A simple assay for inhibitors of mycobacterial oxidative phosphorylation

Oxidative phosphorylation, the combined activities of the electron transport chain (ETC) and ATP synthase, has emerged as a valuable target for antibiotics to treat infection with Mycobacterium tuberculosis and related pathogens. In oxidative phosphorylation, the ETC establishes a transmembrane electrochemical proton gradient that powers ATP synthesis. Monitoring oxidative phosphorylation with luciferase-based detection of ATP synthesis or measurement of oxygen consumption can be technically challenging and expensive. These limitations reduce the utility of these methods for characterization of mycobacterial oxidative phosphorylation inhibitors. Here, we show that fluorescence-based measurement of acidification of inverted membrane vesicles (IMVs) can detect and distinguish between inhibition of the ETC, inhibition of ATP synthase, and nonspecific membrane uncoupling. In this assay, IMVs from Mycobacterium smegmatis are acidified either through the activity of the ETC or ATP synthase, the latter modified genetically to allow it to serve as an ATP-driven proton pump. Acidification is monitored by fluorescence from 9-amino-6-chloro-2-methoxyacridine, which accumulates and quenches in acidified IMVs. Nonspecific membrane uncouplers prevent both succinate- and ATP-driven IMV acidification. In contrast, the ETC Complex III2IV2 inhibitor telacebec (Q203) prevents succinate-driven acidification but not ATP-driven acidification, and the ATP synthase inhibitor bedaquiline prevents ATP-driven acidification but not succinate-driven acidification. We use the assay to show that, as proposed previously, lansoprazole sulfide is an inhibitor of Complex III2IV2, whereas thioridazine uncouples the mycobacterial membrane nonspecifically. Overall, the assay is simple, low cost, and scalable, which will make it useful for identifying and characterizing new mycobacterial oxidative phosphorylation inhibitors.

