The Promises and Limitations of N-Acetylcysteine as a Potentiator of First-Line and Second-Line Tuberculosis Drugs

N-Acetylcysteine (NAC) is most commonly used for the treatment of acetaminophen overdose and acetaminophen-induced liver injury. In patients infected with Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), NAC is given to treat hepatotoxicity induced by TB drugs.

ABSTRACT

N-Acetylcysteine (NAC) is most commonly used for the treatment of acetaminophen overdose and acetaminophen-induced liver injury. In patients infected with Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), NAC is given to treat hepatotoxicity induced by TB drugs. We had previously shown that cysteine, a derivative of NAC, potentiated the activity of isoniazid, a first-line TB drug, by preventing the emergence of INH resistance and persistence in M. tuberculosis in vitro. Here, we demonstrate that, in vitro, NAC has the same boosting activity with various combinations of first- and second-line TB drugs against drug-susceptible and multidrug-resistant M. tuberculosis strains. Similar to cysteine, NAC increased M. tuberculosis respiration. However, in M. tuberculosis-infected mice, the addition of NAC did not augment the activity of first- or second-line TB drugs. A comparison of the activity of NAC combined with TB drugs in murine and human macrophage cell lines revealed that studies in mice might not be recapitulated during host infection in vivo.

INTRODUCTION

Tuberculosis (TB), a disease caused by the bacillus Mycobacterium tuberculosis, remains a serious health threat worldwide. The World Health Organization (WHO) estimated that, in 2019, 10 million people were infected with M. tuberculosis and 1.4 million people died from TB (1, 2). Despite the availability of a vaccine and chemotherapy, TB remains one of the most lethal infectious diseases. Treatment of drug-susceptible TB is complex requiring the use of four drugs—isoniazid (INH), rifampin (RIF), ethambutol, and pyrazinamide—for 2 months, followed by 4 months on INH and RIF. This long therapy ironically called “short-course chemotherapy” often leads to patients dropping out as they are feeling better or due to adverse side effects of the drugs. This is one of the problems leading to the emergence of drug resistance, which is a major hindrance for the control of TB worldwide. In 2019, drug-resistant TB was detected in close to a quarter million people, a 10% increase from the previous year (1).

While drug-susceptible M. tuberculosis infection is treated with the combination of four drugs described above for 6 months, treatment of a patient infected with drug-resistant M. tuberculosis is more complex and requires the use of a combination of second-line TB drugs for up to 2 years. Second-line TB drugs are classified into four groups. The first three groups include group A (fluoroquinolones), group B (injectable), and group C (other core second-line TB drugs) (3). One drug from each group is used to treat multidrug-resistant (MDR) TB. In the past 10 years, novel molecules have been discovered as potent anti-TB drugs and are now included in a fourth group, group D (add-on agents), of second-line TB drugs (3). Not only is the treatment of MDR TB very long, with a success rate of only 57% (1), it also uses a combination of drugs that have serious adverse reactions, such as liver, gastrointestinal, visual, central nervous system, or cardiac toxicity, to cite a few (4).

Along with drug resistance, drug persistence is an underrated problem facing TB control. In contrast to drug resistance, wherein M. tuberculosis will develop a genetic modification to become resistant to a drug, a drug persister organism is still genetically drug sensitive but does not respond to the killing effect of the drug anymore. This is well illustrated in vitro with the first-line TB drug INH. In a culture of M. tuberculosis, INH kills 99 to 99.9% of the bacteria within 3 to 4 days, but then the killing kinetics slow down. This is when the drug persisters are observed (5). From this drug persister population emerges a drug-resistant population (5). We had previously observed that the addition of cysteine to INH-treated M. tuberculosis resulted in sterilization of M. tuberculosis cultures and prevented the emergence of INH resistance and INH persisters in vitro (6). We demonstrated that this sterilization phenotype relied on an increase in M. tuberculosis respiration, which kept M. tuberculosis in an active metabolic state where the bacilli were still susceptible to the effect of the drug and unable to enter into a persistent state (6). Since cysteine is toxic to eukaryotic cells, we were unable to validate these in vitro results with in vivo experiments. N-acetylcysteine (NAC) is an analog of cysteine used in TB patients for the treatment of TB drug-induced toxicity (7). NAC was also tested as an adjuvant to DOTS (directly observed treatment, short course) and shown to significantly cause early sputum negativity and radiological improvements in pulmonary TB patients (8).

We had previously assayed NAC in combination with INH and found the addition of NAC to INH-treated M. tuberculosis cultures resulted in a similar M. tuberculosis death kinetics to the combination INH/cysteine, although sterilization of the M. tuberculosis culture was not observed (6). We decided to test whether NAC could potentiate a combination of TB drugs since TB treatment is a multidrug therapy. Here, we report the adjunct activity of NAC combined with first- or second-line TB drugs in cultures of M. tuberculosis in vitro, in M. tuberculosis-infected macrophage-like cell lines, and in M. tuberculosis-infected mice.

