Tumor necrosis factor-alpha promotes myocarditis in female mice infected with coxsackievirus B3 through upregulation of CD1d on hematopoietic cells

Abstract

Coxsackievirus B3 (CVB3) induces cardiac inflammation (myocarditis) in male but not female C57BL/6 mice. Protection of females correlates with reduced expression of TNF-alpha and IL-1beta at both the mRNA and protein levels in the heart. Treatment of females with 300 ng/mouse of recombinant TNF-alpha on days +1 and +3 relative to infection restores myocarditis susceptibility to levels approximating those of infected male mice, showing that TNF-alpha deficiency is central to disease resistance. Female mice express little CD1d on spleen lymphocytes or cardiac myocytes, while females treated with TNF-alpha show increased CD1d expression in both cell populations. TNF-alpha treatment of male or female CD1d knockout (CD1dKO) mice failed to restore myocarditis susceptibility, demonstrating that of the multiple potential TNF-alpha activities, its ability to upregulate this non-classical major histocompatibility complex antigen is its dominant function in myocarditis susceptibility. Bone marrow chimeric mice were produced between female C57BL/6 and C57BL/6 CD1dKO mice so that either hematopoietic or non-hematopoietic cells were CD1d deficient. TNF-alpha treatment of chimeric mice having wild-type (CD1d+) hematopoietic cells restored myocarditis susceptibility, while TNF-alpha treatment of chimeric mice having CD1dKO hematopoietic cells, but CD1d+ myocytes, failed to develop myocarditis, showing that CD1d expression in lymphoid cells controls disease susceptibility.

FIG. 1.

Susceptibility to CVB3-induced myocarditis correlates with TNF-α expression in male and female mice. Male and female C57BL/6J mice were infected with 10² PFU of CVB3 and killed 7 d later. The hearts were removed and evaluated for (C) myocarditis by image analysis, and (D) cardiac virus titer per gram of heart tissue by plaque-forming assay. The hearts were also homogenized and the supernatants evaluated by (B) ELISA for TNF-α and IL-1β. RNA extracted from heart tissue was evaluated for mRNA expression by real-time PCR (A). Results represent mean ± SEM of 4–6 mice/group. Values for males were significantly different than those for females (*p < 0.05; **p < 0.01).

FIG. 2.

TNF-α increases CD1d expression on cardiac myocytes and spleen lymphocytes. Male and female C57BL/6 and C57BL/6 CD1d KO mice were infected with 10² PFU of CVB3. Half the mice were injected on days + 1 and +3 relative to infection with 300 ng/mouse of recombinant TNF-α. The mice were killed 7 d later and evaluated for CD1d expression in the (A) heart by confocal microscopy, and for (B) CD1d expression on spleen lymphocytes by flow cytometry. For confocal microscopy, the hearts were labeled with mouse anti-α-actinin and Alexa Fluor 488-anti-mouse IgG (green), rat anti-CD1d and Alexa Fluor 568-anti-rat IgG (red), and DAPI (blue nuclei). For flow cytometry, the lymphocyte population was gated by forward versus side scatter, and mean fluorescence intensity (MFI) of the total lymphocyte population was determined. Representative data are from one mouse of the 4–6 mice per experimental group.

FIG. 3.

TNF-α promotes myocarditis susceptibility through CD1d upregulation. Hearts from infected male and female wild-type (WT) C57BL/6 and C57BL/6 CD1d KO mice were evaluated for (A) myocarditis by image analysis, and for (B) cardiac virus titers by the plaque-forming assay on HeLa cells. Half of the mice received 300 ng/mouse of recombinant TNF-α on days +1 and +3 relative to infection. Results represent mean ± SEM of 4–6 mice/group.

FIG. 4.

Images of the representative histology of the mice described in Fig. 3. The hearts were fixed in formalin, sectioned, and stained with hematoxylin and eosin. Areas of inflammation are indicated by the arrows.

FIG. 5.

TNF-α promotes myocarditis susceptibility through its effects on hematopoietic cells. Bone marrow chimeric mice were produced by irradiating recipient female mice and injecting bone marrow from female donors IV into the tail vein. The bone marrow recipients were rested for 6 wk, then injected with 10² PFU of CVB3 IP. Some mice also received 300 ng/mouse of recombinant TNF-α IP on days +1 and +3 relative to infection. All the mice were euthanized 7 d after infection, and the hearts were evaluated by image analysis for myocarditis. The first bar indicates bone marrow donors, and the second bar represents bone marrow recipients (C57BL/6 wild-type [WT] bone marrow injected into irradiated C57BL/6 WT recipients; e.g., WT → WT). Results represent mean ± SEM of 4–5 mice/group (ns, not significant; **p < 0.01).