In Vitro and In Vivo Activity of Gepotidacin against Drug-Resistant Mycobacterial Infections

ABSTRACT

Mycobacterial pathogens, including nontuberculous mycobacteria (NTM) and Mycobacterium tuberculosis, are pathogens of significant worldwide interest owing to their inherent drug resistance to a wide variety of FDA-approved drugs as well as causing a broad range of serious infections. Identifying new antibiotics active against mycobacterial pathogens is an urgent unmet need, especially those antibiotics that can bypass existing resistance mechanisms. In this study, we demonstrate that gepotidacin, a first-in-class triazaacenapthylene topoisomerase inhibitor, demonstrates potent activity against M. tuberculosis and M. fortuitum, as well as against other clinically relevant NTM species, including fluoroquinolone-resistant M. abscessus. Furthermore, gepotidacin exhibits concentration-dependent bactericidal activity against various mycobacterial pathogens, synergizes with several drugs utilized for their treatment, and reduces bacterial load in macrophages in intracellular killing assays comparably to amikacin. Additionally, M. fortuitum ATCC 6841 was unable to generate resistance to gepotidacin in vitro. When tested in a murine neutropenic M. fortuitum infection model, gepotidacin caused a significant reduction in bacterial load in various organs at a 10-fold lower concentration than amikacin. Taken together, these findings show that gepotidacin possesses a potentially new mechanism of action that enables it to escape existing resistance mechanisms. Thus, it can be projected as a potent novel lead for the treatment of mycobacterial infections, particularly for NTM, where present therapeutic interventions are extremely limited.

INTRODUCTION

Mycobacterial pathogens, including Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), and nontuberculous mycobacteria (NTM) such as M. fortuitum and M. abscessus, are some of the most formidable pathogens encountered by modern health care systems worldwide (https://www.who.int/health-topics/tuberculosis#tab=tab_1). This is usually due to their ability to cause complex, chronic infections which are refractory to chemotherapeutic intervention and their ability to easily transmit through aerosol dissemination (e.g., TB). Thus, the discovery and development of new drugs active against drug-resistant mycobacterial pathogens is an urgent, unmet critical need (1).

Fluoroquinolones (FQ) target DNA gyrase, a clinically validated, broad-spectrum drug target (2). Additionally, fluoroquinolones are extensively utilized as front-line therapy for treatment of NTM infections, while they are the second line of treatment for infections caused due to multidrug-resistant (MDR)-TB (3). Indeed, moxifloxacin (MXF), a fourth-generation fluoroquinolone, is under multiple clinical trials as a part of combination therapy to reduce the duration of antimycobacterial treatment (4). However, due to extensive clinical utilization of fluoroquinolones for treating other bacterial respiratory and serious infections, ever-increasing resistance is being clinically encountered, thus nullifying the advantage of fluoroquinolone for treating serious mycobacterial infections (5). Hence, identifying and developing novel chemical entities acting against DNA gyrase would be a welcome addition to the antibacterial chemotherapeutic arsenal, especially for treating drug-resistant mycobacterial infections.

In this context, gepotidacin (GEPO, GSK21409440) is a novel, first-in-class triazaacenapthylene topoisomerase inhibitor with a molecular weight of 448.5 that selectively inhibits bacterial type II topoisomerases by interacting with the GyrA subunit of bacterial DNA gyrase and the ParC subunit of bacterial topoisomerase IV through a unique mechanism that is not utilized by any antibiotic currently in clinical practice (6, 7). GEPO belongs to the novel bacterial topoisomerase inhibitors (NBTIs), a new class of bacterial gyrase and topoisomerase IV inhibitors that are structurally different from fluoroquinolones (6–8). However, the mycobacteria are a unique group of pathogens that lack topoisomerase IV, and its DNA gyrase performs both functions of topoisomerase II (negative supercoiling and decatenation); hence, most of dual-targeting NBTIs show poor activity against mycobacterial pathogens (9, 10). In spite of this limitation, Mycobacterium tuberculosis gyrase inhibitors (MGIs) have been reported (9), where formation of single-stranded DNA breaks is one of the characteristic features of MGIs (10). Fluoroquinolones form a double-stranded DNA cleavage complex with DNA gyrase and generate double-stranded DNA breaks; in contrast, GEPO forms a very stable single-stranded-DNA cleavage complex, leading to single-stranded DNA breaks (11). This difference in mechanisms is primarily due to GEPO’s binding site, which is close to but distinct from that of fluoroquinolones (7, 11) and which enables GEPO to exhibit potent activity against a variety of bacterial pathogens, including fluoroquinolone-resistant strains (6).

