, and A-type ATPases and to other bio- 4.2. Experimental system logical molecular machines such as the bacterial flagellar motor. Following a ten-year search [summarized in Ref. 7], the anion in- Depicted in Fig. 1 is the purified F F -ATP synthase reconstituted volved in biological energy coupling has been identified as a dicarboxyl- 1 O into a tight phospholipid membrane, the biochemist's choice of experi- ic acid anion by the torsional mechanism – succinate in mitochondrial mental system. The system was first introduced by Ephraim Racker in OX PHOS, malate in PHOTO PHOS by chloroplasts, and fumarate in the the early 1970s [36] and was therefore available during Mitchell's bacterial flagellar motor. This is a major experimental accomplishment time; however, today, liposome technology has advanced to a level of the work. The identity of the countercation to H+ (in certain bacterial where routine preparations contain a single molecule of the enzyme and other systems) was found to be either Na+ or K+.HencetheF F - 1 O per vesicle embedded in the membrane and plugged through it in a de- ATP synthase is a proton-dicarboxylic acid monoanion cotransporter in fined orientation (F sector facing out in Fig. 1) [37]. No other contami- mitochondria and chloroplasts, and not simply an electrogenic H+ 1 nating protein is present in the liposomal system of Fig. 1. Both “inside” conductor. and “outside” aqueous phases are macroscopic bulk phases that are spa- tially of large extent, and may be figuratively likened to the Atlantic and 4. The issue of electrical neutrality and its biological implications for Pacific oceans respectively, pursuing the analogy first adumbrated by energy coupling and ATP synthesis Williams [4,9]. This system enables study of ATP synthesis in a single enzyme and single vesicle mode, although for our purposes in this invited 4.1. The principle of electrical neutrality review, no difference exists even if the experimental system contains more than one F F molecule per vesicle or organelle. Such experimen- The principle of overall electrical neutrality in condensed media has 1 O tal systems have been shown to produce physiological rates of ATP syn- been covered in various textbooks of thermodynamics and biophysical thesis per F F molecule, e.g. when incubated typically in a succinic acid chemistry [29–31]. For a bulk liquid phase it can be expressed by the 1 O bath, and the characteristics of the system have been documented in a condition number of previous reports [38–41].Thisexperimentalobservationcre- ates insurmountable hurdles for the operation of a proton-only ∑ ¼ ð Þ zini 0 2 Mitchellian chemiosmotic mechanism (Fig. 2a). i

where zi is the charge number of the ionic species i, and ni is the number 4.3. Violation of the principle of electrical neutrality by Mitchell's chemios- of moles of the ith species. The electroneutrality principle is a corner- motic theory stone in generalization of Wyman's classical linked functions [32] and was used by Record and Anderson to arrive at a theory of preferential in- Fig. 2a shows that uncompensated H+ movement that lies at the teractions applicable to electrolyte ions [33–35]. The principle has major heart of the explanation of ATP synthesis by the chemiosmotic theory implications for molecular mechanisms of ATP synthesis and offers will violate the electroneutrality of both the inside and outside bulk 52 S. Nath / Biophysical Chemistry 224 (2017) 49–58

Fig. 2. Simpliﬁed representation of various models of energy coupling in OX PHOS and PHOTO PHOS based on a) the chemiosmotic theory, b) the torsional mechanism of energy transduction and ATP synthesis, with the bold arrow denoting the primary dicarboxylic acid monoanion translocation, and the dashed arrow showing the succeeding secondary proton translocation at the a-c stator-rotor interface in the membrane-bound FO portion of ATP synthase, and c) a mechanism in which overall electroneutrality is maintained by transporting the anion through the membrane lipid bilayer.

aqueous phases, as also pointed out by Wray [28], and earlier by Green the system, for instance in the form of light or redox energy, and the [12]. Thus, the outside bulk aqueous phase will incessantly accumulate thermodynamic system containing only the tight phospholipid coupling + H ions, while the inside bulk aqueous phase will contain an excess of membrane, the F1FO molecule(s), and different concentrations of un- OH– ions (Fig. 2a), and both bulk phases will violate the principle of charged and anionic components of the dicarboxylic acid and protons electrical neutrality as given by Eq. (2). One can perhaps imagine that in the two bulk compartments “in” and “out” is well-defined, and the if sufficient energy is supplied to the system, it might be possible to problem is well-posed. Such gross, spontaneous violation of electroneu- move protons in an uncompensated way. However, it should be trality of bulk aqueous phases that is required by chemiosmosis and is stressed that in the experiment, no external energy is being supplied to central to its view of energy coupling is unprecedented in nature. In S. Nath / Biophysical Chemistry 224 (2017) 49–58 53 fact, the author is not aware of any natural process that violates bulk is created, which was the view of Tedeschi and colleagues [43–45];anoth- electroneutrality in this way and to this extent. Hence the chemiosmotic er was a completely electroneutral transport across the membrane, which theory cannot be the correct theory of energy coupling in OX PHOS and would also lead to zero potential. However, the present author found it PHOTOPHOS. Therefore, an alternative theory of energy coupling that is very difficult to conceive how a mechanical torque could be produced in consistent with the principle of electrical neutrality is needed. the FO motor of the ATP synthase in the absence of a local electrical potential generated at or in the vicinity of the strictly conserved binding sites of the two 4.4. Other important consequences for the chemiosmotic theory ions (the a-subunit Arg-210 and the c-subunit Glu/Asp-61 for the anion and the proton respectively) in the half-access channels at the a-c interface

4.4.1. Protonmotive force, Δp is not an essential intermediate in ATP of FO, which he considered to be an obligatory requirement for rotation. synthesis Hence, both the above alternatives were considered highly unlikely, if ATP synthesis by the clean experimental liposome system of Fig. 1 not impossible, and therefore eliminated from further consideration. has several other important consequences for the chemiosmotic theory. The logically promising alternative suggested itself that the electrical

