Pro-inflammatory role of natural killer cells in the development of allergic airway disease

Abstract

Background: Natural Killer (NK) cells have been implicated in the development of allergic airway inflammation. However, the in vivo role of NK cells has not been firmly established due to the lack of animal models with selective deficiencies in NK cells.

Objective: To determine the specific contribution of NK cells in a murine model of allergic airway disease (AAD).

Methods: The role of NK cells in AAD was studied using NK-deficient (NKD) mice, perforin(-/-) mice, and mice depleted of Ly49A/D/G(+) NK cell subsets in an ovalbumin-induced model of allergic airway disease (OVA-AAD).

Results: Induction of OVA-AAD in C57BL/6 wild-type (WT) mice resulted in the expansion of airway NK cells and the development of pronounced airway eosinophilia. In the absence of NK cells or specific subsets of NK cells, either in NKD mice, or after the depletion of Ly49A/D/G(+) NK cells, the development of OVA-AAD was significantly impaired as seen by decreased airway inflammation and eosinophilia, decreased secretion of the Th2 cytokines IL-4, IL-5 and IL-13 and diminished OVA-specific antibody production. Furthermore, while OVA-exposure induced a dramatic expansion of dendritic cells (DCs) in WT mice, their induction was significantly attenuated in NKD mice. Development of OVA-AAD in perforin(-/-) mice suggested that the proinflammatory role of NK cells is not dependent on perforin-mediated cytotoxicity. Lastly, induction of allergic disease by OVA-specific CD4 T cells from WT but not NK-depleted or NKD mice in RAG(-/-) recipients, demonstrates that NK cells are essential for T cell priming.

Conclusions and clinical relevance: Our data demonstrate that conventional NK cells play an important and distinct role in the development of AAD. The presence of activated NK cells has been noted in patients with asthma. Understanding the mechanisms by which NK cells regulate allergic disease is therefore an important component of treatment approaches.

Keywords: allergic airway disease; allergic inflammation; asthma; natural killer T cells; natural killer cells.

Figure 1. Inflammation and eosinophilia in NKD mice induced with OVA-AAD

(A) BAL differential count indicating the total numbers of inflammatory cells as assessed by modified Giemsa stain. (B) Total numbers of T and B cells in BAL. (C) Measurement of total BAL IL-5, IL-13 and IFN-γ by ELISA 3 days after OVA-aerosol challenge. (D) IL-4 production by OVA peptide stimulated splenic cells measured by ELISPOT assay 7 days after OVA-aerosol challenge. (E) serum OVA-specific IgE and (F) IgG1 measured by ELISA. Three or more independent experiments were performed. n = 7–10 mice/group. * = p < 0.01, ** = p < 0.05.

Figure 2. Elevated numbers of NK and NKT cells in the airways of WT OVA-OVA mice

(A) BAL FACS profile of WT OVA-OVA mice depicting percentages of NK and NKT cells. Sufficient BAL cells could not be obtained from WT controls and NKD mice for FACS analysis. (B) Absolute numbers of NK and NKT cells in the lungs of naïve and OVA-OVA WT mice as enumerated by flow cytometry. Three or more independent experiments were performed. n=5 mice/group. *= p ≤ 0.0005.

Figure 3. Lung histopathology of wildtype and NK deficient mice

The top panels represent naïve non-OVA exposed control mice for (A) WT and (B) NKD mice. The bottom panels represent (A) WT and (B) NKD mice exposed to OVA 7 days post-OVA aerosol challenge. Three or more independent experiments were performed.

Figure 4. Inhibition of airway inflammation in mice depleted of Ly49 A/D/G⁺ NK cells

(A) WT mice were injected with rat anti-Ly49 A/D/G every 8 days. Percentages of Ly49⁺ NK cells in lung tissue by flow cytometry is shown: (left) undepleted; (right) depleted. Control mice were either untreated or treated with nonspecific rat IgG. (b,c) BAL cell distribution in OVA-OVA mice 7 days after OVA-aerosol challenge is shown. (B) total numbers of individual inflammatory cells, and (C) relative abundance of the respective populations as assessed by modified Giemsa stain. (D) serum OVA-specific IgE determined by ELISA. (E) The top panel represents control WT OVA-OVA exposed mice treated with non-specific rat IgG. The bottom panel represents Ly49 A/D/G depleted groups 7 days after OVA- aerosol challenge. Two independent experiments were performed. n=7–10. *=p<0.01, **=p<0.05.

Figure 5. NKT cells are activated after sensitization with OVA-alum

(A&B) WT mice were injected with OVA-alum i.p. and 2 and 12 hrs. later CD69⁺ NK and NKT cells were evaluated by flow cytometry. Gated NK1.1⁺ cells are shown in the peripheral blood (PBL) and lung respectively. Two or more independent experiments were performed.

Figure 6. Evaluation of DC numbers and antigen presentation capacity in allergic WT and NKD mice

Total number of CD11c+ DCs in (A) the spleen and (B) the lungs of WT and NKD mice 3 days after OVA-aerosol challenge. (C) Total numbers of airway DCs 7 days after OVA-aerosol challenge. Lung tissue from several mice was pooled and DCs were isolated using CD11c magnetic beads. (D) Expression of surface molecules on airway DCs shown in C. (E) Proliferation of naïve CD4 T cells from BALB/c mice to spleen or lung DCs from naïve or 7 day OVA-OVA WT and NKD mice is shown. Allogeneic T cells were isolated using magnetic beads and positive selection. Controls are syngeneic CD4 T cells from WT and NKD OVA-OVA spleen and lungs stimulated with corresponding DCs. Two independent experiments were performed. n=4–5 mice/group. *=p<0.01.

Figure 7. Perforin-deficient mice develop AAD similar to WT mice

(A) BAL responses of WT and perforin-deficient OVA-OVA mice 7 days after OVA-aerosol challenge. Differential counts and relative abundance of individual populations are as assessed by modified Giemsa stain. Cytokines (IL-5 and IFN-γ) and serum OVA-specific IgE were measured by ELISA 3 and 7 days after OVA-aerosol challenge respectively. (n=7) (B) Lung histology of perforin-deficient mice: the Left panel represents non-OVA exposed control perforin-deficient mice. The right panel represents perforin-deficient OVA-OVA mice after 7 days of OVA-aerosol challenge.

Figure 8. Adoptive transfer of OVA-specific T cells from WT, NK-depleted and NKD mice into RAG^−/− recipients

Numbers and percentages of BAL cells in OVA-aerosol challenged RAG^−/− mice adoptively transferred with T cells from either naïve WT mice or OVA-alum sensitized (primed) WT and NKD mice are shown. Donor OVA-alum sensitized mice were either treated with anti-NK1.1 monoclonal antibodies before the priming phase (WT, Priming) or untreated (WT, untreated). The WT, Priming + Challenge group represents RAG^−/− mice that received T cells from NK depleted OVA-alum sensitized mice and that were continuously depleted of NK cells during OVA-aerosol challenge. CD4 T cells were isolated using magnetic beads and negative selection. (A) BAL differential count indicating the total numbers of individual inflammatory cells, and (B) the relative distribution of macrophages (Mac), lymphocytes (lymph), neutrophils (PMN), and eosinophils (eosin) as assessed by modified Giemsa stain. **=p<0.05 compared to WT (untreated). n = 5 mice/group.