RESULTS

NAC potentiates the activity of first-line and second-line TB drugs in vitro against drug-susceptible and drug-resistant M. tuberculosis strains.

In vitro treatment of M. tuberculosis H37Rv with the two first-line TB drugs INH (7 μM) and RIF (1 μM) resulted in sterilization of the culture within 28 days (Fig. 1A). The addition of NAC (4 mM) to the INH/RIF treatment sterilized the M. tuberculosis culture within 21 days (Fig. 1A). We set as a benchmark for comparison between groups the time required for the drug combination to kill 10⁶ M. tuberculosis bacilli. The INH/RIF/NAC combination reached that benchmark 10 days earlier (11 days) than the INH/RIF combination (21 days). Next, the adjunct activity of NAC was assessed in combination with second-line TB drugs. Treatment of drug-resistant TB involves the use of multiple drugs composed of at least one fluoroquinolone (group A), one injectable (group B) and one other second-line TB drug (group C or D). The combinations tested were: (i) ofloxacin (group A) plus kanamycin (group B) plus ethionamide (group C) (OKE) (Fig. 1B and D) and (ii) moxifloxacin (group A) plus amikacin (group B) plus clofazimine (group D) (MAC) (Fig. 1C, E, and F). These combinations were tested against the drug-susceptible M. tuberculosis strain H37Rv (Fig. 1B and C), the MDR strain V2475 (9) (Fig. 1D and E) and the extensively drug-resistant (XDR) strain TF275 (9) (Fig. 1F). TF275, an XDR strain from Kwa-Zulu Natal, South Africa, was resistant to ofloxacin, kanamycin, and ethionamide but fully susceptible to clofazimine and showed low-level resistance to moxifloxacin and amikacin (4-fold higher MIC compared to the drug-susceptible, Kwa-Zulu Natal strain V4207 [9]). Therefore, only the MAC combination was tested against TF275. In both instances, the benchmark was reached quicker with the addition of NAC to the drug combinations: 10 days quicker for OKE/NAC and 5 days quicker for MAC/NAC. This increased killing of M. tuberculosis was significant (P < 0.02) when NAC was added to INH/RIF or OKE treatment of drug-susceptible and MDR M. tuberculosis stains, but not in the combination with MAC except for the XDR strain.

FIG 1.

Adjunct activity of NAC in vitro. The drug-susceptible M. tuberculosis H37Rv was treated at day 0 with NAC (4 mM), INH (7 μM)/RIF (1 μM), or INH/RIF/NAC (A); ofloxacin (O, 14 μM)/kanamycin (K, 41 μM)/ethionamide (E, 150 μM) with or without NAC (4 mM) (B); or moxifloxacin (M, 13 μM)/amikacin (A, 8 μM)/clofazimine (C, 11 μM) with or without NAC (4 mM) (C). The MDR strain V2475 was treated at day 0 with ofloxacin (O, 14 μM)/kanamycin (K, 41 μM)/ethionamide (E, 150 μM) with or without NAC (4 mM) (D) or moxifloxacin (M, 13 μM)/amikacin (A, 8 μM)/clofazimine (C, 11 μM) with or without NAC (4 mM) (E). (F) The extensively drug-resistant strain TF275 was treated at day 0 with moxifloxacin (M, 13 μM)/amikacin (A, 8 μM)/clofazimine (C, 11 μM) with or without NAC (4 mM). At the indicated times, aliquots were removed, serially diluted, and plated to determine the CFU/ml. Means with the standard errors of the mean (SEM) are plotted (n = 2 to 4). The benchmark of 10⁶ bacterial killing is indicated by the dotted lines. *, significant differences (P < 0.05) in benchmarks between NAC/drug-treated and drug-treated groups.

NAC itself decreased the growth of M. tuberculosis H37Rv initially but the growth delay caused by NAC was not significant and disappeared after the first week of treatment (Fig. 1A). This NAC impact on the early growth of M. tuberculosis was mostly observed in the drug-susceptible M. tuberculosis H37Rv and less in the MDR and XDR M. tuberculosis strains. Notably, the killing of drug-susceptible and drug-resistant M. tuberculosis strains by first-line or second-line TB drugs followed similar kinetics for the first 3 to 7 days of treatment whether NAC was present or not. After that period, the viability of M. tuberculosis decreased more sharply with NAC present in the cultures. Interestingly, MAC was the most efficient drug combination against the drug-susceptible H37Rv strain and the MDR strain V2475 resulting in a 6-log decrease in M. tuberculosis viability (Fig. 1C and E) within a week, while the combination INH/RIF and OKE required at least 3 weeks to observe the same level of CFU decrease (Fig. 1A, B, and D). The XDR strain TF275 did not respond as well to MAC, which may be due to its low-level resistance to moxifloxacin and amikacin. The addition of NAC to MAC did result in a faster sterilization, even in the XDR TF275 strain. In all the conditions tested against drug-susceptible and drug-resistant M. tuberculosis strains, NAC reduced the time necessary to achieve sterilization of M. tuberculosis treated with TB drugs.