Interestingly, in these and other experiments [7,42] no protonmotive potential is created at the a-c interface within the FO portion of the ATP syn- force, Δp[Eq.(1)] has been imposed on the system; yet ATP is synthe- thase enzyme molecule itself (in other words, the electrical potential, Δψ is sized at physiological rates under speciﬁc experimental conditions, for local)byanindependent source other than protons. Energetic consider- instance when use of a succinate or malate bath is made. As is well- ations indicated that ΔpH supplied only part of the energy requirement known, the chemiosmotic theory postulates uncompensated protons (~50%) for ATP synthesis; therefore the rest had to be supplied by a locally in an electrogenic mode of translocation as the source of the present but independent source of Δψ. The overall driving forces for ATP protonmotive force [3], not compensated protons, as for example in synthesis are the ion concentration gradients due to protons and counter- the form of succinic acid or malic acid [7,18–20,38–42] imposed by ions (anions transported through symsequenceport or cations transported the experimentalist as gradient in a liposomal system (Fig. 1). Such ex- through antisequenceport), and a dynamically electrogenic but overall periments do not provide proof of the driving power of a protonmotive electroneutral mode of ion transport was proposed in this context [17]. force, as generally believed, because in reality, no protonmotive force This mode of ion transport involves a permeant anion (succinate in mito- was applied in these experiments. In fact, since physiological ATP syn- chondria, malate in chloroplasts, fumarate in the bacterial ﬂagellar motor) thesis occurs without application of a protonmotive force, these bio- that is translocated in the same direction as the proton (Fig. 2b), or a cation chemical experiments in reconstituted liposomes prove that, contrary (Na+ in certain bacterial ATPase/redox systems [46,47],orK+ in the pres- to chemiosmotic dogma, the protonmotive force is not an obligatory in- ence of valinomycin in vitro) being transported in a direction opposite to termediate in ATP synthesis. Hence Δp is not the primary form in which the direction of proton translocation [7]. However, both proton and energy is conserved in these experiments and it is absolutely essential anion (or countercation) do not move together or simultaneously or con- to consider alternatives to Δp as the intermediate in energy coupling. currently as in electroneutral ion transport mechanisms but sequentially (Fig. 2b). Hence, the ion transport is step-wise or dynamically electrogenic, 4.4.2. Delocalized electrical potential, Δφ is not an essential intermediate in but overall electroneutral. The discrete, sequential ion translocations in the

ATP synthesis half-access channels of FO transiently generate (for a time interval between A delocalized electrical potential between two bulk aqueous phases, the primary and secondary ion translocation events) a local Δψ,thatisim- Δφ created upon electrogenic H+ pumping by the redox complexes mediately destroyed by the succeeding (H+) translocation, and thus in- (Complexes I–IV) in animal mitochondria and bacteria at the expense volve a change in local electrical potential, Δ(Δψ)asanintermediatestep of free energy provided by the oxidation process in OX PHOS is a strict for energy transduction and rotation of the c-rotor in this novel mode of requirement of the chemiosmotic theory. Yet ATP synthesis occurs in ion transport. Yet, bulk electroneutrality is not violated and the energy of the purified reconstituted system of Fig. 1 in a succinic acid/malic acid both the ion gradients are harnessed and jointly utilized for rotation by bath even in the absence of redox enzymes or photosystems, complexes the FO motor. These central concepts can be written in the form of mathe- that are supposedly responsible for the creation of the delocalized Δφ matical equations that govern the torsional mechanism and describe its according to the chemiosmotic theory! Hence the delocalized electrical operation, and also include the constraint imposed on the overall process potential, Δφ is not an essential intermediate in ATP synthesis. Further, by the principle of electrical neutrality (Section 4.1). once the central thesis of this article is accepted – that the electroneu- According to the torsional mechanism, the overall driving force trality of bulk aqueous phases cannot be violated by an electrogenic (d.f.), i.e. the energy supplied to synthesize one molecule of ATP by mode of ion transport involving incessant uncompensated proton trans- coupled ion translocation [6,7] is given by location from one bulk phase to another (Section 4.3) – then, increasing : : ¼ Δμ̃ þ Δμ̃ ð Þ the magnitude of ΔpH between the phases cannot compensate for the d f nH H nA=C A=C 3 absence of the delocalized Δφ, and the ΔpH alone cannot sustain ATP synthesis and act as its sole driving force. Hence a new paradigm was where nH and nA/C represent the number of proton and anion/ countercation translocations required to make one ATP molecule, and needed to understand the energetics of energy coupling in the FO por- Δμ̃ Δμ̃ tion of ATP synthase. H and A/C are the electrochemical potential differences of proton and anion/countercation respectively. Eq. (3) is general and is valid 4.5. A dynamically electrogenic but overall electroneutral mode of ion under nonequilibrium conditions at which coupled proton and anion/ transport as a new paradigm in bioenergetics countercation transport occurs in biological systems. Since overall electroneutrality has to be maintained during the As discussed above, physiological ATP synthesis has been shown to coupled ion translocation process, occur under specific experimental conditions with the enzyme molecule n ¼ n = ¼ n ð4Þ purified and reconstituted into liposomes, which do not contain any H A C redox/photo complexes. Further, since bulk electroneutrality is inviolable, fl an extensive study was needed in order to provide new insights into ener- Hence, in general, for nonequilibrium ow conditions, gy coupling in OX PHOS and PHOTO PHOS. In particular, a reappraisal of d:f : ¼ n Δμ̃ þ Δμ̃ = ð5Þ the mode of ion transport across the membrane through the FO portion H A C of ATP synthase was felt to be absolutely necessary when the development of the torsional mechanism of energy transduction and ATP synthesis was Eq. (5) is the principal equation, as per the torsional mechanism, for aworkinprogress[17–19]. One possibility was that no electrical potential describing physiological ATP synthesis by the F1FO-ATP synthase. 54 S. Nath / Biophysical Chemistry 224 (2017) 49–58

