Interpretation of the mechanism of action of antituberculosis drug bedaquiline based on a novel two‐ion theory of energy coupling in ATP synthesis

Abstract Tuberculosis (TB) claims the lives of 1.3 million people each year, more than any other bacterial infection. Hence great interest was generated in health communities upon the recent introduction of the new diarylquinoline anti‐TB drug, bedaquiline. Bedaquiline acts by binding to the c‐subunit in the membrane‐bound FO portion of the F1FO‐adenosine triphosphate (ATP) synthase, the universal enzyme that produces the ATP needed by cells. However, the mechanism of killing by bedaquiline is not fully understood. Recent observations related to the bactericidal effects of bedaquiline, which show that it is a potent uncoupler of respiration‐driven ATP synthesis in Mycobacterium smegmatis are summarized. These observations are then interpreted from the standpoint of Nath's two‐ion theory of energy coupling in ATP synthesis (Nath, Biophys. Chem. 2017; 230:45–52). Especial importance is given to the interpretation of biochemical fluorescence quenching data, and the differences between the uncoupling induced by bedaquiline from that by the classical anionic uncouplers of oxidative phosphorylation are highlighted. Suggestions for new experiments that could lead to a better understanding of the uncoupling mechanism are made. A model of uncoupling action by the drug is presented, and the biochemical basis underlying uncoupling of ATP synthesis and lethality in mycobacteria is elucidated. The major biological implications arising from these novel insights are discussed. It is hoped that the analysis will lead to a more fundamental understanding of biological energy coupling, uncoupling and transduction, and to an integrated view for the design of novel antimicrobials by future research in the field.


| INTRODUCTION
Developing countries are faced with their own peculiar afflictions and diseases, such as human tuberculosis (TB), that also provide opportunities for new drug development. If these developments could be Mycobacterium tuberculosis urgent and important. [4][5][6] The last drug for treating TB, rifampin, was introduced in 1971. Hence, great interest was generated when controlled clinical phase 2 trials revealed a potent bactericidal effect of the novel anti-MDR TB drug, bedaquiline fumarate (generally called bedaquiline, and marketed as Sirturo) accompanied by a considerably shortened treatment time. 7 The drug was granted accelerated Food and Drug Administration Approval. 8 Bedaquiline is a member of the chemical class of diarylquinolines ( Figure 1) and has been shown to exert its effects by binding to the csubunit in the membrane-bound F O portion of the F 1 F O -ATP synthase. 9,10 The F 1 F O -ATP synthase is the universal enzyme in animal mitochondria, plant chloroplasts and bacteria that produces the ATP needed by the cell to fuel its catabolic and anabolic reactions [11][12][13] via the central process of oxidative phosphorylation (OXPHOS) or photophosphorylation. [14][15][16] The F 1 F O -ATP synthase has been shown to be essential for growth in Mycobacterium smegmatis. 17 Furthermore, the mycobacteria catalyze ATP synthesis solely and do not operate in the reverse ATP hydrolysis mode. The mycobacterial ATP synthase has unique properties and selectivity to bedaquiline compared to its eukaryotic homolog. 18 Detailed knowledge of the function of the F 1 F O -ATP synthase is therefore central to a complete understanding of how targeting of the ATP synthase due to bedaquiline challenge leads to lethality of mycobacteria and for determination of the mode of action of bedaquiline.

| OBSERVATIONS RELATED TO DELAYED EFFECTS OF BEDAQUILINE AND CELLULAR ATP DEPLETION
Bedaquiline shows in vitro activity and inhibits the growth of both drug-sensitive and drug-resistant M. tuberculosis strains and also in vivo activity against mycobacteria in TB patients. On treatment of growing or non-growing mycobacterial cells with bedaquiline, timedependent killing has been observed. 9 However, the mechanism of killing is unclear, but it does not involve collapse of the membrane potential, Δψ. 19 A dose-dependent decrease of intracellular ATP has been observed on treatment of M. tuberculosis cells with bedaquiline and this ATP depletion may be a determinant of the observed delayed onset of killing. 19 However, mycobacterial cells can be depleted of ATP via deenergization methods, yet their viability is not compromised 20 implying that cell death is not fully explained by ATP depletion. In fact, novel mechanisms alternative to ATP depletion are necessary to explain the potent bactericidal effects of bedaquiline.

