Contribution of virus-specific CD8+ cytotoxic T cells to virus clearance or pathologic manifestations of influenza virus infection in a T cell receptor transgenic mouse model

Abstract

The ability of influenza virus to evade immune surveillance by neutralizing antibodies (Abs) directed against its variable surface antigens provides a challenge to the development of effective vaccines. CD8+ cytotoxic T lymphocytes (CTLs) restricted by class I major histocompatibility complex molecules are important in establishing immunity to influenza virus because they recognize internal viral proteins which are conserved between multiple viral strains. In contrast, protective Abs are strain-specific. However, the precise role of effector CD8+ CTLs in protection from influenza virus infection, critical for understanding disease pathogenesis, has not been well defined. In transgenic mice with a very high frequency of antiinfluenza CTL precursors, but without protective Abs, CD8+ CTLs conferred protection against low dose viral challenge, but exacerbated viral pathology and caused mortality at high viral dose. The data suggest a dual role for CD8+ CTLs against influenza, which may present a challenge to the development of effective CTL vaccines. Effector mechanisms used by CD8+ CTLs in orchestrating clearance of virus and recovery from experimental influenza infection, or potentiation of lethal pathology, are discussed.

Figure 1

Protection of mice against influenza A virus infection was observed by reduction of viral titers in lungs and survival rate, correlating with levels of antiviral Abs. F5-Tg or C57BL/10 control mice were infected with A/NT/60/68 (A and B, E and F, and I and J) or X31 (C and D, G and H, and K and L) influenza A viruses. (A–D) Mice were infected with 10⁷ PFU (filled circles) or ≤10⁶ PFU (filled triangles) of influenza A virus, and percent survival is shown for groups of 10–15 mice. (E–H) Titer of antiviral IgG (open circles) or IgM (filled circles) in serum of mice infected with 10⁶ PFU of influenza A virus was determined. Values shown for Ab activity are mean log (ELISA titer) ± SEM of three mice. (I–L) Virus in lungs was measured in separate groups of mice infected with 10⁶ PFU of influenza A virus. The virus titer is shown as mean log10 TCID50 per gram of lung ± SEM of three to five mice.

Figure 2

In the absence of protective antiviral Abs, CTLs in the lung tissue conferred protection against low viral challenge, but exacerbated disease at high viral doses. F5–RAG-1^−/− (filled circles) or RAG-1^−/− (open circles) mice were infected with A/NT/60/68 (left) or X31 (right) influenza A viruses. Percent survival is shown for groups of 10–15 mice. The virus was administrated intranasally at a dose of (A) 10⁷, (B) 10⁶, (C) 10⁵, (D) 10⁴, or (E) 10² PFU. Percent survival was significantly greater in F5–RAG-1^−/− mice infected with 10⁷ (P <0.0001 by Wilcoxon test) or 10⁵ PFU (P <0.0001) of A/NT/60/68 than in control infected RAG-1^−/− mice (A and B, left). No significant differences (P >0.09) were observed between the same strains of mice infected with X31 (A and B, right).

Figure 3

CTLs are protective against lethal influenza if they are able to control viral replication in lung tissue at the onset of viral infection. Virus titers in lung tissue of F5–RAG-1^−/− (filled circles) or RAG-1^−/− (open circles) mice infected with A/NT/60/68 (left) or X31 (right) influenza A viruses are expressed as mean log10 TCID50 per gram of lung of three to five mice. Mice are infected intranasally at doses of (A) 10⁷, (B) 10⁵, (C) 10³, or (D) 10 PFU.

Figure 4

Kinetics of total inflammatory cells versus transgenic CTLs in BAL of mice infected with influenza A virus. F5–RAG-1^−/− mice were infected intranasally with (A) 10⁷, (B) 10⁴, or (C) 10² PFU of A/ NT/60/68, or with (D) 10⁷ PFU of control X31 virus (squares). In addition, (D) RAG-1^−/− mice infected with 10⁷ PFU of A/ NT/60/68 were included as controls (triangles). The numbers of inflammatory cells in BAL (open symbols) are indicated as mean ± SEM log10 per lung of three to five mice. BAL samples (total volume 1 ml per lung) containing <10⁴ cells per ml (the limit of detection of our hemocytometer counting assay) were estimated as 10⁴ cells per lung. Tg-CTLs (filled symbols) were detected in the same samples by staining cells with Abs specific for Vβ11, CD8 and analysis by flow cytometry. Absolute numbers of Tg-CTLs were calculated as percent transgenic positive cells by flow cytometry multiplied by total cell number. Populations <1% were considered undetectable.

Figure 5

Histological examination of lung tissue. Lungs were taken on day 6 or 8 from F5–RAG-1^−/− transgenic mice infected with (A) 10² PFU of A/NT/60/68, (B) 10⁷ PFU of A/NT/60/68, or (C) 10⁷ PFU of X31 influenza A virus. (D) Lung tissue of a control uninfected F5–RAG-1^−/− mouse.

Figure 6

The effect of administration of anti–IFN-γ on mortality, elimination of pulmonary virus, and numbers of Tg-CTLs versus total cell counts in the BAL of influenza virus–infected F5–RAG-1^−/− mice. F5– RAG-1^−/− mice were infected intranasally with a lethal dose of 10⁷ PFU (left) or a sublethal dose of 10² PFU (right) of A/NT/60/68. Mice were treated with anti–IFN-γ mAb (filled circles) as described in Materials and Methods, or with PBS as control (open circles). (A and B) Percent survival is shown for groups of 5–10 mice. (C and D) Virus titers in lung tissue were expressed as mean log10 TCID50 per gram of lung of three to five mice. (E and F) Tg-CTLs were detected in the same samples by staining cells with Abs specific for Vβ11, CD8 and analysis by flow cytometry, and were expressed as mean ± SEM log10 per lung of three mice. (G and H) The numbers of inflammatory cells in BAL were indicated as mean ± SEM log10 per lung of three mice.