For the dead-end condition of equilibrium, the net d.f. = 0, ion trans- a-c interface of the FO portion of the F1FO-ATP synthase transporter mol- port ceases to occur, and ATP is not synthesized. The general Eq. (5) re- ecule itself (Fig. 2b), these workers postulated that the succinate is duces to the special case of transported through the membrane (Fig. 2c). While this model over- came the problem of violation of bulk electroneutrality, it created a host of other serious difficulties. A major difficulty for this variant is ΔμH̃ þ Δμ̃ = ¼ 0 ð6Þ A C that the lipid bilayer is known to be impermeable to ions; hence postulating a flux of succinate through the membrane (Fig. 2c) runs counter Eq. (6), though not useful under the prevailing physiological condi- to our knowledge of membrane biophysics. Basically, the lipid bilayer tions of nonequilibrium steady states, can however be utilized to obtain is a barrier to all charged species, and the real question evolutionarily the ionic distributions under conditions of thermodynamic equilibrium. was how to circumvent such a tenacious barrier to charged species, However, in order to extract energy from the anion/countercation, it which is why nature devised membrane-bound transporters. Further, is important to understand the temporal sequence of ion translocation allowing such bulk-to-bulk passage of the succinate monoanion will events. The possibilities of simultaneous transport of proton and anion collapse the delocalized Δφ created by primary proton movement and (or countercation) or proton transport preceding anion (or dissipate the driving force of chemiosmosis in OX PHOS. Moreover, countercation) translocation through access pathways at the a-c sta- such ion translocation (Fig. 2c) is inconsistent with a central postulate tor-rotor interface in FO are ruled out because in either case, the energy of the chemiosmotic theory, as discussed previously [p. 302 of Ref. 6]. of the anion (or countercation) gradient is not made available to the Such ion translocation would anyhow require modification of the fun- proton, and therefore complete rotation of the c-rotor in the FO portion damental postulates of the chemiosmosis, or the proposal of a new the- of ATP synthase cannot occur. Thus anion (or countercation) transloca- ory of energy coupling. Further, the acute mechanistic difficulties of tion (bold arrow in Fig. 2b) must precede proton translocation (dashed transmitting the driving force to the FO sites in the a-stator and c- arrow in Fig. 2b) through the proton half-access channels at the a-c in- rotor, where they are required, are not addressed in a model such as terface in FO. In other words, anion or countercation translocation is pri- that shown in Fig. 2c. These difficulties do not arise in the proposals of mary and proton translocation is secondary on the ATP side; primary the torsional mechanism (Fig. 2b) because the molecular driving forces and secondary events are interchanged on the redox/photo sides are created by anion and proton binding/unbinding to/from the strictly where proton translocation in response to electron transfer is the pri- conserved Arg-210 and Glu/Asp-61 binding sites in the a-stator and c- mary event, followed by secondary anion or countercation transloca- rotor respectively; such collimated molecular driving forces act exactly tion. A strictly conserved Arg-Leu-Asn-Ala (RLNA) sequence on helix 4 at the sites where they are required for torque production and rotation of the a-subunit has been identified as the signature of an anion-binding in the FO motor, and therefore they can be readily utilized for sequence in mitochondria and chloroplasts, with the essential aArg-210 performing useful work. Finally, postulating that the energy of succinate forming the dicarboxylate anion binding site in the immediate vicinity translocation through the membrane is not employed for performance of the cGlu/Asp-61 proton-binding site at the a-c interface of FO [7]. of useful work but serves only as a means to maintain electroneutrality In summary, the huge problem of violation of electrical neutrality of is not adequate either: it essentially implies that the succinate ion gradi- bulk aqueous phases is avoided in Nath's torsional mechanism of ATP ent is dissipated, and that the efficiency of energy conversion is halved. synthesis by postulating a dynamically electrogenic but overall Such a model is also incompatible with comprehensive nonequilibrium electroneutral mode of ion transport. This unique mode of ion transport thermodynamic analyses of ATP synthesis [14,21]. has been quantified, and the principal result [Eq. (5)] is shown to be In summary, the only three models of energy coupling possible in valid for general nonequilibrium states, and incorporates within it the the context of ATP synthesis are clearly delineated and depicted in Fig. constraint of overall electroneutrality arising from the coupling of pro- 2. Following the arguments made in Sections 4.3–4.4,andSection 4.6 re- ton and anion/countercation translocation. Because of the sequential spectively, the alternatives shown in Fig. 2(a) and Fig. 2(c) appear inde- translocation of anion (or countercation) and proton through contigu- fensible to the author. From the detailed analysis in Section 4.5,the ous transport access channels in the a- and c-subunits respectively to/ theory of energy coupling proposed by Nath's torsional mechanism of from their binding sites at the a-c stator-rotor interface of FO,atransient energy transduction and ATP synthesis (Fig. 2b) emerges as the only local field is created and subsequently destroyed, leading to the genera- available theory that is consistent with the principle of electrical neu- tion of a mechanical torque that is ultimately transduced through a cas- trality, and it is by far the best conceptual framework available for logi- cade of events into torsional energy in the γ-subunit of the F1 portion of cal interpretation of experimental data and guidance in the design of ATP synthase [15–20], and hence the name of the mechanism. This tor- future experiments in this highly interdisciplinary field. sional energy is used at the three catalytic β-sites in F1 to competently bind ADP and Pi, forcibly condense them to form ATP, and release 5. Possible chemical logic for the selection of dicarboxylic acids as in- preformed ATP respectively by a novel catalytic cycle in which all termediates of energy coupling in OX PHOS and PHOTO PHOS three catalytic sites are occupied by bound nucleotide and every elementary step requires energy [17], contradicting the fundamental te- Following the detailed analysis in Section 4, the obvious question nets of Boyer's binding change mechanism [48].Thefifteen novel that arises in the reader's mind is what the underlying chemical reasons predictions made in 2002 [see pp. 79–80 of Ref. 17] for the physiologi- could be for the evolutionary selection of dicarboxylic acids as interme- cally important mode of ATP synthesis by the F1FO-ATP synthase (as op- diates in biological energy coupling. A possible and attractive chemical posed to ATP hydrolysis by F1-ATPase, a wasteful process in both explanation can be offered thanks to the pioneering classical thermody- mitochondria and chloroplasts) were well ahead of their time and are namic research work of Christensen, Izatt, and Hansen. These workers still relevant today. conducted a broad research program to measure the thermodynamic quantities associated with proton ionization in dilute aqueous solution 4.6. Sites of entry and exit of the dicarboxylic acid anion from various donor atom types. Initially, these studies were performed at 298 K [49], but later, temperature studies were also carried out by cal- Depending on the sites of entry and exit of the dicarboxylic acid