Repurposing Vanoxerine as a new antimycobacterial drug and its impact on the mycobacterial membrane

Mycobacterium tuberculosis is a deadly pathogen, currently the leading cause of death worldwide from a single infectious agent through tuberculosis infections. If the End TB 2030 strategy is to be achieved, additional drugs need to be identified and made available to supplement the current treatment regimen. In addition, drug resistance is a growing issue, leading to significantly lower treatment success rates, necessitating further drug development. Vanoxerine (GBR12909), a dopamine re-uptake inhibitor, was recently identified as having anti-mycobacterial activity. Repurposing vanoxerine or its analogues to treat tuberculosis infections may allow a faster route to clinical use than novel drug discovery. However, its effects on Mycobacteria were not well characterised. Herein, we report vanoxerine as a disruptor of the membrane potential, inhibiting mycobacterial efflux and survival, with an undetectable level of resistance. This study suggests a mechanism of action for vanoxerine, which will allow for its continued development and optimisation for pre-clinical testing.


Introduction
Tuberculosis is a major cause of death worldwide, accounting for approximately 1.6 million deaths in 2020 (WHO, 2022a). Mycobacterium tuberculosis, the causative agent, is a slow growing pathogenic species responsible for this deadly lung infection in over 10 million individuals every year (WHO, 2022a). Drug resistance to the current treatment regimen is a growing issue, representing 3% of new cases and 17% of reinfections (WHO, 2021). The treatment success rate against multi-drug resistant (MDR) tuberculosis is only 60% globally, necessitating new treatment options to combat this pandemic (WHO, 2022a;Pai et al., 2016).
For treatment of drug susceptible tuberculosis, a combination of four front-line drugs, ethambutol, isoniazid, pyrazinamide, and rifampicin, are taken for two months.
Followed by isoniazid and rifampicin for a further four months (Nahid et al., 2016;Zumla et al., 2013). The combination therapy presents both financial and health burdens to patients, so any shortening or simplification of this treatment would provide a direct benefit to millions of patients every year. In addition, the treatment of MDR tuberculosis is more complicated, only recently being shortened to six-months with three to four drugs, from six drugs for up to 24 months (Nahid et al., 2016;Zumla et al., 2013;WHO, 2022b;Conradie et al., 2020).
In recent years, there has been some progress in the development of novel antituberculosis drugs, with bedaquiline, delamanid and pretomanid all being approved for use (Zumla et al., 2013;Andries et al., 2005;Matsumoto et al., 2006;Skripconoka et al., 2012;Stover et al., 2000). However, all three drugs are restricted for use against MDR tuberculosis (Skripconoka et al., 2012;Centers for Disease Control and Prevention, 2013), and so do not allow for the shortening or simplification of the front-line drug regimen. In addition, some resistance has already been identified against these new drug compounds, necessitating further drug development (Andries et al., 2014;Hartkoorn et al., 2014).
