The 7-phenyl benzoxaborole series is active against Mycobacterium tuberculosis

We identified a series of novel 7-phenyl benzoxaborole compounds with activity against Mycobacterium tuberculosis. Compounds had a range of activity with inhibitory concentrations (IC90) as low as 5.1 μM and no cytotoxicity against eukaryotic cells (IC50 > 50 μM). Compounds were active against intracellular mycobacteria cultured in THP-1 macrophages. We isolated and characterized resistant mutants with mutations in NADH dehydrogenase (Ndh) or the regulatory protein Mce3R. Mutations suggest that Ndh may be the target of this series.


A B S T R A C T
We identified a series of novel 7-phenyl benzoxaborole compounds with activity against Mycobacterium tuberculosis. Compounds had a range of activity with inhibitory concentrations (IC 90 ) as low as 5.1 μM and no cytotoxicity against eukaryotic cells (IC 50 > 50 μM). Compounds were active against intracellular mycobacteria cultured in THP-1 macrophages. We isolated and characterized resistant mutants with mutations in NADH dehydrogenase (Ndh) or the regulatory protein Mce3R. Mutations suggest that Ndh may be the target of this series. Tuberculosis (TB) represents a serious threat to global health as one of the top 10 causes of death worldwide with increasing rates of new infections [1]. Current treatments suffer from long treatment time and a growing problem of multidrug resistance. There is therefore a need for new molecules with anti-tubercular activity as well as new targets for the development of treatments.
The combination of two privileged structures [2], the biphenyls and the benzoxaboroles, identified a novel 7-phenyl benzoxaborole series with activity against Mycobacterium tuberculosis in vitro. Three compounds with a 7-phenyl benzoxaborole structure were tested for activity against Mycobacterium tuberculosis ( Fig. 1 and Supplementary Information). Compounds had activity in both liquid and solid medium (Table 1). We determined inhibitory concentrations under aerobic conditions as previously described [3]. Compounds were solubilized in DMSO and tested as 10-point two-fold serial dilutions. Bacteria were cultured in Middlebrook 7H9 medium supplemented with 10% v/v OADC (oleic acid, albumin, dextrose, catalase) and 0.05% w/v Tween-80. Assay plates were inoculated with M. tuberculosis H37Rv LP (ATCC 25618) [4] and incubated for 5 days at 37°C. Growth was measured using OD 590 and plotted using the Levenberg-Marquart algorithm. The IC 90 was defined as the concentration at which 90% growth was inhibited. All three compounds were active against replicating bacteria. AN11987 had good activity, with an IC 90 of 5.1 μM (Table 1). AN6288 and AN6291 had less activity, with IC 90 of 80 μM and 55 μM respectively (about 10-fold less active). Compounds were also active on solid medium. We determined the MIC 99 (defined as the concentration required to prevent 99% growth) for each compound on Middlebrook 7H10 containing 10% v/v OADC. Again, AN11987 was the most active (MIC 99 = 5 μM) as compared to AN6288 and AN6291 (MIC 99 = 12.5 μM), although the difference was less pronounced (Table 1).
We also measured compound activity against intracellular bacteria. We infected the THP-1 human monocytic cell line (ATCC-TIB202). Briefly THP-1 cells were maintained in RPMI-1640, 10% v/v FBS, 2 mM Corning glutagro, 1 mM sodium pyruvate and differentiated using 80 nM phorbol 12-myristate 13-acetate. Cells were infected at a multiplicity of infection of 1 with M. tuberculosis constitutively expressing luciferase [5], washed to remove extra-cellular bacteria and seeded into 96-well plates containing compounds. Infected cells were cultured for 72 h and bacterial viability assessed by RLU. The cytotoxicity of compounds was determined against uninfected THP-1 cells exposed to compounds for 72 h; cell viability was measured using CellTiter-Glo (Promega). For both assays, curves were fitted using the Levenberg-Marquardt algorithm. The IC 50 and IC 90 were defined as the concentrations that reduced cell/bacterial viability by 50% or 90% respectively. All three compounds showed good intracellular activity with no cytotoxicity (IC 50 > 50 μM) ( Table 1). The activity trend seen in axenic culture was preserved against intracellular bacteria, with AN11987 being the most active, IC 90 = 1.8 μM AN6288 was the least active, with IC 90 = 30 μM. Although the target for the 3-aminomethyl benzoxaboroles in M. tuberculosis is leucyl-tRNA synthetase (LeuRS) [6] the target and/or mechanism of resistance for these 7-phenyl benzoxaboroles compounds is unknown. We isolated resistant mutants on solid medium [7]. M. tuberculosis was plated onto solid medium containing 5X or 10X the MIC 99 . Colonies were isolated and confirmed for resistance in solid or liquid medium. We isolated and confirmed resistant mutants for AN6291 (7 mutants), AN6288 (2 mutants) and AN11987 (3 mutants). We selected three mutants resistant to AN6291 and two mutants resistant to AN6288 for whole genome sequencing.
