Tetrahydropyrazolo[1,5-a]Pyrimidine-3-Carboxamide and N-Benzyl-6′,7′-Dihydrospiro[Piperidine-4,4′-Thieno[3,2-c]Pyran] Analogues with Bactericidal Efficacy against Mycobacterium tuberculosis Targeting MmpL3

Mycobacterium tuberculosis is a major human pathogen and the causative agent for the pulmonary disease, tuberculosis (TB). Current treatment programs to combat TB are under threat due to the emergence of multi-drug and extensively-drug resistant TB. As part of our efforts towards the discovery of new anti-tubercular leads, a number of potent tetrahydropyrazolo[1,5-a]pyrimidine-3-carboxamide (THPP) and N-benzyl-6′,7′-dihydrospiro[piperidine-4,4′-thieno[3,2-c]pyran] (Spiro) analogues were recently identified against Mycobacterium tuberculosis and Mycobacterium bovis BCG through a high-throughput whole-cell screening campaign. Herein, we describe the attractive in vitro and in vivo anti-tubercular profiles of both lead series. The generation of M. tuberculosis spontaneous mutants and subsequent whole genome sequencing of several resistant mutants identified single mutations in the essential mmpL3 gene. This ‘genetic phenotype’ was further confirmed by a ‘chemical phenotype’, whereby M. bovis BCG treated with both the THPP and Spiro series resulted in the accumulation of trehalose monomycolate. In vivo efficacy evaluation of two optimized THPP and Spiro leads showed how the compounds were able to reduce >2 logs bacterial cfu counts in the lungs of infected mice.


Introduction
Despite the existence of treatments for Tuberculosis (TB), nine million people are currently infected and one and a half million die, each year [1], [2]. The disease also represents an escalating threat for global health, with the increasing prevalence of Multi Drug Resistant (MDR) and Extensively Drug Resistant (XDR) TB strains [3]. While the misuse of current anti-tuberculars is one of the contributing driving forces behind this trend, this situation is also a consequence of the nature of the treatment: a combination of at least three different drugs with known side-effects that must be taken for 6 months or longer. This situation often leads to poor patient compliance with a consequent rise in drug resistant strains and infection relapse cases. The WHO sponsored implementation of Directly Observed Treatment Short course [DOTS] [1] in which treatment compliance is monitored by healthcare workers has been successful when appropriately implemented (cure rate .90%). Despite this, the development of new drugs with novel modes of action (MoA) for the treatment of TB would likely still be the most cost-effective way of tackling the pandemic. Specifically, any new drug should be able to shorten the duration of treatment, avoid significant drug-drug interaction with current regimens, treat MDR as well as XDR-TB patients and be competitive in terms of cost with current drugs [4].
A quick historical analysis of the most commonly applied drug discovery strategies shows how 'target based' approaches have lately dominated drug discovery with the antibacterial and antitubercular fields being no exception. These bottom-up approaches have had limited success considering the amount of investment and the number of late stage clinical assets generated [5]. In the case of TB drug discovery these attrition rates become even more alarming given the limited number of validated targets that have gone on to be exploited for drug discovery purposes. Consequently, many researchers are turning towards phenotypic screens, where promising compounds are identified in antibacterial whole cell assays. As examples of this approach, a number of molecules currently exist at different stages of late pre-clinical and clinical development, the most advanced of them being SQ-109 [6], PA-824 [7] and Bedaquiline [8]. In addition to these, a number of promising leads [9], [10], [11] and candidates [12] further validate this shift from target-based programs to phenotypic screens. The often attractive anti-tubercular properties of these molecules have also spurred interest in mode of action (MoA) studies, leading to the postulation of new hypotheses for benzothiazinones [12], bedaquiline [8], PA-824 [13], SQ-109 [14], BM212 [15], adamantyl ureas [16] and benzimidazole C215 [17]. This, in time, could lead to new target-based screening efforts that will hopefully not be affected by the historically high attrition rates in antibacterial drug discovery.
As part of our continuous efforts towards the discovery of new anti-tubercular leads, a number of potent tetrahydropyrazolo[1,5a]pyrimidine-3-carboxamide (THPP) and N-benzyl-69,79-dihydrospiro[piperidine-4,49-thieno [3,2-c]pyran] (Spiro) analogues were recently identified against Mycobacterium tuberculosis and Mycobacterium bovis BCG [11]. Herein, we describe the attractive in vitro and in vivo pharmacological profiles of both lead series and demonstrate how these molecules interfere with trehalose dimycolate (TDM) production through mutations in the essential mmpL3 gene.