| RECENT OBSERVATIONS RELATED TO THE POTENT BACTERICIDAL ACTION OF BEDAQUILINE
Recently, it was shown that bedaquiline is a potent uncoupler of redox-driven ATP synthesis in M. smegmatis. 21 Structural analysis revealed that bedaquiline does not function as a general conductor of protons from one bulk aqueous phase to another; rather, it exerts its effects by binding to the c-subunit in the F O portion of ATP synthase. 10 Classical uncouplers of OXPHOS dissipate both the ΔpH and the Δψ which are then recreated by respiratory activity leading to a futile cycle of ion transport uncoupled from ATP synthesis. 13,22 Uncoupling activity induced by the dinitrophenols has been quantitatively modeled using the principles of thermodynamics and transport phenomena by Jou and colleagues. 23,24 Such uncoupling is not lethal for Escherichia coli. However, for reasons that are not well understood, the uncoupling arising from bedaquiline challenge helps cause lethality in mycobacteria.

| NOVEL TWO-ION THEORY OF ENERGY COUPLING IN ATP SYNTHESIS
According to Mitchell's chemiosmotic theory, 25 a delocalized electrical potential across bulk aqueous phases is an essential intermediate that is communicated far away and its energy used by the F O -nanomotor to synthesize ATP by retro proton translocation. However, physiological ATP synthesis was demonstrated in the 1990s with single molecules of the ATP synthase purified and reconstituted into liposomes that contained no redox/photosystem complexes. Therefore, either no electrical potential is created, 26 or a local potential, Δψ is created at the a-c interface in the membrane-bound F O portion of ATP synthase. 12,13,27 A more detailed alternative Nath's torsional mechanism of ATP synthesis and two-ion theory of energy coupling went beyond Mitchell's chemiosmotic theory. 13 The theory was quantitative and contained mathematical equations that described the overall driving force of ATP synthesis arising from discrete proton and anion/countercation translocations in the membrane-bound F O portion of F 1 F O -ATP synthase. 28 A 10-year experimental search 22 established the identity of the elusive anion involved in energy transduction and ATP synthesis as a C4 dicarboxylic acid anion-succinate in animal mitochondria and malate in plant chloroplasts-although in certain bacteria a countertransport of Na + or K + was also proposed to be involved. 14,16,29-35 FIGURE 1 The chemical structure of the diarylquinoline drug, bedaquiline fumarate, generally known as bedaquiline, and marketed as SIRTURO. The first stereo-label refers to the carbon atom with the phenyl group, while the second stereo-label refers to the carbon atom harboring the hydroxyl group The recent important experimental observations 21 and the availability of a novel two-ion theory of energy coupling in ATP synthesis 28 provides further mechanistic insights into the bactericidal mode of action of bedaquiline, and helps in explanation and interpretation of the antibiotic's bactericidal effects. This ought to also lead to a more holistic view of biological energy coupling and transduction for the design of novel antimicrobials by future research in the field. subunit that these ion-protein interactions induce. 12 Hence the acceleration of cellular respiration that results in order to restore the ionic gradients by the redox side of OXPHOS, and how this phenotypic outcome arises originally from a drug-target interaction in F O is also clarified.
The explanation for the dose-dependent reversal of fluorescence quenching of acridine orange by bedaquiline is similar to that of CCCP uncoupler given above except that owing to its hydrophobic properties, bedaquiline enters and binds to its site on the c-subunit in the membrane without competing for entry and binding with physiological succinate monoanions (or K + ). This reversal of fluorescence quenching was completely abolished in IMVs of the D32V mutant upon bedaquiline treatment due to the inability of bedaquiline to bind to the mutant c-subunit, thus impeding its ability to capture protons. As CCCP entry and binding to a different site on the a-or c-subunit is not inhibited in the D32V mutant, 50 μM CCCP caused complete reversal of the fluorescence of acridine orange, exactly as in wild-type membranes. It should, however, be noted that formation of the neutral form of the CCCP uncoupler also dissipates the Δψ created by uncoupler anion translocation and binding unlike during the uncoupling action of bedaquiline, which is a key difference and distinguishing feature of bedaquiline as compared to the classical uncouplers of OXPHOS.
We now turn to the data on quenching of Oxonol VI fluorescence and its reversal. The explanation for the rapid quenching of dye fluorescence is similar to that for acridine dyes except that the absorption of protons by the dye is a measure of the creation of Δψ, given the "voltmeter-type" characteristics of the dye and its sensitivity to local fields. Treatment with bedaquiline and its binding to the c-subunit in wild-type membranes does not interfere with the con- Hence bedaquiline has no effect on the magnitude of the local Δψ, as observed in both IMVs and whole cells. As the a-c interface is disrupted in the D32V mutant, the lack of significant effect of valinomycin-K + in the reversal of Oxonol VI fluorescence quenching is also explained, given that K + needs to translocate through the same interface to exert its effects. However, the model for uncoupling given by Hards et al, 21  Regardless, the ability of bedaquiline to dissipate both the ΔpH and the ΔpK due to its activity as a genuine H + /K + antiporter, but not the Δψ arising from succinate translocation (Figure 2A

