Mtb H37Rv (ATCC 27294) and the clinical isolates (17 MDR-TB and 4 XDR-TB, all from the National Clinical Laboratory on Tuberculosis) were grown at 37 ℃ in 5% ambient CO2 for 14 days in Middlebrook 7H9 broth (Becton Dickinson, Sparks, MD, USA) supplemented with 0.2% (v/v) glycerol (Sigma, St. Louis, MO, USA), 10% (v/v) oleic acid-albumin-dextrose-catalase (OADC; Becton Dickinson) and 0.05% (v/v) Tween 80 (Sigma). The bacteria were washed, suspended in phosphate-buffered saline, and passed through a filter (8-µm pore size) to eliminate clumps. The filtrates were aliquoted and stored at -80 ℃ for use within 30 days.
Chemicals were purchased in pure form from various manufacturers: levofloxacin (Lfx) from Shuanghe Pharmaceutical Co., Ltd. (Beijing, China); clofazimine (CFZ) from Nanjing Liye Pharmaceutical Co., Ltd. (Nanjing, Jiangsu, China); aminosalicylate (PAS), protionamide (Pto) and capreomycin (CPM) from Shanghai Xinyi Pharmaceutical Co., Ltd. (Shanghai, China); rifampin (RIF), pyrazinamide (PZA) and isoniazid (INH) from Sigma; linezolid (LZD) from Pfizer (Andover, MA, USA); BP from Zhengda Tianqing Pharmaceutical Co., Ltd. (Jiangsu, China); and CL from Aladdin Industrial Corp. (Calhoun, GA, USA).
Initial stock solutions of Lfx, CPM, PAS, BP/CL, Pto, and CFZ were constituted in dimethyl sulfoxide (DMSO) at concentrations of 10 or 20 g/L. Dilutions were made from the stock solutions in 7H9 broth. To avoid solvent effects, the highest concentration of DMSO was 0.5%. BP and CL were dissolved just before use. The MICs of the antimicrobial agents were determined against H37Rv strain using a microplate Alamar Blue assay with serial two-fold dilutions of the drugs. The MICs of the combinations of BP/CL and other second-line anti-TB agents were determined in 96-well microtiter plates using the dynamic checkerboard method. Initial concentrations of Lfx, CPM, PAS, CFZ, LZD, and Pto were 1, 4, 1, 0.48, 1, and 2 µg/mL, respectively, and the final concentrations of the drugs (two-fold dilutions) ranged from 1/8× to 1× of the MIC of each drug.
The data were interpreted by calculating the fractional inhibitory concentration (FIC) as following formula: FIC = (A/MICA) + (B/MICB), where A and B represent the MICs of each drug combination, and MICA and MICB represent the MICs of each drug alone. The FIC results reflect synergy for FIC < 0.5[13, 14].
After the cultured macrophage cell line J774A.1 was suspended in antibiotic-free Dulbecco's Modified Eagle Medium (DMEM and 10% fetal bovine serum), the cell suspension was added to 24-well tissue culture plates at a cell density of 5 × 105 CFU/mL. The plates were incubated at 37 ℃ for 12 h in 5% CO2. The macrophages were infected with a 100-µL bacterial suspension containing 5 × 106 CFU/mL Mtb H37Rv and incubated for 4 h. The cells were washed with prewarmed serum-free DMEM to remove extracellular bacteria from the adherent cells. Next, 1 mL of DMEM with or without drugs was added, and the plates were incubated in 5% CO2 as described above for 2 days. Wells containing cells without added drugs served as negative controls. The BP concentrations were 40 and 20 µg/mL, and the concentrations of CL and INH were 2.5 and 2 µg/mL, respectively. Due to the instability of BP and CL in medium, spent medium was replaced with new medium containing freshly constituted BP/CL twice each day. Two days after infection, each macrophage monolayer was lysed with 0.2 mL of 7H9 medium containing 0.1% SDS. Next, the lysates were serially diluted and dilutions were inoculated onto OADC-enriched 7H11 agar medium for enumeration of CFU.
Male 6-week-old BALB/c mice (Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) were infected with the H37Rv strain using a Glas-col inhalation exposure system (Glas-col Inc., Terre Haute, IN, USA) and a fresh log-phase broth culture to implant 3 log10 CFU into the lungs of each mouse. Five mice were sacrificed the following day to determine the number of CFUs implanted in the lungs. All animal procedures were approved by the Animal Care and Use Committee of the Beijing Tuberculosis and Thoracic Tumor Institute (No. 2017009).