Identification and description of THPP's and Spiros as anti-TB agents
As a result of GSK's continued anti-mycobacterial phenotypic screening efforts, two new chemical scaffolds, namely tetrahydropyrazolo[1,5-a]pyrimidine-3-carboxamides (THPP's) and N-benzyl-69,79-dihydrospiro[piperidine-4,49-thieno [3,2-c]pyrans] (Spiro's), were identified against M. bovis BCG (Figure 1) [11]. While in the case of THPP's, closely related analogues have been previously described in phenotypic screenings [18], the antitubercular properties of the novel Spiro hit have not been previously reported. A preliminary profiling of both hit structures (Table 1) showed how the compounds were endowed with potent and selective anti-tubercular activity. Compound 1 did not show initial signs of cytotoxicity at the maximum concentration tested where this compound was observed to be in solution. On the other hand, the Tox50 value determined for more soluble Spiro 2 was 36 mM. Interestingly, both representative molecules showed noticeable intracellular in vitro anti-tubercular activity against infected murine macrophages in addition to constant MIC values when assayed against an extensive panel of M. tuberculosis clinical isolates (MIC90). As primary issues, THPP 1 showed very low solubility in different media and Spiro 2 offered poor intrinsic clearance values in both mice and human microsomes.
In vitro TB profile -killing curves, MDR/XDR, frequency of resistant mutants The bactericidal potential of both 1 and 2 were assessed in killing rate experiments with Linezolid and Moxifloxacin as bacteriostatic and bactericidal controls, respectively ( Figure 2). Bactericidal activity against M. tuberculosis was defined as a 99.9 % reduction of the initial inoculum after seven days of incubation with the compound. The results show that both the THPP and Spiro hits were able to reduce .3 log cfu counts after one week of treatment.
The frequency of spontaneous resistant mutant generation was also assessed. In both instances the frequency was within the same order of magnitude, being 3.6610 28 mutant/cfus for compound 2 and 6.4610 28 mutant/cfus for compound 1. Both compounds were tested against M. tuberculosis H37Rv in the presence of serum and no significant shift was observed in the MIC values (data not shown). No major shifts in the in vitro MIC values were either observed against a panel of 125 different clinical isolates, including MDR and XDR strains ( Table 2).