The clinical safety of GEPO has been evaluated in healthy human subjects and subjects with renal impairment in phase I clinical trials (ClinicalTrials registration no. NCT02202187, NCT02729038, respectively) by both intravenous and oral formulations and has been found to be safe (12–15). GEPO is currently under various clinical trials for the treatment of sexually transmitted infections (STI) caused by Neisseria gonorrhoeae, as well as acute bacterial skin and skin structure infections (ABSSSI) (16, 17). Despite the similarity of GEPO with MGIs (i.e., generating single-stranded DNA breaks), its activity against mycobacterial pathogens, including NTM, has not been explored yet. Here, in this study, we demonstrate the promising activity of GEPO against various clinically relevant mycobacterial pathogens, including fluoroquinolone-resistant mycobacterial pathogens.

RESULTS AND DISCUSSION

GEPO is potently active against various mycobacterial pathogens.

The MIC of GEPO was determined against a mycobacterial panel including M. tuberculosis; various clinically significant NTMs and MIC results are tabulated in Tables 1 and 2. As can be seen in Table 1, GEPO exhibits a significantly lower MIC (2 mg/L) against fast-growing NTM such as M. fortuitum ATCC 6841, M. chelonae ATCC 35752, and fluoroquinolone-resistant M. abscessus ATCC 19977, and its MIC is comparable to that of amikacin, which is one of the potent therapeutic interventions against NTM infections (18). However, the MIC of GEPO (2 to 32 mg/L) was significantly higher than that of fluoroquinolones (moxifloxacin and levofloxacin MIC range, 0.03 to 2 mg/L) against most mycobacterial pathogens; additional structural modification and/or optimization may be needed to further enhance its activity against mycobacteria.

TABLE 1.

MIC of GEPO and other comparator drugs against NTM strains

Antibiotic	MIC (mg/L)a
	Fast-growing NTMs			Slow-growing NTMs
	M. fortuitum ATCC 6841	M. chelonae ATCC 35752	M. abscessus ATCC 19977	M. avium ATCC 19698	M. gordonae ATCC 14470	M. nonchromogenicum ATCC 19530	M. kansasii ATCC 12478	M. intracellulare ATCC 13950
Gepotidacin	2	2	2	16	16	32	32	16
Amikacin	2	2	8	2	2	2	4	1
Ceftazidime	>64	>64	>64	NT	NT	NT	NT	NT
Ceftriaxone	>64	>64	>64	NT	NT	NT	NT	NT
Levofloxacin	0.06	0.12	2	NT	NT	NT	NT	NT
Moxifloxacin	0.03	0.03	2	2	1	1	1	0.5
Clarithromycin	1	1	0.5	0.25	0.25	0.25	0.25	0.12

^aNT, not tested.

TABLE 2.

MIC of GEPO and other comparator drugs against M. tuberculosis-resistant isolates

Antibiotic	MIC (mg/L)
	M. tuberculosis H37Rv ATCC 27294	INH-resistant ATCC 35822	RIF-resistant ATCC 35838	STR-resistant ATCC 35820	EMB-resistant ATCC 35837
Gepotidacin	0.5	1	0.5	0.25	0.25
Isoniazid	0.03	>64	0.03	0.03	0.03
Rifampicin	0.06	0.03	64	0.03	0.03
Streptomycin	1	1	1	>64	1
Ethambutol	1	2	2	1	>64
Levofloxacin	0.25	0.5	0.12	0.25	0.5
Amikacin	0.12	0.25	0.12	0.25	0.12

Interestingly, the MIC of GEPO against tested mycobacterial pathogens is in line with the reported MIC of GEPO against various other bacterial pathogens (≤0.06 to 2 mg/L) for which GEPO is currently being evaluated in multiple clinical trials (19). However, GEPO is much less potent against slow-growing NTM such as M. avium ATCC 19698, M. gordonae ATCC 14470, M. kansasii ATCC 12478, M. intracellulare ATCC 13950 and M. nonchromogenicum ATCC 19530 (MIC, 16 to 32 mg/L). At the same time, amikacin is almost equipotent across NTM except for M. abscessus ATCC 19977, where its activity is inferior to that of GEPO (MIC, 8 mg/L). This is significant because M. abscessus ATCC 19977 is one of the most drug-resistant mycobacterial pathogens for which there is an acute shortage of chemotherapeutic options available; thus, identification of new drugs active against it is an unmet medical need (20, 21).

When tested for in vitro anti-M. tuberculosis activity against M. tuberculosis H37Rv ATCC 27294, GEPO showed very promising activity (MIC, 0.5 mg/L), and it has equipotent activity against a panel of clinical, single-drug-resistant M. tuberculosis strains (Table 2). The MIC of GEPO against various mycobacterial strains compares very well with reported MIC against other bacterial pathogens, including those resistant to fluoroquinolones (19). The equipotent activity against multiple clinical, mycobacterial pathogens could be attributed to GEPO’s ability to bind a distinct site in bacterial DNA gyrase that is currently not being targeted by other topoisomerase II inhibitors, thus exhibiting a new mechanism of action not influenced by existing drug resistance mechanisms (10, 11).

GEPO exhibits concentration-dependent killing against mycobacterial pathogens.

