An update of the chemiosmotic theory as suggested by possible proton currents inside the coupling membrane

We may seek for the proton plausible pathways inside the respiring membrane, even if it is a plasma-membrane. As only five alpha-helices have been found in the complex I structure, in order to allow for proton translocation, it was necessary to postulate the existence of two hemi channels: one from the matrix (n) side to the centre of the respiratory complex, referred here to as the ‘proton entrance hemi channel’, and the other from the centre to the periplasmatic (p) side, here indicated as ‘proton exit hemi channel’. However, only the latter was well identifiable in complex I [4]. Instead, the entry pathway has not been identified with certainty, so we only can talk about putative pathways labelled with ‘?’ [4]. Furthermore, it clearly emerges from the X-ray studies that there is an obvious proton tunnelling at the centre of the complex I. Comprehensive studies on the protonic movement inside complex I have been carried out by the Helsinki Bioenergetic Group of M. Wikström, which highlighted uncertainty margins on the stoichiometry of protonic extrusion that appear closer to 3 H+/2e− [84] instead to the classic 4 H+/2e−. Also, in the review of Verkhovskaya & Bloch [85] (of Helsinki Bioenergetic Group), four mechanisms for proton translocation are proposed and the ‘proton entrance half channel’ is not identified with certainty, while the ‘proton exit half channel’ is clearly identifiable. Emblematically, on the website of the Helsinki Bioenergetic Group it is written (www.biocenter.helsinki.fi/bi/hbg/cl/complex_I_main.html) that ‘the mechanism of proton transfer in complex I remains completely enigmatic’.

As far as complex IV (cytochrome c oxidase) is concerned, the proton translocation of has been studied in depth [86]. In their recent review, M. Wikström & V. Sharma talk of an ‘anniversary’: ‘Proton pumping by cytochrome c oxidase—A 40 year anniversary’ [87]. Among the many works cited in this review the Chemical Review article of the M. Wikström group stands out [8]. It goes into the details of the possible molecular processes carried out by complex IV [8], thereby including proton translocation. The topic is complex, and more putative pathway of protons are well developed in the review, which we cannot here detail here. For the ‘proton entrance half channel’, for each molecule of oxygen reduced to water they need 4 H+, and it is hypothesized that another 4 H+ (for a total of 8) enter through this channel. It seems unlikely that there exists equivalence between protons that exist as particles and link to water molecules and protons that should move as a charge, according to the Grotthuss mechanism.

Taking into account what is reported above, it is possible to trace a plausible proton pathway within the respiring membrane. In 2006 a direct proton transfer was proposed to couple the respiratory pathway of complexes I, III, IV with ATP synthase [88]. However, in a recent review [21] the concept of transmembrane proton motor force to move the ATP synthase is reinforced, although it is noted that many aspects of coupling are not yet clarified. Clear-cut consideration was proposed some years ago (1991) by Akeson & Deamer [89] about speed of proton translocation through putative proton channel as a limiting step for ATP synthesis by ATP synthase. For the sake of simplicity, we only examine the coupling between the respiratory complex I and the ATP synthase. The existence of a proton pathway at the centre of complex I is quite clear, and we can hypothesize that the protons are sent to the well-identified ‘exit half channel’, as shown in figure 3. The proton donor at the centre of complex I has already been tentatively identified in the phospholipid cardiolipin (CL) [78], essential for OXPHOS. However, for example, phosphatidylethanolamine (PE) can effectively sustain the functioning of OXPHOS [90]. Data from C. von Ballmoos's group [91,92] unconventionally clarify such a role of phospholipids. In fact, pure phosphatidylcholine (PC) is excellent for the coupling, increasing when the membrane is formed by PC + PE. By contrast, coupling is dramatically inhibited if the membrane is formed by PC + CL, incredibly diverging from the traditional role assigned to CL. This research is also important because experiments are performed with a reconstructed system, more adherent to the ‘H+/ATP coupling’ or ‘second coupling’ (figure 1).

Here, the driving force that feeds ATP synthase was not the rapid transfer of K+ generated by valinomycin, a widely used method, but the bo3 oxidase of Escherichia coli, analogous to the respiratory complex I of vertebrates. It emerges that all membrane phospholipids can act as mobile proton transporters, hindering proton relocation in the near-aqueous medium, so that these are never free. The exergonic process of proton solvation would release an impressive amount of energy (262 400 Cal mol−1) [62], which, if not transferred on a generic acceptor, would generate heat devastating not only the membrane but also the cell integrity. Complex I would transfer protons to the p side of the membrane, but these would not be dispersed in the aqueous bulk. In fact it is well documented that there exists a barrier of water molecules attached to the membrane, determining the lateral displacement of protons on the phosphate heads of phospholipids [68,93,94] to meet a sink that is the ATP synthase subunit which, through a proton wire [95,96], would lead the proton to the rotor (figure 3), accommodating on the central glutamic or aspartic residue (depending on the species) of the c subunit, which in variable numbers from 8 to 15 make up the rotor Fo [97]. As Fo rotates it is conceivable that the proton returns to the centre of complex I, a highly hydrophobic environment, thus closing the protonic circuit. This last passage, from the rotor exit to the centre of the respiratory complex, appears plausible for the operativity of the Brownian motion (diffusion) of particles in a highly anisotropic environment that can occur efficiently [45,98] and for a theoretically possible direct passage of the proton exiting the subunit at the entrance to the complexes. Moreover the investigations of the C.

von Ballmoos group have produced exhaustive studies with reconstructed systems [91,92,99], in which a direct transfer of proton on the p-side of the membrane from complex I to ATP synthase is evident. Their recent paper also demonstrates that proximity to the membrane between the respiratory complex and ATP synthase is required for the ‘proton coupling’ [92]. However, the fact remains that the last part of the protonic circuitry (i.e. from Fo moiety of ATP synthase to respiratory complexes) is only hypothetical.