Identification of MmpL3 as a target of THPP 1 and Spiro 2
The elucidation of the cellular target is an important step towards the validation of the inhibitor family as a suitable therapeutic anti-tubercular agent. 1 and 2 exhibited encouraging MIC values against M. tuberculosis (Table 1) and thus were selected for use in strategies towards the identification of their target. Using these values, M. tuberculosis spontaneous resistant mutants were  isolated at 106 MIC of each compound (Protocol S2). A total of 67 mutants were studied, 33 in respect of 1 and 34 for 2. The resistant mutants contained one or more mutations allowing them to survive in the presence of the inhibitor, by either affecting drug accumulation within the cell or by preventing the inhibitory impact of the drug on the target itself. Thus identification of the mutation locus within a gene could indicate the drug target. Whole-genome sequencing analysis of 6 mutants (3 from each class) detected a single nucleotide polymorphism (SNP) compared to the M. tuberculosis H37Rv wild type sequence. In all cases, this SNP was found in mmpL3 conferring a predicted amino acid alteration in the protein sequence. For 1 we observed A249P and A677V, and for 2 we found F255L and Y252C. For the remaining 61 mutants, mmpL3 was specifically sequenced and 6 new SNPs were identified (Table 3). According to TraSH analysis, mmpL3 is an essential gene [19]. As first attempt to confirm MmpL3 as the target of the two compounds and to establish cross-resistance, resistant mutants were tested against both chemical series ( Table 4).
The results clearly demonstrate that each M. tuberculosis resistant strain with a specific mutation raised against its respective inhibitor 1 and 2 was confirmed to be resistant. Despite that all mutations were located in mmpL3, no clear cross-resistance pattern was observed between THPP and SPIRO, compounds 1 and 2 (Table 4). Finally, it is also noteworthy all isolated mutants were found susceptible to classic antitubercular drugs (Moxifloxacin, Ethambutol, Linezolid).
1 and 2 treatments give rise to mutations in MmpL3 and lead to an accumulation of trehalose monomycolate (TMM) in mycobacteria Recently, spontaneous mutants resistant to BM212, SQ109, adamantyl ureas and benzimidazole C215, compounds with anti-TB activities, were shown to contain mutations mapping to mmpL3 [15], [16], [17]. Bacilli exposed to SQ109 and adamantyl ureas were shown to accumulate TMM suggesting that MmpL3 was involved in the export of TMM [14], [15]. To address the putative role of MmpL3 in TMM export, we assessed the effects of increasing concentrations of 1 and 2 on lipid metabolism in M. bovis BCG. Cultures of M. bovis BCG were grown in broth in the presence or absence of inhibitor for 8 hours, then labelled with [ 14 C]-acetate for a further 8 hours and then subjected to an organic extraction procedure. In contrast to the cultures grown in the absence of inhibitor, [ 14 C]-labelled cultures grown either in the presence of 1 and 2 showed an accumulation of TMM ( Figure 3). To rule out non-specific effects, other classes of lipids, in particular phospholipids characteristic of mycobacteria were shown to be unaffected upon exposure to 1 and 2 ( Figure 3). The accumulation of TMM by 1 and 2 suggests that biochemically they target MmpL3 and lead to a characteristic phenotypic profile as observed for other inhibitor families of MmpL3.

Lead optimization summary, new leads profile
Despite the interesting in vitro anti-tubercular profile stated in Table 1, the initial hits from both series were affected by a number of compound development liabilities (in vitro Drug metabolism and pharmacokinetics DMPK and physicochemical properties) that make difficult their progression to the available in vivo acute murine TB efficacy model. Further medicinal chemistry efforts (details to be reported elsewhere) were hence dedicated to the identification of optimized compounds that were able to retain or even improve the anti-tubercular profile while simultaneously meets good oral pharmacokinetic profile. These efforts led to the identification of 3 and 4 as new optimized lead representative structures for both chemical series ( Figure 4). Both compounds retained or even improved the previous leads in vitro potency values against H37Rv and were able to reach acceptable exposure levels in blood after a single oral administration at 50 mg/kg ( Figure 5). Cross resistance experiments with previously isolated mutants proved that the observed TB activity for both new lead compounds was target related (Table 3). Therefore, compounds 3 and 4, despite the fact that some properties still remained to be optimized (high ClogP), fulfilled our minimum requirements in terms of in vitro potency (MIC ,0.5 mM) and PK properties (AUC after oral administration 50 mg/kg .1 mg*h/mL) to be progressed to in vivo proof of concept efficacy studies in an acute TB murine infection model.

In vivo efficacy of 3 and 4 in acute TB murine model
Efficacy studies were performed as previously described [20]. Under these conditions, both compounds showed a clear bactericidal effect, demonstrating a more than 1.5 log CFU reduction with respect to bacterial burdens at start of treatment ( Figure 6). Compound 3 showed a reduction of 1.5 log and 2.2 log units in lung bacterial load when administered at recommended tolerable doses of 100 and 300 mg/kg respectively. In the case of compound 4, it showed a reduction of 2.3 and 2.7 log units in the lungs of infected mice at recommended tolerable doses of 50 and 100 mg/ kg respectively.