After infection, the mice were randomly assigned to one of the nine treatment groups described in Table 1. The first group was a negative control group that included infected mice receiving no treatment. The last group was a positive control group of mice treated with the standard RIF-INH-PZA (RHZ) TB-treatment regimen. Mice in the remaining groups were treated with additional antibiotics. The treatment began 15 days after infection (on day 0) to attain a high initial bacterial burden within the lungs of mice. Treatment was administered 5 days per week thereafter. For consistency, 4 weeks were considered equivalent to 1 month of treatment. Drugs were administered in the doses shown in Table 2. Based on the areas under the concentration (AUC)-time curves, these doses were chosen due to equipotency with the usual doses administered to humans and were similar to doses used in previous studies[16, 17]. All regimens were administered for 2 months. BP and CL were given by subcutaneous injection twice each day, 5 days per week. The remaining drugs were administered by gavage.
Group Drug Alone or Combinationa No. of Mice Sacrificed at Indicated Time Pointb Total No. of Mice D -14 D0 M1 M2 A Untreated 5 5 5 5 20 B BP/CL 5 5 10 C PAS 5 5 10 D Pto 5 5 10 E BP/CL + PAS 5 5 10 F BP/CL + Pto 5 5 10 G Lfx + Pto + PAS + PZA 5 5 10 H BP/CL + Lfx + Pto + PAS + PZA 5 5 10 I RIF + INH + PZA 5 5 10 Total No. of mice 5 5 45 45 100 Note. aDrugs were administered 5 days per week at the following doses: BP (100), CL (100), PAS (750), Pto (100), Lfx (200), RFP (10), INH (25), and PZA (150). bMice were sacrificed at the following times: 1 day after infection (D -14); 15 days after infection [day 0 (D0)]; 1, 2 months after treatment (M1, M2, respectively).
Table 1. Experimental Design Used in This Study
Agents Dosage (mg/kg) Agents Dosages (mg/kg) BP 100 Lfx 200 CL 100 RFP 10 PAS 750 INH 25 Pto 100 PZA 150
Table 2. Dosages of Drugs Administered to the Mice
To provide baseline values, a group of five infected but untreated mice were euthanized per day on days 1 and 15 after infection (days designated: -14 and 0 in relation to initiation of treatment). The treated mice were euthanized after 1 and 2 months of treatment to determine lung CFU counts. The numbers of CFUs within the lungs were determined on days -14 and 0 and after 1 month of treatment by plating four serial 10-fold dilutions of homogenized suspensions onto OADC-enriched 7H11 agar medium. After 2 months of treatment, the entire suspension prepared from each individual organ, predicted to contain few bacilli, was plated without dilution onto OADC-enriched 7H11 agar medium. The results for the various cultures were recorded after a 4-week incubation at 37 ℃. The bactericidal effect of each treatment was defined as a significant decrease in the mean number of CFUs compared to the corresponding infected untreated control tissue CFU values.
CFU counts were converted to log10 values before analysis, and each was expressed as CFU log10 ± standard error. Experimental group means were compared to means of the untreated infected group by two-way analysis of variance using Fisher's exact test. Differences were considered significant at a P-value < 0.05. The data analysis was performed using the Statistical Package for the Social Sciences software (SPSS 17.0; IBM, Armonk, NY, USA).
Bacterial Strains and Culture Conditions
Drugs and Chemicals
Determination of Minimum Inhibitory Concentration (MIC) and Synergy
Activity of BP/CL in Macrophages
Assessment of Infection and Treatment
The MICs of BP/CL against H37Rv, the 17 MDR, and the four XDR isolates were determined according to a standard two-fold agar dilution method under defined conditions as described above. The MIC of BP/CL against H37Rv was 2 µg/mL and ranged between 0.5 and 8 µg/mL for the 17 MDR strains and between 1 and 4 µg/mL for the four XDR strains (Figure 1). No differences were observed between the MICs of BP/CL when compared with H37Rv, MDR-, or XDR-TB (P > 0.05).
Figure 1. Minimum inhibitory concentrations (MICs) of biapenem/clavulanate (BP/CL) against extensive drug-resistant tuberculosis (XDR-TB), multidrug-resistant tuberculosis (MDR-TB), and the H37Rv strain in vitro. #Fisher's exact test. *No difference between the MICs of BP/CL against XDR-TB or H37Rv. **No difference between the MICs of BP/CL against XDR-TB or MDR-TB. ***No difference between the MICs of BP/CL against MDR-TB or H37Rv.
