ATP synthesis during exogenous NADH oxidation. A reappraisal.

This paper reports a reinvestigation on the pathway for mitochondrial oxidation of exogenous NADH and on the related ATP synthesis, first reported 30 years ago (Lehninger, A.L. (1951) J. Biol. Chem. 190, 345-359). NADH oxidation, both in intact and in water-treated mitochondria, is 90% inhibited by mersalyl, an inhibitor of the outer membrane NADH-cytochrome b5 reductase, and 10% inhibited by rotenone. The mersalyl-sensitive, but not the rotenone-sensitive, portion of NADH oxidation is stimulated by exogenous cytochrome c. Part of ATP synthesis is independent of exogenous NADH and cytochrome c, and is inhibited by rotenone and antimycin A, and is therefore due to oxidation of endogenous substrates. Another part of ATP synthesis is dependent on exogenous NADH and cytochrome c, is insensitive to rotenone and antimycin A, and is due to operation of cytochrome oxidase. It is concluded that (i) oxidation of exogenous NADH in the presence of cytochrome c proceeds mostly through NADH-cytochrome b5 reductase and cytochrome b5 on the outer membrane and then through cytochrome oxidase via the cytochrome c shuttle, and (ii) ATP synthesis during oxidation of exogenous NADH is partly due to oxidation of endogenous substrates and partly to operation of cytochrome oxidase receiving electrons from the outer membrane via cytochrome c.

Electroneutral H+-K+ exchange in liver mitochondria. Regulation by membrane potential.

The paper analyzes the factors affecting the H+-K+ exchange catalyzed by rat liver mitochondria depleted of endogenous Mg2+ by treatment with the ionophore A23187. The exchange has been monitored as the rate of K+ efflux following addition of A23187 in low-K+ media. (1) The H+-K+ exchange is abolished by uncouplers and respiratory inhibitors. The inhibition is not related to the depression of delta pH, whereas a dependence is found on the magnitude of the transmembrane electrical potential, delta psi. Maximal rate of K+ efflux is observed at 180-190 mV, whereas K+ efflux is inhibited below 140-150 mV. (2) Activation of H+-K+ exchange leads to depression of delta pH but not of delta psi. Respiration is only slightly stimulated by the onset of H+-K+ exchange in the absence of valinomycin. These findings indicate that the exchange is electroneutral, and that the delta psi control presumably involves conformational changes of the carrier. (3) Incubation in hypotonic media at pH 7.4 or in isotonic media at alkaline pH results in a marked activation of the rate of H+-K+ exchange, while leaving unaffected the level of Mg2+ depletion. This type of activation results in partial 'uncoupling' from the delta psi control, suggesting that membrane stretching and alkaline pH induce conformational changes on the exchange carrier equivalent to those induced by high delta psi. (4) The available evidence suggests that the activity of the H+-K+ exchanger is modulated by the electrical field across the inner mitochondrial membrane.

Permeability of inner mitochondrial membrane and oxidative stress.

The mechanism of increase in the inner membrane permeability induced by Ca2+ plus Pi, diamide and hydroperoxides has been analyzed. (1) The permeability increase is antagonized by oligomycin and favoured by atractyloside. The promoting effect of atractyloside is strongly reduced if the mitochondria are simultaneously treated with oligomycin. (2) Addition of the free-radical scavenger, butylhydroxytoluene, results in a complete protection of the membrane with respect to the permeability increase. (3) Although membrane damage and depression of the GSH concentration are often associated, there is no direct correlation between extent of membrane damage and concentration of reduced glutathione. Abolition of the permeability increase by butylhydroxytoluene or by oligomycin is not accompanied by maintenance of a high GSH concentration in the presence of diamide or hydroperoxides. The membrane damage induced by Ca2+ plus Pi is not accompanied by a depression of the GSH concentration. (4) It is proposed that a variety of processes causing an increased permeability of the inner mitochondrial membrane merge into some ultimate common steps involving the action of oxygen radicals.

Proton electrochemical gradient and phosphate potential in mitochondria.