Discussion
The M. tuberculosis genome encodes 13 members of the MmpL (mycobacterial membrane protein, large) family [21]. Our previous studies have shown that mmpL3 specifically plays a key role in mycolic acid transport and 'gene depletion' in a conditional mutant resulted in the intracellular accumulation of TMM [19]. These studies were supported by mmpL3-inhibitor related studies [14], [15] which resulted in chemical phenotype identical to our genetic phenotype observed in a M. smegmatis conditional mutant [19]. The identification of mmpL3 as an essential membrane protein-encoding gene involved in mycolate transport opens up a new avenues for targeting this essential and under-exploited mycobacterial pathway for developing new anti-TB drugs. In this present study, following our initial hit-to-lead identification of two novel series of anti-TB compounds, THPP and Spiro, and through the isolation of spontaneous resistant M. tuberculosis mutants followed by subsequent whole genome sequencing identified a number of mutations in mmpL3 (Rv0206c) suggesting a likely mode of resistance to the THPP and Spiro agents. In all cases, the observed mutation is a single base change in mmpL3 conferring a predicted amino acid alteration in the protein sequence. The majority of the mutations identified in mmpL3 (57 out of 67) map to the first six transmembrane segments of the predicted MmpL3   (Table S1). The fact that most of the mutated amino acids are located within predicted transmembrane segments is not surprising given the hydrophobicity of the THPP and Spiro series of compounds. However, if these amino acids do interact directly with THPP and Spiro they might well inhibit MmpL3's ability to act as a TMM transporter. This is supported by our observations of THPP and Spirotreatment of M. bovis BCG leading to an accumulation of TMM. As initial step to support MmpL3 as the target of the two compounds and to establish possible cross-resistance, resistant mutants were tested against each of the respective compounds. The results in this study concluded that both series did not possess significant cross-resistance against each other, pointing potentially to a different mechanism of resistance/action.
In order to characterise the in vitro bactericidal behaviour of leads 1 and 2, the anti-tubercular kill rate of both molecules was monitored over a period of 7 days. Both compounds were able to reduce cfu counts at rates superior to Linezolid and comparable to Moxifloxacin and their behaviour could be clearly defined as bactericidal. Reassuringly, both compounds showed stable MIC values against a panel of susceptible, mono-resistant, MDR and XDR strains. While this panel was not intended to be representative of the genetic diversity within the M. tuberculosis complex, it gave an early indication of the therapeutic potential of this mode of action to tackle TB.
While a number of putative MmpL3 inhibitors with potent antitubercular in vitro activity have been previously reported [15], [16], [17], we have been able to identify and optimize two novel families of inhibitors (THPP and Spiro's) with lead representatives also capable of producing a significant efficacious response in a murine model of acute TB infection after oral administration. According to this model, the compounds are competitive when compared in terms of dosing to Moxifloxacin, an anti-tubercular drug candidate currently in Phase III clinical trials [20].
Despite the early optimisation of potency and microsomal clearance, compound 4 was found to still suffer from a high clogP value, with the consequent potential liabilities for further development [22]. In the case of compound 3, potential issues associated were the mechanisms of potential drug-drug interaction and toxicity of benzodioxolane containing compounds which have been thoroughly reported elsewhere [23], [24] and highlights the need to develop new inhibitors where the presence of this derivative is avoided. Current synthetic efforts are focused on the optimization of these parameters and will be reported in due course. The mice were kept in air-conditioned facilities with fifteen air changes per hour. Room temperature and relative humidity were 2263uC and 40-70%, respectively. The mice were accommodated in groups of up to five individuals in TecniplastH type IV cages with autoclaved dust free corncob bedding (Panlab, Barcelona, Spain). The mice were maintained under a twelve hours light/ dark period. Autoclaved tap water and irradiated pelleted diet were provided ad libitum. The compounds used in these in vivo studies were prepared as suspensions in 1% Methyl Cellulose. The antitubercular standards used in the efficacy study were: Moxifloxacin (Sequoia Research Products Ltd) prepared as solution in 20% Captisol(R)/water).