Whole genome sequencing and mutation identification was conducting as described in Ioerger et al. 2010 [4] using an Illumina HiSeq 2500 and either 72 × 72bp paired-end reads or 54 × 54 paired-end reads. We identified a number of single nucleotide polymorphisms (SNPs) and a large deletion. Resistant mutants raised against AN6288 contained mutations within ndh, while resistant mutants raised against AN6291 contained mutations affecting Mce3R including an amino acid substitution, a stop codon and a large deletion (Table 2). Mutations were confirmed by PCR amplification and sequencing from genomic DNA. We sequenced ndh and mce3R in the remaining isolates. All the strains isolated against AN6291 had mutations in mceR3, whereas the strains isolated against AN6288 and AN11987 had mutations in ndh ( Table 2).
We tested selected strains for cross-resistance; MICs were determined on solid medium using a modified version of the MIC 99 , using 96-well plates and a smaller inoculum to conserve compound. Again, we saw the same trend of activity with AN11987 being most active (MIC = 0.2 μM) ( Table 3). Cross-resistance to all three compounds was observed in strains containing either ndh or mce3R mutations.
Ndh is an essential oxidoreductase, which catalyzes electron transfer from NADH to menoquinone as part of the electron transport chain [8]. Ndh is the target of several small molecule series including the phenothiazines [9] and quinolinyl pyrimidines [10]. Mutations in Ndh found in clinical isolates have also been associated with resistance to isoniazid (R13C, V18A, T110A, A226E, R268H, G313R and N316K) [11][12][13]. The mutations we identified (P223T, F421S, V332 M and A333V) have not previously been reported. However, alignment of M. tuberculosis Ndh with Staphylococcus aureus (pdb 4XDB) and Caldalkalibacillus thermarum (pdb 4NWZ) structures revealed that V332 and A333 correspond to residues involved in the quinone binding pocket [14,15]. V332 (T315 in C. thermarum and T318 in S. aureus) is involved in NADH binding while A333 is (A316 in C. thermarum and A319 in S.   Table 2 Mutations identified in resistant isolates. Strains were isolated on 5X or 10X MIC on solid medium and were confirmed as resistant by determining MICs. AN6291-RM1 had a large deletion encompassing mce3R. Mutations were identified by whole genome sequencing in five strains (indicated by an asterisk). Mutations in the remaining strains were identified by PCR amplification and sequencing of the gene.
Compound Strain Gene(s) Mutation Amino Acid change aureus) is part of the conserved AQxAxQ motif common amongst NDH-2 analogs. Given that these Ndh mutations conferred resistance to all three compounds, we propose that Ndh is likely the target of this compound series, and that mutations lead to loss of binding. Further work to express, purify and co-crystallise the protein would address this question.
Mce3R is a TetR transcriptional repressor that negatively regulates genes involved in lipid metabolism and redox reactions [16]. The Mce3R mutations were inactivating mutations, including premature stop codons, a frameshift and a large deletion covering mce3R (Table 2). This suggests Mce3R might control genes involved in resistance, since deletion of Mce3R activity would lead to upregulation of regulon members. Genes controlled by Mce3R include a putative epoxide hydrolase (Rv1938), a monooxygenase (Rv1936), a methyltransferase (Rv1498c) and a succinate-semialdehyde dehydrogenase (Rv0234c) as well as several synthetases and conserved hypotheticals of unknown function. Since MceR3 is a regulatory protein, we propose that a regulon member might be involved in compound degradation or metabolism. It is interesting to note that the addition of 6-methyl group to AN6291 yielded a more potent compound, while the mutations all mapped to ndh not mce3R, which suggest that this methyl group might interfere with this resistance mechanism.
In conclusion, we identified a new benzoxaborole series, the 7phenyl benzoxaboroles, that had activity against M. tuberculosis in vitro and may target Ndh rather than LeuRS. The 7-phenyl benzoxaboroles are effective against M. tuberculosis in axenic culture and in macrophage culture, with no cytotoxicity. We characterized two routes to resistance, one of which is via gene regulation (Mce3R) and the other in Ndh. It is possible that Ndh represents the target of the 7-phenyl benzoxaborole compounds while Mce3R has some effect on the target or on the compounds through activity of a gene within its regulon. We propose that this is a series with potential for further development and target validation work.
We thank James Ahn and Yulia Ovechkina for technical assistance.