To determine the killing kinetics of GEPO against various mycobacteria, GEPO was tested at 1× and 10× MIC against NTMs and M. tuberculosis along with control drugs, and kill curves were plotted in Fig. 1A to C. As shown, GEPO exhibits concentration-dependent bactericidal activity against all three mycobacterial strains tested, where its 10× MIC was bactericidal against M. fortuitum, M. abscessus, and M. tuberculosis. At 1× MIC, GEPO was bacteriostatic against M. abscessus (Fig. 1B) and M. tuberculosis (Fig. 1C) but caused a reduction of ~5 log₁₀ CFU/mL against M. fortuitum ATCC 6841.

FIG 1.

Time-kill kinetics of gepotidacin and comparator antibiotics against various mycobacterial strains. (A) M. fortuitum ATCC 6841; (B) M. abscessus ATCC 19977; (C) M. tuberculosis H37Rv ATCC 27294.

In contrast, 10× MIC of GEPO eliminated culture (~9 log₁₀ CFU/mL reduction) in 48 h with no observed regrowth compared to the untreated control (Fig. 1A). This concentration-dependent killing is comparable to that of levofloxacin, which caused a reduction of ~8.5 log₁₀ CFU/mL in 24 h at 10× MIC, while its 1× MIC caused a reduction of ~5.2 log₁₀ CFU/mL in 24 h. In contrast, fluoroquinolone-resistant M. abscessus ATCC 19977 is far less susceptible to levofloxacin (MIC, 2 mg/L), with 10× MIC of levofloxacin taking 48 h to reduce ~8.5 log₁₀ CFU/mL. In contrast, the killing kinetics of GEPO remain undisturbed even in the presence of apparent fluoroquinolone resistance (Fig. 1B). These results indicate that GEPO exhibits concentration-dependent bactericidal kinetics that is unaffected by the fluoroquinolone susceptibility of strain.

Following the same trend, GEPO exhibits concentration-dependent bactericidal activity against M. tuberculosis H37Rv ATCC 27294. The 10× MIC of GEPO caused a reduction of ~8.0 log₁₀ CFU/mL in 120 h with no regrowth afterward, whereas 1× MIC of GEPO led to a reduction of only ~1.7 log₁₀ CFU/mL in 120 h compared to the untreated control (Fig. 1C). Similar concentration-dependent killing kinetics is observed in isoniazid, where 1× MIC and 10× MIC of isoniazid lead to an ~3.6 log₁₀ CFU/mL reduction and complete elimination of culture (reduction of ~5.7 log₁₀ CFU/mL), respectively, in 24 h, as can be seen in Fig. 1C. Taken together, GEPO exhibits concentration-dependent killing kinetics similar to that of levofloxacin for NTMs and isoniazid for M. tuberculosis.

GEPO synergizes with multiple antibiotics against mycobacterial pathogens.

The treatment of complex, recalcitrant mycobacterial infections often utilizes various drug combinations, which help in suppressing the emergence of drug resistance as well as lead to a faster clearance of infection, thus improving patient compliance while simultaneously reducing drug-related toxicity (https://asm.org/Articles/2018/September/Combination-Antibiotic-Testing-When-2-Drugs-are-Be; 22). This situation demands that any new drug candidate be tested for its ability to synergize with other approved drugs. The ability of GEPO to interact and/or synergize with drugs utilized for treatment of mycobacterial pathogens was determined by the checkerboard method, which facilitated calculation of the fractional inhibitory concentration (FIC). As can be seen in Table 3, GEPO synergizes with meropenem, moxifloxacin, ciprofloxacin, and linezolid, whereas it has no interaction with vancomycin and amikacin against M. fortuitum ATCC 6841. On the other hand, GEPO synergized with moxifloxacin, ciprofloxacin, and amikacin against M. abscessus ATCC 19977. Synergistic interaction of GEPO with fluoroquinolones (moxifloxacin and ciprofloxacin) could be attributed to the fact that both the molecules target DNA gyrase, causing more pronounced gyrase inhibition and massive DNA damage leading to a lethal SOS response. Additionally, GEPO synergized with isoniazid against M. tuberculosis, whereas it did not interact with rifampicin, ethambutol, or streptomycin.

TABLE 3.