NAC increases M. tuberculosis respiration in vitro.

We had previously demonstrated that the NAC analog cysteine increased M. tuberculosis respiration, which we hypothesized kept M. tuberculosis into an active metabolic state and prevented M. tuberculosis from entering a persister state (6). To assess whether NAC had a similar activity in M. tuberculosis, the effect of NAC on the oxygen consumption (respiration) of M. tuberculosis grown in logarithmic phase was measured (Fig. 2A). Upon addition of NAC, the rate of oxygen consumption by M. tuberculosis doubled (from 0.95- to 1.2-fold increases in M. tuberculosis respiration observed in three independent experiments). This increase was much smaller than with its analog cysteine (Fig. 2B). The addition of cysteine to log-phase M. tuberculosis resulted in a 4- to 5-fold increase in the oxygen consumption rate. Cysteine is a stronger oxygen consumption booster for M. tuberculosis than NAC and a more efficient potentiator of INH activity against M. tuberculosis in vitro (6), but cysteine mammalian cell toxicity prevents any experiment in a mouse model of M. tuberculosis infection. Although NAC showed limited potential in increasing M. tuberculosis oxygen consumption, its ability to sterilize M. tuberculosis cultures in vitro treated with a first-line or a second-line TB drug combination led us to test whether NAC would have any potentiating activity on TB drugs in a mouse model of M. tuberculosis infection.

FIG 2.

NAC stimulates M. tuberculosis respiration. M. tuberculosis strain mc²6230 was placed in the chamber of an Oroboros Oxygraph-2k respirometer. Once the oxygen consumption of mc²6230 was stable, 4 mM NAC (A) or 4 mM cysteine (B) was added to the chamber, as indicated by the arrow. The graph shows the oxygen (O₂) concentration (in blue) and the oxygen consumption rate (in red) in the chamber in real time. The graph shows a single replicate representative of three independent experiments.

NAC does not potentiate first- and second-line TB drugs in mice.

NAC toxicity in mice depends on its mode of delivery. NAC lethal dose LD50 is 400 mg/kg by intraperitoneal injection and 4.4 g/kg orally. The option of NAC administration via oral gavage was therefore chosen to allow for delivery of high concentrations (0.5 to 1 g/kg, daily, 5 times a week) of NAC for a 4-week period.

The acute model of infection where M. tuberculosis is actively growing in the lungs of immunocompetent mice was used. CBA/J mice were infected with M tuberculosis H37Rv via the aerosol route at a low dose (∼100 CFU in lungs). After infection, treatment was started 2 weeks later and lasted for 4 weeks. The mice were treated by oral gavage with either the combination of first-line TB drugs INH and RIF (Fig. 3A) or with the second-line TB drugs moxifloxacin and ethionamide (Fig. 3B). NAC was given by oral gavage at a concentration of 0.5 g/kg. Lungs, liver, and spleen were harvested at 2, 4, 5, and 6 weeks postinfection. No bacteria were detected in the spleens and livers of the treated mice at weeks 4, 5, and 6 in all treatments (data not shown). Although the lung bacterial load was 4-fold lower in mice cotreated with INH, RIF, and NAC compared to mice treated with INH and RIF after 2 weeks of treatment, that difference was reversed after 3 weeks of treatment, and there was no difference between the two treatment groups after 4 weeks of treatment. Increasing the concentration of NAC to 1 g/kg did not change the outcome (data not shown). For the M. tuberculosis-infected mice treated with second-line TB drugs, the addition of NAC to the TB drug regimen resulted in an enhanced reduction of the bacterial load in the lung (8-fold lower in the mice receiving NAC) after 3 weeks of treatment (week 5), but again, this improvement disappeared at the later time point. In both experiments, the potentiation of the TB drugs by NAC was short-lived.

FIG 3.

Adjunct activity of NAC in vivo. CBA/J mice were infected via the aerosol route with M. tuberculosis H37Rv (initial doses average, 100 bacilli/lung). The mice were rested for 2 weeks before treatment started. Mice received INH (10 mg/kg) and RIF (10 mg/kg) (A) or moxifloxacin (Moxi, 100 mg/kg) and ethionamide (ETH, 50 mg/kg) (B) by oral gavage once a day, five times a week, for 4 weeks. NAC (0.5 g/kg) was given by oral gavage at the same time than the TB drugs. At the indicated times, mice were euthanized. Left lungs were collected and homogenized, and the lysates were serially diluted to determine the burden of M. tuberculosis. Each dot represents the burden of M. tuberculosis in the lung of one mouse.