MIC determination against mycobacteria
The measurement of the Minimum Inhibitory Concentration (MIC) against M. tuberculosis strains for each tested compound was performed in 96-well flat-bottom, polystyrene microtiter plates in a final volume of 100 ml. Ten two-fold drug dilutions in neat DMSO starting at 50 mM were performed. Drug solutions were added to Middlebrook 7H9 medium (Difco) and Isoniazid (INH) (Sigma Aldrich) was used as a positive control with two-fold dilutions of INH starting at 160 mg/ml. The inoculum was standardized to approximately 1610 7 cfu/ml and diluted 1 in 100 in Middlebrook 7H9 broth (Difco). This inoculum (100 ml) was added to the entire plate but G-12 and H-12 wells were used as blank controls. All plates were placed in a sealed box to prevent drying out of the peripheral wells and incubated at 37uC without shaking for six days. A Resazurin solution was prepared by dissolving one tablet of resazurin (Resazurin Tablets

MIC against Clinical strains
The BACTEC MGIT 960 System (Becton Dickinson) was used to MIC determination in clinical isolates (Institute Carlos III and Hospital Val d'Hebron) following the manufacturer instructions.

MIC intracellular determination
Confocal automated plate reader (PE Opera) was used to assess the intracellular growth and proliferation of GFP expressing M. tuberculosis H37Rv in Raw264.7 macrophages as well as the mortality of host cells. In this experiment macrophages were infected as a batch with the batch preparation of mycobacteria (MOI = 1) for 2 hours and then treated with amikacin to remove extracellular bacilli. Infected cells were plated into 384 well microtiter plates containing compounds and incubated for 5 days. After incubation cell-permeable Syto60 was added to each well for macrophage nuclei staining. Images of cells in each well (several image fields per well) were recorded on an automated Opera platform located in a BSL3 facility. The percent of infected cells containing GFP-expressing bacteria were used as primary readout. The decrease in % of GFP containing cells was reflecting the ability of compounds to reduce bacterial growth, proliferation and infectivity. The total number of macrophages, determined by quantification of Syto60 stained nuclei, was used as a second independent readout in assay. Ten point dose titrations will be used for pharmacological characterization of compounds. The decrease of bacteria load with simultaneous increase in number of surviving host macrophages determined the minimal inhibitory concentration and effective concentration of inhibitors. Positive controls wells were used with INH and Rifampin at MIC 100 concentration and negative control wells with 1% DMSO on every plate in a screening run. These controls allowed normalizing data to 100% of inhibition on the plate-based level thus avoiding variability of results during multiple days of screening.

General antimicrobial activity assay
Whole-cell antimicrobial activity was determined by broth microdilution using the Clinical and Laboratory Standards Institute (CLSI) recommended procedure, Document M7-A7, ''Methods for Dilution Susceptibility Tests for Bacteria that Grow Aerobically''. Some compounds have been evaluated against a panel of Gram-positive and Gram-negative organisms, including Enterococcus faecium, Enterococcus faecalis, Haemophilus influenzae, Moraxella catarrhalis, Streptococcus pneumoniae, Escherichia coli and Streptococcus pyogenes. The MIC was determined as the lowest concentration of compound producing a .80 % reduction in fluorescence observed.