The paper reports an analysis of the relationship between deltamuH the proton electrochemical potential difference, and deltaGp, the phosphate potential. Depression of deltamuH and deltaGp has been obtained by titration with: (a) carbonylcyanide trifluoromethoxyphenylhydrazone; (b) nigericin (+ valinomycin); (c) KCl (+ valinomycin); and (d) rotenone. The uncoupler depresses deltamuH more than nigericin (+ valinomycin), KCl (+ valinomycin) and rotenone at equivalent deltaGp. The deltaGp/deltamuH ratio is about 3 at high values of deltamuH. When deltaGp and deltamuH are depressed by nigericin (4 valinomycin) the deltaGp/deltamuH ratio remains constant. When deltaGp and deltamuH are depressed by uncouplers, the deltaGp/deltamuH ratio increases hyperbolically tending to infinity while deltamuH tends to zero. The absence of constant proportionality between deltaGp and deltamuH indicates that the proton gradients driving ATP synthesis presumably operate within microscopic environments.

Mechanism of active shrinkage in mitochondria. II. Coupling between strong electrolyte fluxes.

1. Addition of succinate to valinomycin-treated mitochondria incubated in KCL causes a large electrolyte penetration. The process depends on a steady supply of energy and involves a continuous net extrusion of protons. Rates of respiration and of electrolyte penetration proceed in a parallel manner. 2. A passive penetration of K+ salt of permeant anions occurs in respiratory-inhibited mitochondria after addition of valinomycin. Addition of succinate at the end of the passive swelling starts an active extrusion of anions and cations with restoration of the initial volume. The shrinkage is accompanied by a slow reuptake of protons. The initiation of the active shrinkage correlates with the degree of stretching of the inner membrane. The extrusion of electrolytes is inhibited by nigericin, while it is only slightly sensitive to variations of the valinomycin concentration larger than two orders of magnitude. 3. Passive swelling and active shrinkage occurs also when K+ is replaced by a large variety of organic cations. The rate of organic cation penetration is enhanced by tetraphenylboron, while the rate of electrolyte extrusion is insensitive to variation of the tetraphenylboron concentration. 4. Active shrinkage, either with K+ or organic cation salts, is inhibited by weak acids. The phosphate inhibition is removed by SH inhibitors. The active shrinkage is also inhibited by mersalyl to an extent of about 60%. 5. Three models of active shrinkage are discussed: (a) mechanoprotein, (b) electrogenic proton pump, and (c) proton-driven cation anion pump.

ATP synthase-mediated proton fluxes and phosphorylation in rat liver mitochondria: dependence on delta mu H.

The dependence of the proton flux through the ATP synthases of rat liver mitochondria on a driving force composed mainly of a potassium diffusion potential was determined and compared with the relationship between rate of phosphorylation and delta mu H given by titrations with the respiratory inhibitor malonate. The two functions are in good agreement in the lower part of the delta mu H range covered. However, the maximal proton fluxes through the ATP synthases are much lower than needed to account for the rate of State 3 phosphorylation sustained by the same mitochondria oxidizing succinate. Possible reasons for this behavior are discussed.

Studies on the relationship between ATP synthesis and transport and the proton electrochemical gradient in rat liver mitochondria.

The effect of ATP synthesis on delta mu H in rat liver mitochondria has been analyzed by separating the steps of adenine nucleotide translocation and ATP synthesis in the matrix. Either exchange of ATP, synthesized by substrate level phosphorylation in the matrix of oligomycin-treated mitochondria, for external ADP, or activity of the membrane-bound ATP synthase complex results in delta mu H depression with respect to resting state levels. This depression appears to be more pronounced, under strictly comparable conditions, when arsenate is used to stimulate ATP synthase activity than when the ornithine-citrulline conversion reaction is used for the same purpose.

Molecular slipping in redox and ATPase H+ pumps.

The titration of the mitochondrial ATPase H+ pump with oligomycin has been compared with the titration of the redox H+ pump with antimycin. In both cases there is extensive inhibition of the pumps without significant depression of delta muH. The two pumps exhibit 'nonohmic' behavior in different ranges of delta muH. This discrepancy favors the hypothesis of nontightly coupled or 'slipping' H+ pumps with respect to that of a steep dependence of the membrane 'leak' conductance for H+ on delta muH.

Mechanism of active shrinkage in mitochondria. I. Coupling between weak electrolyte fluxes.