One drug development approach which has been undertaken previously is drug repurposing (Corsello et al., 2017;Huang et al., 2011;Maitra et al., 2016), taking drugs with known activity and using their anti-mycobacterial properties for treatment of tuberculosis. Repurposing of drugs has been used for treating MDR tuberculosis in the past, including the use of linezolid, clofazimine and fluoroquinolones (Zumla et al., 2013). Currently, one-quarter of all ongoing clinical trials, for drug validation against tuberculosis, are focussed on repurposed drugs (WHO, 2022a). However, these drugs have all been known antibacterial compounds (Zumla et al., 2013). A recent study focussed on screening the Prestwick library, which consists of approved drugs against a large range of clinical implications, for activity against Mycobacteria (Kanvatirth et al., 2019). One of the drugs identified was vanoxerine (GBR12909), which showed activity against Mycobacteria smegmatis, Mycobacteria bovis BCG and M tuberculosis H37Rv (Kanvatirth et al., 2019).
Vanoxerine has been tested in several clinical trials, for two different clinical applications and has been well-tolerated by healthy volunteers (Søgaard et al., 1990;Preti, 2000;Lacerda et al., 2010;Laguna Pharmaceuticals, Inc., 2015). It has been identified as a dopamine re-uptake inhibitor (Lewis et al., 1999;Schmitt et al., 2008), for treatment of cocaine dependency, passing a phase I clinical trial. Vanoxerine has also been tested for use as an antiarrhythmic drug and passed a phase II clinical trial (Lacerda et al., 2010;Laguna Pharmaceuticals, Inc., 2015). It's activity in this context was blocking the hERG potassium channel, which had no adverse effects on healthy volunteers but allowed treatment of atrial fibrillation (Lacerda et al., 2010;Cakulev et al., 2011;Obejero-Paz et al., 2015). However, the follow-up Phase III clinical trials was stopped due to adverse effects on the heart (ventricular proarrhythmia) in the treatment group (Piccini et al., 2016). These issues specifically occurred in patients with structural heart disease, which represented two thirds of the patients enrolled on the trial, hence, it was terminated early. If it is to be repurposed successfully, further trials are required to assess its safety against other patient groups (Piccini et al., 2016;Geng et al., 2020). Due to the knowledge of the human targets of vanoxerine, any future applications could monitor these targets for side effects. Future development of vanoxerine would also involve searching for analogues which retain their antimycobacterial effects but have reduced impact on dopamine reuptake.
Following the initial discovery of vanoxerine's antimycobacterial properties (Kanvatirth et al., 2019), further characterisation of this drug's effects on Mycobacteria was required. Herein, we have studied the impact of vanoxerine on mycobacteria, both phenotypic effects and transcriptomic impacts. Work has been undertaken with the aim of deconvoluting the target of vanoxerine and we suggest it targets the mycobacterial membrane, causing loss of the membrane electric potential and preventing efflux. In future, the presented work may allow the repurposing of vanoxerine or its analogues for treatment of tuberculosis. Complex, M. bovis BCG), unless stated otherwise. Bacterial cultures were grown at 37 °C and 180 rpm (or static for BCG) until mid-log (OD600 = 0.5-1.0), kanamycin (50 µg/ml) was added when required.
The resazurin fluorescence (excitation -544 nm, emission -590 nm) was measured in a BMG Labtech POLARstar Omega plate reader, normalised and cell survival calculated. For A. baumanii, C. glutamicum, E faecium, K. pneumoniae, P. aeruginosa and S. aureus, the OD600 was measured following 24-hours incubation, to calculate the percentage growth. For Checkerboard MICs, each well contained two drug compounds or DMSO, for a maximum 2% final concentration of DMSO. The fractional inhibitory concentration (FIC) was calculated (MIC99A in combination/MIC99A alone + MIC99B in combination/MIC99B alone) = FIC. If the FIC < 0.5, this represents synergy, 0.5 < FIC < 4, is indifference and FIC > 4, is antagonism (Odds, 2003).