HepG2 cytotoxicity assay
Cytotoxicity assay was performed measuring the viability of immortalized human cell line culture (HepG2) after 48 h of incubation with a full curve concentration of the test substances in culture medium. For read out of viability of cells we used Resazurine (BDHH). Resazurin is used as an oxidation-reduction indicator that yields a colorimetric change and a fluorescent signal in response to metabolic activity. As cell grow, metabolic activity results in a chemical reduction of Resazurin indicated by a change from non-fluorescent blue to the reduced fluorescent pink form. The degree of Resazurin fluorescence is therefore, an indicator of the number of viable cells in the culture system. The cytotoxicity results were expressed as Tox50 values (IC50). This was done by calculating the concentration of test compounds that would result in a decrease in Resazurin reduction equivalent to 50% of the concurrent control values. The protocol we followed is described below briefly.
Cell line HepG2 (HB-8065) were cultured with fresh medium (Essential Minimum Eagle Medium, EMEM, supplemented with 5% fetal calf serum and 2 mM L-glutamine) the day before subculturing the plates. On the day of the assay, cells (10,000 cells/ well) were seeded in a black 96-well collagen coated microplate with clear bottom, (Becton DickinsonH) except in column 11, that was dispensed only 100 mL of culture medium.
Stock solution from each test substances was prepared in 100% DMSO. Ten serial 1:2 dilutions were prepared of each test compound and, finally, a 1:200 dilution was made, in medium, to achieve a final concentration of 0.5% of DMSO. Resazurin tablets (MerckH) were dissolved in phosphate buffer saline at a concentration of 0.0042%.
After 24 h of incubation of the cells (37uC, 5% CO 2 , 95% relative humidity), a volume of 150 mL of culture medium containing the appropriate test concentrations of the compounds dilutions were added to cells in two replicates. In column 12, only 150 mL of 0.5% DMSO was added (blank control). Then, cells were exposed to test compounds for 48 h. After that, medium was removed and Resazurin solution was added to each well and incubated for further 1.5 h. Fluorescence was measured at an excitation wavelength of 515 nm and an emission wavelength of 590 nm in a Microplate reader 1420 Multilabel HTS counter, Victor 2, (WallacH) The fluorescence value of each well is corrected by subtracting the background value (average of column 11) from the absolute value. The percentages of inhibition are calculated relatively to the DMSO control wells (average of column 12). For each compound, the average value of the duplicate samples is calculated. Data were processed using an Excel spreadsheet and GraphPad Software analysis. Tox50 values were calculated from the Sigmoidal dose-response (variable slope) curves by nonlinear regression analysis.

Microsomal fraction stability experimental procedure
Pooled mouse, rat, dog and human liver microsomes were purchased from Xenotech. Microsomes (final protein concentration 0.5 mg/ml, MgCl 2 (final concentration 5 mM) and test compound (final substrate concentration 0.5 mM; final DMSO concentration 0.5 %) in 0.1 M phosphate buffer pH 7.4 were preincubated at 37uC prior to the addition of NADPH (final concentration 1 mM) to initiate the reaction. The final incubation volume was 600 ml. Control incubation was included for each compound tested where 0.1 M phosphate buffer pH 7.4 was added instead of NADPH (minus NADPH). One control compound was included with each species. All incubations were performed singularly for each test compound. Each compound was incubated for 30 minutes and samples (90 ml) of incubate were taken at 0, 5, 10, 20 and 30 minutes. The control (minus NADPH) was sampled at 0 and 30 minutes only. The reactions were stopped by the addition of sample to 200 ml of acetonitrile:methanol (3:1) containing an internal standard. The terminated samples were centrifuged at 3700 rpm for 15 minutes at 4uC to precipitate the protein. Quantitative analysis: following protein precipitation, the samples were analyzed using specific LC-MS/MS conditions. Data analysis: from a plot of ln peak area ratio (compound peak area/internal standard peak area) against time, the gradient of the line was determined. Subsequently, half-life and intrinsic clearance were calculated using the equations below: Elimination rate constant (k) = (-gradient) Half life t 1=2 À Á min ð Þ~0:6937k Intrinsic Clearance (CLint) (ml/min/g protein) = V |0:6937t1=2 where V = Incubation volume ml/g microsomal protein.

Killing kinetics assay
Bacteria were grown at 37uC in 7H9 broth ADC Tween to midexponenetial phase and then diluted in 10 ml fresh Middlebrook 7H9 to an 56105 cfus/ml. Incubation was continued after the addition of compounds at 20X the MIC. At specified time points, aliquots of cultures were withdrawn, serially diluted in 7H9 broth Tween and plated on solid culture medium. Plates were then incubated at 37uC and CFU were counted after 3 to 4 weeks.

