T cells are essential for bacterial clearance, and gamma interferon, tumor necrosis factor alpha, and B cells are crucial for disease development in Coxiella burnetii infection in mice

Abstract

Coxiella burnetii, the etiological agent of Q fever, has two phase variants. Phase I has a complete lipopolysaccharide (LPS), is highly virulent, and causes Q fever in humans and pathology in experimental animals. Phase II lacks an LPS O side chain, is avirulent, and does not grow well in immunocompetent animals. To understand the pathogenicity of Q fever, we investigated the roles of immune components in animals infected with Nine Mile phase I (NM I) or Nine Mile phase II (NM II) bacteria. Immunodeficient mice, including SCID mice (deficient in T and B cells), SCIDbg mice (deficient in T, B, and NK cells), nude mice (deficient in T cells), muMT mice (deficient in B cells), bg mice (deficient in NK cells), mice deficient in tumor necrosis factor alpha (TNF-alpha(-/-) mice), and mice deficient in gamma interferon (IFN-gamma(-/-) mice), were compared for their responses to infection. SCID, SCIDbg, nude, and IFN-gamma(-/-) mice showed high susceptibility to NM I, and TNF-alpha(-/-) mice showed modest susceptibility. Disease caused by NM I in SCID, SCIDbg, and nude mice progressed slowly, while disease in IFN-gamma(-/-) and TNF-alpha(-/-) mice advanced rapidly. B- and NK-cell deficiencies did not enhance clinical disease development or alter bacterial clearance but did increase the severity of histopathological changes, particularly in the absence of B cells. Mice infected with NM II showed no apparent clinical disease, but T-cell-deficient mice had histopathological changes. These results suggest that T cells are critical for clearance of C. burnetii, either NM I or NM II, that IFN-gamma and TNF-alpha are essential for the early control of infection, and that B cells are important for the prevention of tissue damage.

FIG. 1.

Body weight changes of mice infected with the low dose of NM I or NM II. (A) wt (CB17) mice; (B) SCID mice; (C) SCIDbg mice. Data for SCIDbg mice infected with the low dose of NM I at 28 DPI are not available because of death of the mice. The results (means ± standard deviations) are expressed as the body weight index. One asterisk indicates that the P value is <0.05 and two asterisks indicate that the P value is <0.001 for comparisons of infected mice with PBS-inoculated control mice.

FIG. 2.

Spleens of mice infected with the low dose of NM I or NM II: splenomegaly (A) and bacterial genome numbers (B) in the spleen at 28 DPI. The results (means ± standard deviations) are expressed as the percentage of spleen weight compared with the body weight (A) and the number of C. burnetii genomes in the whole spleen (B). One asterisk indicates that the P value is <0.05, and two asterisks indicate that the P value is <0.001. Double dagger, not detected.

FIG. 3.

Histopathology of SCIDbg mice infected with the low dose of NM I at 28 DPI. (A to E) Heart; (F to J) lung; (K to O) liver; (P to T) spleen. Panels A, F, K, and P show the results for control mice from the PBS group. HE staining or IHC staining (C, E, H, J, M, O, R, and T) was used. (A to H, K to M, and P to R) Original magnification, ×100; (S and T) original magnification, ×200; (I, J, N, and O) original magnification, ×400. Severe cellular infiltrations associated with C. burnetii antigen (brown granules with IHC staining) were observed in all tissues from infected mice. Subepicardial lesions (B and C), subendocardial lesions (D and E), and peribronchial and some perialveolar lesions (G to J) were also observed, and extensive cellular infiltrations obscured the tissue architecture in the liver and spleen associated with diffuse antigen distribution (L to O and Q to T).

FIG. 4.

Spleens of mice infected with the high dose of NM II: splenomegaly (A) and bacterial genome numbers (B) in the spleen at 28 DPI. The results (means ± standard deviations) are expressed as the percentage of spleen weight based on the body weight (A) and the number of C. burnetii genomes in the whole spleen (B). An asterisk indicates that the P value is <0.05. SCID and SCIDbg mice were significantly different from PBS-inoculated groups (P < 0.05) for splenomegaly (A). Double dagger, not detected.

FIG. 5.

Liver histopathology of mice infected with the high dose of NM II at 28 DPI. (A) SCIDbg mice had hepatocellular coagulative necrosis (arrowhead) and severe cellular accumulation (arrows). (B) SCID mice had granuloma-like cellular accumulation (arrows). (C) wt (CB17) mice had mild cellular coagulation (arrow). HE staining was used. Original magnification, ×100.

FIG. 6.

Body weight changes of mice infected with the high dose of NM I or NM II. (A) Nude mice; (B) IFN-γ^−/− mice. The double dagger indicates that the data for IFN-γ^−/− mice infected with the high dose of NM I at 14, 21, and 28 DPI are from a single mouse that survived (three of four mice died before 14 DPI). The results (means ± standard deviations) are expressed as the body weight index. An asterisk indicates that the P value is <0.05 for comparisons of infected mice with PBS-inoculated control mice.

FIG. 7.

Spleens of mice infected with the high dose of NM I or NM II: splenomegaly (A) and bacterial genome numbers (B) in the spleen at 28 DPI. The results (means ± standard deviations) are expressed as the percentage of spleen weight based on the body weight (A) and the number of C. burnetii genomes in the whole spleen (B). One asterisk indicates that the P value is <0.05 and two asterisks indicate that the P value is <0.001 for the comparison of infected mice with wt mice infected with the same inoculum. (A) Every mouse group infected with NM I had significant differences from the control group of the same mouse strain treated with PBS (P < 0.05 for wt [B6], muMT, and bg mice; P < 0.001 for nude and TNF-α^−/− mice). Double dagger, not detected.

FIG. 8.

Histopathology of mice infected with the high dose of NM I at 28 DPI. (A to C) Heart; (D to I) lung; (J to P) liver; and (Q to X) spleen. (D and Q) Tissue section of PBS group. (A to C, J to L, and Q to U) HE staining; (D to I, M to P, and V to X) IHC staining. (A to E, G, H, J to L, O, Q to W) Original magnification, ×100; (M and N) original magnification, ×200; (F, I, P, and X) original magnification, ×400. (A and B) Hearts of wt mice had few small areas of cellular coagulation in the myocardium (arrowhead). (C) Hearts of TNF-α^−/− mice had extensive cellular infiltration in the myocardium. (D) C. burnetii antigen was not detectable in PBS groups, and there was no nonspecific IHC staining. (E and F) Lungs of IFN-γ^−/− mice had C. burnetii antigen distributed in the parenchyma and not accumulated in perivascular areas. (G and M) C. burnetii antigen in wt mice was barely detectable. (H and I) Lungs of nude mice had abundant antigen distributed primarily in perivascular areas. (J and M) Livers of wt mice had perivascular, periportal, and random formation of small cellular accumulations (arrowheads) but no detectable C. burnetii antigen. (K and N) Livers of muMT mice had perivascular, periportal, and random cellular accumulation but no detectable C. burnetii antigen. (L, O, and P) Livers of IFN-γ^−/− mice had diffuse cellular infiltration and antigen distribution. (Q) Spleens of nude mice had smaller follicles than wt mice but no characteristic pathological changes. (R and V) Spleens of nude mice had diffuse cellular infiltration and antigen distribution that obscured tissue architecture. (S, W, and X) Spleens of IFN-γ^−/− mice had random cellular accumulations and scattered C. burnetii antigen distribution. (T) Spleens of muMT mice had marked random cellular accumulations (arrowheads).