orimetry and the results tabulated in a Handbook [50]. Inspection of monoanion, other ways of energy coupling are perhaps possible. Kaim their results shows that weak dicarboxylic acids such as succinic and and Dimroth [41] showed experimentally that succinate monoanion is malic acids are among the few chemical systems that have a fairly indeed translocated in the purified reconstituted system of Fig. 1.How- small standard enthalpy of ionization, ΔH° (compared to water or ever, unlike the torsional mechanism that localized the translocation of strong acids) and their ionic fractions at the operating conditions and succinate monoanion through the aqueous access half-channels at the the values of the first and second ionization enthalpies are tuned such S. Nath / Biophysical Chemistry 224 (2017) 49–58 55 that the net ΔH° approaches zero at 310 K [21,49,50]. In other words, 6.2. Energy transduction by a Mitchellian fuel cell versus a Nathean biolog- virtually no input of external enthalpy is needed to dissociate these di- ical molecular machine carboxylic acids; the first proton is almost completely disassociated in aqueous solution at physiological pH and the enthalpy change for disas- In the Mitchellian view [51] where energy-transducing membranes sociation of the second proton is near zero. Gibbs free energies for disas- in mitochondrial OX PHOS are fuel cells that split water, transfer of a sociation of carboxylic acids are almost entirely entropic. large number of electrons (8000·2e− cycles) and protons (80,000 H+ By contrast, the protons in Mitchell's chemiosmotic theory originate per mitochondrion) are required to attain a Δp competent to synthesize from water [51], which has a large standard enthalpy of ionization of ATP. Transfer of 2e− creates only a negligible delocalized Δφ and Δpin 57.3 kJ/mol. Thus the mitochondrial respiratory chain needs to addition- Mitchell's model, and consequently does not lead to ATP synthesis. By ally expend 57.3 × 10 = 573 kJ per pair of electrons in order to generate contrast, a Nathean biological molecular machine is characterized by 10H+ and 10 OH– ions (assuming the consensus value of stoichiometry the involvement of nonequilibrium states and an irreversible mode of of 10H+/2e− in animal mitochondria) [21,52] from water by the hypo- operation [17,18] (though the molecular machine/enzyme can be re- thetical fuel cell in mitochondria that release H+ on one side of the versed by imposing a different set of initial and boundary conditions, membrane and the OH– on the other side and ensure the “charge sepa- and in that sense the system is indeed reversible [20]), and a mecha- ration across the membrane” required by chemiosmosis [3,4,51].How- nism of e− transfer and ATP synthesis where ΔG is equal to or close to ever, a pair of electrons moving down the respiratory chain in OX PHOS ΔH. In a Nathean model where the potential energy depends on micro- can generate 220 kJ/mol only. Moreover, in addition to synthesizing scopic events and the driving forces of ATP synthesis are molecular in ATP, the energy of 220 kJ/mol per 2e− also needs to compensate for ac- nature, the translocation of only a few protons and anions/ tive transport losses in the electrogenic H+ pumps, leaks in the mem- countercations (arising from a single cycle of 2e− transfer) is sufficient brane, respiratory slip, and other losses at the coupling sites. The to reach a potential energy level compatible with the Gibbs free energy extremely steep costs of ionization/dissociation of water (that have of phosphorylation and produce a nonequilibrium energized state of a not been included previously by any energy balance of the OX PHOS submolecular component of a single F1FO-ATP synthase molecule in- process) are inconsistent with the nonequilibrium thermodynamics of volved in the energy transduction. This stored internal energy within energy coupling [14,21,52,53]. Needless to say, inclusion of the high the single molecule is subsequently utilized to synthesize ATP. The thermodynamic costs owing to the large ionization enthalpy of water two alternative models (fuel cell vs. biological molecular machine) are renders impossible the operation of a Mitchellian charge-separating diagrammatically compared in Fig. 3. In the molecular energy model chemiosmotic engine in which water is the origin of the translocated of Nath's torsional mechanism, first the c-subunits of the c-rotor in FO protons. These difficulties do not arise in the torsional mechanism of en- and subsequently the γ-subunit in F1 store an internal potential energy ergy transduction and ATP synthesis according to which a weak dicar- (of ~54–58 kJ/mol as twist/torsion). This is contrasted with the equiva- boxylic acid is the source of the proton and the anion, and the lent energy of the Δp attained by a macroscopic electrogenic process in energies of both proton and anion translocations in the membrane- Mitchell's chemiosmosis model (Fig. 3). Hence, in conclusion, it is unre- bound FO portion of ATP synthase are collaboratively utilized to synthe- alistic to retain the analogy of energy-transducing membranes as water- size ATP in the hydrophilic F1 headpiece. The above thermodynamic ar- splitting Mitchellian fuel cells, and the analogy of the ATP synthase to a guments lend us even greater confidence in the superiority of the Nathean biological molecular machine is far superior. concept of energy coupling proposed in the framework of the torsional mechanism of ATP synthesis over other alternatives. 6.3. The issue of H+/2e− and H+/ATP stoichiometries

− 6. Further details Mitchell postulated an H+/2e stoichiometry of 6 and an H+/ATP stoichiometry of 2 in his theory, stoichiometries that he obdurately While it is satisfying to be assured by one of the referees that hung on to until his dying day even in the face of adverse experimental “bulk electroneutrality is indeed broken in Mitchell's theory” and evidence that unequivocally showed higher values of redox and ATPase by another that “use of the principle of electrical neutrality has con- stoichiometries in mitochondrial OX PHOS (Section 5) [13]. However, vincingly shown the ﬂaw in Mitchell's chemiosmotic theory,” it is he never gave any explanation, or offered even a hint of the reasons nonetheless important to quantify the extent of electrogenic charge for his deﬁance. Following the detailed thermodynamic considerations transfer postulated by chemiosmosis under physiological conditions, in Sections 5 and 6.2, it can be easily calculated that Mitchell's water- as suggested. splitting fuel cell model [51] in the mitochondrion could provide only 220/57.3 or approximately four H+ per 2e−, with another 2H+/2e− + − arising from the QH2 → Q+2H +2e redox reaction, i.e., he could ob- − 6.1. Estimation of electrogenic H+ charge transfer required by tain at most 6H+/2e from his theory. Acceptance of the chemiosmosis in mitochondrial OX PHOS