Polar and apolar lipid extraction of [14C]-labelled inhibitor-treated cultures of M. bovis BCG
A 5 ml culture of M. bovis BCG was grown until mid-log (OD600 0.6), then inhibitor (1 or 2) added at 36 MIC and the culture grown for a further 8 hours at 37uC. At this point, the cells were labelled with 10 mCi ml-1 [1,2-14C]acetate (57 mCi mmol-1, GE Healthcare, Amersham Bioscience), and incubated at 37uC for a further 8 hours. The [14C]-labelled cells were harvested by centrifugation followed by washing with PBS. Polar and apolar lipids were then extracted according to the procedures described by Dobson et al. [25]. Briefly, freeze-dried M. bovis BCG cells were treated in 2 ml of methanolic saline (CH3OH:0.3% NaCl, 10:1, v/v) and 2 ml of petroleum ether for 2 h [25]. The suspension was centrifuged and the upper layer containing apolar lipids was separated. An additional 2 ml of petroleum ether was added, mixed and centrifuged as described above. The two upper petroleum ether fractions were combined and dried. For polar lipids, 2.3 ml CHCl3/CH3OH/0.3% NaCl (9:10:3, v/v/v) was added to the lower aqueous phase and stirred for 4 h. The mixture was centrifuged, separated and the residue re-extracted twice with 0.75 ml of CHCl3/CH3OH/0.3% NaCl (5:10:4, v/v/v). Equal amounts of CHCl3 and 0.3% NaCl (1.3 ml of each) were added to the combined filtrates and stirred for 1 h. The mixture was centrifuged and the lower layer containing the polar lipids recovered, dried, and resuspended in 200 ml of CHCl3:CH3OH (2:1, v/v), and incorporation of [14C] acetate quantified by liquid scintillation counting using 5% of the lipid fraction in 5 ml of EcoScint A (National Diagnostics). The polar lipid extracts were examined by two dimensional thin-layer chromatography (2D-TLC) on aluminum backed plates of silica gel 60 F254 (Merck 5554) by spotting equal counts of polar lipid extracts (50, 000 cpm), which were then developed using solvent systems D and E as described [25]. System D: chloroform:methanol:water (100:14:0.8) in direction 1 and chloroform:acetone:methanol:water (50:60:2.5:3) in direction 2; and System E: chloroform:methanol:water (60:30:6) in direction 1 and chloroform:acetic acid:methanol:water (40:25:3:6) in direction 2. Lipids were visualized by 48 h exposure on X-ray films by autoradiography (Kodak Biomax MR film).

Pharmacokinetic studies
Experimental compounds were administered by oral gavage at 50 mg/kg single dose at a volume of 20 ml/kg to n = 5 mice. All mice received treatment in the fed state. Peripheral total blood was the compartment chosen for the establishment of compound concentrations: 25 ml of blood was taken from the lateral tail vein for each mouse at 15, 30 and 45 minutes, 1, 2, 4, 8 and 24 hours. LC-MS was used as the analytical method of choice for the establishment of compound concentration in blood with a sensitivity of LLQ = 1-5 ng/ml in 25 ml blood. The non-compartmental data analysis (NCA) was performed with WinNonlin Phoenix 6.3 (Pharsight, Certara L.P) and supplementary analysis was performed with GraphPad Prism 5 (GraphPad Software, Inc).

In vivo efficacy assessment
The experimental design has been previously described [20]. In brief, mice were intratracheally infected with 100.000 CFU/ mouse (M. tuberculosis H37Rv strain). Products were administered for 8 consecutive days starting one day after infection. Lungs were harvested 24 hours after the last administration. All lung lobes were aseptically removed, homogenized and frozen. Homogenates were plated in 10% OADC-7H11 medium for 14 days at 37uC. Homogenates from compound 3 and 4 treated mice were incubated for 18 days at 37uC in plates supplemented with 0.4% (wt, vol) activated charcoal (Sigma Aldrich) to prevent the effect of product carryover.

Supporting Information
Protocol S1 Compounds synthesis. This section provides a detailed description of the protocols, experimental procedures and specifics of how compounds 1-4 were obtained. It also includes information on materials used in the study and references. (DOCX) Protocol S2 Growth and generation of M. tuberculosis spontaneous resistant mutants. This section provides a detailed description of the protocols, experimental procedures and specifics of how mutants were obtained. It also includes sequencing methods.