| MAJOR BIOLOGICAL IMPLICATIONS
The above has major biological implications. The key similarities and differences between the action of bedaquiline as uncoupler and the classical anionic uncouplers of OXPHOS, such as the dintrophenols can now be summarized. Like the classical uncouplers that dissipate both ΔpH and Δψ across the F 1 F O -ATP synthase, the ionophoric properties of bedaquiline that cause it to function as an H + /K + exchanger ( Figure 2) lead to an "uncoupler-like" mode of action upon bedaquiline challenge that dissipates the transmembrane H + and K + concentration gradients. However, the Δψ arising from succinate translocation (Figure 2A) is left virtually intact at its original value. In other words, bedaquiline treatment does not have the power to dissipate the Δψ in whole cells or IMVs. This key difference may lie at the heart of the bactericidal mode of action of bedaquiline. This implies that, upon bedaquiline treatment (unlike due to the action of the classical anionic uncouplers of OXPHOS), a local field will continue to exist in the aqueous a-c access channels of F O (negative polarity inside). This maintained field arises from the noncollapsible electrical potential due to succinate monoanion translocation from outside to inside, or what is electrically equivalent, from a translocation of K + from inside to outside. It should be noted that events within the energy-transducing membrane can be rapidly communicated to the bulk aqueous phases. In fact, since the local field discussed above is within the aqueous access channels normal to the plane of the membrane, a Δψ will be "measured" across the bulk aqueous phases by use of permeant ions or dyes/probes from the measured distribution ratio of the motive probe and calculation based on the Nernst equation.
What are the mechanistic consequences of the above features for lethality? It has generally been thought that dissipation of ΔpH and Δψ is lethal to mycobacteria. 39 However, based on the above interpretations, the origin of lethality in mycobacteria may be attributed to the continued presence of the Δψ. Continued redox-linked coupled proton and succinate translocation, will interfere with the homeostasis of succinate, causing its depletion inside and accumulation outside, ultimately leading to death of the mycobacteria.
A final note on the electrochemical measurement of the restoration of the ΔpH and Δψ/ΔpA by the redox enzymes during physiological functioning is in order. In response to electron transfer, a primary ion translocation (say of H + ) will create a Δψ that can be destroyed by a dye or permeant ion (e.g., K + in the presence of valinomycin) as probe, and one can claim that a Δψ exists. However, the "measured" Δψ would have anyhow been destroyed subsequently by the secondary translocation (say of succinate ions) across the membrane. In other words, the existence of Δψ was only transient and would have been replaced by a succinate concentration gradient, or ΔpA, and the Δψ would disappear. All current theories of energy coupling (except Nath's two-ion theory) postulate that only a "single ion" (e.g., H + ) is the generator of both ΔpH and Δψ. However, the new theory proposes that ΔpH and Δψ are created by two independent agents, that FIGURE 2 Model for the uncoupling action of bedaquiline and mycobacterial cell death resulting from subsequent metabolic consequences. The vast majority of ATP in a neutrophilic mycobacterial cell is produced by OXPHOS. Due to its high hydrophobicity, bedaquiline (shown here as U) binds from the outside bulk aqueous medium to its binding site on the c-subunit at the rotorstator a-c interface in the membrane-bound F O portion of the F 1 F O -ATP synthase. Bedaquiline (U) captures protons translocated from the periplasm along their concentration gradient from the vicinity of their Glu/Asp-61 (E. coli numbering) binding site on the c-subunit, chelates K + ions translocated from the cytoplasm along their concentration gradient, and mediates an electroneutral H + /K + exchange, depicted by the circle in the schematic. The activity of bedaquiline as an H + /K + ionophore releases H + ions into the inside bulk aqueous medium, and K + ions into the outside space, thereby dissipating the ΔpH and the ΔpK and creating a futile cycle of the ions that is uncoupled from ATP synthesis. As a result of the uncoupling of ATP synthesis caused by such ionophoric action of bedaquiline as an H + /K + antiporter, respiration will be accelerated compared to the physiological case in the absence of bedaquiline, when protons and physiological succinate counteranions A − /K + countercations translocate through F O , bind/ unbind to/from their respective binding sites, and induce conformational changes in the c-subunits of the c-ring by means of ion-protein interactions. 12 However, bedaquiline does not interfere with the Δψ created by succinate translocation, A − (A) or K + translocation in the opposite direction (B) to their respective binding sites on the a-subunit at the a-c interface in F O . Hence, the Δψ is not dissipated by bedaquiline, in contrast to the uncoupling induced by the classical anionic uncouplers of OXPHOS, such as the dinitrophenols, that dissipate both ΔpH and Δψ. In the model, the maintained Δψ in mycobacteria is postulated to be ultimately responsible for death of the organism by adversely impacting succinate homeostasis due to its continual translocation outwards coupled with protons by the redox reactions but blocked translocation inwards through the F O portion of ATP synthase, thereby causing a depletion of internal succinate required for the physiological functioning of the Krebs cycle and the redox-side of OXPHOS in cells. For further details, please refer to the text. The case of the combined operation of (a) and (B) has already been discussed and quantitatively analyzed during the conception and detailed formulation of Nath's torsional mechanism of energy transduction and ATP synthesis (e.g., in Reference 37, see especially table 2 of section 3.4 and eqs. 1-9) 37 is, H + and succinate/K + , respectively. Failure to consider the critical role of the "second ion" that is also involved in biological energy transduction along with protons leads to inconsistencies and incorrect interpretations and has been shown to be the fundamental reason underlying past and present difficulties. 28 To illustrate the above concept further for the redox reactions in animal mitochondria in the physiological mode of operation, let us call the cytochrome bc1 Complex III-produced driving force as

