Reduced virulence of an extensively drug-resistant outbreak strain of Mycobacterium tuberculosis in a murine model

Abstract

Bacterial drug resistance is often associated with a fitness cost. Large outbreaks of multidrug-resistant (MDR) and extensively drug-resistant (XDR) TB have been described that predominately affect persons with HIV infection. We obtained four closely-related Mycobacterium tuberculosis strains (genotype F15/LAM4/KZN) from an outbreak in KwaZulu-Natal (KZN), South Africa, including drug-sensitive, MDR, and XDR clinical isolates. We compared the virulence of these strains in a murine model of aerosol M. tuberculosis infection for four phenotypes: (1) competitive in vivo growth in lung and spleen, (2) non-competitive in vivo growth in lung and spleen, (3) murine survival time, and (4) lung pathology. When mixtures of sensitive, MDR, and XDR KZN strains were aerosolized (competitive model), lung CFUs were similar at 60 days after infection, and spleen CFUs were ordered as follows: sensitive > MDR > XDR. When individual strains were aerosolized (non-competitive model), modest differences in lung and spleen CFUs were observed with the same ordering. C57BL/6, C3H/FeJ, and SCID mice all survived longer after infection with MDR as compared to sensitive strains. SCID mice infected with an XDR strain survived longer than those infected with MDR or sensitive strains. Lung pathology was reduced after XDR TB infection compared to sensitive or MDR TB infection. In summary, increasing degrees of drug resistance were associated with decreasing murine virulence in this collection of KZN strains as measured by all four virulence phenotypes. The predominance of HIV-infected patients in MDR and XDR TB outbreaks may be explained by decreased virulence of these strains in humans.

Figure 1. Necrosis and apoptosis induction by KZN M. tuberculosis strains in vitro.

Erdman and H37Rv reference strains were compared to sensitive (S), multidrug-resistant (M), and extensively drug-resistant (X) KZN strains. (A) For necrosis measurement, alveolar epithelial cells (A549) were infected with the KZN strains at MOI of 10 for 96 h and the supernatant was assayed for lactate dehydrogenase (LDH). Percentage cytotoxicity was calculated by the following formula: [release of LDH from infected cells (OD490)-release of LDH from uninfected control/maximum LDH release (OD490)] X 100. Data from four independent experiments. (B) Infected A549 cells were lysed and plated to monitor bacterial survival. (C) The percentage of apoptosis of THP1 macrophages infected with the KZN strains at an MOI of 10 was determined by TUNEL staining of DNA fragmentation. Values are means with error bars indicating standard error. Statistically significant differences in necrosis relative to that of Erdman are shown: *, P<0.05; **, P<0.001.

Figure 2. Competitive growth in murine aerosol model.

Mice were aerosolized with combinations of two (n = 10) or three (n = 12) sensitive (S), multidrug-resistant (M), and extensively drug-resistant (X) KZN strains. Mice were sacrificed 60 days later, and bacterial burden determined for each strain based on CFU on selective antibiotic media. Groups were compared by the Wilcoxon signed rank test. The data shown here are from a representative experiment repeated twice. Statistically significant differences between groups are shown: *, P<0.05; **, P<0.001.

Figure 3. Bacterial burden and murine survival after aerosol infection.

C57BL/6 mice (A, B, C) and C3H/FeJ mice (D, E, F) were infected with individual KZN strains by aerosol. The aerosolized infectious dose was determined by homogenates from 4 mice in each group at 24 hours post aerosolization. At indicated time points, four mice from each group were sacrificed and lung (A, D) and spleen (B, E) homogenates were plated on 7H10 media. CFU data are shown as means and standard errors. Spleen CFU were not assessed at week 0, and no bacteria were isolated from spleens from any mice at week 1. The survival of 8 mice from each group was compared among the infected groups (C, E). Statistically significant differences compared with V9124 (S) are shown: *, P<0.05; **, P<0.001; ***, P<0.0001.

Figure 4. Bacterial burden and murine survival after aerosol infection.

(A) C57BL/6 mice were infected with individual KZN strains by aerosol. CFU of homogenates of lungs (A) and spleens (B) were enumerated at indicated time points. Spleen CFU were not assessed at week 0, and no bacteria were isolated from spleens from any mice at week 1. Values represented are means and standard errors. (C) Cross-sections of lungs at 4 and 8 weeks post infection were stained with hematoxylin & eosin from each group. Four lungs of each group, and at each time point, were examined. One entire section of each mouse lung was evaluated. Representative lung pathology pictures are shown. Statistically significant differences between groups are shown: V9124 (S) vs TF275 (X), **, P<0.001; V2475 (M) vs TF275 (X), ##, P<0.001.

Figure 5. Survival of SCID mice infected with M. tuberculosis KZN clinical isolates.

Mice were infected by aerosol with 100–200 bacilli per mouse, which was determined by lung CFU at 24 hours post infection. (A) Survival was compared among the drug-sensitive KZN strain V9124 (S), the MDR strain V2475 (M), the lab strain Erdman, and (B) the XDR strain TF275 (X). Each group has 8 mice.