Activity of drugs against dormant Mycobacterium tuberculosis.

Objective/background: Heterogeneous mixtures of cellular and caseous granulomas coexist in the lungs of tuberculosis (TB) patients, with Mycobacterium tuberculosis (Mtb) existing from actively replicating (AR) to dormant, nonreplicating (NR) stages. Within cellular granulomas, the pH is estimated to be less than 6, whereas in the necrotic centres of hypoxic, cholesterol/triacylglycerol-rich, caseous granulomas, the pH varies between 7.2 and 7.4. To combat TB, we should kill both AR and NR stages of Mtb. Dormant Mtb remodels lipids of its cell wall, and so lipophilic drugs may be active against NR Mtb living in caseous, lipid-rich, granulomas. Lipophilicity is expressed as logP, that is, the logarithm of the partition coefficient (P) ratio P/P. In this study, the activity of lipophilic drugs (logP>0) and hydrophilic drugs (logP⩽0) against AR and NR Mtb was measured in hypoxic conditions under acidic and slightly alkaline pHs.

Methods: The activity of drugs was determined against AR Mtb (5-day-old aerobic cells: A5) and NR Mtb (12- and 19-day-old hypoxic cells: H12 and H19) in a Wayne dormancy model of Mtb H37Rv at pH 5.8, to mimic the environment of cellular granulomas. Furthermore, AR and NR bacilli were grown for 40days in Wayne models at pH 6.6, 7.0, 7.4, and 7.6, to set up conditions mimicking the caseous granulomas (hypoxia+slightly alkaline pH), to measure drug activity against NR cells. Mtb viability was determined by colony-forming unit (CFU) counts.

Results: At pH 5.8, lipophilic drugs (rifampin, rifapentine, bedaquiline, PA-824, clofazimine, nitazoxanide: logP⩾2.14) reduced CFU of all cells (H12, H19, and A5) by ⩾2log. Among hydrophilic drugs (isoniazid, pyrazinamide, ethambutol, amikacin, moxifloxacin, metronidazole: logP⩽0.01), none reduced H12 and H19 CFUs by ⩾2log, with the exception of metronidazole. When Mtb was grown at different pHs the following Mtb growth was noted: at pH 6.6, AR cells grew fluently while NR cells grew less, with a CFU increase up to Day 15, followed by a drop to Day 40. AR and NR Mtb grown at pH 7.0, 7.4, and 7.6 showed up to 1 log CFU lower than their growth at pH 6.6. The pHs of all AR cultures tended to reach pH 7.2-7.4 on Day 40. The pHs of all NR cultures remained stable at their initial values (6.6, 7.0, 7.4, and 7.6) up to Day 40. The activity of drugs against H12 and H19 cells was tested in hypoxic conditions at a slightly alkaline pH. Under these conditions, some lipophilic drugs were more active (>5 log CFU decrease after 21days of exposure) against H12 and H19 cells than clofazimine, nitazoxanide, isoniazid, pyrazinamide, amikacin (<1 log CFU decrease after 21days of exposure). Testing of other drugs is in progress.

Conclusion: Lipophilic drugs were more active than hydrophilic agents against dormant Mtb in hypoxic conditions at pH 5.8. The Wayne model under slightly alkaline conditions was set up, and in hypoxic conditions at a slightly alkaline pH some lipophilic drugs were more active than other drugs against NR Mtb. Overall, these models can be useful for testing drug activity against dormant Mtb under conditions mimicking the environments of cellular and caseous granulomas.