N-Acetylcysteine amide, an NAC analog, has in vitro and in vivo activities similar to NAC.

NAC has low bioavailability which may impede its in vivo effect (10). The amide derivative of NAC, the N-acetylcysteine amide (NACA), has higher lipophilicity, better membrane permeability, and better bioavailability in mice (11). NACA is also better tolerated than NAC. NACA is an NAC prodrug as NACA is converted into NAC upon oral administration (11). When tested in vitro, M. tuberculosis cultures were sterilized faster when NACA was added to a combination of first-line (INH/RIF) or second-line (OKE or MAC) TB drugs (Fig. 4A). Like what was observed with NAC, the addition of NACA to M. tuberculosis cultures resulted in a nonsignificant growth delay during the first week of treatment, which then disappeared. There were also no significant differences between the drug-treated groups with and without NACA for the first 4 to 7 days of treatment. After that, the viability of M. tuberculosis decreased more sharply when NACA was present. The benchmark of 10⁶ killed M. tuberculosis bacilli was reached 7 to 9 days earlier with INH/RIF/NACA, OKE/NACA, and MAC/NACA than when NACA was not present, although this was significant only for the OKE/NACA combination. Sterilization of the M. tuberculosis cultures was obtained between 10 and 14 days posttreatment with NACA and between 21 and 28 days when NACA was not present in the drug-treated cultures.

FIG 4.

The NAC analog NACA shows in vitro and in vivo activities similar to NAC. (A) M. tuberculosis H37Rv was treated at day 0 with NACA (4 mM), INH (7 μM)/RIF (1 μM), or INH/RIF/NACA; ofloxacin (O, 14 μM)/kanamycin (K, 41 μM)/ethionamide (E, 150 μM) with or without NACA (4 mM); or moxifloxacin (M, 13 μM)/amikacin (A, 8 μM)/clofazimine (C, 11 μM) with or without NACA (4 mM). At the indicated times, aliquots were removed, serially diluted, and plated to determine the CFU/ml. Means with the SEM are plotted (n = 2). The benchmark of 10⁶ bacterial killing is indicated by the dotted lines. *, significant differences (P < 0.05) in the benchmarks between NACA/drug-treated and drug-treated groups. (B) CBA/J mice were infected via the aerosol route with M. tuberculosis H37Rv (infection dose average, 117 bacilli/lung). The mice were rested for 2 weeks before treatment started for 4 weeks. Treatment was given by oral gavage once a day, five times a week at the following concentrations: INH (10 mg/kg), RIF (10 mg/kg), and NACA (1 g/kg). Left lungs were collected and homogenized, and the lysates were serially diluted to determine the burden of M. tuberculosis. Each dot represents the burden of M. tuberculosis in the lung of one mouse. (C) H&E staining of lung tissues procured from INH/RIF-, INH/RIF/NAC-, or INH/RIF/NACA-treated or untreated M. tuberculosis-infected CBA/J mice. The experimental setup was as described in the legends to Fig. 3A and 4B. Inflammatory infiltrations were most conspicuous in the untreated group at both 2- and 4-week intervals (arrow). The INH/RIF and INH/RIF/NAC treatment groups displayed minimal perivascular/peribronchiolar infiltrates (arrowhead) at all time points studied. The INH/RIF/NACA-treated mice exhibited at the 2-week interval more prominent inflammatory infiltrations (arrow) relative to the INH/RIF and INF/RIF/NAC treatment groups; the level of inflammation, however, diminished significantly at the 4-week time point to minimal perivascular/peribronchiolar infiltrates (arrowhead). Compared to the untreated mice, the level of inflammation of the INH/RIF/NACA-treated group is apparently less abundant.

NACA was then tested in a mouse model of acute M. tuberculosis infection where the M. tuberculosis-infected mice were treated with INH/RIF or with INH, RIF, and NACA. NACA was given by oral gavage at a concentration of 1 g/kg. Lungs, liver, and spleen were harvested at 2, 4, 5, and 6 weeks postinfection. No bacteria were detected in the spleens and livers of the treated mice at week 4, 5, and 6 in both treatments (data not shown). Similar to what we observed in M. tuberculosis-infected mice treated with INH, RIF, and NAC, the lung bacterial load was 4-fold lower in mice treated with INH, RIF, and NACA compared to mice treated with INH and RIF after 2 weeks of treatment. Again, we observed a reversed pattern where the drug-only treated group had a significantly lower lung titer than the INH/RIF/NACA group after 3 weeks of treatment. After 4 weeks of treatment, there was no significant difference between the two treated groups (Fig. 4B). The similarities between the lung burden of the mice treated with INH/RIF/NAC and INH/RIF/NACA suggest that NACA is quickly converted to NAC in vivo. Although the replacement of NAC by NACA did not lead to an increased reduction in the lung bacterial burden of the treated mice, we examined whether NACA had an impact on the disease itself by comparing lung tissues of mice treated with INH/RIF, INH/RIF/NAC, and INH/RIF/NACA after 2 and 4 weeks of treatment (weeks 4 and 6 postinfection) (Fig. 4C). No differences were found in the lung tissues of mice treated with INH/RIF or INH/RIF/NAC at weeks 4 and 6. Both lung tissues revealed minimal changes with rare small perivascular/peribronchiolar lymphoid infiltrates. In contrast, the lung tissues of mice treated with INH/RIF/NACA revealed the presence of inflammatory aggregates at week 4, also observed in the lung tissues of untreated mice, but those signs of inflammation were considerably diminished at week 6 postinfection. The lung tissue pathology of the mice treated with INH/RIF/NACA suggests that NACA addition to INH/RIF treatment resulted in more signs of TB disease compared to INH/RIF or INH/RIF/NAC. Although NACA has been hailed as a better substitute to NAC, in our experiments, we did not observe any significant improvement in the killing of M. tuberculosis by TB drugs with NACA present in vitro and in vivo.