1. A passive penetration of (NH4)2 HPO4 or of K2HPO4+nigericin occurs in respiratory-inhibited liver mitochondria. Addition of succinate at the end of the passive swelling initiates a shrinkage phase which leads to restoration of the initial mitochondrial volume. The rate and time of onset of the active shrinkage depend on the degree of stretching of the mitochondrial membrane. The rate of active shrinkage increases proportionally to the concentration of nigericin while it is strongly inhibited by valinomycin. 2. A number of SH inhibitors such as N-ethylmaleimide, p-chloromercuribenzoate, p-chloromercuriphenylsulphonate, dithiobisnitrobenzoate, exert a marked enhancing effect on the rate of shrinkage. The enhancing effect parallels titration of the phosphate carrier and inhibition of the passive phosphate efflux. In contrast, mersalyl is a powerful inhibitor of the rate of active shrinkage. The inhibition parallels that on phosphate passive efflux and requires higher mersalyl concentrations in respect to inhibition of phosphate influx. 3. The active shrinkage is discussed in terms of (a) a mechanoenzyme, (b) an electrogenic proton pump and (c) a proton-driven Pi pump.

Analysis of mechanisms of free-energy coupling and uncoupling by inhibitor titrations: theory, computer modeling and experiments.

The rates of ATP synthesis and of ATP-driven NAD reduction have been measured in bovine heart submitochondrial particles as a function of the fraction of inhibited redox pumps (in titrations with either antimycin or rotenone) and of the fraction of inhibited ATPases (in titrations with DCCD). The flux control coefficients of the redox and ATPase proton pumps on the rates of ATP synthesis and of ATP-driven NAD reduction have been derived and found to be equal to 1 for both pumps; i.e., both pumps appear to be 'completely rate limiting'. A theoretical analysis of the inhibitor titration approach based on kinetic models of chemiosmotic coupling and on the theory of metabolic control is presented. This analysis (i) shows that the results of the single inhibitor titrations are incompatible with a delocalized chemiosmotic mechanism of energy coupling if the proton conductance of the membrane is sufficiently low with respect to the conductances of the pumps; and (ii) suggests an experimental approach based on the determination of the P/O and the respiratory control ratios at different degrees of inhibition of the proton pumps to establish the origin of the 'loose coupling' of submitochondrial particle preparations. Three independent types of observation show that the 'loose coupling' of the particle preparation is not mainly due to an increased membrane proton conductance. The same and other independent observations are consistent with the view that the loose coupling of submitochondrial particle preparation is due mainly to inhomogeneity, i.e. to the presence of a subpopulation of highly leaky non-phosphorylating vesicles respiring at maximal rate. The results as a whole together with the simulations and analysis presented lead to the conclusion that the mechanism of free-energy coupling in submitochondrial particles is not completely delocalized.

Mechanism of nitrofurantoin toxicity and oxidative stress in mitochondria.

5-Nitrofuran derivatives change the inner mitochondrial membrane permeability as indicated by the transmembrane potential, the rate of spontaneous K+ efflux and the basal respiratory rate: (a) at low concentrations nitrofurantoin prevents the increase of inner membrane permeability due to hydroperoxides or to diamide; (b) at higher concentrations or after longer times of incubation, nitrofurantoin enhances the membrane damage due to hydroperoxides or to diamide; the damage due Ca2+ plus Pi is enhanced by nitrofurantoin at all concentrations; (c) higher nitrofurantoin concentrations cause membrane damage independently of the presence of hydroperoxides or of diamide. The effect of nitrofurantoin is cancelled by the addition of free-radical scavengers. The above effects of nitrofurantoin are compatible with the observations of Mason and colleagues that nitrofurantoin is reduced by a NADPH nitroreductase to a nitro anion radical which can then undergo subsequent reactions, among which are (a) initiation of a free-radical reaction chain and (b) reduction of hydroperoxides and diamide.

Flow-force relationships during energy transfer between mitochondrial proton pumps.

