|Home | About | Journals | Submit | Contact Us | Français|
Diethyldithiocarbamate (DETC) and pyrrolidine dithiocarbamate (PDTC) were highly active against tubercle bacilli, with MICs of 8 μg/ml and 0.13 μg/ml, respectively. DETC and PDTC were active against old cultures, enhanced pyrazinamide or pyrazinamide/rifampin activity, and had serum inhibitory titers of 1:2 and 1:4, respectively, in mice given 100 mg/kg orally.
The increasing emergence of multiple-drug-resistant and extensively drug-resistant Mycobacterium tuberculosis (1) emphasizes the need for new tuberculosis (TB) drugs. Drugs that have activity against nongrowing persisters are considered important for shortening the TB therapy (8). To identify such new drugs, we screened the Johns Hopkins clinical compound library against M. tuberculosis in a persister model.
A 2-month-old M. tuberculosis H37Ra standing culture grown in 7H9 medium (Difco) with 10% albumin-dextrose-catalase (ADC) and 0.05% Tween 80 (pH 6.8) was resuspended in acid pH 5.5 7H9 medium without ADC and was used as the inoculum for drug screens. A clinical compound library consisting of 3,360 pharmacologically active compounds that had been screened in different assays (2, 3) was used. The compounds diluted from master stock solution (10 mM in dimethyl sulfoxide) to 10 μM (final concentration) were incubated with the bacilli in 200 μl in acid pH 5.5 7H9 medium without ADC in 96-well plates for 3 days without shaking under 1% oxygen in a hypoxic chamber (Coy Laboratory), a condition that did not permit the growth of the bacilli and was used as a persister model for the drug screen. The screen was done in duplicate. Rifampin (RIF) (5 μg/ml) was used as a positive control. After 3 days of drug exposure, the viability of the bacilli was determined by adding 20 μl of 1 mg/ml XTT [2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide] and incubating at 37°C for 7 days, when the optical density at 450 nm of the plates was read. The screen identified three related compounds: sodium diethyldithiocarbamate (DETC), pyrrolidine dithiocarbamate (PDTC), and disulfiram [1-(diethylthiocarbamoyldisulfanyl)-N,N-diethyl-methanethioamide] (an oxidized derivative of DETC) that had significant activity against nongrowing bacilli.
The MICs and minimum bactericidal concentrations (MBCs) of DETC, PDTC, and disulfiram for M. tuberculosis H37Rv were determined as described previously (7). DETC had an MIC of 4 to 8 μg/ml and an MBC of 8 to 16 μg/ml (Table (Table1).1). Disulfiram had an MIC of 8 to 16 μg/ml. PDTC had an MIC of 0.125 μg/ml; the MBC and the concentration that kills 99% of the inoculum with no further reduction in CFU were 0.5 μg/ml. PDTC was less active against Mycobacterium smegmatis mc26 and Escherichia coli DH5α, with MICs of 16 μg/ml and greater than 64 μg/ml, respectively.
We next examined the effect of PDTC or DETC alone or in combination with antituberculosis drugs on the survival of a 3-week-old (young) culture and a 2-month-old (old) culture of H37Ra at pH 5.5 and pH 7.0 after a 3-day drug exposure in 1 ml 7H9 in 24-well plates followed by CFU determination. PDTC (20 μM, equivalent to 3.3 μg/ml) and DETC (20 μM, equivalent to 4.5 μg/ml) had activity against both young and old cultures under neutral or acidic conditions (P < 0.01) (Table (Table2,2, condition A). DETC was more active against the old culture in acidic conditions (P < 0.01) than was PDTC (P > 0.05). PDTC exhibited the greatest activity for the young culture at neutral pH (P < 0.01). The combination of PDTC (5 μg/ml) and pyrazinamide (PZA) (100 μg/ml) led to more killing than either drug alone for the young culture (P < 0.01) but not for the old culture (P > 0.05) (Table (Table2,2, condition B); in contrast, DETC enhanced the activity of PZA against old cultures (P < 0.01). PDTC and DETC also enhanced the activity of RIF plus PZA during a 3-day exposure (Table (Table2,2, condition C). PZA (100 μg/ml) and RIF (10 μg/ml) led to a greater-than-16-fold reduction in CFU compared with the control, but addition of 1 μg/ml PDTC and 1 μg/ml DETC to PZA and RIF decreased CFU by another 19-fold (P < 0.01) and 81-fold (P < 0.01), respectively, compared to PZA and RIF.
We also determined the effect of oxygen concentrations and the energy inhibitors dicyclohexylcarbodiimide (DCCD) (an F1F0 ATPase inhibitor) and azide (a cytochrome c oxidase inhibitor) on PDTC and DETC activity for a 25-day old H37Ra culture in a 3-day drug exposure assay (Table (Table3).3). Oxygen at 1% or 21% and DCCD (0.05 mM) or sodium azide (0.5 mM) alone had little effect on bacterial viability. Exposure to RIF, relative to control, led to drop of 3 log10 CFU in aerobic condition and of about 0.6 log10 CFU in hypoxia. PDTC (20 μM) or DETC (20 μM) alone led to a reduction in CFU compared with the control at both oxygen concentrations (P < 0.01), with the greater effect at 21% oxygen. Further reductions in CFU were observed with combination of DCCD with either PDTC or DETC mainly at 21% oxygen (P < 0.01). Azide also enhanced the activity of PDTC and DETC but not as much as DCCD.
To determine if PDTC and DETC can reach inhibitory concentrations in mice, 6-week-old female Swiss mice received, by oral gavage, sterile water (control) or water containing PDTC (100 mg/kg), DETC (100 mg/kg), or isoniazid (INH; 25 mg/kg). Blood was collected at 1 h (INH) or 2 h (PDTC and DETC) following administration. Serial twofold dilutions of serum were made in 7H9 broth inoculated with 106 H37Rv bacilli. Serum inhibitory titers determined after 2 weeks of incubation were 1:4 for PDTC, 1:2 for DETC, and 1:16 for the INH positive control. The discrepancy between the MICs/MBCs of PDTC and DETC and their serum inhibitory titers may be attributed to differences in absorption.
DETC is the primary in vivo metabolite of disulfiram, which is used in treatment of alcohol addiction. DETC and PDTC are antioxidants and metal chelators, which are used as inhibitors of NF-κB activation (5) and may inhibit metalloproteinases. PDTC was shown to kill E. coli and Staphylococcus through a metal-chelating mechanism (6). DETC was found to enhance the antimycobacterial activity of human monocytes (4) with an MIC of 1 μg/ml against M. tuberculosis (4), a value lower than the one we observed, presumably due to differences in methods of susceptibility testing. The mechanism by which PDTC and DETC kill M. tuberculosis may also be related to their role as a metal chelator. The observations that PDTC and DETC had activity against persisters and could enhance the activity of RIF and PZA are of particular interest, as activity against persisters may be important for shortening TB therapy (8). Further studies are needed to determine if PDTC and DETC can shorten TB treatment in vivo.
Y.Z. was supported by NIH grants AI44063 and AI49485. We acknowledge the support from the Johns Hopkins Malaria Research Institute (to J.O.L. and D.J.S.), the Department of Pharmacology of Johns Hopkins School of Medicine, the Keck Foundation, the Flight Attendant Medical Research Institute fund, and the Fund for Medical Discovery from Johns Hopkins (to J.O.L.) for establishing the Johns Hopkins clinical compound library used in this study.
Published ahead of print on 17 September 2007.