Mt-AroB protein expression and purification
The M. tuberculosis aroB gene was cloned into the pET28a vector using HiFi cloning.

Resistant mutant generation
Agar plates (5 ml, 7H11 + OADC) containing 2x, 2.5x, 5x and 10x MIC of vanoxerine were each inoculated with 1x10 8 cells of M. bovis BCG. The plates were then incubated at 37 ˚C until the plates dried out, approximately 3 months.
The cells were either added directly to the 96-well plate or supplemented with glucose (final concentration 0.4% (w/v)) before addition. In both assays, the fluorescence was then measured, in a BMG Labtech POLARstar Omega plate reader, every 60 seconds for 1 hour (emission -544 nm, excitation -590 nm), at 37 ˚C. Method adapted from (Rodrigues et al., 2021).

Lipid extraction and thin-layer chromatography (TLC) analysis:
M. smegmatis cultures were split into flasks (5 ml, OD600 = 0.3) containing either no drug, 13, 26 or 65 ug/ml vanoxerine, alongside C 14 acetic acid (0.5 µCi per ml). These cultures were incubated at 37 ˚C for 24-hours, before cells were harvested by centrifugation (3,000 g, 10 minutes). The cell samples were stored at -20 ˚C until required. The pelleted cells were thawed and CH3Cl/MeOH/H2O (10:10:3, v/v/v, 2 ml) was added. The resuspension was incubated (2 hours, 50 ˚C), followed by addition of Vanoxerine exhibited no activity against Gram-negative species (Table 1), including species with EDTA-permeabilised outer membranes (data not shown), suggesting no equivalent target is present. Amongst Gram-positive species, Enterococcus faecium was inhibited by the drug, but no effect was found against Staphylococcus aureus, indicating some specificity amongst Gram-positive species. While the liquid MIC99 for M. bovis BCG was much higher than M. tuberculosis, the solid agar MIC99 was determined to be 15.2 µg/ml.

AroB is not the mycobacterial target of Vanoxerine
In addition to reporting the anti-mycobacterial activity of vanoxerine, previous work   Figure 1B). Overall, these results suggested AroB is not the target of vanoxerine, and hence additional work was undertaken to identify vanoxerine's mechanism of action.

No resistance could be generated against Vanoxerine in M. bovis BCG
In an effort to find the true Mycobacterial target of vanoxerine, spontaneous resistant mutant generation was attempted in M. bovis BCG (Abrahams and Besra, 2020).
However, no resistant mutants could be generated, with no growth occurring at even 2x MIC99 of vanoxerine (30 µg/ml). This suggested the rate of resistant generation is below 10 8 . In addition, an M. bovis BCG strain with a recG mutation was used, due to its increased mutation rate (Ley et al., 2019;Batt et al., 2015). However, the ∆recG strain also failed to acquire any spontaneous resistant mutations, suggesting a rate of resistance even lower than 10 8 . This is promising for the future use of vanoxerine; however, this approach did not allow a target to be identified.

Vanoxerine inhibits ethidium bromide efflux
Many current antimycobacterial drugs target either the cell envelope or cellular energetics (Andries et al., 2005, p.203;Pethe et al., 2013;Zumla et al., 2013) . Hence, this led us to investigate the impact of vanoxerine on the cell envelope, to study the impact on membrane integrity and energy requiring processes such as efflux. Ethidium bromide uptake and efflux in Mycobacteria has previously been described (Rodrigues et al., 2021) and (3.27 µg/ml) led to more ethidium bromide accumulation than the DMSO control. The effect was also concentration dependent, with 2x MIC99 (52 µg/ml) of vanoxerine causing 2.6x more fluorescence than DMSO and 1.2x more fluorescence than 0.5x MIC99 (13 µg/ml) across the 60-minutes, due to ethidium bromide accumulation. As there could be several explanations for these results, including membrane disruption/lysis, efflux inhibition or cell energetic effects, further work was undertaken.
To study inhibition of efflux by vanoxerine, Mycobacterial cells were allowed to accumulate ethidium bromide in the presence of a non-toxic efflux inhibitor verapamil.
The verapamil MIC99 against M. smegmatis and M. bovis BCG was greater than 100 bovis BCG, but with lower rates of accumulation and efflux (Supplementary Figure 3B).