The effect of inhibitors of proton pumps, of uncouplers and of permeant ions on the relationship between input force, delta mu H+, and output flows of the ATPase, redox and transhydrogenase H(+)-pumps in submitochondrial particles was investigated. It is concluded that: (1) The decrease of output flow of the transhydrogenase proton pump, defined as the rate of reduction of NADP+ by NADH, is linearily correlated with the decrease of input force, delta mu H+, in an extended range of delta mu H+, independently of whether the H(+)-generating pump is the ATPase or a redox pump, or whether delta mu H+ is depressed by inhibitors of the H(+)-generating pump such as oligomycin or malonate, or by uncouplers. (2) The output flows of the ATPase and of the site I redox H(+)-pumps exhibit a steep dependence on delta mu H+. The flow-force relationships differ depending on whether the depression of delta mu H+ is induced by inhibitors of the H(+)-generating pump, by uncouplers or by lipophilic anions. (3) With the ATPase as H(+)-consuming pump, at equivalent delta mu H+ values, the output flow is more markedly inhibited by malonate than by uncouplers; the latter, however, are more inhibitory than lipophilic anions such as ClO4-. With redox site I as proton-consuming pump, at equivalent delta mu H+ values, the output flow is more markedly inhibited by oligomycin than by uncouplers; again, uncouplers are more inhibitory than ClO4-. (4) The results provide further support for a delocalized interaction of transhydrogenase with other H(+)-pumps.

Proton electrochemical gradient and rate of controlled respiration in mitochondria.

The correlation between deltamuH, the proton electrochemical potential difference, and the rate of controlled respiration is analyzed. deltamuH (the proton concentration gradient) is measured on the distribution of [3H]acetate, and deltapsi (the membrane potential) on the distribution of 86Rb+, 45Ca2+ and [3H]triphenylmethylphosphonium used either alone or simultaneously. The effects of the addition of ADP + hexokinase (state-3 ADP) and of carbonylcyanide trifluoromethoxyphenylhydrazone (state-3 uncoupler) on respiration and deltamuH are not equivalent: the uncoupler depresses deltamuH more than ADP at equivalent respiratory rates. The effects of the additions of nigericin-valinomycin and of ionophore A23187 (state-3 cation transport) and of carbonylcyanide trifluoromethoxy-phenylhydrazone (state 3-uncoupler) on respiration and deltamuH are also not equivalent: the uncoupler depresses deltamuH more than A23187 and nigericin + valinomycin at equivalent respiratory rate. A23187 is very efficient in stimulating respiration with negligible deltamuH changes.

The effect of respiration on the permeability of the mitochondrial membrane to ions.

1. The rates of cation uptake, for either organic cations such as tetrapropylammonium, TPA+, at variable tetraphenylboron concentrations, TPB-, or inorganic cations such as Mn2+, or K+ plus valinomycin, have been measured in mitochondria either respiring, under uncoupler titrations, or non-respiring, under variable K+ diffusion potentials. 2. The flow-force relationship for the respiration-coupled ion fluxes during titrations with uncouplers is almost identical to that obtained for the K(+)-diffusion driven fluxes. Similar results are obtained when TPA+ is replaced with inorganic cations, either monovalent such as K+ (+valinomycin), or divalent such as Mn2+. 3. By applying the Eyring analysis, as developed by Garlid et al. (Garlid, K.D., Beavis, A.D. and Ratkje, S.K. (1989) Biochim. Biophys. Acta 976, 109-121), from the flux-voltage relationships the values for the permeability coefficients and for the energy barriers have been obtained for the transport of the ion pair TPA(+)-TPB-, of Mn2+ and of K+ plus valinomycin, in non-respiring and in respiring, coupled and uncoupled, mitochondria. 4. The findings that the rates of respiration-coupled ion fluxes, at all values of membrane potential, are similar to the rates of the K+ diffusion potential-coupled ion fluxes and the similar pattern of the flux-voltage relationships during the titrations with uncouplers and artificial gradients indicate that the membrane permeability for ions is not modified by respiration.

The effect of the protonmotive force on the redox state of mitochondrial cytochromes.

In the absence of kinetic limitations, as determined either by high substrate concentrations or by absence of respiratory chain inhibitors, we have observed that: (a) the relationship between the percentage reduction of the cytochromes and the protonmotive force is linear in the case of cytochrome c and biphasic in the case of cytochrome b, (b) the redox state of cytochrome c depends only on the membrane potential and not on the total proton motive force and (c) the alkalinization of the matrix enhances the extent of cytochrome c reduction because of the marked inhibitory effect on the cytochrome oxidase activity. Thus, although the redox states of the b, c and aa3 mitochondrial cytochromes depend on the protonmotive force, the quantitative correlation between the two parameters and the relative effects of the electrical and chemical components of the force differ among the various cytochromes.

The nature of uncoupling by n-hexane, 1-hexanethiol and 1-hexanol in rat liver mitochondria.