NAC potentiates TB drugs in the human THP1 cell line but not in the murine J774 cell line.

There is clinical evidence that TB patients receiving chemotherapy supplemented with NAC reached sputum negativity faster and show better clearing of infiltration and reduction in cavity size than TB patients receiving TB chemotherapy alone (8). To test whether NAC adjunct activity could be species-specific, we compared the adjunct activity of NAC in M. tuberculosis-infected murine versus human macrophage cell lines. The murine macrophage cell line J774 (Fig. 5A) and the human macrophage cell line THP1 (Fig. 5B) were infected with M. tuberculosis H37Rv at a multiplicity of infection (MOI) of 2 for 4 h prior to treatment with INH, RIF, and NAC. The addition of NAC to INH and RIF treatment did not improve the intracellular killing of M. tuberculosis by INH and RIF in the murine cell line J774 (Fig. 5A). There was no difference in M. tuberculosis killing between the INH/RIF and the INH/RIF/NAC treatments during the course of the experiment (7 days). In contrast, in the human cell line THP1, the addition of NAC to the INH and RIF treatment resulted in a small (2-fold) but significant (P = 0.005) increase in the intracellular killing of M. tuberculosis at day 6 (Fig. 5B). There was no potentiation of INH and RIF killing of intracellular M. tuberculosis by NAC at the earlier time points (days 1, 2, and 3), which is a phenotype we had also observed in in vitro M. tuberculosis cultures where no significant NAC potentiation of INH/RIF was observed before day 7. The killing of intracellular M. tuberculosis by INH/RIF was notably arrested by day 6, suggesting that the intracellular M. tuberculosis population might have entered a persistent state. To test this hypothesis, we repeated the INH/RIF/NAC treatment of intracellular M. tuberculosis in gamma interferon (IFN-γ)-treated THP1 cells (Fig. 5C) since Liu et al. demonstrated that activation of murine macrophages with IFN-γ and LPS induced tolerance to INH or RIF in M. tuberculosis (12). In THP1 treated with IFN-γ, a similar reduction in the rate of killing was observed between days 2 and 5 in INH/RIF- or INH/RIF/NAC-treated M. tuberculosis. After day 5, killing of M. tuberculosis by INH and RIF stopped, while the loss of M. tuberculosis viability continued in the INH/RIF/NAC-treated M. tuberculosis-infected THP1 cells (Fig. 5C). One hypothesis for the increased M. tuberculosis killing in THP1 cells by INH/RIF/NAC compared to INH/RIF at the later time points was that NAC affected THP1 viability. An MTT assay was performed in THP1 cells infected with heat-killed M. tuberculosis (to prevent MTT reduction to formazan by viable M. tuberculosis cells) and treated with NAC, INH/RIF, or INH/RIF/NAC. No significant difference in the viability of THP1 between cells treated with INH/RIF and cells treated with INH/RIF/NAC was found (Fig. 5D). The addition of NAC to INH/RIF treatment did not decrease the viability of THP1 cells.

FIG 5.

Adjunct activity of NAC in cell lines. Murine (J774) (A) and human (resting THP1 [B] and IFN-γ-treated THP1 [C]) macrophage-like cell lines were infected with M. tuberculosis H37Rv for 4 h at an MOI of 2 before treatment started. The concentrations used were as follows: NAC (2 mM), INH (7 μM), and RIF (1 μM). At the indicated times, macrophages were lysed, and the lysates were serially diluted to determine the CFU/ml. Graphs represent two biological replicates performed in duplicates. *, significant differences (P < 0.05) between NAC/drug-treated and drug-treated groups. (D) The viability of THP1 cells treated with NAC, INH/RIF, or INH/RIF/NAC was measured with an MTT assay. THP1 cells were infected with heat-killed M. tuberculosis and treated with NAC (2 mM) and/or INH (7 μM)/RIF (1 μM) for 7 days. At the indicated time points, MTT was added to the cells, and the absorbance at 540 nm was read. The graph represents the means and the SEM of three biological replicates.