Vanoxerine impacted the membrane potential (∆y)
As vanoxerine was able to inhibit efflux of ethidium bromide, the voltage sensitive dye DiOC2(3) was used to monitor the electric potential (∆y) of the membrane in response to vanoxerine treatment, as ∆y disruption would prevent the majority of efflux (Remm et al., 2022). DiOC2(3) partitions across phospholipid membranes proportionally to the ∆y present (Chen et al., 2018;Chawla et al., 2012;Hudson et al., 2020;Li et al., 2019;Novo et al., 1999), fluorescing red inside cells and green in solution. It was investigated whether vanoxerine disrupts the electric potential of the membrane by measuring DiOC2 (3)  with DiOC2(3) for two hours prior to drug addition, to allow the dye to equilibrate across the membrane proportional to the ∆y. The red/green fluorescence ratio fluctuated around 1.75 in the absence of compound, reducing to 1.60 across the 60-minute assay. Vanoxerine was able to disrupt the dye partitioning in a concentration dependent manner, with 52 µg/ml vanoxerine reducing the red/green fluorescence to 1.09 after 50 minutes, suggesting disruption of the ∆y (Figure 3). The membrane potential was still disrupted at sub-MIC amounts of vanoxerine, with 13 µg/ml (0.5x MIC) leading to a red/green fluorescent ratio of 1.30. CCCP is a protonophore which is known the disrupt both the ∆y and proton gradient (∆pH) of the membrane (Chen et al., 2018) and had the lowest red/green fluorescent ratio in this assay, of 0.95.
Vanoxerine disrupted the DiOC2(3) dye partitioning at a slower rate compared to  (3) for 2 hours to allow dye partitioning across the membrane, the fluorescence was then measured for 10 minutes. Then either vanoxerine or controls were added, and the fluorescence was measured for a further 50 minutes. N=3.
CCCP. The assay was specific to the ∆y disruption, rather than general proton motive force (PMF) disruption. Bedaquiline confirmed the specificity as it is known to only disrupt the ∆pH of the PMF, and hence had a response similar to DMSO in this assay, a red/green fluorescent ratio of 1.63 (Feng et al., 2015). In addition, kanamycin had no impact on DiOC2(3) partitioning, with a red/green fluorescent ratio of 1.63 after 50 minutes of incubation, suggesting a general bactericidal response is less likely to explain the dyes disruption.

Vanoxerine may potentiate the activity of other anti-mycobacterial drugs
As vanoxerine appears to inhibit efflux from Mycobacteria, it was hypothesised that any anti-mycobacterial drugs which are known to be pumped out by the cell, may have These FIC values were all over 0.5 and so indicate no interactions were occurring; hence, further work needs to be undertaken to determine if the observed MIC99 reductions are due to the presence of the vanoxerine or due to experimental set-up. In contrast, the FIC value for bedaquiline was 1.13, also suggesting no interaction, but corresponded to no difference to the bedaquiline MIC99 in the presence of 10.3 µg/ml vanoxerine ( Figure 4A).

Vanoxerine induced clear transcriptomic changes in M. bovis BCG
As the inhibition of efflux and disruption of electric potential occurred at sub-inhibitory concentrations of vanoxerine ( Figure 2B, Figure 3 The significantly dysregulated genes at 30 µg/ml vanoxerine were functionally annotated and clustered using the DAVID server (Sherman et al., 2022;Huang et al., 2009), to identify pathways or biological processes that were either up-or downregulated in response to vanoxerine ( Figure 4C). In relation to up-regulated evidence that vanoxerine's mechanism of action inhibits these processes.
As only 31 genes were significantly dysregulated in a concentration dependent manner, these were investigated in more detail (Supplementary Table 2). There was little consensus of function among the significantly upregulated genes, with the majority being of unknown function. The efflux pump MmpL5 was significantly upregulated, which could be due to an increasing impact on efflux at higher vanoxerine concentrations. The majority of significantly down-regulated transcripts were part of the mycolic acid biosynthetic pathway. Explanations for this result include the high energetic costs associated with mycolic acid production, indirect inhibition of MmpL3, or vanoxerine directly impacting this pathway.