The genus Mycobacterium includes numerous pathogenic bacteria, the most notable of which being Mycobacterium tuberculosis, the causative agent of the disease tuberculosis (TB).Although mycobacteria can survive for an extended time in hypoxic conditions (1), they are considered obligate aerobes.Oxygen serves as the final electron acceptor in oxidative phosphorylation, the combined activities of the electron transport chain (ETC) and ATP synthase.The membraneembedded protein complexes of the ETC couple oxidation of nutrients to the transport of protons across the mycobacterial inner membrane, producing an electrochemical proton motive force (PMF) that powers ATP synthase.
Unlike canonical ETCs found in mitochondria and many bacteria, the mycobacterial ETC is highly branched (reviewed in Ref. (2)).NADH is oxidized by at least two different NADH dehydrogenases: the proton pumping respiratory Complex I as well as one or more nonproton pumping NDH-2 enzymes.Both Complex I and NDH-2 reduce menaquinone to menaquinol within the mycobacterial inner membrane.Two forms of respiratory Complex II (Sdh1 and Sdh2), and possibly fumarate reductase functioning in reverse (3), serve as succinate:quinone oxidoreductases.Reduction of fumarate has been detected in inverted membrane vesicles (IMVs) from the slow growing organism Mycobacterium bovis BCG but not from the fast growing Mycobacterium smegmatis (4).A malate:quinone oxidoreductase (Mqo) can also contribute to the pool of reduced menaquinol in the membrane (5).Sdh1, Sdh2, fumarate reductase, and Mqo are not thought to contribute to the PMF directly, and Sdh2 may even harness the PMF to drive the endergonic reduction of menaquinone by succinate (6).Although mycobacteria can use electrons from menaquinol to reduce nitrate, fumarate, and hydrogen (7), oxygen is essential for mycobacterial growth (8,9).In the mycobacterial ETC, electrons are transferred from menaquinol to oxygen by two terminal oxidases: a supercomplex of respiratory Complexes III and IV (also known as cytochrome bcc-aa 3 or CIII 2 CIV 2 ) and cytochrome bd.The structures of CIII 2 CIV 2 (10,11) and cytochrome bd (12,13) suggest that they translocate four protons per electron and a single proton per electron, respectively.
The diarylquinoline drug bedaquiline was discovered in a phenotypic screen for compounds that inhibit growth of M. smegmatis (14).Bedaquiline inhibits mycobacterial ATP synthase by binding with low affinity to the ring of c subunits in the enzyme's rotor subcomplex (14,15), with two highaffinity binding sites at the interface of subunit a with the c ring (16).Bedaquiline has become instrumental for the treatment of drug-resistant TB (17).Furthermore, its discovery revealed that targeting oxidative phosphorylation can kill mycobacteria to treat TB.Subsequently, the second-generation diarylquinolines TBAJ-876 and TBAJ-587 were developed, which exhibit improved binding to mycobacterial ATP synthase, reduced inhibition of the human Ether-à-go-go-Related Gene channel, and decreased cLogP (18).These compounds are currently undergoing clinical trials.The imidazopyridine compound telacebec (also known as Q203) was discovered in a phenotypic screen of M. tuberculosis-infected macrophages and found to inhibit CIII 2 CIV 2 (19).CIII 2 CIV 2 activity can be replaced, at least in part, by cytochrome bd (20,21), and although treatment of M. tuberculosis-infected marmosets with a telacebec analog controlled disease progression and reduced lesion-associated inflammation, it led to most lesions becoming cavitary (22).Nonetheless, treatment with telacebec decreased viable mycobacterial sputum load in humans in a phase 2 clinical trial (23).
The activity of detergent-solubilized and purified complexes involved in oxidative phosphorylation can be measured in multiple ways.These methods include following the oxidation of substrates like NADH spectrophotometrically or monitoring reduction of oxygen with a Clark electrode.However, experiments with purified complexes are complicated by the tendency of soluble menaquinone analogs to autoxidize (10,11,24,25).Furthermore, purifying ETC complexes is often time consuming and resource intensive, presenting a barrier to use of enzyme assays in the characterization of inhibitors of oxidative phosphorylation.Alternatively, oxidative phosphorylation can be measured in whole cells or isolated mycobacterial membranes without purification of individual protein complexes, which can provide more physiologically relevant insight into the properties of inhibitors (26).A whole-cell screen for depletion of intracellular ATP, conducted in the presence of the cytochrome bcc-aa 3 inhibitor telacebec, led to identification of the cytochrome bd inhibitor ND-011992 (27).Isolated membranes readily form IMVs upon resuspension in buffer.These IMVs include intact ETCs capable of NADH- (28) or succinate-driven (15) ATP synthesis.ATP synthesis activity can be followed with the enzyme luciferase, which uses ATP and oxygen to oxidize D-luciferin, producing oxyluciferin, adenosine monophosphate, carbon dioxide, and light, with the latter detected using a luminometer.However, even with IMVs, monitoring oxygen consumption is technically challenging and monitoring ATP synthesis requires either expensive reagents such as luciferase and luciferin (28) or complex procedures to denature proteins and quantify the remaining ATP (29).Furthermore, distinguishing between inhibitors of ATP synthase and nonspecific membrane uncouplers that dissipate the PMF to prevent ATP synthesis requires additional experiments.Despite these complications, ATP synthesis assays have been used by the company AstraZeneca to identify new mycobacterial ATP synthase inhibitors (30).
In an alternative assay for mycobacterial ETC activity, IMVs can be added to a buffer containing the inexpensive fluorophores 9-amino-6-chloro-2-methoxyacridine (ACMA) or N,N,N 0 ,N 0 -tetramethylacridine-3,6-diamine (acridine orange).Incubating the vesicles with an electron source, either succinate (15), NADH (28), malate (5), or fumarate (31) leads to the ETC pumping protons into the IMVs (Fig. 1A, upper).This acidification results in concentration of the fluorophore within the IMVs and quenching of its fluorescence, which can be restored by addition of an ionophore like nigericin to collapse the PMF.IMV acidification assays have been used extensively to demonstrate mycobacterial ETC activity (29,(31)(32)(33)(34), but in the absence of a complementary assay for ATP synthase activity, they cannot identify ATP synthase inhibitors or distinguish ETC inhibitors from uncouplers.
Vesicles can also be acidified by ATP-driven proton pumps such as V-type ATPases (35) or mitochondrial ATP synthases working in reverse as ATPases (36).ATP hydrolysis by an ATP synthase can be detected by quantifying the release of free phosphate (37,38) or with an assay that couples ATP hydrolysis to the oxidation of NADH, which can be monitored spectrophotometrically (39).However, like many bacterial ATP synthases (reviewed in Ref. (40)), mycobacterial ATP synthases are inhibited from working as ATP-powered proton pumps (29).In mycobacterial ATP synthase, ATP hydrolysis is inhibited by C-terminal extensions of the α subunits that prevent rotation of the rotor subcomplex in the hydrolysis direction (16).Modification of the genome of M. smegmatis to truncate the inhibitory extensions of the α subunits allows the enzyme to function as an ATPase (16), and chemical inhibition of ATP hydrolysis by the purified truncation mutant can be monitored with traditional ATPase assays (16,31).
Here, we show that ATP-driven acidification of IMVs from an M. smegmatis strain with truncated α subunits allows straightforward detection of mycobacterial ATP synthase inhibitors such as bedaquiline.Furthermore, inhibitors of CIII 2 CIV 2 , such as telacebec, can be detected by their inhibition of succinate-driven IMV acidification, whereas nonspecific uncouplers of the PMF prevent both succinate-and ATP-driven acidification.Surprisingly, NADH-driven acidification of IMVs is less sensitive to telacebec than succinate-driven acidification, even though both NADH and succinate should provide menaquinol to CIII 2 CIV 2 .As a result, the assay cannot distinguish succinate dehydrogenase inhibitors from CIII 2 CIV 2 inhibitors.With the assay, we show that lansoprazole sulfide (LPZS) blocks succinate-driven acidification of IMVs at submicromolar concentrations.LPZS is a metabolite of the gastric proton-pump inhibitor lansoprazole (LPZ) and was previously shown to inhibit CIII 2 CIV 2 and have antimycobacterial activity (41).The assay also supports the finding that thioridazine (THZ), a first-generation antipsychotic drug that was found to inhibit mycobacterial NDH-2 (42)(43)(44), functions as an uncoupler of the PMF at micromolar concentrations (45).Owing to its simplicity and low cost, the assay should prove useful for identification and characterization of new inhibitors of mycobacterial oxidative phosphorylation.