DISCUSSION

The first clinical use of NAC was reported more than 50 years ago when NAC’s mucolytic properties were demonstrated in cystic fibrosis patients (13). In the 1970, NAC was successfully used clinically for the treatment of acetaminophen/paracetamol poisoning (14). Since then, the potential therapeutic benefits of NAC have been described for a variety of diseases such as pulmonary diseases, cardiovascular diseases, diabetes, or neuropsychiatric disorders (15). More recently, NAC was administered intravenously to severe COVID-19 patients, which resulted in improved clinical outcomes (16).

In TB patients, NAC is used for the treatment of TB drug toxicity (7). The coadministration of NAC with TB standard drug regimen has also shown some potential benefits. In a randomized, placebo-controlled, double-blinded study in India on 48 newly diagnosed, sputum-positive TB patients treated with standard TB drug regimen, one group received two NAC tablets (600 mg each) daily for 2 months, while the other group received placebo tablets. Of 24 TB patients receiving NAC, 23 achieved sputum culture negativity within 3 weeks (14/24 for the placebo group). Radiological improvements were also achieved with significant clearing of infiltration at 2 months for 21/24 of TB patients receiving NAC (8/24 for placebo group) and increased reduction in cavity size for the NAC group (8). A recent phase II trial in Brazil determined that the addition of NAC to standard TB regimen was not unsafe in hospitalized patients with severe TB and coinfected with the human immunodeficiency virus (HIV) (17). NAC (600 mg twice daily) was given for 8 weeks. There was no difference in adverse events and hepatotoxicity between the placebo (21 patients) and NAC (18 patients) groups. Similarly, the authors observed no difference between the two groups in the time to achieve sputum culture negativity or in radiological responses. Nevertheless, the authors emphasized that their study was based on a small number of patients severely ill and may not apply to nonsevere TB patients.

In our study, NAC enhanced the killing of M. tuberculosis by first- and second-line TB drugs in vitro. This outcome was not observed in M. tuberculosis-infected mice treated with first- or second-line TB drugs for 4 weeks. Although we did observe a small decrease in lung burden in mice cotreated with NAC and TB drugs during the second or third weeks of treatment, this effect did not last. Amaral et al. have shown that the lung burden of M. tuberculosis-infected mice treated with only NAC (400 mg/kg, gavage daily for 6 days) had decreased by 0.5 log₁₀ at day 7 compared to untreated mice (18). This effect was significant and reproducible. The timing of the NAC treatment might be an important factor in its antimycobacterial activity as reported by Amaral et al. (18) or in its ability to potentiate TB drugs. In our experiments with NAC in M. tuberculosis in vitro cultures, macrophages or mouse infection, the effect of NAC was time dependent. In in vitro cultures, an increase in the killing rate of M. tuberculosis by the combination NAC/TB drugs was observed after the first 4 to 7 days of treatment compared to M. tuberculosis treated with TB drugs only. Similarly, in the human macrophage cell line THP1 (activated or resting), the potentiation of TB drugs by NAC was not seen before day 6 of treatment. In contrast, the transient effect of NAC on M. tuberculosis killing by TB drugs in mice was observed earlier in the treatment. This might be the result of the complex immune response developed during the mice infection with M. tuberculosis. The different timelines render difficult any interpretation as to what component(s) of the immune system if any may play a role in the NAC transient in vivo activity.

This lack of NAC effect in vivo might be specific to the host as we did observe potentiation of INH and RIF activity by NAC in the human THP1 cell line but not in the murine J774 cell line. As previously shown with cysteine (6), NAC also stimulated M. tuberculosis respiration, which we hypothesize keeps M. tuberculosis metabolically active, susceptible to the action of the TB drugs and unable to enter into a quiescent state. This is the NAC effect measured on the bacterium itself, but previous studies on NAC also demonstrated that NAC affected the host. It is postulated that in humans, NAC acts as both a free radical scavenger and as a precursor of cysteine, a substrate for the synthesis of glutathione (19). In acetaminophen poisoning, NAC replenished the host with glutathione that had been depleted by acetaminophen (20). Glutathione, glutathione reductase, and glutathione peroxidase levels are also lower in TB patients, suggesting that TB infection elicits an oxidative stress which might be offset by NAC (21). Mahakalkar et al. (8) found that the addition of NAC to TB standard drug regimen treatment increased glutathione peroxidase levels in TB patients. These authors hypothesized that this glutathione peroxidase increase might reduce ROS production and tumor necrosis factor alpha (TNF-α) mRNA expression, as previously described (22). The combination of NAC effects on both the pathogen and the host might be required to observe a clinical benefit.