The dynamic checkerboard method was used to evaluate the effects of BP/CL combined with either CLF, PAS, Pto, LZD, Lfx, or CPM on MTb CFUs (Table 3). The FIC of BP/CL and pairwise combinations with either PAS, Pto, or CPM was 0.375, indicating that synergy existed between each pair of drugs tested. Combinations of BP/CL with each of the other three drugs only exhibited additive effects.
Agents BP/CL LZD BP/CL Pto BP/CL Lfx BP/CL PAS BP/CL CPM BP/CL CLF MICalonea 2 0.5 2 2 2 0.5 2 0.25 2 2 2 0.24 MICcombinationb 1 0.0625 0.25 0.5 1 0.0625 0.25 0.0625 0.5 0.25 1 0.06 FIC 0.625 0.375* 0.625 0.375* 0.375* 0.75 Note. aMICalone: MIC of agent alone against H37Rv; bMICcombination: MIC of agents combined against H37Rv; *FIC < 0.5 means these two drugs were synergistic.
Table 3. Fractional Inhibitory Concentration (FICs) of Biapenem/Clavulanate (BP/CL) Combinations with Other Second-line Anti-tuberculosis (TB) Drugs in vitro
The activity of BP/CL by day 2 of treatment against H37Rv within macrophages (infected at a MOI = 1) is presented in Figure 2. BP/CL (40 and 20 µg/mL) demonstrated 1.3- and 0.9-log reductions in bacterial numbers, respectively, compared to results for untreated infected control macrophages. INH decreased CFU by 2.1-log and BP/CL40, BP/CL20, and INH were each active against H37Rv within macrophages compared to infected untreated macrophage controls (infected with H37Rv). Although a significant difference was observed between BP/CL40 and INH activities (P = 0.03), no difference was observed between the BP/CL40 and BP/CL20 treatments.
Figure 2. Efficacy of regimens against Mycobacterium tuberculosis (Mtb) H37Rv in macrophages after 2 days of treatment. Fisher's Exact Test. Biapenem/clavulanate (BP)/(CL)40: BP and CL concentrations were 40 and 25 μg/mL, respectively; BP/CL20: BP and CL concentrations were 20 and 25 μg/mL, respectively; *P < 0.05, compared with the control group; three agents decreased CFU significantly; **No difference between these two groups.
On the day after the aerosol infection, the lung CFU count (mean ± standard error) was 3.24 ± 0.06 log10, with a mean lung CFU count at the initiation of treatment (15 days after infection) of 4.98 ± 0.10 log10.
To evaluate the synergy of BP/CL with PAS and Pto in vivo, we determined the CFUs in lungs of mice after treatment with BP/CL, PAS, Pto, BP/CL + PAS; and BP/CL + Pto. After 2 months, the CFU result of the BP/CL + PAS group was no different from the results of the BP/CL, PAS and BP/CL + Pto groups. Moreover, no drug interactions were observed for any of the drug combinations administered in vivo (Figure 3A and 3B).
Figure 3. Efficacy of regimens against Mycobacterium tuberculosis (Mtb) H37Rv in mice after 1 and 2 months of treatment. Control values used in panels A, B, and C were obtained from the same control group. (A) CFU reduction by PAS, BP/CL, and PAS + BP/CL in mice; (B) CFU reduction by Pto, BP/CL, and Pto + BP/CL in mice; (C) CFU reduction of regimen (Lfx + PZA + Pto + PAS), regimen + BP/CL, RHZ in mice; *In Figure C: P = 0.0049, significant difference in CFUs between the regimen group and the regimen + BP/CL group.
At the same time, the activities of three regimens (RHZ; Lfx + PZA + Pto + PAS; and Lfx + PZA + Pto + PAS + BP/CL) against TB were determined in mice. After 2 months of treatment, the CFUs in the lungs of all groups declined significantly compared with control untreated infected lungs (Figure 3C). Lfx + PZA + Pto + PAS reduced the number of bacilli in the lungs by 4.06 log10. When BP/CL was combined with the Lfx + PZA + Pto + PAS regimen, CFUs significantly decreased further by 4.51 log10 (P = 0.0049). However, no difference was observed between the Lfx + PZA + Pto + PAS + BP/CL and RHZ groups (P = 0.2125).