Epetraborole Is Active against Mycobacterium abscessus

ABSTRACT Benzoxaboroles are a new class of leucyl-tRNA synthetase inhibitors. We recently reported that the antitubercular 4-halogenated benzoxaboroles are active against Mycobacterium abscessus. Here, we find that the nonhalogenated benzoxaborole epetraborole, a clinical candidate developed for Gram-negative infections, is also active against M. abscessus in vitro and in a mouse model of infection. This expands the repertoire of advanced lead compounds for the discovery of a benzoxaborole-based candidate to treat M. abscessus lung disease.

M ycobacterium abscessus lung disease is notoriously difficult to treat due to the bacterium's high intrinsic drug resistance (1,2). In addition to resistance to all first-line tuberculosis (TB) drugs, M. abscessus displays resistance to macrolides (3,4), threatening the current macrolide-based treatment regimens (2,5). Therefore, new antibiotics with novel targets and mechanisms of action are needed to treat this disease (6).
Benzoxaboroles are a class of boron-heterocyclic antimicrobials that target leucyl-tRNA synthetase (LeuRS) (7). Acting through the oxaborole tRNA-trapping (OBORT) mechanism (8), these compounds form adducts with uncharged tRNA Leu molecules that subsequently bind to the LeuRS editing domain, blocking protein synthesis. Following the discovery of tavaborole (7,8), a benzoxaborole with antifungal activity, this compound class was optimized for antibacterial activity. Addition of a 3-aminomethyl group to the benzoxaborole core improved interactions with the editing domain of Escherichia coli LeuRS, while a 7-Opropanol substituent added a novel interaction with the phosphate backbone of tRNA Leu (9). Combining these modifications yielded epetraborole (Fig. 1), a clinical candidate with potent activity against a broad range of Gram-negative bacteria (9,10). The subsequent addition of a 4-halogen group (particularly Cl or Br) improved antituberculosis activity (11)(12)(13).
Recently, we reported that the antituberculosis 4-halogen benzoxaborole EC/11770 ( Fig. 1) is active against M. abscessus in vitro and in vivo in a mouse infection model (14). Here, we asked whether the anti-Gram-negative, nonhalogenated benzoxaborole epetraborole ( Fig. 1) is active against M. abscessus. We first measured the MIC of this compound against our screening strain M. abscessus subsp. abscessus Bamboo (15) in Middlebrook 7H9 medium using 96-well plates, as previously described (14). Surprisingly, epetraborole showed activity similar to that of the antitubercular EC/11770 (Table 1). Epetraborole retained activity against culture collection reference strains for each of the three subspecies of the M. abscessus complex and a panel of M. abscessus clinical isolates (16,17) ( Table 1). Taken together, the anti-Gram-negative, nonhalogenated benzoxaborole epetraborole was active against the M. abscessus complex in vitro.
To confirm that epetraborole indeed exerts its antimycobacterial activity by targeting M. abscessus LeuRS, we selected for epetraborole-resistant M. abscessus mutants ( Table 2).
Adapting our previously described method (14), M. abscessus ATCC 19977 culture was plated on Middlebrook 7H10 agar containing 16.5mM epetraborole, the lowest concentration suppressing the emergence of wild-type colonies. After 5 days of incubation, apparent resistant colonies were confirmed by restreaking on epetraborole-containing agar. Based on two independent selections, we calculated the frequency of resistance to epetraborole to be 5.4 Â 10 28 /CFU. This frequency of resistance was on the lower end of a range determined for epetraborole in several Gram-negative bacterial species (3.8 Â 10 28 /CFU to 8.1 Â 10 27 /CFU) (9) and was comparable to what we reported for EC/11770 in M. abscessus (3.9 Â 10 28 /CFU) (14). MIC profiling of nine epetraborole-resistant mutants (RM1 to 29) showed high-level resistance to epetraborole ( Table 2). Sequencing of leuS (MAB_4923c) showed that RM1 to 29 all had missense mutations in the LeuRS editing domain (residues V292 to K502) ( Table 2). These results suggest that epetraborole retains LeuRS as its target to exert its anti-M. abscessus activity (8,9). Development of epetraborole for the treatment of complicated urinary tract infections caused by Gram-negative bacteria was discontinued after rapid emergence of drug resistance in a phase II clinical trial (18). Determination of spontaneous resistance frequencies for epetraborole in the current study, and for EC/11770 previously (14), suggest low propensity for the development of resistance against benzoxaboroles in M. abscessus. However, it is to note that we needed to carry out selection of resistant mutants on agar containing high concentrations of the drugs (50 to 100Â broth MIC), as lower concentrations did not suppress outgrowth of wild-type bacteria. Thus, it cannot be excluded that the spontaneous resistance frequency of M. abscessus against the benzoxaboroles would be higher than the observed 4 Â 10 28 to 5 Â 10 28 /CFU when lower drug concentrations could be used. Such resistant strains, presumably displaying  low level resistance, would have been missed in our selection experiments. In any case, given the use of multidrug chemotherapy in M. abscessus treatment (2,5), the risk of benzoxaborole resistance emerging in this bacterium would be reduced significantly.
To determine whether epetraborole is active against M. abscessus in vivo, we evaluated the efficacy of this compound in a previously established murine model of M. abscessus infection (17). All experiments involving live animals were approved by the Institutional Animal Care and Use Committee of the Center for Discovery and Innovation, Hackensack Meridian Health. NOD SCID mice were infected intranasally with M. abscessus K21. At day 1 postinfection, the lung bacterial burden of the mice reached ;10 6 CFU ( Fig. 2A). Beginning on day 1, clarithromycin (formulated in 0.5% carboxymethyl cellulose-0.5% Tween 80-sterile water), epetraborole (formulated in sterile phosphate-buffered saline [PBS]), or vehicle (sterile PBS) was administered by oral gavage once per day for 10 days. Based on a previous efficacy study using a Pseudomonas aeruginosa mouse infection model (9), epetraborole was administered at 150 and 300 mg/kg body weight. The lung bacterial burden remained unchanged in mice that received the drug-free vehicle control ( Fig. 2A, day 11). Mice that received epetraborole at 300 mg/kg showed a statistically significant 1-log reduction in lung CFU that was comparable to that after treatment with clarithromycin at 250 mg/kg ( Fig. 2A). A similar pattern of CFU reduction was observed in the spleen (Fig. 2B). Thus, epetraborole was active against M. abscessus in vivo. It is interesting to note that epetraborole, despite having similar in vitro activity as the previously characterized benzoxaborole EC/11770 (Table 1) (14), required with 300 mg/kg a 30-fold higher dosing to achieve a similar (;10-fold) reduction in bacterial lung burden. The basis for this difference remains to be determined but may be due to differences in the pharmacokinetic properties of the two compounds, including oral bioavailability (9,14).  In conclusion, we show that epetraborole, an advanced nonhalogenated 3-aminomethyl benzoxaborole developed for Gram-negative infections, is also active against M. abscessus in vitro and in a mouse model of infection. This agrees with a recent publication that identified epetraborole in a screen of the MMV pandemic response box for anti-M. abscessus activity and reported this compound's efficacy against M. abscessus in a zebrafish infection model (19). Our findings reaffirm leucyl-tRNA synthetase as an attractive target against M. abscessus and expand the repertoire of advanced lead compounds for the discovery of a benzoxaborole-based candidate for the treatment of M. abscessus lung disease.