A clinical trial (NCT03702738) is currently ongoing in Tanzania to determine whether the addition of NAC (1,200 mg twice a day) to standard TB regimen (2 months on INH/RIF/PZA/EMB and 4 months on INH/RIF) would hasten sputum culture conversion of drug-sensitive TB patients with moderately advanced or far-advanced pulmonary disease. The patients will receive NAC during the first 4 months of treatment. This clinical trial will determine the safety and efficacy of NAC adjunct therapy in TB patients. We will then have a better indication as to whether NAC could be a useful adjunct compound to shorten chemotherapy in TB patients.

The development of new TB drugs to shorten TB chemotherapy remains a daunting task. The past 20 years of TB research aiming at finding new TB drugs efficacious against active, latent, and drug-resistant M. tuberculosis strains have yielded a few candidates that may be used in specific TB cases (23, 24). Clinical trials are under way to determine combinations of novel TB drugs to shorten TB chemotherapy. The Nix-TB trial (NCT02333799) successfully tested the combination of pretomanid, bedaquiline and linezolid for 6 months in nonresponsive MDR, XDR, and treatment-intolerant TB patients and was approved by the U.S. Food and Drug Administration for use in this specific category of TB patients. Another avenue of research would be to identify boosters of M. tuberculosis respiration that potentiate the activity of TB drugs in use or in clinical trials; novel molecules that could be used at a much lower concentration that the known M. tuberculosis respiration inducers NAC and vitamin C (6). The need to shorten TB chemotherapy especially for MDR TB cases is dire, and M. tuberculosis respiration boosters could be a new tool in the arsenal to defeat TB.

MATERIALS AND METHODS

Bacterial strains, media, and materials.

Chemicals and biologicals were obtained from Sigma (St. Louis, MO) or Thermo Fisher Scientific (Waltham, MA) unless otherwise stated. N-Acetylcysteine amide was obtained from BOC Sciences (Shirley, NY). The M. tuberculosis strains and cell lines used in this study were obtained from laboratory stocks. M. tuberculosis mc²6230 (H37Rv ΔRD1 ΔpanCD), a biosafety level 2-safe laboratory strain, was used in the Oroboros experiments for safety reasons. The M. tuberculosis strains were grown shaking at 37°C in Middlebrook 7H9 medium supplemented with 10% (vol/vol) OADC (oleic acid-albumin-dextrose-catalase) enrichment (Difco, Sparks, MD), 0.2% (vol/vol) glycerol, and 0.05% (vol/vol) tyloxapol. Plating for CFU was done on Middlebrook 7H10 medium supplemented with 10% (vol/vol) OADC enrichment and 0.2% (vol/vol) glycerol. The plates were incubated at 37°C for 4 to 6 weeks. Plating of undiluted, treated M. tuberculosis cultures was done on Middlebrook 7H10 medium supplemented with 10% (vol/vol) OADC enrichment, 0.2% (vol/vol) glycerol, and 0.4% carbon.

In vitro drug treatment of M. tuberculosis.

M. tuberculosis strains were grown to exponential phase (optical density at 600 nm [OD₆₀₀] ≈ 0.8 to 1) and then diluted 1/100. Drugs and compounds were added once at time zero. Aliquots were taken at the indicated times. Tenfold serial dilutions were performed using Dulbecco phosphate-buffered saline (DPBS) and plated as described above.

Oxygen consumption measurement.

M. tuberculosis mc²6230 was grown in Middlebrook 7H9 supplemented with 10% (vol/vol) ADS enrichment (50 g of albumin, 20 g of dextrose, and 8.5 g of sodium chloride in 1 liter of water), pantothenate (24 mg/liter) and 0.05% (vol/vol) tyloxapol to exponential phase (OD₆₀₀ ≈ 0.6). The rate of oxygen consumption was measured at 37°C using the Oroboros Oxygraph-2k respirometer (Oroboros Instruments, GmbH, Innsbruck, Austria). mc²6230 (2 ml) was placed in the chamber and, once its oxygen consumption had stabilized, NAC (4 mM) or cysteine (4 mM) was added via the injection ports. The oxygen concentration in the chamber and the rate of oxygen consumption were recorded in real time.

Mouse experiments.