We have analyzed the effects of n-hexane, 1-hexanethiol, and 1-hexanol on the coupled respiration of rat liver mitochondria. Incubation of mitochondria with n-hexane, 1-hexanethiol and 1-hexanol resulted in a stimulation, at low concentrations, and an inhibition, at high concentrations, of the state 4 mitochondrial respiration. Three criteria, all based on the comparison with the effect of DNP, have been used to establish whether the stimulation of respiration, at low concentrations of n-hexane, 1-hexanethiol, and 1-hexanol, depends on protonophoric mechanisms. First, the quantitative relationship between the extents of respiratory stimulation and membrane potential depression: a strong decrease of membrane potential was induced by increasing concentrations of DNP and a negligible depression by increasing concentrations of n-hexane or 1-hexanethiol. Only a slight decrease was induced by 1-hexanol. Second, the quantitative relationship between the extents of respiratory stimulation and of proton conductance increase: at equivalent rates of respiration, the enhancement of the proton conductance induced by DNP was very marked, by n-hexane and 1-hexanethiol practically negligible, and by 1-hexanol much smaller than that induced by DNP. Third, in titrations with redox inhibitors of the proton pumps, the pattern of the relationship between proton pump conductance and membrane potential was markedly different from protonophoric and non-protonophoric uncouplers: almost linear in the case of DNP, highly non-linear in the case of n-hexane, 1-hexanethiol and 1-hexanol. These three criteria support the view that n-hexane, 1-hexanethiol, and partially 1-hexanol, uncouple mitochondrial respiration by a non-protonophoric mechanism.

Proton electrochemical gradient and phosphate potential in submitochondrial particles.

The aerobic uptake of inorganic ions, such as 86Rb+ or 125I-, by submitochondrial particles, is about one order of magnitude lower than the uptake of organic ions, such as acridines or 8-anilino-1-naphthalene sulphonate. The values of deltapH, the transmembrane pH differential, and deltapsi, the transmembrane membrane potential are between 60 and 100 mV when calculated on the inorganic ions and between 150 and 240 mV when calculated on the organic ions. The discrepancy between the deltapH and deltapsi values from organic and inorganic ions is large at high but not at low ion/protein ratios. 2. In the absence of weak bases and strong acids the values of deltamuH, the proton electrochemical potential difference, are close to 100 mV and the magnitude of deltapH and deltapsi are similar. Weak bases decrease deltapH and enhance deltapsi. Strong acids decrease deltapsi and enhance deltapH. Interchangeability of deltapH with deltapsi occurs at low concentrations of weak bases and strong acids. High concentrations of weak bases and strong acids cause depression of deltamuH. 3. Concentrations of weak bases capable of abolishing deltapH, do not affect ATP synthesis. Concentrations of strong acids capable of abolishing deltapsi affect only slightly ATP synthesis. Concentrations of weak bases and strong acids capable of causing a decline of deltapH + deltapsi inhibit ATP synthesis. 4. Depression of deltamuH is paralleled by inhibition of ATP synthesis and decline of deltaGp, the phosphate potential. Abolition of ATP synthesis occurs only when deltamuH is below 20 mV. The deltaGp/deltamuH ratio increases hyperbolically with the decrease of deltamuH.

Proton electrochemical potential in steady state rat liver mitochondria.

Delta approximately muH has been determined in steady state mitochondria by measuring the magnitude of delta pH on the distribution of acetate and of deltapsi on the distribution of K+, tetraphenylphosphonium, Ca2+, Sr2+ and Mn2+. (1) The matrix concentration of divalent cations has been calculated from the total cation uptake, from the increase of matrix volume and from the ESR sextet signal of Mn(H2O)L2+. The [cat2+]i based on osmotic data is about five times higher than that based on ESR measurements. The [cat2+]i based on total uptake is much higher than that based on osmotic data at low cation/protein ratios. (2) In the presence of 10 mM acetate the maximal deltapsi on Ca2+ is about 130 mV and on Sr2+ is 95 mV. Deltapsi on Mn2+ is 91 or 109 mV, according to whether [cat2+-a)i is calculated from ESR or osmotic data. Under the same conditions, deltapH is about 60 mV. Hence delta approximately muH on divalent cations is between 151 and 190 mV. (3) Deltapsi on K+, in valinomycin treated mitochondria with 10 mM acetate or 2 mM Pi, drops from 200 mV, at low [K+]0 to almost zero parallel to the increase of [K+]0. DeltapH is 30 mV at low [K+]0 and about 42 mV at 600 muM K+. Hence delta approximately muH drops from 22m mV lower values with the increase of [K+]0. (4) Maximal deltapsi on triphenylmethylphosphonium is 140 mV. (5) When delta approximately muH is measured simultaneously on divalent cations and on K+, the values on K+ tend to approach those on Ca2+ while those on Sr2+ are about 50 mV lower. (6) It is concluded that the steady state mitochondrial energy potential is equivalent to a delta approximately muH between 150 and approx. 190 mV.