Fluorescence recovery allows quantification of IMV acidification activity
The absolute fluorescence measured from IMVs in a 96-well plate depends on fluorimeter settings and is sensitive to slight differences in the composition of the sample as well as changes in the sample volume.Notably, increasing the sample volume in the 96-well plate can cause an apparent increase in fluorescence owing to the sample surface being closer to the plate reader probe.Therefore, to investigate whether IMV acidification can be used to detect inhibition of oxidative phosphorylation, we first set out to find a way to compare fluorescence quenching between experiments.We did not attempt to quantify the pH within IMVs but simply to measure the relative activity of the protein complexes that drive acidification.Dilution of IMVs led to differences in the rate and extent of succinate-and NADH-driven fluorescence quenching (Fig. S1).We found that this activity could be quantified most reproducibly by adding substrate, allowing 15 min for IMV acidification to quench fluorescence, and then measuring the recovery of fluorescence after adding a small volume of the potent H + /K + antiporter nigericin to the sample (Fig. 1A, green double-headed arrow).Fluorescence curves are plotted with the fluorescence immediately before addition of nigericin set to zero.Using this quantification strategy and succinate-driven acidification, dilution of IMVs led to the expected decrease in fluorescence recovery (Fig. 1B, upper), although the fluorescence recovery was not linear with IMV concentration (Fig. 1B, lower).The recovery of fluorescence correlated with the concentration of vesicles better than other parameters, such as the initial rate of fluorescence quenching, particularly in experiments shown later.This difference is likely because the rate of fluorescence quenching depends on multiple parameters, including the rates of the enzymes that are pumping protons and the concentration of vesicles that can accumulate the fluorophore.In our protocol for succinate-driven acidification, each sample in the 96-well plate contained 10 μl of IMVs prepared by resuspending membranes from 4 l of M. smegmatis culture in 10 ml and diluting twofold before use.Consequently, with these conditions, a 4 l growth of bacteria provides enough material to perform up to 2000 assays (or 500 assays/l of bacteria cultured).