| CONCLUSIONS
Based on the interpretation of recent biophysical fluorescence quenching data and other biochemical experiments by a novel twoion theory of energy coupling in ATP synthesis, it has been concluded that the anti-TB drug bedaquiline fumarate exerts its bactericidal function due to metabolic effects that arise from and follow its uncoupling mode of action. However, on further examination, it has been found that the mechanism of uncoupling by bedaquiline shows key differences from that of the classical anionic uncouplers of OXPHOS, which have been postulated to lie at the heart of the mechanism of lethality in mycobacteria. The classical anionic OXPHOS uncouplers have been shown to collapse both the driving forces ΔpH and Δψ, while bedaquiline only collapses the ΔpH but leaves the Δψ intact. It has been concluded that the results make logical sense only if the two driving forces ΔpH and Δψ are created by two independent agents, as predicted by the two-ion theory of energy coupling in ATP synthesis by the F 1 F O -ATP synthase, and not solely by protons, as in current dogma. The critical role of succinate co-anion or K + countercation translocation in energy transduction and coupling, the crucial ionophoric function of bedaquiline in H + /K + antiport, the potent and specific uncoupling that results from localization of the drug to its target c-ring in F O , and the consequent stimulation of respiration has been highlighted. The results show how stereospecific drug-target interactions in F O lead to an acceleration of cellular respiration and to lethality of mycobacteria arising from a perturbation in succinate homeostasis over time (Figure 2). New experiments that offer further insights into the uncoupling mechanism by bedaquiline have been suggested. The experimental information has been integrated into a consistent model of the uncoupling and mycobacterial cell death actions of bedaquiline. The major biological implications of these fundamental insights for the interdisciplinary fields of bioenergetics, bacterial motility, respiration, and the design of antimicrobials have been discussed.

ACKNOWLEDGMENTS
The author would like to thank Dr. C. Channakeshava for his several helpful discussions and the Editor-in-Chief for his kind invitation to contribute to the Special Issue.