CBA/J female and male mice (6 to 8 weeks old) were obtained from Envigo (Indianapolis, IN). The animal protocol AUP#20180310 was approved by the Einstein Animal Institute, which is accredited by the American Association for the Use of Laboratory Animals (DHEW publication no. [NIH] 78-23, revised 1978), and accepts as mandatory the NIH “Principles for the Use of Animals.” M. tuberculosis H37Rv was grown to exponential phase (OD₆₀₀ ≈ 0.8 to 1), spun down, washed twice with DPBS, and sonicated for 10 s before the measuring OD₆₀₀ to estimate culture titer (OD₆₀₀ = 1 corresponds to 3 × 10⁸ CFU/ml, determined in the laboratory). The mice were infected with M. tuberculosis H37Rv via the aerosol route. At 2 weeks after infection, treatment was started and continued for 4 weeks. The drugs and compounds were given by oral gavage once a day, five times a week, for 4 weeks at the following concentrations: INH, 10 mg/kg; RIF, 10 mg/kg; moxifloxacin, 100 mg/kg; ethionamide, 50 mg/kg; NAC, 0.5 or 1 g/kg; and NACA, 1 g/kg. At the indicated times, mice were euthanized, and their spleens, livers, and lungs were removed. The organs were homogenized in DPBS, serially diluted in DPBS, and plated (see above) to determine CFU/organ. For pathology, lungs were fixed in 10% buffered formalin phosphate. After paraffin embedment, lung tissues were sectioned and stained with hematoxylin and eosin (H&E). Tissues were analyzed by a pathologist at the Histopathology and Comparative Pathology facility at the Albert Einstein College of Medicine.

Cell line experiments.

M. tuberculosis H37Rv was grown to exponential phase (OD₆₀₀ ≈ 0.8 to 1), spun down, washed twice with DPBS, and sonicated for 10 s before measuring OD₆₀₀ to estimate CFU/ml (OD₆₀₀ = 1 corresponds to 3 × 10⁸ CFU/ml). J774 cells were obtained from laboratory stocks and grown in Dulbecco modified Eagle medium (DMEM; Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gemini Bio-Products, West Sacramento, CA). The cells were seeded at a concentration of 8 × 10⁵ cells in 24-well plates. J774 cells were infected with M. tuberculosis H37Rv (MOI of 2) for 4 h at 37°C in 5% CO₂ to allow for mycobacterial uptake. Cell monolayers were washed twice with DPBS to remove extracellular M. tuberculosis bacteria. DMEM/FBS (1 ml) was added to each well with the tested compounds at the following concentrations: INH, 7 μM; RIF, 1 μM; and NAC, 2 mM.

THP1 cells were obtained from laboratory stocks and grown in RPMI 1640 containing 10% FBS and 1× GlutaMAX. The cells (0.5 ml containing 0.04 mg of phorbol 12-myristate 13-acetate [PMA]) were seeded overnight at 37°C in 5% CO₂ at a concentration of 2 × 10⁵ cells in 24-well plates. THP1 cells were infected with M. tuberculosis (MOI of 2 [see the preparation of the culture described above]) for 4 h at 37°C in 5% CO₂ to allow for mycobacterial uptake. Cell monolayers were washed twice with DPBS. RPMI/FBS (1 ml) was added to each well with the tested compounds at the following concentrations: INH, 7 μM; RIF, 1 μM; and NAC, 2 mM. For the IFN-γ THP1 experiment, THP1 cells were treated with recombinant human IFN-γ (100 U/ml; Peprotech, Rocky Hill, NJ) 24 h prior to M. tuberculosis infection.

The M. tuberculosis-infected J774 or THP1 cells were incubated at 37°C in 5% CO₂. At specific time points, the medium was removed. The cells were washed once with DPBS and then lysed with 0.05% aqueous sodium dodecyl sulfate (1 ml). The lysates were serially diluted and plated onto Middlebrook 7H10 plates (see above) to determine CFU. Media with all supplements (INH, RIF, NAC, and IFN-γ when required) was replenished at every time point.

THP1 viability assay was performed as previously described (25). THP1 cells (0.1 ml containing 0.008 mg PMA) were seeded overnight at 37°C in 5% CO₂ at a concentration of 1 × 10⁴ cells in 96-well plates. THP1 cells were infected with heat-killed M. tuberculosis (0.1 ml) prepared as follows: M. tuberculosis H37Rv was grown to exponential phase (OD₆₀₀ ≈ 0.8 to 1), spun down, washed twice with DPBS, sonicated for 10 s, resuspended in DPBS for a final concentration of 5 × 10⁵ CFU/ml (based on the OD₆₀₀ measurement), and heated at 100°C for 1 h. The infected THP1 cells were incubated for 4 h at 37°C in 5% CO₂ to allow for mycobacterial uptake. Cell monolayers were washed twice with DPBS. RPMI/FBS (0.1 ml) containing INH (7 μM), RIF (1 μM), and NAC (2 mM) was added to each well. The plate was incubated at 37°C in 5% CO_2. The medium was replenished at days 2, 5, 6, and 7. For the MTT assay, the medium was removed from the tested wells and replaced with 0.11 ml RPMI/FBS containing 120 nmol of MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide]. The plate was incubated for 4 h at 37°C in 5% CO₂. Then, 0.08 ml of medium was removed from the tested wells, and 0.05 ml of DMSO was added to each tested well. The plate was incubated for 10 min at 37°C in 5% CO₂ prior to reading the absorbance at 540 nm.

Statistics.

Differences between groups were analyzed by a two-tailed, unpaired t test using GraphPad Prism 7.05 (San Diego, CA).