In order to achieve a bulk-to-bulk delocalized Δφ of 200 mV by the macroscopic process of chemiosmotic coupling, a charge transfer of 0.8 μeq H+/g protein in rat liver mitochondria was estimated [3]. Classical morphological data shows that 1 mg mitochondrial protein contains 7.2 × 109 mitochondria [54]. Hence, in order to reach a delocalized Δφ of 240 mV required by the chemiosmotic theory during OX PHOS, the number of protons translocated per mitochondrion corresponds to [0.8 × (240/200) × 10−9 × 6.0 × 1023]/7.2 × 109 = 80,000 H+/mitochondrion. Thus, in chemiosmosis, where the potential energy depends on a macroscopic charge transfer process, 8000 cycles of transfer of a pair of electrons in OX PHOS are required in a single mitochondrion for the formation of a protonmotive force, Δp compatible with the thermodynamic requirements imposed by the Gibbs free ener- Fig. 3. Comparison of macroscopic fuel cell and molecular energy transduction models of gy of phosphorylation. energy coupling. Consult the text in Section 6.2 for details. 56 S. Nath / Biophysical Chemistry 224 (2017) 49–58 experimentally obtained higher OX PHOS redox stoichiometries of 10 formation was electrolytic field driven and is impossible” [see p. 148, H+/2e− created an energy crisis for the chemiosmotic theory, because col. 1 in Ref. 55]and“Both Mitchell and I failed to give a good description it meant that water could not be the source of the coupling protons, of the ATP synthetases” [p.149,col.2inRef.55]. Energy conservation re- that the chemiosmotic theory would violate the first law of thermody- quires that useful work be done at the expense of a reduction in electri- namics, and that his (hypothetical) fuel cell could not operate in a man- cal field strength, in which case the stationary field has to take on a ner competent to generate the obligatory delocalized intermediate, Δp lower value of field strength. This reduction in the electrical potential envisaged in chemiosmosis. Mitchell would have known that accepting can only be brought about by an actual ion current of differing character the higher consensus stoichiometry would strike at the root of his redox from the one that created the field in the first place. Hence localized loop concept, and definitely invalidate his chemiosmotic theory because models [9,57,58] where protonic charge build-up is confined to a thin of the steep energy shortfall. surface region adjacent to the membrane are also inadequate, not be- Similarly, a stoichiometry on the ATP-side of 2 H+/ATP is required by cause they violate electroneutrality (since a membrane surface can ac- and central to the direct coupling between H+ and ATP formation pos- cumulate ionic charge and a membrane potential can be built up), but tulated by Mitchell [4]. In this model (contrary to a model [16,17] in- because they cannot be coupled to rotation in FO and performance of volving conformational changes in F1), it is envisaged that, at the F1 useful work. active site, binding of ADP and phosphate is arranged so that the sub- The above difficulties are avoided in the torsional mechanism be- strates are oriented perpendicular to the plane of membrane at the in- cause only closely bound H+ and A− at their respective binding sites terface between FO and F1, thereby permitting the terminal oxygen can be coupled to rotation due to destructive interference by an ion of op- + atom from phosphate to accept 2H from FO, i.e. two positive charges posite charge from the one that created the transient electrical field, and move inward to finally reside in water. Acceptance of higher values of further, the electrical imbalance and its return to a balanced charge con- three to four for the H+/ATP ratio would also have invalidated direct figuration occurs only at the restricted site of the ion binding/unbinding coupling schemes between H+ and ATP formation postulated by the process. chemiosmotic theory. Electrostatic interaction between the essential a-c residues upon Hence, we have explained for the first time the underlying reasons electrical imbalance produced by ion binding/unbinding is the driving for Mitchell's apparently irrational and dogmatic refusal to accept the force for c-ring rotation and ATP synthesis in the torsional mechanism. higher stoichiometry values definitively revealed by various experi- Since inception, it was predicted that the geometry of the a- and c-sub- ments in bioenergetics. The new higher stoichiometry values have units should be such that “while c is a complete cylinder, the a-subunit now been accepted as the “consensus” values [21], but it is ironical (containing the key Arg-210 and His-245 residues) is part of a cylinder that the theory of energy coupling remains of the old, and one that can- coaxial to c, covering two subunits of c (each containing a Glu/Asp-61 not be reconciled with the new experiments. H+ binding site). Thus, the interacting region of a- and c-subunits can be considered as the surfaces of two coaxial cylinders in close proximity 6.4. Usage of the term “electrogenic” and stationary vs. transient electrical to each other” [15], a prediction supported by the recent cryoelectron fields microscopy (cryo-EM) structure of the a-c interface in an F-type ATPase [59]. In this paper, the word “electrogenic” refers to the internal potential The a-subunit access channels translocate anions to their binding differences appearing inside FO due to discrete sequential proton and site located at the a-c interface. This would bring the dicarboxylate anion/countercation translocation events to and from their respective anion very close to the proton, as repeatedly emphasized by the tor- binding sites in the a-c half-access channels. It should not be confused sional mechanism and italicized (see pp. 2222–2223 of Ref. 19;p. with the transfer of electrical charge across the membrane or to the 433 of Ref. 7]. Interestingly, the important side chains in the horizon- build-up of electrical potential across the membrane in the bulk or at tal helices of the a-subunit have also been interpreted to approach the membrane surface. the protonation site in the c-subunit more closely (than would be It should also be understood that the chemiosmotic process of pro- the case if the helices were vertical) in the cryo-EM structure, and ton transfer does not cause acidification of an external medium or also enable the Arg-210 and His-245 to interact with two adjacent bulk region per se as happens extensively for instance in PHOTO c-subunits simultaneously with a constant distance of interaction PHOS; however for acidification and measurement of the pH difference [59], exactly as predicted by Nath's torsional mechanism of energy by probes that always access a larger space than the immediate mem- transduction and ATP synthesis since 1998. The close approach pre- brane surface, a second ion such as a dicarboxylic acid anion or chloride dictions of the torsional mechanism were made to account for the in- etc. is also necessary. teraction of the anion with the proton and thereby enable their An interesting aspect of the liposome experiments (Fig. 1; Section coupled symsequenceport translocation as separate charges without 4.2) is that it is