Succinate, NADH, fumarate, malate, and ATP can all drive IMV acidification
With the same preparation of vesicles used for succinatedriven acidification (Fig. 1B), we found that fumarate,

Succinate Nigericin
Fluorescence   malate, and NADH demonstrate the same concentrationdependent acidification of IMVs (Fig 2, A-C).NADH:menaquinone oxidoreductase activity is likely the result of NDH-2 only, because the proton-pumping Complex I is almost undetectable in the conditions we used to culture M. smegmatis (46,47).Malate oxidation by Mqo also produces menaquinol (5).As seen previously (31), fumarate drives IMV acidification through a mechanism that is not completely clear but could involve a contaminating fumarase producing malate.However, when we tested for contaminating cytoplasmic enzymes by adding 5 mM glucose to the IMV preparation, we could not observe any glucose-driven IMV acidification.Compared with succinate, addition of NADH, malate, and fumarate all result in faster and more extensive IMV acidification, as judged by the IMV dilution needed to produce a similar fluorescence recovery: twofold dilution for succinate, fourfold for fumarate and malate, and eightfold for NADH (Figs. 1B and 2).
Next, we investigated whether ATP-driven acidification could be observed with IMVs prepared from an M. smegmatis strain where the α subunits of the ATP synthase had been truncated to allow ATP hydrolysis (16) (Fig. 2D).We found that this proton-pumping activity could indeed be observed (Fig. 2E, upper).ATP-driven IMV acidification required using a higher concentration of IMVs in each well than succinatedriven acidification and was measured with 10 μl of IMVs prepared by resuspending membranes from a 3 l growth of M. smegmatis in 8 ml of buffer without further dilution.Although this condition requires more material per well than succinate-driven acidification, it still allows for 270 assays/l of bacteria cultured.This quantity of IMVs was used in all subsequent ATP-driven acidification assays.As with succinate-, NADH-, malate-, and fumarate-driven IMV acidification, fluorescence recovery from ATP-driven acidification also decreases as the IMVs are diluted (Fig. 2E, lower).

The CIII 2 CIV 2 inhibitor telacebec inhibits succinate-driven IMV acidification
To determine if succinate-driven IMV acidification can be used to detect CIII 2 CIV 2 inhibitors, we performed experiments with the well-characterized inhibitor telacebec (Q203) added to the sample at a range of concentrations.This experiment (Fig. 3, A and B, with a replicate inhibitor dilution series in Fig. S2A) showed that telacebec can block succinate-driven IMV acidification, yielding an IC 50 of 140 nM.This IC 50 is comparable to the IC 50 of 53 nM reported previously from measurement of purified CIII 2 CIV 2 activity with an oxygensensitive electrode (24).The near-complete inhibition of acidification by telacebec indicates that most of the acidification results from CIII 2 CIV 2 activity rather than cytochrome bd activity.The same concentrations of telacebec did not inhibit ATP-driven acidification of IMVs (Fig. 3C), as expected from the compound's specific inhibition of CIII 2 CIV 2 rather than ATP synthase.Telacebec's inhibition of succinate-driven acidification but not ATP-driven acidification also confirms that its effect does not arise from uncoupling of the PMF.Surprisingly, while NADH drives robust acidification of IMVs (Fig. 2C), this acidification could only be inhibited completely with high concentrations of telacebec and with IMVs diluted 16-fold relative to the concentration used for the succinate experiments.As described above, in the growth conditions for the bacteria used to prepare these IMVs, NADH:menaquinone oxidoreductase activity is due almost entirely to the nonproton pumping NDH-2 (46,47).Whether menaquinol is produced by succinate dehydrogenase or NDH-2, it should contribute to the PMF in the same way via CIII 2 CIV 2 and to a lesser extent cytochrome bd.Therefore, it is not clear why telacebec can block succinate-driven acidification almost completely, while only slightly inhibiting NADH-driven acidification.Unfortunately, the inability to detect inhibition of NADH-driven IMV acidification means that the assay cannot distinguish succinate dehydrogenase inhibitors from CIII 2 CIV 2 inhibitors, although other assays exist for that purpose (10,11,24,48).

