Bacterial Strain-Dependent Dissociation of Cell Recruitment and Cell-to-Cell Spread in Early M. tuberculosis Infection

Abstract

In the initial stage of respiratory infection, Mycobacterium tuberculosis traverses from alveolar macrophages to phenotypically diverse monocyte-derived phagocytes and neutrophils in the lung parenchyma. Here, we compare the in vivo kinetics of early bacterial growth and cell-to-cell spread of two strains of M. tuberculosis: a lineage 2 strain, 4334, and the widely studied lineage 4 strain H37Rv. Using flow cytometry, live cell sorting of phenotypic subsets, and quantitation of bacteria in cells of the distinct subsets, we found that 4334 induces less leukocyte influx into the lungs but demonstrates earlier population expansion and cell-to-cell spread. The earlier spread of 4334 to recruited cells, including monocyte-derived dendritic cells, is accompanied by earlier and greater magnitude of CD4⁺ T cell activation. The results provide evidence that strain-specific differences in interactions with lung leukocytes can shape adaptive immune responses in vivo. IMPORTANCE Tuberculosis is a leading infectious disease killer worldwide and is caused by Mycobacterium tuberculosis. After exposure to M. tuberculosis, outcomes range from apparent elimination to active disease. Early innate immune responses may contribute to differences in outcomes, yet it is not known how bacterial strains alter the early dynamics of innate immune and T cell responses. We infected mice with distinct strains of M. tuberculosis and discovered striking differences in innate cellular recruitment, cell-to-cell spread of bacteria in the lungs, and kinetics of initiation of antigen-specific CD4 T cell responses. We also found that M. tuberculosis can spread beyond alveolar macrophages even before a large influx of inflammatory cells. These results provide evidence that distinct strains of M. tuberculosis can exhibit differential kinetics in cell-to-cell spread which is not directly linked to early recruitment of phagocytes but is subsequently linked to adaptive immune responses.

Keywords: Beijing strain; Mycobacterium tuberculosis; T cell activation; T cell priming; dendritic cells; innate immunity; innate response; macrophage; strain diversity; tuberculosis.

FIG 1

M. tuberculosis 4334 grows more rapidly but induces less cell recruitment than H37Rv in early infection. C57BL/6 mice were infected by aerosol with 200 to 300 CFU/ of M. tuberculosis H37Rv, H37RvΔRD1 (ΔRD1), or 4334. Lungs were harvested at day 1, 3, 6, 8, 10, and 14 post infection, and processed to single cell suspensions. (A) Total CFU isolated from the lungs of mice infected with the individual strains during the first 14 days of infection. (B) Total number of cells isolated from the lungs of mice infected with the individual strains. Results are shown as mean ± SD in 1(ΔRD1), 2 (4334), and 6 (H37Rv) pooled experiments, with n = 5 mice per strain per day in each experiment. Statistical significance was assessed by multiple unpaired T-test for each day and Holm-Sidak multiple comparisons correction, with a 95% confidence interval and *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. For clarity, asterisks above H37Rv and 4334 are comparing only these 2 strains, and below are ΔRD1 compared to H37Rv.

FIG 2

Differential strain-dependent spread of M. tuberculosis from CD11b^neg/loCD11c^pos alveolar macrophages to recruited lung myeloid cells. Mice were infected and lungs harvested as in Fig. 1 (A) Cells were stained and analyzed by flow cytometry for quantitation of CD11b^neg/loCD11c^pos alveolar macrophages, CD11b^posCD11c^pos monocyte-derived dendritic cells, Gr-1^hiCD11c^neg neutrophils, Gr-1^intCD11c^neg monocytes, and Gr-1^negCD11c^neg monocyte-derived recruited macrophages in the lungs (see Fig. S1 and Materials and Methods for gating strategy). (B) Total CFU in each leukocyte subset after live flow cytometry sorting and plating of sorted cell populations on 7H11 solid media. Results are shown as mean ± SD in 1 (ΔRD1), 2 (4334), and 6 (H37Rv) and pooled experiments, with n = 5 mice per strain per day in each experiment. Statistically significant difference between strains was assessed multiple unpaired T-test for each day and Holm-Sidak multiple comparisons correction, with a 95% confidence interval and *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. For clarity, asterisks above H37Rv and 4334 are comparing only these 2 strains, and below are ΔRD1 compared to H37Rv.