Determination of the interaction of GEPO with front-line antimycobacterial drugs against M. fortuitum ATCC 6841, M. abscessus ATCC 19977, and M. tuberculosis ATCC 27294

Strain and drugs	MIC of drug alone (mg/L)	MIC of drug in presence of gepotidacin (mg/L)	MIC of gepotidacin alone (mg/L)	MIC of gepotidacin in presence of drug (mg/L)	FIC index	Indication
M. fortuitum ATCC 6841
Amikacin	0.5	0.25	1	0.125	0.62	No interaction
Meropenem	1	0.25	1	0.06	0.31	Synergistic
Moxifloxacin	0.06	0.0075	1	0.125	0.25	Synergistic
Clarithromycin	1	0.5	1	0.25	0.75	No interaction
Ciprofloxacin	0.06	0.015	1	0.125	0.37	Synergistic
Linezolid	1	0.06	1	0.125	0.18	Synergistic
Vancomycin	0.25	0.25	1	0.06	1.06	No interaction
M. abscessus ATCC 19977
Amikacin	2	0.5	2	0.125	0.31	Synergistic
Clarithromycin	0.5	0.5	2	1	1.5	No interaction
Moxifloxacin	2	0.5	2	0.06	0.28	Synergistic
Ciprofloxacin	1	0.125	2	0.125	0.18	Synergistic
M. tuberculosis ATCC 27294
Isoniazid	0.03	0.0075	1	0.015	0.265	Synergistic
Rifampicin	0.06	0.03	1	0.5	1	No interaction
Ethambutol	1	0.5	1	0.5	1	No interaction
Streptomycin	1	0.5	1	0.25	0.75	No interaction

This observed synergy was further tested by combination time-kill analysis, and the results are plotted in Fig. 2A to C. As shown in Fig. 2A, a combination of 1× MIC of linezolid and GEPO was dramatically more effective than either linezolid and GEPO alone and caused a reduction of ~7.5 log₁₀ CFU/mL of M. fortuitum ATCC 6841 in 48 h. At the same time, combination of 1× MIC of GEPO and moxifloxacin is indistinguishable from 1× moxifloxacin alone and eliminates all culture in 24 h compared to the untreated control. The combination of 1× MIC of GEPO and meropenem was more effective than 1× MIC of meropenem alone, but it could not be distinguished from killing by 1× MIC of GEPO, and both of them reduced M. fortuitum ATCC 6841 by ~4.5 log₁₀ CFU/mL in 48 h.

FIG 2.

(A to C) Time-kill kinetics of gepotidacin in combination with (A) (i) linezolid, (ii) moxifloxacin, and (iii) meropenem against M. fortuitum ATCC 6841, with (B) (i) amikacin and (ii) moxifloxacin against M. abscessus ATCC 19977 and with (C) isoniazid against M. tuberculosis ATCC 27294.

The combination of 1× MIC of amikacin and GEPO was much more potent than 1× MIC of GEPO alone and caused a reduction of ~7.5 log₁₀ CFU/mL in 48 h for fluoroquinolone-resistant M. abscessus ATCC 19977. The combination efficacy of 1× MIC of GEPO and moxifloxacin was indistinguishable from that of 1× MIC of moxifloxacin alone and eliminated culture in 36 h, while 1×s GEPO alone did not (Fig. 2B).

In the case of M. tuberculosis, a combination of 1× MIC of isoniazid and GEPO was more effective than either isoniazid or GEPO alone and eliminated all culture (~6.5 log₁₀ CFU/mL) in 120 h, which is faster than isoniazid (144 h), with 1× MIC GEPO alone causing a reduction of ~5.5 log₁₀ CFU/mL in 168 h. No regrowth was observed for any mycobacterial pathogen under any conditions. Collectively, the studies indicate that identified combinations could potentially be utilized as a part of a multidrug therapy to treat drug-resistant, recalcitrant mycobacterial infections (Fig. 2C).

GEPO potentiates increased clearance of intracellular M. fortuitum ATCC 6841.

Since mycobacteria are significant intracellular pathogens, the ability of GEPO to clear intracellular mycobacterial infection was determined. Briefly, J774A.1 cells were infected with M. fortuitum ATCC 6841. GEPO and amikacin were added to culture at multiple MICs, followed by cell lysis and plating on Middlebrook 7H11 (MB7H11) supplemented agar to determine the remaining CFU. As can be seen in Fig. 3, GEPO at 5× MIC significantly reduced the mycobacterial load in macrophages by ~4 log₁₀ CFU/mL, whereas amikacin at 5× MIC reduced it by ~4.5 log₁₀ CFU/mL in 96 h compared to the untreated control. This demonstrates the potent ability of GEPO to control and eliminate intracellular mycobacterial growth, an effect which is comparable to that of amikacin, a frontline antimicrobial typically utilized for the treatment of serious NTM infections, thus boding well for its utilization as a potent antimycobacterial therapeutic.

FIG 3.

Intracellular activity of gepotidacin against M. fortuitum ATCC 6841 at 96 h (****, P < 0.0001). All data are presented as the mean ± standard deviation (SD).

GEPO possess a long in vitro postantibiotic effect (PAE) against M. fortuitum ATCC 6841.

Possessing a long PAE is an asset for any antibacterial molecule under consideration, as it helps in minimizing the dosages required for therapeutic clearance. In this context, GEPO exhibited a prolonged PAE of 18 h at 10× MIC and 12 h at 1× MIC, which is better than levofloxacin (12 h at 10× MIC, 6 h at 1× MIC) but comparable to amikacin (Table 4). This in vitro PAE of GEPO against M. fortuitum is comparable to earlier reports of ~3 to 12 h against Staphylococcus aureus under in vivo conditions (23) and could be a consequence of relatively more stable gyrase-DNA cleavage complex formation by GEPO (6, 7, 11). Thus, GEPO exhibits concentration-dependent bactericidal activity with prolonged PAE that is comparable to amikacin.