The ATP synthase inhibitor bedaquiline inhibits ATP-driven IMV acidification
We next tested whether IMV acidification could be used to detect inhibition of mycobacterial ATP synthase.Using IMVs with hydrolytically competent ATP synthase, we tested a range of concentrations of bedaquiline for inhibition of ATP-driven IMV acidification (Fig. 4, A and B, with a replicate inhibitor dilution series in Fig. S2B).This analysis provided an IC 50 of 34 nM, which is somewhat higher than the reported IC 50 of 2.5 nM in an ATP synthesis assay with M. smegmatis IMVs (28) or the nanomolar inhibition of ATP hydrolysis by purified hydrolytically competent M. smegmatis ATP synthase ( 16).The higher IC 50 values in the IMV acidification assay suggest that it is not as sensitive as the other assays, but it is still capable of easily detecting an ATP synthase inhibitor like bedaquiline.Bedaquiline did not inhibit succinate-driven IMV acidification, consistent with the compound being a specific inhibitor of mycobacterial ATP synthase that does not inhibit mycobacterial CIII 2 CIV 2 (Fig. 4C).The lack of inhibition of succinate-driven IMV acidification also demonstrates that the inhibition of the ATP-driven acidification is not the result of uncoupling the PMF.
THZ, a first-generation antipsychotic drug, was proposed to be an inhibitor of NDH-2 (43).THZ has been used in combination with first-line antibiotics to treat extensively drugresistant TB (50).We found that THZ inhibited both succinate-driven IMV acidification (Fig. 6, A and B) and ATPdriven IMV acidification (Fig. 6, C and D).Although this result cannot exclude the possibility that THZ inhibits NDH-2, it shows that the present assay is not capable of distinguishing NDH-2 inhibition by THZ from nonspecific uncoupling of the PMF.The micromolar uncoupling activity of THZ is consistent with reports that it uncouples oxidative phosphorylation in mitochondria at similar concentration (51).Uncoupling of the PMF is an area of active investigation in the study of compounds that sterilize M. tuberculosis infection (32,33,42).

Discussion
As described previously, micromolar concentrations of the CIII 2 CIV 2 inhibitor telacebec appear to block succinate-driven IMV acidification completely.This finding is somewhat surprising because the cytochrome bd oxidase should also contribute to acidification, albeit with fewer protons translocated per electron.One explanation for the full inhibition by telacebec is that there is insufficient cytochrome bd activity in the IMVs for it to contribute detectably to acidification.This explanation would suggest that cytochrome bd makes only a minor contribution to the PMF in the M. smegmatis growth conditions we used.However, this conclusion is inconsistent with our finding that telacebec inhibits NADH-driven acidification only slightly.An alternative explanation for these  findings is that there could be some form of channeling that occurs in the mycobacterial ETC: with electrons from succinate dehydrogenase preferentially passing to CIII 2 CIV 2 and electrons from NDH-2 preferentially passing to cytochrome bd.Regardless of the cause of this inconsistency, the inability to detect CIII 2 CIV 2 inhibition with NADH-driven acidification leads to a limitation of the assay.Comparing NADH-and succinate-driven acidification would allow distinguishing between succinate dehydrogenase, NDH-2, and CIII 2 CIV 2 inhibitors.Without this ability, the assay cannot distinguish succinate dehydrogenase inhibitors from CIII 2 CIV 2 inhibitors.It is not clear if the assay can be used to detect NDH-2 inhibitors because the NDH-2 inhibitor that we tested uncoupled the PMF.Furthermore, the intrinsic fluorescence or absorbance from some compounds may interfere with ACMA fluorescence.For these compounds, it may be possible to use an alternative fluorophore, such as acridine orange (31).
Despite these limitations, the assay presented here could have utility in characterizing CIII 2 CIV 2 inhibitors and ATP synthase inhibitors or even for performing high-throughput screens.The success of bedaquiline in treating TB supports the value of targeting the ATP synthase, whereas other studies suggest limitations in targeting CIII 2 CIV 2 (20,22).Compared with existing ATP synthase assays, the assay described here is inexpensive and easy to perform, facilitating its use in largescale studies.There are many remaining fundamental questions about mycobacterial oxidative phosphorylation and how it adapts to different growth conditions (5,7,46).The assay presented here could also serve as a tool to study these aspects of mycobacterial biology.