FIG 3

M. tuberculosis strains 4334 and H37Rv do not produce PGL. PGL and PDIM analysis of M. tuberculosis H37Rv, HN878, and 4334 in 7H9 cultures, using electrospray ionization-quadrupole time-of-flight-mass spectrometry (ESI-QTOF-MS). Shown are representative images of (A) extracted ion chromatograms of PGL [C114H212O18+NH4]⁺ at m/z 1887.601 and PDIM [C91H180O5+NH4]⁺ at m/z 1371.417, and (B) mass spectra. Strain HN878 was included in the analysis as a known PGL producing positive control.

FIG 4

M. tuberculosis strain 4334 replicates to a greater extent than H37Rv in alveolar macrophages. (A) Cultured alveolar macrophages were infected for 48 or 72h with an MOI of 1 with M. tuberculosis H37Rv or 4334. Results are shown as mean CFU fold-increase over initial inoculum. (B) Representative flow cytometry analysis of cultured alveolar macrophages infected in vitro for 48h with M. tuberculosis H37Rv or 4334, each expressing DsRed fluorescent protein and stained with markers of necrosis (Zombie Aqua, ZA) and apoptosis (Annexin V-APC). The upper panels show the frequency of infected cells (DsRed⁺) and lower panels staining with ZA and AV for cells infected with the indicated strain 48h post infection. (C) Percentage of DsRed⁺ cultured alveolar macrophages as quantitated from flow cytometry analysis using the gating shown in panel B. (D) Frequency of cultured alveolar macrophages that are viable (ZA-AV-), necrotic (ZA⁺AV-), early apoptotic (ZA-AV⁺) or late cell death (ZA⁺AV⁺) 48h after infection with the indicated strain expressing DsRed fluorescent protein; bystanders are defined as DsRed negative cells. Results are shown as mean ± SD for n = 3 per time point and per strain. Data between cells infected with respective strains were analyzed by unpaired Student's t test for each day and Holm multiple comparisons correction, with a 95% confidence interval. Asterisks H37Rv versus 4334 (*P < 0.05, **P < 0.01, ****P < 0.0001), hashtag uninfected versus H37Rv or 4334 (^#P < 0.05, ^##P < 0.01, ^### P < 0.001).

FIG 5

Earlier dissemination of M. tuberculosis strain 4334 to lymph nodes is associated with a greater magnitude of Ag85B-specific CD4 T cell priming. (A) Bacterial load in mediastinal lymph nodes (MdLN) of mice infected 14 days earlier with M. tuberculosis H37Rv or 4334. Results are shown as mean CFU fold-increase over initial inoculum. Flow cytometry assessment (B) and absolute quantitation (C) of the proliferation of CellTrace Violet-labeled adoptively transferred naive P25TCR-Tg/CD45.1 CD4⁺ T cells in the MDLN 10-, 14-, and 17-days post infection with H37Rv or 4334. Percentages shown in panel B are the frequencies of adoptively transferred P25TCR-Tg CD4⁺ T cells in which CTV was diluted as a result of cell division (CTV^dilute). Results in panel C are shown as mean ± SD, for n = 4 mice per day and per strain. (D) Quantitation of IFNγ⁺ endogenous and P25TCR-Tg (as defined by congenic markers CD45.1/2) CD4⁺ T cells in the lungs of mice infected for 14 or 17 days with H37Rv or 4334. Results are shown as mean ± SD, for n = 4 mice per day and per strain. Flow cytometry analysis for these results is shown in Fig. S4B Differences between M. tuberculosis strains were assessed by Student's t test for each day and Holm multiple comparisons correction, with a 95% confidence interval and *P < 0.05, and **P < 0.01.