Mycolic acid down-regulation occurred following vanoxerine treatment, but direct inhibition is unlikely
To gain further evidence for the impact on mycolic acid biosynthesis, in addition to the four genes which were transcriptionally downregulated in a concentration dependent manner, the effect of vanoxerine on all the genes in the pathway was investigated (Table 2). Treatment with vanoxerine led to transcriptional repression of mycolic acid biosynthesis, including all the enzymes in FAS-I and FAS-II. The only gene transcript which was not downregulated was mmpL3, although its regulation was not significantly different from the DMSO control, even using 30 µg/ml vanoxerine. To investigate whether this transcriptional down-regulation translated into loss or reduction of mycolic acids in vitro; lipid extraction was undertaken on M. smegmatis treated with vanoxerine. The lipids were labelled using C 14 acetic acid at the same time as drug treatment. The extracted lipids were separated using thin-layer chromatography (TLC) and visualised using X-ray film ( Figure 5A). Loss of trehalose mono-mycolate (TMM) occurred at 0.5x MIC of vanoxerine and trehalose di-mycolate M. smegmatis wild-type was treated with vanoxerine for 24-hours, followed by a total lipid extraction. Thin layer chromatography was performed (CH3Cl/MeOH/H2O, 80:20:2, v/v/v), followed by exposure to an X-ray film to image. MIC = 26 µg/ml GPL = Glycopeptidolipids, TDM = Trehalose dimycolates, TMM = Trehalose monomycolates, PI = Phosphatidylinositol, PIMs = phosphatidylmannosides. Representative image from N=4. (B) Densiometric analysis of TMM loss following vanoxerine treatment. The same silica plates were exposed to a storage phosphor screen and then scanned to quantify the (TDM) at 2.5x MIC of vanoxerine. The loss of TMM was quantified by comparing the radioactive counts on the TLC plate in the presence or absence of vanoxerine ( Figure   5B). A drop in TMM from 7% of the total counts to less than 1%, confirmed it is not just lower growth in the presence of vanoxerine which is caused this decrease. As the mycolic acids are essential to mycobacteria, this might be another mechanism of vanoxerine inhibition.
To investigate further, we chose to study vanoxerine's impact on C. glutamicum, a species where the mycolic acids are not essential. As vanoxerine could inhibit the growth of C. glutamicum, with an MIC99 of 15.5 µg/ml (Table 1), comparable to the MIC99 of M. tuberculosis, suggesting mycolic acid biosynthesis is not the only target of vanoxerine. A C. glutamicum ∆pks mutant, which does not synthesise mycolic acids, was compared to the wild-type strain for its survival following vanoxerine treatment ( Figure 5C). No shift in % growth curve or difference in MIC99 was observed, suggesting mycolic acid biosynthesis inhibition is not the main mode of inhibition of vanoxerine. In addition, the over-expression of several genes in the mycolic acid biosynthetic pathway was undertaken in Mycobacteria (Supplementary Figure 6).
However, no differences in survival to vanoxerine treatment were found, including for MmpL3 over-expression, providing further evidence that this is not a direct target of vanoxerine.
radioactivity of each spot. The counts for the TMM spot were compared to the total counts in each lane, to generate a %counts per mm 2 . N=4. (C) %Growth comparing C. glutamicum WT vs ∆pks13 mutant in the presence of vanoxerine. C. glutamicum was incubated with vanoxerine for 24 hours, before OD600 was measured. The % survival was calculated compared to DMSO only and rifampicin controls. N=3.