TABLE 4.

Determination of postantibiotic effect (PAE) of GEPO against M. fortuitum ATCC 6841

Treatment	Time taken for 1 log10 growth (h)	PAE (h)
Untreated	12	0
Amikacin 1× MIC	24	12
Amikacin 10× MIC	30	18
Levofloxacin 1× MIC	18	6
Levofloxacin 10× MIC	24	12
Gepotidacin 1× MIC	24	12
Gepotidacin 10× MIC	30	18

GEPO does not induce resistance in M. fortuitum ATCC 6841.

Considering the unrelenting rise in antimicrobial resistance (AMR), possessing a low resistance frequency is a must-have property for any investigational new anti-infective agent. When GEPO was tested for its ability to induce resistance in M. fortuitum ATCC 6841, we were unable to isolate any stable mutants resistant to GEPO. In contrast, under same conditions, M. fortuitum ATCC 6841 was able to generate stable resistance to levofloxacin and clarithromycin. The resistance frequency for GEPO was calculated to be ~10⁻⁹, which is substantially lower than the resistance frequency of levofloxacin (~10⁻⁵).

Our observed resistance frequency for levofloxacin against M. fortuitum (Table 5) is in line with the previously reported resistance frequency for ciprofloxacin (~10⁻⁵) against Escherichia coli (24). However, we did not observe any changes in the number of spontaneous, resistant mutant colonies when we extended the incubation time, as was observed by Cirz et al. (24). We think it is more likely due to the difference in method and organisms, rather than drug stability.

TABLE 5.

Frequency of spontaneous resistance mutants and mutant prevention concentration (MPC) for levofloxacin and gepotidacin against M. fortuitum ATCC 6841

Drugs	MIC (mg/L)	Mutant frequency and fold MIC						Mutant prevention concn (MPC)
		8×	16×	32×	64×	128×	256×
Levofloxacin	0.06	0.8 × 10−5	1.8 × 10−5	3.6 × 10−7	3.5 × 10−7	2.8 × 10−8	0	256× MIC (15.36 mg/L)
Gepotidacin	1	1.1 × 10−7	3.7 × 10−9	0	0	0	0	32× MIC (32 mg/L)

Additionally, we performed susceptibility assays for in vitro resistant mutant isolates of M. fortuitum recovered from levofloxacin and clarithromycin containing MB7H11 supplemented agar plates and did not observe any cross-resistance to GEPO in levofloxacin- and clarithromycin-resistant isolates (Table 6). Taking these findings together, GEPO exhibits concentration-dependent killing kinetics with potent activity against fluoroquinolone-resistant mycobacterial pathogens with a long PAE and does not induce resistance.

TABLE 6.

MIC of different antibiotics against in vitro spontaneously resistant mutant isolates of M. fortuitum recovered from agar plates containing LVX and CLR^a

Strain/isolate	Source	MIC range (mg/L) for:
		GEPO	LVX	CLR	MEM	LZD
M. fortuitum ATCC 6841	Wild type	2	0.06	1	2	2
M. fortuitum LVX resistant	Isolated from LVX-containing agar plates inoculated with wild type	2	4	0.5	4	2
M. fortuitum CLR resistant	Isolated from CLR-containing agar plates inoculated with wild type	2	0.12	>64	2	2

^aGEPO, gepotidacin; LVX, levofloxacin; CLR, clarithromycin; MEM, meropenem; LZD, linezolid.

GEPO outcompetes amikacin in a murine model of M. fortuitum ATCC 6841 infection.

Since GEPO exhibited potent activity against various mycobacterial pathogens, we proceeded to determine whether the in vitro activity of GEPO translates in vivo. To test GEPO’s in vivo activity, we utilized a murine neutropenic M. fortuitum bacteremia model, since it mimics clinical presentations of NTM infections (25); the results are plotted in Fig. 4A to C. As can be seen in Fig. 4A to C, GEPO (10 mg/kg) significantly reduced bacterial counts in lungs (~2 log₁₀ CFU/mL), spleen (~2 log₁₀ CFU/mL), and kidney (~1.2 log₁₀ CFU/mL), which is comparable to the reduction seen in lungs (~1.5 log₁₀ CFU/mL), spleen (~2 log₁₀ CFU/mL), and kidney (~1.2 log₁₀ CFU/mL) in the levofloxacin-treated (10 mg/kg) group, while amikacin (100 mg/kg) reduced bacterial counts in lungs (~1.6 log₁₀ CFU/mL), spleen (~2 log₁₀ CFU/mL), and kidney (~0.54 log₁₀ CFU/mL) at 18 dpi. This more potent activity exhibited by GEPO is quite important since it is at a 1/10 concentration of amikacin, which is well known to inflict nephrological and neurological/ototoxicity damage upon sustained administration, a hallmark of NTM treatment (26). Taken together, GEPO exhibits potent in vivo activity against M. fortuitum ATCC 6841 compared to amikacin, and at a 10-fold lower concentration.