not possible to take refuge in explanations based on cre- recombination in the membrane, and for energy addition, joint ener- ation of a localized or delocalized field by other redox/photosystem gy utilization, and ATP synthesis. However, in the absence of any role pumps present in natural experimental mitochondria/chloroplast sys- ascribed to the anion, the rationale for such a close approach (of tems or other transporters present in partially purified systems, and col- 1.1 nm – 1.35 nm depending on the diameter of the c-ring in differ- lapse of the field by other agents such as K+ in the presence of ent organisms) is very difficult to explain. Hence a functional role valinomycin or a general H+ movement on the ATP-side. Of course, and significanceforthecloseapproachofthesidechainsofthees- H+ translocation can generate a field in the liposome system of Fig. 1, sential Arg-210 and His-245 a-subunit residues located on the long but without destroying the field it is not possible to synthesize ATP. horizontal helix hairpin with the Glu/Asp-61 protonation site on Moreover, the same source (e.g. H+) moving unidirectionally (“in” to the c-subunit is provided by the torsional mechanism. “out” in Fig. 2a) cannot be the agent that creates as well as destroys the electrical field in a liposome system containing only the ATP syn- 6.5. Relationship to the Nernst equation thase (Fig. 1, Fig. 2a). The above analysis also implies that a stationary electrical field Δφ For an uncoupled process involving a single ion, assuming that the associated with a concentration gradient of ions cannot be converted ion species distributes at electrochemical equilibrium,theNernst into useful work and therefore direct field-driven chemistry equation can be applied and a value of the electrical potential can cannot take place as postulated in the theory of chemiosmosis. be calculated. However, the field stays as such and cannot be coupled This point was also emphasized by R. J. P. Williams [55] in his last to rotation and mechanical torque (Section 6.4). In a nonequilibrium paper on the subject [56]. In his words, “Mitchell's mechanism of ATP transport process involving coupling of proton and anion flows, S. Nath / Biophysical Chemistry 224 (2017) 49–58 57

where two ions are involved in energy-linked transport, it is invalid occluded by the β-catalytic site 3 in F1), the anion and proton access to assume that any one givenionicspeciesisinsole equilibrium channels in the FO portion of ATP synthase are closed, and the energy with the membrane potential, and thus there is no solid theoretical of the ion gradients is not wasted. The electrical imbalance created basis for using the Nernst equation in this case. It would therefore upon binding of MgADP− is the signal for the anion pathway to be quite impossible to calculate the presumed transmembrane elec- open and conduct an anion, and an interesting inter-access channel trical potential from the concentration gradients of a particular ion communication develops that leads to anion and proton cotransport, without knowing for certain that one ion or the other (or none!) is that continues upon subsequent phosphate binding, and only ceases in passive equilibrium with the electrical potential. In fact, as upon restoration of electrical neutrality after release of product 2− − explained in Sections 4.5 and 6.4,atransient field is created and MgATP from the F1 catalytic site along with OH . The self-regula- then destroyed by discrete ion translocation, and hence the electrical tory cycle repeats, with the transient departure from and restoration potential is not long-lasting, and therefore it is incorrect to interpret back to electrical neutrality acting as an ordering principle for the el- the calculated electrical potential obtained on application of the ementary steps of transport and chemical reaction. Nernst equation as a stationary electrical potential that remains as In order to prevent dissipation of ion gradient energies until the sub- such and continues to exist at all times. This important aspect strate(s) is/are bound in F1, Boyer's binding change mechanism has pro- requires a separate and more detailed analysis based on the princi- posed long-range interactions between β-catalytic sites and site-site ples of nonequilibrium thermodynamics and transport [53] that is cooperativity in F1 [48] based on oxygen exchange studies. However, beyond the scope of this work. However, it can be safely concluded owing to the complexity of the intermediate Pi-HOH and intermediate that failure to consider the second ion in an ion-coupled mechanism ATP-HOH exchange reactions and the difficulties in their interpretation of transport is a lacuna of all previous theories of biological energy [18], supramolecular structural effects and coupling between FO and F1 coupling. Therefore it is recommended that the torsional mechanism mediated by the energy-transducing membrane could provide an alter- be utilized for scientific and technological progress in the field as a native explanation. Such an explanation for a number of puzzling fea- matter of course, instead of considering only outmoded and tures of the mitochondrial intermediate exchange reactions [18] is misinformed theories that were postulated more than half-a-centu- more elegant and attractive, and one that is more likely to be correct ry ago and have now been surpassed by a more detailed and superior in the light of a dynamically electrogenic but overall electroneutral alternative. mode of ion transport. Hence the incorporation additionally of key reg- ulatory aspects within the self-consistent conceptual framework of en- 6.6. Transient departure from electrical neutrality as an ordering principle ergy coupling in the torsional mechanism can be considered to be for transport and reaction steps and the stringency of electrical neutrality another merit of the new biochemical theory.

The regulating principle that emerges at the molecular level (the 7. Conclusions level at which driving forces are defined in the torsional mechanism of ATP synthesis) is that charge imbalance can (and needs to) be created A systematic reappraisal of the modes of ion transport proposed by at conserved, restricted transport sites in the membrane for biological various theories of energy coupling in oxidative phosphorylation and energy transduction, but that the imbalance cannot be sustained for photosynthetic phosphorylation and molecular mechanisms of ATP long. Hence a discrete step-by-step mechanism of transport has evolved synthesis has been undertaken for the routinely-used reconstituted li- in which translocation of a counterion is favored over translocation of posome experimental system containing purified F1FO-ATP synthase. another co-ion, which implies that the requirement of electrical neutral- The chemiosmotic theory involving uncompensated proton transloca- ity is very stringent. In the overall sense, the entire transport process is tion has been shown to violate the principle of electrical neutrality of initiated by the concentration