Tracking of proton flow during transition from anaerobiosis to steady state in rat liver mitochondria.

(1) The hydrophobic pH indicator Bromthymol blue and the hydrophilic pH indicator Phenol red have been used to follow the redox-pump-linked proton flows during transition from anaerobiosis to static head. The domains monitored by the pH indicators, whether external or internal, and the localization of the dye, whether free or membrane bound, have been identified by recording the absorbance changes following addition of nigericin or valinomycin to anaerobic or aerobic mitochondria and the effects of permeant and impermeant buffers. (2) After addition of the H+/K+ exchanger, nigericin, to anaerobic mitochondria. Phenol red and Bromthymol blue record an alkalinization and an acidification, respectively, indicating that while the hydrophilic pH indicator faces an external domain, the hydrophobic pH indicator faces, at least partly, an internal domain. The latter effect is sensitive to phosphate and to phosphate carrier inhibitors. On the other hand, addition of nigericin to aerobic mitochondria leads to an increased Bromthymol blue absorbance, which reflects an alkalinization, indicating that the pH indicator faces an external domain. The reorientation of the dye from the internal to the external domain is a function of the uncoupler concentration and thus of the membrane potential (cf. Mitchell et al. (1968) Eur. J. Biochem. 4, 9-19). (3) The amount of oxygen required for the transition from anaerobiosis to static head has been determined by following in parallel the extent of oxidation of cytochrome aa3 and the rise of delta mu H+. With succinate as substrate, 50% levels of cytochrome oxidation are obtained at 0.125 ngatom oxygen/mg and 50% of Safranine response at about 0.2 ngatom oxygen/mg. These amounts of oxygen correspond to an H+ displacement of about 0.8-1.2 ngatom/mg on the basis of the H+/O stoichiometry. It is concluded that mitochondria are in presteady state below, and in static head above, displacement of 2-3 ngatom H+/mg. This figure is very close to the original calculation of Mitchell (Mitchell, P. (1966) Biol. Rev. 41, 445-502). (4) Transition, by oxygen pulses, of EGTA-supplemented mitochondria from anaerobiosis to either presteady state or static head state results in a response of the hydrophilic pH indicator, Phenol red, which is negligible in amount and/or kinetically unrelated to the delta mu H+ rise. The fact that H+ extrusion in the bulk aqueous phase is negligible also in presteady state excludes proton cycling as an explanation. Addition of oxygen pulses to Sr2(+)-supplemented anaerobic mitochondria results in an H+ extrusion whose amount and rate is proportional to the Sr2+ concentration.(ABSTRACT TRUNCATED AT 400 WORDS)

A minimal hypothesis for membrane-linked free-energy transduction. The role of independent, small coupling units.

Experimental data are reviewed that are not in keeping with the scheme of 'delocalized' protonic coupling in membrane-linked free-energy transduction. It turns out that there are three main types of anomalies: (i) rates of electron transfer and of ATP synthesis do not solely depend on their own driving force and on delta mu H, (ii) the ('static head') ratio of delta Gp to delta mu H varies with delta mu H and (iii) inhibition of either some of the electron-transfer chains or some of the H+-ATPases, does not cause an overcapacity in the other, non-inhibited proton pumps. None of the earlier free-energy coupling schemes, alternative to delocalized protonic coupling, can account for these three anomalies. We propose to add a fifth postulate, namely that of the coupling unit, to the four existing postulates of 'delocalized protonic coupling' and show that, with this postulate, protonic coupling can again account for most experimental observations. We also discuss: (i) how experimental data that might seem to be at odds with the 'coupling unit' hypothesis can be accounted for and (ii) the problem of the spatial arrangement of the electrical field in the different free-energy coupling schemes.