FIG 4.

In vivo efficacy of gepotidacin against M. fortuitum ATCC 6841. Bacterial load in various organs in the murine bacteremia model. The bacterial load in various organs infected with a nonlethal dose of M. fortuitum ATCC 6841 (~5 × 10⁶ CFU) decreased significantly after treatment with gepotidacin (10 mg/kg), levofloxacin (10 mg/kg), and amikacin (100 mg/kg). (****, P < 0.0001). All data are presented as the mean ± SD.

Conclusion. Intrinsic and acquired drug resistance in mycobacteria is one of the factors responsible for a major increase in NTM infections observed worldwide (27). This is especially troubling since most NTM infections occur in immunocompromised patients, thus negatively affecting their mortality (27). Similarly, the proportion of drug-resistant TB (DR-TB) continues to increase worldwide (https://www.who.int/health-topics/tuberculosis#tab=tab_1; 5), despite the availability of an approved multidrug regimen and an approved and clinically utilized vaccine. Thus, identifying novel drugs active against mycobacterial pathogens that escape existing drug resistance mechanisms is an urgent, unmet medical need.

In this context, we have systematically explored the antimycobacterial potential of GEPO, a novel, first-in-class triazaacenapthylene molecule that inhibits bacterial type II topoisomerases by interacting with the GyrA subunit of bacterial DNA gyrase and the ParC subunit of bacterial topoisomerase IV through a unique mechanism to exhibit exquisite activity against various mycobacterial pathogens, including those resistant to fluoroquinolones. When tested in a time-kill assay, GEPO exhibits concentration-dependent bactericidal activity against both NTM and M. tuberculosis and synergizes with meropenem, moxifloxacin, ciprofloxacin, linezolid, and amikacin against NTMs and with isoniazid against M. tuberculosis, while not interacting with rifampicin, ethambutol, or streptomycin against M. tuberculosis. GEPO exhibits potent intracellular infection-clearing potential against NTMs and shows a strong PAE of 12 h, which is comparable to that of amikacin. Furthermore, M. fortuitum ATCC 6841 does not generate resistance to GEPO, while under the same conditions, it generates stable resistance to levofloxacin and clarithromycin, a cornerstone for treatment of NTM infections. Additionally, when tested in the murine neutropenic bacteremia model, GEPO is more potent than amikacin in reducing bacterial load in various organs, at a 10-fold lower concentration. This is very pertinent since GEPO has so far not demonstrated any neurological or nephrological toxicities.

Interestingly, Bulik et al. have demonstrated that GEPO exhibits an in vivo PAE of ~3 to 12 h against S. aureus, which compares very well with the in vitro PAE reported by us (23, 28). Additionally, Bulik et al. identified free-drug plasma AUC/MIC ratio as the pharmacokinetic/pharmacodynamic (PK-PD) index most associated with in vivo efficacy of GEPO, with a reported maximum concentration of drug in serum (C_max) of 4.5 to 14.3 mg/L in fasted and fed states, respectively, which is very much in agreement with the MIC against mycobacterial pathogens, including FQ-resistant M. abscessus ATCC 19977 (23).

Taken together, these data show that GEPO is a broad-spectrum antibacterial agent with a novel mechanism of action that has demonstrated a safety profile consistent with that of other approved antibiotics in phase I clinical trials. PK-PD profiling of GEPO is further warranted for treating respiratory infections caused by mycobacterial and other pathogens (19). These characteristics highlight GEPO’s stature as a promising, new antibacterial molecule targeting recalcitrant, drug-resistant mycobacterial infections with a unique mode of action that is unaffected by existing resistance mechanisms as well as its inability to induce resistance.

MATERIALS AND METHODS

Media and chemicals.

All bacterial media and supplements, including Middlebrook 7H9 broth (MB7H9 broth), Middlebrook 7H11 agar (MB7H11 agar), ADC (albumin, dextrose, and catalase) and OADC (oleic acid, albumin, dextrose, and catalase) supplements were purchased from BD (Franklin Lakes, NJ, USA). All other chemicals and antibiotics were procured from Sigma-Aldrich (St. Louis, MO, USA). GEPO was purchased from MedChemExpress, NJ, USA.

Bacterial cultures.

The bacterial strains, including drug-susceptible M. tuberculosis H37Rv ATCC 27294, isoniazid (INH)-resistant M. tuberculosis ATCC 35822, rifampicin (RIF)-resistant ATCC 35838, streptomycin (STR)-resistant ATCC 35820, ethambutol (ETB)-resistant ATCC 35837, M. fortuitum ATCC 6841, M. abscessus ATCC 19977, and M. chelonae ATCC 35752, were procured from ATCC (Manassas, USA). The mycobacteria were propagated in Middlebrook 7H9 broth supplemented with ADC and 0.05% Tween 80 at 37°C. Middlebrook 7H11 agar supplemented with 0.2% glycerol and 10% OADC plates were used for CFU enumeration.