Discussion
This study has highlighted that vanoxerine has a limited spectrum of antibacterial activity, mainly targeting the Mycobacteriales, alongside some other Gram-positive species. Contrary to previous claims (Kanvatirth et al., 2019) Table 1). The evidence provided indicates vanoxerine does not interact with AroB and it is unlikely to be the target.
Previous attempts to generate resistant mutants to vanoxerine were performed in M.
smegmatis (Kanvatirth et al., 2019), as a standard mode-of-action determination method used for anti-mycobacterial drugs, followed by whole-genome sequencing of the mutants (Abrahams and Besra, 2020). This approach was repeated in M. bovis BCG. However, during the course of this work, several attempts were made to generate resistant mutants to vanoxerine, but no mutants could be isolated, including use of a recG mutant (Batt et al., 2015). The lack of spontaneous resistance is promising for the longevity of the drug, due to the increasing levels of MDR tuberculosis (WHO, 2021), and resistance to the most recently approved antimycobacterial drugs, bedaquiline, delamanid, and pretomanid (Zumla et al., 2013).
The lack of in vivo resistance may also indicate pleotropic effects on the cell, reducing resistance development.
The retention of ethidium bromide by Mycobacteria provides strong evidence that vanoxerine inhibits efflux. Cell lysis or pore formation are unlikely to be the mechanism of action, as a faster decrease in ethidium bromide fluorescence would be the expected outcome. Efflux inhibition could occur via several mechanisms, including inhibition of ATP synthesis, disruption of membrane energetics, or direct efflux pump inhibition (Remm et al., 2022). Direct efflux pump inhibition is less likely, due to several types of efflux pumps being involved in ethidium bromide efflux (Remm et al., 2022;Johnson et al., 2020). The use of efflux inhibitors for the treatment of tuberculosis, to complement and enhance existing treatment options has been discussed in numerous papers (Remm et al., 2022;Pule et al., 2016;Gupta et al., 2014;Laws et al., 2022;Szumowski et al., 2013). These studies have focussed upon verapamil, CCCP or plant natural products (Chen et al., 2018;Pule et al., 2016;Gupta et al., 2014), however, no efflux inhibitors are currently used clinically against tuberculosis (Pule et al., 2016). This is due to either a lack of safety data, for plant natural products, or toxic effects of inhibitors on eukaryotic systems, such as CCCP (Pule et al., 2016). Verapamil has a better safety profile, but causes serious adverse effects at higher concentrations (Pule et al., 2016). In contrast, vanoxerine has passed Phase I clinical trials without safety concerns arising in health volunteers (Obejero-Paz et al., 2015). Any adverse effects from vanoxerine were in patients with underlying structural heart disease and hence could be screened out during clinical trial testing and future use (Piccini et al., 2016).
The voltage sensitive dye DiOC2 (3) is an indicator of membrane potential disruption (Chen et al., 2018;Chawla et al., 2012;Hudson et al., 2020;Li et al., 2019). Based on its use as a proxy, disruption of the electric potential (∆y) is more likely to be a mechanism of action of vanoxerine. The ∆y disruption would cause the PMF of the cell to be dissipated, interfering with the energetics of the Mycobacterial cell, and hence leading to cell death (Chen et al., 2018;Feng et al., 2015). The dissipation of the PMF also would prevent the activity of efflux pumps, as many rely on the PMF to function (Remm et al., 2022), explaining the lack of ethidium bromide efflux.
Vanoxerine is a cationic amphiphile at physiological pH, with a pKa of 8.2 and a cLogP value of 5.3. Compounds with these properties have been shown to insert into lipid membranes and uncouple the PMF in bacterial inverted membrane vesicles, while having low mitotoxicity (Chen et al., 2018;Feng et al., 2015). In addition, membrane uncouplers have previously been reported to have several mechanisms of action (Feng et al., 2015) and this study does not preclude vanoxerine also having several mechanisms-of-action. The ability to interfere with cellular energetics has been shown to kill latent M. tuberculosis (Rao et al., 2008;Manjunatha et al., 2009), hence, vanoxerine should be tested for bactericidal effects in Mycobacteria during latency.
