The efflux pump inhibitor timcodar improves the potency of antimycobacterial agents

Abstract

Previous studies indicated that inhibition of efflux pumps augments tuberculosis therapy. In this study, we used timcodar (formerly VX-853) to determine if this efflux pump inhibitor could increase the potency of antituberculosis (anti-TB) drugs against Mycobacterium tuberculosis in in vitro and in vivo combination studies. When used alone, timcodar weakly inhibited M. tuberculosis growth in broth culture (MIC, 19 μg/ml); however, it demonstrated synergism in drug combination studies with rifampin, bedaquiline, and clofazimine but not with other anti-TB agents. When M. tuberculosis was cultured in host macrophage cells, timcodar had about a 10-fold increase (50% inhibitory concentration, 1.9 μg/ml) in the growth inhibition of M. tuberculosis and demonstrated synergy with rifampin, moxifloxacin, and bedaquiline. In a mouse model of tuberculosis lung infection, timcodar potentiated the efficacies of rifampin and isoniazid, conferring 1.0 and 0.4 log10 reductions in bacterial burden in lung, respectively, compared to the efficacy of each drug alone. Furthermore, timcodar reduced the likelihood of a relapse infection when evaluated in a mouse model of long-term, chronic infection with treatment with a combination of rifampin, isoniazid, and timcodar. Although timcodar had no effect on the pharmacokinetics of rifampin in plasma and lung, it did increase the plasma exposure of bedaquiline. These data suggest that the antimycobacterial drug-potentiating activity of timcodar is complex and drug dependent and involves both bacterial and host-targeted mechanisms. Further study of the improvement of the potency of antimycobacterial drugs and drug candidates when used in combination with timcodar is warranted.

FIG 1

Codosing with TIM potentiates the activity of RIF in a high-dose, acute infection model of tuberculosis after 4 weeks of treatment. (A) Dose-response of TIM administered orally at 10, 50, 100, or 200 mg/kg/day with or without RIF at 10 mg/kg/day. TIM was dosed 6 h before RIF. Mice were infected with 2.0 × 10⁶ CFU of M. tuberculosis Erdman per mouse and randomly assigned to groups of six mice each. Data for the early control and late control groups are shown. (B) Results for TIM at 200 mg/kg/day twice a day (TIM bid; left) and for RIF at 10 mg/kg/day once per day (RIF), RIF at 10 mg/kg/day once per day plus TIM at 200 mg/kg once per day (RIF + TIM qd), or RIF at 10 mg/kg/day once per day plus TIM at 200 mg/kg twice per day (RIF + TIM bid) (right) are shown. Scatter plots of the log₁₀ number of CFU recovered from the lungs of infected mice are shown, and statistical significance is noted on the graphs as 0.05 > P > 0.01 (*) or 0.01 > P > 0.001 (**).

FIG 2

TIM in combination with BDQ in a high-dose, acute infection model of tuberculosis with 8 days of treatment. BDQ (25 mg/kg/day) and RIF (10 mg/kg/day) were administered alone or in combination with TIM (200 mg/kg/day). TIM was dosed 6 h before BDQ and RIF. Results for untreated early controls at 1 day postinfection and late controls infected with M. tuberculosis Erdman at 10⁶ CFU are shown separately. Scatter plots of the log₁₀ number of CFU recovered from the lungs of infected mice are shown, and statistical significance is noted on the graphs as P values of <0.05 (*).

FIG 3

TIM reduced bacterial relapse after 12 weeks of treatment (Trmt) in a long-term, chronic infection model of M. tuberculosis infection. RIF at 10 mg/kg/day and INH at 25 mg/kg/day were dosed with TIM at 200 mg/kg/day, and mice were treated for either 9 weeks or 12 weeks after infection with 10² CFU of the M. tuberculosis Erdman isolate. TIM was dosed 6 h before RIF and INH were dosed; RIF and INH were dosed together. Separate groups of six mice each were used during each phase of treatment and were then left untreated for an additional 8 weeks during an observation phase (Obs) to determine the extent of relapse of infection. Scatter plots of the log₁₀ number of CFU recovered from the lungs of infected mice are shown, and statistical significance is noted on the graphs as P values of <0.05 (*).