Antibacterial susceptibility testing.

Antibacterial susceptibility testing was carried out utilizing a broth microdilution assay according to CLSI guidelines (29). Stock solutions (10 mg/mL) of test and control compounds were prepared in dimethyl sulfoxide (DMSO) and stored at −20°C. Bacterial cultures were inoculated in MB7H9 supplemented broth (MB7H9 broth supplemented with 10% ADC, 0.2% glycerol, 0.05% Tween 80) with a density of ~10⁶ CFU/mL based on the optical density at 600 nm (OD₆₀₀). The compounds were tested from 64 to 0.5 mg/L in a 2-fold serial-diluted fashion with 2.5 μL of each concentration added per well of a 96-well round bottom microtiter plate. Later, 97.5 μL of bacterial suspension was added to each well containing test compound along with controls and incubated at 37°C for 7 days for slow-growing mycobacteria and 48 h for fast-growing NTM. MIC is defined as the lowest concentration of the compound where there is no visible growth. MIC assays were performed in duplicates, and each assay was carried out independently three times.

Bacterial time-kill kinetics of GEPO against mycobacterial pathogens.

The bactericidal activity of GEPO was assessed by the time-kill method (30). M. fortuitum ATCC 6841, M. abscessus ATCC 19977, and M. tuberculosis H37Rv ATCC 27294 were diluted to achieve ~10⁶ CFU/mL in a total volume of ~0.3 mL and added to 96-well plates along with GEPO and controls at 1× and 10× MIC, followed by incubation at 37°C for 7 days for M. tuberculosis and 48 h for M. fortuitum and M. abscessus. For evaluating the reduction in CFU, a 0.03-mL sample was removed at various time points and serially 2-fold diluted in 0.27 mL normal saline, and 0.1 mL of dilution was spread on an MB7H11 supplemented agar plate (MB7H11 agar supplemented with 10% OADC and 0.2% glycerol). Then the plates were incubated at 37°C for 72 h for NTMs (M. fortuitum and M. abscessus) and 4 weeks for M. tuberculosis, and after that, colonies were enumerated. The kill curves were constructed by counting the colonies from plates and plotting the CFU/mL of surviving bacteria at each time point in the presence and absence of compound. Each experiment was performed in duplicate and repeated three times independently, and the mean data were plotted.

Synergy of GEPO with front-line antimycobacterial drugs.

Determination of the interaction of GEPO with typically utilized antibiotics for mycobacterial treatment, including amikacin, meropenem, moxifloxacin, ciprofloxacin, linezolid, clarithromycin, and vancomycin, was tested by the checkerboard method as per CLSI guidelines. Serial 2-fold dilutions of each drug were freshly prepared prior to testing. GEPO was 2-fold diluted along the ordinate (8 dilutions), while antibiotics were serially diluted along the abscissa (12 dilutions) in a 96-well microtiter plate. Then, 95 μL of ~10⁵ CFU/mL of M. tuberculosis, M. fortuitum ATCC 6841, and M. abscessus ATCC 19977 was added to each well, and plates were incubated at 37°C for 48 h for NTM and 7 days for slow-growing mycobacteria. After the incubation period was over, ΣFICs (fractional inhibitory concentrations) were calculated as follows: ΣFIC = FIC A + FIC B, where FIC A is the MIC of drug A in combination/the MIC of drug A alone, and FIC B is the MIC of drug B in combination/the MIC of drug B alone. The combination is considered synergistic when ∑FIC is ≤0.5, additive or indifferent when ∑FIC is >0.5 to 4, and antagonistic when ∑FIC is >4 (31).

Activity of GEPO against intracellular mycobacteria.

M. fortuitum ATCC 6841 was grown overnight in MB7H9 supplemented broth. The bacterial inoculum was prepared from mid-log-phase bacteria and diluted to ~10⁷cfu/mL. Then, ~10⁵ cells/well of J774A.1 cells were seeded in 6-well flat-bottom plates and infected with M. fortuitum at a 1:5 multiplicity of infection (MOI). After 4 h of infection, cells were washed twice with 1× phosphate-buffered saline (PBS), replenished with fresh RPMI medium containing GEPO and amikacin, and further incubated for 48 h. To estimate the initial count, 3 wells were lysed 4 h postinfection, serially diluted, and plated on MB7H11 supplemented agar plates. After 96 h of incubation, 3 wells were lysed, serially diluted, plated on MB7H11 supplemented agar plates, and incubated at 37°C for 72 h to estimate the CFU. Kill curves were constructed by counting the colonies from plates and plotting the CFU/mL. Each experiment was repeated three times in duplicate, and the mean data were plotted.