M. tuberculosis infections are always treated using a combination therapy, both to increase treatment efficacy and reduce levels of drug resistance development (Nahid et al., 2016;Zumla et al., 2013;Berry and Kon, 2009). Any new anti-mycobacterial drugs would be used within a combination therapy and thus, need to complement drugs within the current treatment regimen or novel drugs in development. Vanoxerine showed no interactions with the current or in development drugs clofazimine, FNDR-20081 or rifampicin, based on the FICs determined, although some shifts in the MIC99 were found (Kaur et al., 2021;Gopal et al., 2013). A reduction in MIC99 may allow lower doses of drugs to be used during treatment, hence, reducing associated side effects. Potential drug interactions may be masked by the concurrent upregulation of the MmpL5 efflux pump, suggested to efflux clofazimine and FNDR-20081, and inhibition of Mycobacterial efflux by vanoxerine (Hartkoorn et al., 2014;Andries et al., 2014;Remm et al., 2022;Kaur et al., 2021). Further testing is required to evaluate whether vanoxerine could potentiate the effects of mycolic acid or arabinogalactan biosynthesis inhibitors. In contrast, vanoxerine did not alter the MIC99 of bedaquiline, which may be due to these drugs having analogous mechanisms, both disrupting the PMF and hence having a similar cellular effect (Andries et al., 2005;Feng et al., 2015).
Comparing the RNA-sequencing data to other efflux inhibitors and uncouplers has given more perspectives. Phenothiazines have been shown to target the NADH dehydrogenase II, disrupting the electron transport chain, hence stopping efflux through PMF and ATP depletion (Remm et al., 2022). Transcriptomic data has shown phenothiazines to cause an increase in the transcript levels of ndh, nuoE-G and icd1 (Dutta et al., 2010;Boshoff et al., 2004). Conversely, vanoxerine treatment did not affect the expression of ndh and icd1, while nuoE-G were all significantly downregulated, suggesting NADH dehydrogenase II is not a target of vanoxerine. In contrast, the transcriptomic data showed a high degree of similarity to the 2aminoimidazole class of compounds, which have been shown to dissipate the PMF and block the electron transport chain (Jeon et al., 2019(Jeon et al., , 2017. Treatment with both 2B8 (an 2-aminoimidazole) and vanoxerine resulted in upregulation of mprA, sigB, sigE, mmpL5, mmpL8, mmpL10, rv3160c (bcg_3184c) and rv3161c (bcg_3185c), as responses to membrane stress, increasing membrane transporter/efflux and a putative dioxygenase and its regulator, respectively (Jeon et al., 2017). In addition, both compounds caused downregulation of both the mycolic acid (fasI and fasII) and peptidoglycan biosynthesis (mur) genes (Jeon et al., 2017). The main difference was vanoxerine did not induce transcription of the propionate detoxification genes, prpC and prpD or the sigK regulon (sigK, rv0449c, mpt83, dipZ), except for mpt70. Overall, the high similarity of the transcriptional responses of Mycobacteria to vanoxerine and 2B8, provides further evidence that vanoxerine is impacting the PMF and cellular energetics (Jeon et al., 2019(Jeon et al., , 2017. Although the lipid analysis of M. smegmatis showed loss of TMM and TDM following vanoxerine treatment, the evidence suggests that vanoxerine does not directly target mycolic acid biosynthesis. This is supported by the fact that vanoxerine inhibits C. glutamicum, for which mycolic acids are not essential, with a comparable MIC99 to M. tuberculosis. Vanoxerine also had equal activity against C. glutamicum wild-type and the ∆pks mutant. It is more likely that the PMF dissipation caused by vanoxerine has an indirect effect on the PMF-dependent transporter MmpL3, which transports TMM across the inner membrane (Su et al., 2019). This indirect inhibition is in contrast to the direct MmpL3 inhibition, which has been observed for many other antimycobacterial drugs (Li et al., 2019, p.3;Degiacomi et al., 2020, p.3). The observed reduction in the expression of the mycolic acid biosynthesis genes could also be an indirect result of MmpL3 inhibition leading to the accumulation of TMM and precursors in the cytoplasm. Loss of mycolates and downregulation of mycolic acid biosynthesis genes was also observed for the 2-aminoimidazole compounds, which are known to target the PMF (Jeon et al., 2017).
In summary, vanoxerine has been confirmed as an antimycobacterial drug with the ability to disrupt the membrane potential of Mycobacteria and should be included during pre-clinical testing of novel combination therapies. Future directions of research could include confirmation of electric potential disruption in M. tuberculosis, testing vanoxerine within a macrophage infection model, or against Mycobacteria in a hypoxia-induced latent state. In addition, analogues of vanoxerine could be synthesised to increase their potency against Mycobacteria and ideally reduce their effects on other human targets.