Experimental procedures M. smegmatis strains and growth
For ATP-driven acidification assays, IMVs were prepared from M. smegmatis strain GMC_MSM2 (16) where a 3×FLAG affinity tag truncates the α subunits of the ATP synthase following residue Ser518.For succinate-, NADH-, malate-, and fumarate-driven acidification assays, IMVs were prepared from M. smegmatis strain QcrB-3×FLAG (24), which is identical to GMC_MSM2 except that the ATP synthase α subunits are intact and the 3×FLAG tag is at the C terminus of the QcrB subunit of the CIII 2 CIV 2 supercomplex.M. smegmatis strains were grown in Middlebrook 7H9 broth (4.7 g/l) supplemented with 10 g/l tryptone, 2 g/l glucose, 0.8 g/l NaCl, and 0.5% (v/v) Tween-80.Each 1 l culture was grown in a 2.8 l Fernbach flask at 30 C for 72 h with shaking at 180 rpm.Bacteria were harvested by centrifugation for 20 min at 6500g and 4 C and were sometimes frozen at −80 C before use.These strains are available from the Biodefense and Emerging Infections (BEI) Resources (www.beiresources.org) as Items #NR-59698 and #NR-59699.

Preparation of IMVs
To prepare IMVs from M. smegmatis strain QcrB-3×FLAG, 4 l cultures were grown.Cell pellets were resuspended with a Dounce homogenizer in 40 ml lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM MgSO 4 , 5 mM benzamidine hydrochloride, and 5 mM 6-aminocaproic acid) per 1 l of starting cell culture.DNase I in water and phenylmethylsulfonyl fluoride in ethanol were added to the suspension to final concentrations of 100 μg/ml and 100 μM from 100 mg/ml and 100 mM stocks, respectively.The cell suspension was then filtered through four layers of Miracloth (Millipore) to remove clumps of cells, and cells were lysed with four passes through an Avestin homogenizer operating at 20,000 to 25,000 psi.Insoluble debris was removed by centrifugation for 30 min at 39,000g and 4 C.The membrane fraction from cells was collected by centrifugation for 1 h at approximately 199,269g and 4 C using a Beckmann Ti70 ultracentrifuge rotor.To form IMVs, pelleted membranes were resuspended with a Dounce homogenizer in 2.5 ml resuspension buffer A (50 mM Tris [pH 7.5], 150 mM NaCl, 5 mM MgSO 4 , 5 mM benzamidine hydrochloride, 5 mM 6aminocaproic acid, 20% [v/v] glycerol) per 1 l of original cell culture.This concentration corresponds to 19 mg/ml total protein in the IMV suspension, as determined by bicinchoninic acid assay (Pierce) without solubilizing the IMVs.IMVs were divided into aliquots and stored at −80 C.These IMVs were diluted before use, as described later.IMVs from M. smegmatis strain GMC_MSM2 were prepared in the same way as from the QcrB-3×FLAG strain with the minor modification that membranes were resuspended at 2.7 ml/l of starting culture in resuspension buffer B (20 mM Hepes-KOH [pH 7.5], 50 mM NaCl, 50 mM KCl, 5 mM MgSO 4 , and 20% [v/v] glycerol).This concentration corresponds to 15 mg/ml total protein in the IMV suspension, as determined by bicinchoninic acid assay without solubilizing the IMVs.