Determination of postantibiotic effect (PAE) of GEPO.

To determine the PAE of GEPO, M. fortuitum ATCC 6841 was diluted in MB7H9 supplemented broth to achieve ~10⁵ CFU/mL and exposed to 1× and 10× MIC of GEPO, amikacin, and levofloxacin and incubated at 37°C for 1 h. Following the incubation period, the culture was washed 2 times with prewarmed MB7H9 supplemented broth to remove any traces of antibiotics. Finally, cells were resuspended in drug-free, MB7H9 supplemented broth and incubated further at 37°C. After every 3 h, samples were taken, serially diluted, and plated on MB7H11 supplemented agar plates for enumeration of CFU. The PAE was calculated as PAE = T – C, where T is the difference in time required for a 1-log₁₀ increase in CFU after drug exposure, and C is time difference for a 1-log₁₀ increase in CFU in similarly treated drug-free control (32).

In vitro determination of resistant frequency.

Isolation of in vitro spontaneous resistance mutants of GEPO and determination of its resistance frequency were performed as reported (24, 33). M. fortuitum (~10⁸ CFU) was spread on MB7H11 supplemented agar plates containing 8 to 256× MIC of GEPO, clarithromycin, and levofloxacin, along with MB7H11 supplemented agar plates without antibiotics and incubated at 37^°C. For calculating the mutant prevention concentration (MPC), ~10¹⁰ CFU of M. fortuitum was spread on drug-containing MB7H11 supplemented agar plates (34). The plates were observed daily from the 3rd to the14th day for the appearance of colonies. The resistant colonies from these plates were streaked over fresh drug-containing MB7H11 supplemented agar plates, and the number of days taken to form colonies on fresh drug plates was compared with day of appearance of that colony in the previous step. The resistance frequency (R_f) is calculated as R_f = R_N/T_N, where R_N is number of resistant colonies on the drug plate and T_N is the CFU of inoculum (i.e., CFU on drug-free plate × dilution factor).

A similar experiment was performed on 14-day-old drug-containing MB7H11 supplemented agar plates stored at 37°C to compare the mutant generation frequency on fresh drug-containing MB7H11 supplemented agar plates with 14-day-old plates to rule out emergence of resistance due to stability issues or possible degradation of drug. Furthermore, we transferred individual resistant mutant colonies to drug-free MB7H9 supplemented broth and grew them to the mid-log phase (OD₆₀₀, 0.4 to 0.8), followed by 5 passages on drug-free MB7H9 supplemented broth. Afterward, the MIC of different antibiotics against these in vitro resistant mutants were determined as described above.

Animal experiments.

Animal experiments were performed on six 8-week-old BALB/c mice procured from the National Laboratory Animal Facility of the Central Drug Research Institute (CSIR), Lucknow, India. The Institutional Animal Ethics Committee approved the experimental protocols.

In vivo evaluation of GEPO in a murine neutropenic bacteremia model of M. fortuitum.

Since the majority of NTM infections occur under immunocompromised conditions, we used a murine neutropenic M. fortuitum infection model to mimic clinical conditions (25). Briefly, 6-week-old (22 to 24 g) male BALB/c mice were randomly divided into 4 groups: untreated control (nine mice, group 1), GEPO-treated (six mice, group 2), levofloxacin-treated (six mice, group 3), and amikacin-treated (six mice, group 4). Neutropenia was induced in mice by intraperitoneal (i.p.) doses of cyclophosphamide, 150 mg/Kg body weight, administered 4 days and 1 day before infection. Infection in mice was initiated by an intravenous (i.v.) injection of ~5 × 10⁶ CFU M. fortuitum ATCC 6841 in the tail vein. The infected mice were treated daily with a single dose of GEPO (orally, 10 mg/Kg in 5% acacia gum solution) in group 2, levofloxacin (orally, 10 mg/kg in 5% acacia gum solution) in group 3, and amikacin (i.p., 100 mg/Kg in PBS) in group 4 for 15 days, while the control mice in group 1 were given a placebo (i.p., orally, 5% acacia gum solution in water and/or PBS). On the 18th day postinfection, all mice were sacrificed, various organs were resected and homogenized, and the bacterial burden was enumerated by plating on MB7H11 supplemented agar plates. Kill curves were constructed by counting the colonies on the plates and plotting the CFU/mL. Each experiment was done in duplicate, and the CFU of individual mice were plotted.

Statistical analysis.

Statistical analysis was performed using GraphPad 8.0 software (GraphPad Software, La Jolla, CA, USA). Comparison between three or more groups was analyzed using one-way analysis of variance (ANOVA), with post hoc Tukey’s multiple-comparison test. P values of <0.05 were considered significant.

Ethics statement.

The use of mice for infectious studies (IAEC/2019/114) was approved by the Institutional Animal Ethics Committee at CSIR-CDRI, Lucknow.