gradients of the species to which the bulk aqueous phases, and on these grounds it has been concluded to membrane is permeable. These concentration differences are therefore be indefensible, and therefore unsound as a theory of biological energy of fundamental importance and constitute the main factors that differ- coupling. Experiments suggest the movement of succinate/malate entiate the internal and external compartments of the cell or organelle, monoanion in OX PHOS or PHOTO PHOS (or countercations Na+ or and thus all voltages and electrical potentials result from these concen- K+ in other bioenergetic systems) in addition to protons. However, at- tration differences. tempts to save the chemiosmotic theory by postulating movement of The above ideas are also supported in a self-consistent way struc- these anions/countercations through the barrier of the membrane turally by the fact [59] that there exists only one ion-binding site per lipid bilayer have been shown to be plagued by a host of insurmount- a- and c-subunit in an F1FO molecule. We can apply the dynamically able problems, and such models are therefore inadequate, as are other electrogenic but overall electroneutral ion transport to the complete purely electroneutral mechanisms. The dynamically electrogenic but process of oxidative phosphorylation in mitochondria. ATP synthesis overall electroneutral mode of ion transport postulated by Nath's tor- is regulated by its demand in various cellular processes; when ATP4− sional mechanism of energy transduction and ATP synthesis has been is required, it is transported out from the intracristal space to the cy- quantified and shown to be consistent both with the experimental evi- toplasm along its concentration gradient.Thelocal electrical potential dence and the principle of electrical neutrality. This mode of ion trans- thus created drives ADP3− along its concentration gradient to the port involves an ordered and sequential translocation of dicarboxylic mitochondrialintracristalspaceinexchangeforATP4− by the ade- acid monoanion (or countercation) through the aqueous access half- nine nucleotide transporter. The resulting unbalanced local electrical channels known to be located at the a-c stator-rotor interface of the 2− – potential causes HPO4 to move into the crista, and the OH pro- membrane-bound FO portion of ATP synthase and the creation tran- duced during ATP synthesis in the F1 portion of ATP synthase is driv- siently of a local electrical potential, Δψ in these access channels that 2− 2− en out of the mitochondria in exchange for the HPO4 by the HPO4 / can be transduced into mechanical torque and rotation by the FO OH– antiporter, and overall electrical neutrality is restored. The same motor. The energies of the ΔpH and Δψ contributed by the two ions exchanges occur between the intracristal space and the matrix space are jointly utilized for rotation and converted finally into a store of tor- − of the mitochondrion, and thus the substrates MgADP and Pi are sional energy in the γ-subunit of F1 through a mechanoelectrochemical made available to the ATP synthase F1 portion that protrudes into process by ion-protein interactions. It is concluded that the ATP syn- the matrix space. The binding of MgADP− to the β-catalytic site 3 thase functions as a proton-dicarboxylic acid anion cotransporter in in the F1 portion is the rate limiting step of the process according to the processes of OX PHOS or PHOTO PHOS. Several novel mechanistic the torsional mechanism of ATP synthesis, and until MgADP− bind- and thermodynamic insights into biological energy transduction by − ing occurs in F1 (and the single negative charge of MgADP is the ATP synthase in the FO and F1 portions have been offered based on 58 S. Nath / Biophysical Chemistry 224 (2017) 49–58 the new vistas of energy coupling within the conceptual framework of [25] L. Xu, F. Liu, The chemo-mechanical coupled model for F1F0-motor, Prog. Biophys. Mol. Biol. 108 (2012) 139–148. Nath's torsional mechanism of ATP synthesis. These considerations [26] B. Agarwal, Revisiting the ‘chemiosmotic theory’: coupled transport of anion and make the new theory of energy coupling more complete, and take our proton for ATP synthesis, Bioenergetics 2 (2013), e116. . fundamental understanding of these vital life processes a step deeper. [27] N.M. Rao, Medical Biochemistry, second ed. New Delhi, New Age International, 2014 280–281. [28] V. Wray, Commentary on Nath and Villadsen review entitled “oxidative phosphor- Acknowledgements ylation revisited” biotechnol. bioeng. 112 (2015) 429-437, Biotechnol. Bioeng. (112) (2015) 1984–1985. – The funding and constant support of the Department of Science and [29] E.A. Guggenheim, Thermodynamics, 5th rev. ed., 1967 298 299 North-Holland: Amsterdam. Technology, India to the author's research program on the mechanism [30] A. Kotyk, K. Janacek, J. Koryta, Biophysical Chemistry of Membrane Functions, Wiley, and thermodynamics of biological molecular machines by various New York, 1988. grants for the past 25 years (1992–2016) is gratefully acknowledged. [31] C.R. Cantor, P.R. Schimmel, Biophysical Chemistry, Part 3, Freeman, Oxford, 1980. [32] J. Wyman, Linked functions and reciprocal effects in hemoglobin: a second look, The author thanks all three reviewers for their detailed comments and Adv. Protein Chem. 19 (1964) 223–286. constructive suggestions that have greatly helped to clarify and elabo- [33] M.T. Record, C.F. Anderson, Interpretation of preferential interaction coefficients of rate several physicochemical aspects in this work. nonelectrolytes and of electrolyte ions in terms of a two-domain model, Biophys. J. 68 (1995) 786–794. [34] J.A. Schellman, A simple model for solvation in mixed solvents. Application to the References stabilization and destabilization of macromolecular structures, Biophys. Chem. 37 (1990) 121–140. [1] E.C. Slater, Mechanism of phosphorylation in the respiratory chain, Nature 172 [35] E. Di Cera, Preferential interactions: It's as simple as 1, 2, 3, Biophys. J. 68 (1995) (1953) 975–978. 727–728. [2] P. Mitchell, Coupling of phosphorylation to electron and hydrogen transfer by a [36] E. Racker, Reconstitution, mechanism of action and control of ion pumps, Biochem. chemi-osmotic type of mechanism, Nature 191 (1961) 144–148. Soc. Trans. 3 (1975) 785–802. [3] P. Mitchell, Chemiosmotic coupling in oxidative and photosynthetic phosphoryla- [37] B.A. Feniouk, et al., Chromatophore vesicles of Rhodobacter capsulatus contain on av-

tion, Biol. Rev. 41 (1966) 445–502. erage one FOF1-ATP synthase each, Biophys. J. 82 (2002) 1115–1122. [4] P. Mitchell, Bioenergetic aspects of unity in biochemistry: evolution of the concept [38] F.E. Possmayer, P. Gräber, The pHin and pHout dependence of the rate of ATP synthe- + of ligand conduction in chemical, osmotic and chemiosmotic reaction mechanisms, sis catalyzed by the chloroplast H -ATPase, CFoF1, in proteoliposomes, J. Biol. Chem. in: G. Semenza (Ed.), Of Oxygen, Fuels and Living Matter, Part 1, John Wiley, New 269 (1994) 1896–1904.