Vesicle acidification assays
For succinate-driven acidification experiments with inhibitors, IMVs from the QcrB-3×FLAG strain were diluted twofold in resuspension buffer A before use.For NADHdriven acidification experiments with telacebec, IMVs were diluted 32-fold in resuspension buffer A before use.For ATPdriven acidification experiments with inhibitors, IMVs from the GMC_MSM2 strain were used without further dilution.IMV acidification experiments were performed in 96-well plates (BRANDplates pureGrade 96-well black microplates).For each well of the plate, a master solution that had 1.5× the volume needed for the well was prepared in a microcentrifuge tube at room temperature.This solution contained 120 μl ACMA assay buffer (20 mM Hepes-KOH [pH 7.5], 200 mM KCl, and 10 mM MgCl 2 ), 35 μl of MilliQ water, and 4.8 μl of inhibitor in dimethyl sulfoxide at 50× its intended final concentration, or dimethyl sulfoxide alone.To this solution, 0.375 μl of ACMA from a 2 mM stock in ethanol and 15 μl of diluted IMV solution were added.The master solution was mixed by pipetting, and 116.8 μl of the solution was transferred to each well of the 96-well plate and incubated in the dark for 15 min at room temperature prior to starting the experiment.
ACMA fluorescence was followed with a BioTek Synergy Neo2 Multi-mode Assay Microplate reader (Agilent Technologies) with samples held at 25 C during experiments.The samples were excited at 410 nm, and fluorescence was measured at 480 nm with the fluorescence gain set to 90.Fluorescence of the samples was monitored for 2 min to determine a baseline, at which point 40 μl of either 4 mM disodium NADH in water, 20 mM sodium succinate in water, 20 mM fumarate in 40 mM Tris with pH unadjusted (resulting in a pH of 7), 20 mM malate in 40 mM Tris with pH unadjusted (resulting in a pH of 7), or 8 mM disodium ATP in 16 mM Tris with pH unadjusted (resulting in a pH of 7) was added to each well using the instrument's automated injector, which had been primed with these solutions.Adding these solutions resulted in final concentrations of 1 mM NADH, 5 mM succinate, 5 mM fumarate, 5 mM malate, or 2 mM ATP, respectively.Fluorescence from the sample was monitored for 15 min before 3.2 μl of 100 μM nigericin in 1% (v/v) ethanol was added to each well with a multichannel pipettor.The samples were then mixed with a different multichannel pipettor, and recovery of fluorescence signal was monitored for an additional 5 min before ending the experiment.For IMV dilutions, the fluorescence recovery is reported as the absolute value of the difference between the fluorescence (arbitrary units) immediately before and 5 min after addition of nigericin.For inhibitor titrations, the relative fluorescence recovery is presented, which is the fluorescence recovery at each inhibitor concentration normalized by the fluorescence recovery observed with no inhibitor.For visualization, raw fluorescence values were normalized by addition of an offset so that the fluorescence immediately prior to nigericin injection is set to 0.

Figure 1 .
Figure 1.Quantification of succinate-driven acidification of Mycobacterium smegmatis IMVs.A, schematic of the IMV acidification assay with example data from succinate-driven acidification.B, a dilution series (upper) and a plot of fluorescence recovery versus IMV dilution (lower) for succinate-driven acidification of IMVs shows decreased fluorescence recovery with more dilute samples.Open symbols show technical replicates.Filled symbols show the mean from n = 3 technical replicates.Error bars indicate ±SD when shown.IMV, inverted membrane vesicle; MQ, menaquinone; SDH, succinate dehydrogenase.