In this controversial scenario, a decisive contribution is undoubtedly the sophisticated bioengineering experiment that labelled both complex IV and ATP synthase with proteins of the GFP family, to experimentally observe a local ΔpH triggered by the respiratory substrate galactose [41]. Observations were conducted in cultured HeLa cells. The authors concluded [41, p. 1]: ‘the observed lateral variation in the proton-motive force necessitates a modification to Peter Mitchell's chemiosmotic proposal’. In other words, the experimentally proven lateral proton motive force is in line with the hypothesis of localized coupling.

The chemiosmotic theory [1] as it was formulated envisions a process that can only take place in organelles possessing double membrane systems forming closed compartments to entrap protons, such as mitochondrial cristae, bacteria and thylakoids (here, for the sake of simplicity, we have considered the mitochondrial inner membrane). However, in the last years, several authors have described a functional expression of OXPHOS machinery in cellular districts devoid of mitochondria [100,101], suggesting a possible build-up of a transversal proton gradient across the membrane, out of mitochondria. In particular, an extramitochondrial OXPHOS has been described in rod outer segment (OS) discs [102–105], myelin sheath [79,106–110], cell plasma membrane [111–119], platelets [120], and extracellular vesicles shedding from cells such as exosomes and microvesicles [121,122], which seem to carry an unsuspected metabolic signature [121,123]. Notably, since the plasma membrane potential is positive on the outside and negative on the inside, it would favour ATP hydrolysis rather than its synthesis. The extramitochondrial ATP synthesis is in line with the postulated independence of ATP synthase activity from the membrane potential, and therefore it is plausible that this synthesis of ATP depends on the proton intramembranous coupling.

The impressive amount of experimental data cited appears globally in contrast with the above-cited three assumptions underlying the chemiosmotic theory. First of all, it excludes that the protons can accumulate on the coupling membrane surface, whose high permeability would dissipate, since it would correspond to an extreme acidity incompatible with any vital process. Second, since proton diffusion does occur, it is clear that the membrane potential is irrelevant to proton translocation. Third, it can be excluded that the respiratory complexes operate as transmembrane proton transfer from the aqueous bulk, where the proton would exist as hydroxonium ion. This appears to rule out the actual possibility that ATP synthase can overcome the energy barrier, higher than 500 meV [61], required to extract the proton from water, leaving space for the localized coupling hypothesis [63,64], as a dehydration–hydration reaction from hydroxonium ion to free protons would require 262 400 Cal mol−1 [62]. Which are the actual processes acting on the ATP synthase? There is no certainty, as highlighted by the emblematic title of an article by J. Walker: ‘The ATP synthase: the understood, the uncertain and the unknown’ [124].

Indeed, today a constellation of clues leads us to hypothesize the existence of protonic currents internal to the membrane, with the formation of possible circuits travelled by the positive elementary charge, thus realizing a localized coupling that excludes an osmotic nature of the process. To trace this circuit, at least for the possible coupling between respiratory complex I and ATP synthase, it appears realistic that complex I may transfer protons from its central part to the periplasmatic side, allowing them to travel on the membrane surface thanks to the heads of phospholipids [93,94] finally tunnelling inside the a subunit of ATP synthase [95,125] (figure 3). Moreover, several studies show that the membrane is isolated from the aqueous bulk thanks to a layer of water molecules on both sides of the membrane, which consolidates the idea that the membrane is radically distinct and isolated from the liquid phase [59]. Considering the two isolated phases, the only way evolution could pursue to link proton movement to ATP synthesis was a nanomachine connecting the proton movement inside the membrane to the deformation mechanics of the F1 sphere immersed in the aqueous phase. We can reasonably assume that the proton movement inside the membranes occurs as a charge, according to the proton-hopping Grotthuss mechanism, with the establishment of ‘protonic currents’ inside the membrane [126]. In the non-biological field we find a remarkable adherence to this theory for the development of protonic devices (as proton diodes, transistors, memories and transducers) [77]. It is surprising that the biological and the physical–chemical areas have ignored each other.

Since in the history of biology the application of physical–chemical methodologies has led to dramatic advances in biology (we may remind the reader that the resolution of DNA structure [127] was obtained with the fundamental application of X-ray crystallography, developed in 1913 by Williams Henry Bragg and his son Lawrence for the study of inorganic crystals), it is desirable that this fusion of knowledge can be realized in the years to come.

Furthermore, the classical mechanical approach cannot be used to approach these currents. In fact, any time there is a movement of charge bound to a mass, the dualism that cannot be assessed by classical mechanics, instead quantum mechanics must be applied, and in this perspective lay the promising recent quantum-mechanical approaches by Riccardi et al. [76] and Ivontsin et al. [125]. There is a need to take into account the Heisenberg indeterminacy principle as Philip Hunter already highlighted in the title of a 2006 paper: ‘A quantum leap in biology: one inscrutable field helps another, as quantum physics unravels consciousness’ [128].

The authors are very grateful to Professor Giorgio Lenaz, Bologna University (Italy), for the critical reading of the manuscript and for the valuable comments and discussions, and Professor Matthew Stephen Hodgart, University of Surrey, for help in revising the manuscript language.