York 1981, pp. 30–56. [39] G. Kaim, P. Dimroth, ATP synthesis by the F1Fo ATP synthase in Escherichia coli is [5] S. Nath, Beyond the chemiosmotic theory: analysis of key fundamental aspects of obligatorily dependent on the electric potential, FEBS Lett. 434 (1998) 57–60. energy coupling in oxidative phosphorylation in the light of a torsional mechanism [40] S. Fischer, P. Gräber, Comparison of ΔpH- and Δφ-driven ATP synthesis catalyzed by of energy transduction and ATP synthesis – invited review part 1, J. Bioenerg. the H+-ATPases from Escherichia coli or chloroplasts reconstituted into liposomes, Biomembr. 42 (2010) 293–300. FEBS Lett. 457 (1999) 327–332. [6] S. Nath, Beyond the chemiosmotic theory: analysis of key fundamental aspects of [41] G. Kaim, P. Dimroth, ATP synthesis by F-type ATP synthase is obligatorily dependent energy coupling in oxidative phosphorylation in the light of a torsional mechanism on the transmembrane voltage, EMBO J. 18 (1999) 4118–4127. of energy transduction and ATP synthesis – invited review part 2, J. Bioenerg. [42] A.J. Jagendorf, E. Uribe, ATP formation caused by acid-base transition of spinach Biomembr. 42 (2010) 301–309. chloroplasts, Proc. Natl. Acad. Sci. U. S. A. 55 (1966) 170–177. [7] S. Nath, J. Villadsen, Oxidative phosphorylation revisited, Biotechnol. Bioeng. 112 [43] J.T. Tupper, H. Tedeschi, Mitochondrial membrane potentials measured by micro- (2015) 429–437. electrodes: probable ionic basis, Science 166 (1969) 1539–1540. [8] E.C. Slater, The mechanism of the conservation of energy of biological oxidations, [44] H. Tedeschi, Absence of a metabolically induced electrical potential across the mito- Eur. J. Biochem. 166 (1987) 489–504. chondrial semipermeable membrane, FEBS Lett. 59 (1975) 1–2. [9] R.J.P. Williams, Some unrealistic assumptions in the theory of chemi-osmosis and [45] H. Tedeschi, Old and new data, new issues: the mitochondrial ΔΨ, Biochim. Biophys. their consequences, FEBS Lett. 102 (1979) 126–132. Acta 1709 (2005) 195–202. [10] B. Reynafarje, A. Alexandre, P. Davies, A.L. Lehninger, Proton translocation stoichi- [46] K. Schlegel, et al., Promiscuous archaeal ATP synthase concurrently coupled to Na+ ometry of cytochrome oxidase: Use of a fast-responding oxygen electrode, Proc. and H+ translocation, Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 947–952. Natl. Acad. Sci. U. S. A. 79 (1982) 7218–7222. [47] A.P. Batista, et al., Energy conservation by Rhodothermus marinus respiratory com- [11] G. Weber, Energetics of ligand binding to proteins, Adv. Protein Chem. 29 (1975) plex I, Biochim. Biophys. Acta 1797 (2010) 509–515. 1–83. [48] P.D. Boyer, The binding change mechanism for ATP synthase – some probabilities [12] D.E. Green, A critique of the chemosmotic model of energy coupling, Proc. Natl. and possibilities, Biochim. Biophys. Acta 1140 (1993) 215–250. Acad.Sci.U.S.A.78(1981)2240–2243. [49] J.J. Christensen, R.M. Izatt, L.D. Hansen, Thermodynamics of proton ionization in di- [13] J. Prebble, Peter Mitchell and the ox phos wars, Trends Biochem. Sci. 27 (2002) lute solution. VII. ΔH° and ΔS° values for proton ionization from carboxylic acids at 209–212. 25 °C, J. Am. Chem. Soc. 89 (1967) 213–222. [14] S. Nath, A thermodynamic principle for the coupled bioenergetic processes of ATP [50] J.J. Christensen, L.D. Hansen, R.M. Izatt, Handbook of Proton Ionization Heats and Re- synthesis, Pure Appl. Chem. 70 (1998) 639–644. lated Thermodynamic Quantities, Wiley, New York, 1976. [15] H. Rohatgi, A. Saha, S. Nath, Mechanism of ATP synthesis by protonmotive force. [51] P. Mitchell, Proton-translocation phosphorylation in mitochondria, chloroplasts and Curr. Sci. 1998, 75, 716-718. Erratum, Curr. Sci. 78 (2000) 201. bacteria: natural fuel cells and solar cells, Fed. Proc. 26 (1967) 1370–1379. [16] S. Nath, H. Rohatgi, A. Saha, The torsional mechanism of energy transfer in ATP syn- [52] D. Jou, F. Ferrer, A simple nonequilibrium thermodynamic description of some in- thase, Curr. Sci. 77 (1999) 167–169. hibitors of oxidative phosphorylation, J. Theor. Biol. 117 (1985) 471–488.

[17] S. Nath, The molecular mechanism of ATP synthesis by F1F0-ATP synthase: a scrutiny [53] G. Lebon, D. Jou, J. Casas-Vázquez, Understanding Non-equilibrium Thermodynam- of the major possibilities, Adv. Biochem. Eng. Biotechnol. 74 (2002) 65–98. ics: Foundations, Applications, Frontiers, Springer, Berlin, 2008. [18] S. Nath, The torsional mechanism of energy transduction and ATP synthesis as a [54] R.W. Estabrook, A. Holowinsky, Studies on the content and organization of the respi- breakthrough in our understanding of the mechanistic, kinetic and thermodynamic ratory enzymes of mitochondria, J. Biophys. Biochem. Cytol. 9 (1961) 19–28. details, Thermochim. Acta 422 (2004) 5–17. [55] R.J.P. Williams, Chemical advances in evolution by and changes in use of space dur- [19] S. Nath, A novel systems biology/engineering approach solves fundamental molecu- ing time, J. Theor. Biol. 268 (2015) 146–159. lar mechanistic problems in bioenergetics and motility, Process Biochem. 41 (2006) [56] S. Mann, A,J. Thomson, Robert J.P. Williams: founding father of bioinorganic chemis- 2218–2235. try, Angew. Chem. Int. Ed. 54 (2015) 7746. [20] S. Nath, The new uniﬁed theory of ATP synthesis/hydrolysis and muscle contraction, [57] W.R. Harvey, Voltage coupling of primary H+ V-ATPases to secondary Na+-orK+- its manifold fundamental consequences and mechanistic implications and its appli- dependent transporters, J. Exp. Biol. 212 (2009) 1620–1629. cations in health and disease, Int. J. Mol. Sci. 9 (2008) 1784–1840. [58] A.D.N.J. de Grey, Incorporation of transmembrane hydroxide transport into the [21] S. Nath, The thermodynamic efﬁciency of ATP synthesis in oxidative phosphoryla- chemiosmotic theory, Bioelectrochem. Bioenerg. 49 (1999) 43–50. tion, Biophys. Chem. 219 (2016) 69–74. [59] M. Allegretti, N. Klusch, D.J. Mills, J. Vonck, W. Kühlbrandt, K.M. Davies, Horizontal [22] S. Jain, R. Murugavel, L.D. Hansen, ATP synthase and the torsional mechanism: re- membrane-intrinsic α-helices in the stator a-subunit of an F-type ATP synthase, Na- solving a 50-year-old mystery, Curr. Sci. 87 (2004) 16–19. ture 521 (2015) 237–240. [23] J. Villadsen, J. Nielsen, G. Lidén, Bioreaction Engineering Principles, third ed. New York, Springer, 2011 136–145. [24] C. Channakeshava, New paradigm for ATP synthesis and consumption, J. Biosci. 36 (2011) 3–4.