Figure 2 .
Figure 2. IMVs can be acidified with fumarate, malate, NADH, and ATP.A, dilution series (upper) and a plot of fluorescence recovery versus IMV dilution (lower) for fumarate-driven acidification of IMVs.IMVs were prepared from the QcrB-3×FLAG Mycobacterium smegmatis strain.B, dilution series (upper) and a plot of fluorescence recovery versus IMV dilution (lower) for malate-driven acidification of IMVs.IMVs were prepared from the QcrB-3×FLAG M. smegmatis strain.C, dilution series (upper) and a plot of fluorescence recovery versus IMV dilution (lower) for NADH-driven acidification of IMVs.IMVs were prepared from the QcrB-3×FLAG M. smegmatis strain.D, schematic of the ATP-driven IMV acidification assay.E, dilution series (upper) and a plot of fluorescence recovery versus IMV dilution (lower) for ATP-driven acidification of IMVs.IMVs were prepared from the M. smegmatis strain GMC_MSM2.Open symbols show technical replicates.Filled symbols show the mean from n = 3 technical replicates.Error bars indicate ±SD when shown.IMV, inverted membrane vesicle.

Figure 3 .
Figure 3. Telacebec inhibition of CIII 2 CIV 2 can be detected by its effect on succinate-driven IMV acidification.A, telacebec (Q203) at different concentrations inhibits succinate-driven IMV acidification.B, a plot of fluorescence recovery versus inhibitor concentration indicates an IC 50 of 140 nM.C, telacebec does not inhibit ATP-driven acidification of IMVs, showing that it is does not uncouple the PMF.D, telacebec inhibits NADH-driven acidification of IMVs only at high concentrations.Open symbols show technical replicates.Filled symbols show the mean from n = 3 technical replicates.Error bars indicate ±SD when shown.IMV, inverted membrane vesicle; PMF, proton motive force.

Figure 4 .
Figure 4. Bedaquiline (BDQ) inhibition of ATP synthase can be detected by its effect on ATP-driven IMV acidification.A, BDQ at different concentrations inhibits ATP-driven acidification of IMVs.B, a plot of fluorescence recovery versus inhibitor concentration indicates an IC 50 of 30 nM.C, BDQ does not inhibit succinate-driven acidification of IMVs, showing that its effect on ATP-driven acidification is not a result of uncoupling the PMF.Open symbols show technical replicates.Filled symbols show the mean from n = 3 technical replicates.Error bars indicate ±SD when shown.IMV, inverted membrane vesicle; PMF, proton motive force.

Figure 5 .
Figure 5. Lansoprazole sulfide (LPZS) inhibits CIII 2 CIV 2 .A, LPZS at different concentrations inhibits succinate-driven acidification of IMVs.B, a plot of fluorescence recovery versus inhibitor concentration indicates an IC 50 of 860 nM.C, LPZS does not inhibit ATP-driven acidification of IMVs, showing that its effect on succinate-driven acidification is not a result of uncoupling the PMF.D, lansoprazole (LPZ) does not inhibit succinate-driven acidification of IMVs, showing that it is not an inhibitor of CIII 2 CIV 2 .Open symbols show technical replicates.Filled symbols show the mean from n = 3 technical replicates.Error bars indicate ±SD when shown.IMV, inverted membrane vesicle; PMF, proton motive force.

Figure 6 .
Figure 6.THZ uncouples the PMF in IMVs.A, THZ at different concentrations inhibits succinate-driven acidification of IMVs.B, a plot of fluorescence recovery versus inhibitor concentration for succinate-driven acidification.C, THZ at different concentrations also inhibits ATP-driven acidification of IMVs, indicating that its effect on succinate-driven acidification results from uncoupling the PMF.D, a plot of fluorescence recovery versus inhibitor concentration for ATP-driven acidification.Open symbols show technical replicates.Filled symbols show the mean from n = 3 technical replicates.Error bars indicate ±SD when shown.IMV, inverted membrane vesicle; PMF, proton motive force; THZ, thioridazine. (RFU)