Genetically determined susceptibility to tuberculosis in mice causally involves accelerated and enhanced recruitment of granulocytes

Abstract

Classical twin studies and recent linkage analyses of African populations have revealed a potential involvement of host genetic factors in susceptibility or resistance to Mycobacterium tuberculosis infection. In order to identify the candidate genes involved and test their causal implication, we capitalized on the mouse model of tuberculosis, since inbred mouse strains also differ substantially in their susceptibility to infection. Two susceptible and two resistant mouse strains were aerogenically infected with 1,000 CFU of M. tuberculosis, and the regulation of gene expression was examined by Affymetrix GeneChip U74A array with total lung RNA 2 and 4 weeks postinfection. Four weeks after infection, 96 genes, many of which are involved in inflammatory cell recruitment and activation, were regulated in common. One hundred seven genes were differentially regulated in susceptible mouse strains, whereas 43 genes were differentially expressed only in resistant mice. Data mining revealed a bias towards the expression of genes involved in granulocyte pathophysiology in susceptible mice, such as an upregulation of those for the neutrophil chemoattractant LIX (CXCL5), interleukin 17 receptor, phosphoinositide kinase 3 delta, or gamma interferon-inducible protein 10. Following M. tuberculosis challenge in both airways or peritoneum, granulocytes were recruited significantly faster and at higher numbers in susceptible than in resistant mice. When granulocytes were efficiently depleted by either of two regimens at the onset of infection, only susceptible mice survived aerosol challenge with M. tuberculosis significantly longer than control mice. We conclude that initially enhanced recruitment of granulocytes contributes to susceptibility to tuberculosis.

FIG. 1.

Depletion of granulocytes after treatment with either RB6-8C5 or anti-GCSF. Mice were depleted of granulocytes with either 100 μl RB6-8C5 antibody 1 day before and 3 days after infection or 100 μl anti-GCSF antiserum 5 and 3 days before infection. The appropriate amount of IgG was injected as a control. Cells were counted in blood smears from several animals at different time points. Data shown are means ± standard deviations for four mice at 1 day after the last application of the antibody or IgG.

FIG. 2.

Growth of M. tuberculosis in the lungs of different mouse strains. Resistant (C57BL/6, filled circles; BALB/c, filled triangles) and susceptible (DBA/2, unfilled circles; CBA/J, unfilled triangles) mouse strains were infected with ∼1,000 CFU M. tuberculosis (H37Rv) per aerosol. Lungs were removed 14 and 28 days after infection, and numbers of CFU were determined. Means of six mice of each strain (±SD) are shown. ***, P ≤ 0.001; error bars are included but are too small to be seen.

FIG. 3.

mRNA expression of molecules involved in granulocyte function in the lungs of different mouse strains infected with M. tuberculosis. DBA/2 (black bars), CBA/J (dark-gray bars), BALB/c (light-gray bars), or C57BL/6 (white bars) mice were infected with ∼1,000 CFU M.tuberculosis (H37Rv) per aerosol. Lungs were removed 28 days after infection and processed for quantitative RT-PCR of LIX, PI3-kinase delta, IP-10, IL-17 receptor, TNF, and MIG. Means for three mice per group (± SD) normalized for HPRT expression are shown. n.s., not significant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG. 4.

Early enhanced recruitment of granulocytes into the bronchoalveolar space in susceptible mouse strains infected with M. tuberculosis. C57BL/6 and DBA/2 mice were infected aerogenically with ∼1,000 CFU/mouse M. tuberculosis (H37Rv). Lungs were washed five times each at indicated time points, and lavage cells were analyzed flow cytometrically using the FITC-labeled anti-Ly6G antibody RB6-8C5 (Gr1). Granulocytes were determined as Gr1-positive cells in region 1 (R1, A). The number of recovered cells was between 3 × 10⁵ and 6 × 10⁵ depending on the time of infection. Mean percentages of granulocytes from C57BL/6 (white bars) and DBA/2 (black bars) as defined by Gr1-positive cells in R1 are shown (B). Means of two to four independent determinations (±SD) are shown. *, P ≤ 0.05; **, P ≤ 0.01.

FIG. 5.

Immunohistological detection of RB6-8C5-positive cells in the lungs of different mouse strains infected with M. tuberculosis. Resistant C57BL/6 (A) or BALB/c (B) or susceptible DBA/2 (C) or CBA/J (D) mice were aerogenically infected with 1,000 CFU M. tuberculosis H37Rv. Cryosections (2 to 3 μm) were prepared from frozen lungs removed at 28 days postinfection. Immunohistological staining was performed with the monoclonal antibody RB6-8C5. Shown are representative results of four mice per group (magnification, ×100; insert: magnification, ×400).

FIG. 6.

Early and enhanced recruitment of granulocytes in peritoneal exudates of susceptible mice. (A) DBA/2 (black bars), CBA/J (dark-gray bars), BALB/c (light-gray bars), or C57BL/6 (white bars) mice were infected intraperitoneally with 2.5 × 10⁷ CFU M. tuberculosis (H37Rv). After 1, 2, 3, and 4 h, peritoneal cavities were washed with prewarmed PBS. (B) DBA/2 (black bars) or C57BL/6 (white bars) mice were injected with neutrophil chemoattractants NAP 2, LIX, and thioglycolate. After 3 h, peritoneal cavities were washed with prewarmed PBS. Granulocyte detection was performed flow cytometrically using the FITC-labeled anti-Ly6G antibody RB6-8C5. The total number of recovered cells was between 2 × 10⁶ and 7 × 10⁶ cells/ml depending on the time after infection. Means of triplicate determinations (±SD) are shown. n.s., not significant; ***, P ≤ 0.001.

FIG. 7.

Depletion of granulocytes during M. tuberculosis infection prolongs survival only in susceptible mice. DBA/2 (A, C) and C57BL/6 (B, D) mice were depleted of granulocytes by intraperitoneal injection of either RB6-8C5 (A, B) antibody or a sheep anti-GCSF antiserum (C, D) (unfilled circles). As control, mice were injected with the appropriate IgG (filled circles). Subsequently, DBA/2 mice were aerogenically infected with 1,009 CFU (RB6-8C5) or 880 CFU (aGCSF) and C57BL/6 mice with 541 CFU (RB6-8C5) or 1,093 CFU (aGCSF) M. tuberculosis (H37Rv) (5 to 10 mice per group) and observed daily for weight loss. Moribund mice were sacrificed. n.s., not significant; d, day; *, P ≤ 0.05; **, P ≤ 0.01.