Requirement for γδ T Cells in Allergic Airway Inflammation

See allHide authors and affiliations

Science  22 May 1998:
Vol. 280, Issue 5367, pp. 1265-1267
DOI: 10.1126/science.280.5367.1265


The factors that contribute to allergic asthma are unclear but the resulting condition is considered a consequence of a type-2 T helper (TH2) cell response. In a model of pulmonary allergic inflammation, mice that lacked γδ T cells had decreases in specific immunoglobulin E (IgE) and IgG1 and pulmonary interleukin-5 (IL-5) release as well as in eosinophil and T cell infiltration compared with wild-type mice. These responses were restored by administration of IL-4 to γδ T cell-deficient mice during the primary immunization. Thus, γδ T cells are essential for inducing IL-4-dependent IgE and IgG1 responses and for TH2-mediated airway inflammation to peptidic antigens.

Allergic asthma is a chronic inflammatory disease associated with a predominant TH2 response, IgE synthesis, airway infiltration by inflammatory cells, particularly eosinophils, and bronchial hyperreactivity (1). Identification of the mechanisms involved in the in vivo commitment of naive T cells to a TH2 phenotype will aid our understanding of the initiation and maintenance of tissue inflammation. IL-4 drives TH2 responses and promotes IgE synthesis (2), but the nature of the cells that provide this cytokine after in vivo interactions among peptidic antigens, antigen-presenting cells, and TH cells remains largely elusive.

A subset of γδ T cells can produce TH2-type cytokines (3, 4) which suggests their possible participation in the development of TH2 responses and, thus, in the onset of pulmonary allergic reactions. To address this question, we backcrossed 6-week-old mice that were genetically deficient in the δ chain of the T cell antigen receptor (TCR) and developed no γδ T cells (5) to BALB/c mice for 10 generations. These γδ T cell-deficient (γδ−/−) and BALB/c wild-type (γδ+/+) mice were repeatedly immunized intraperitoneally with soluble ovalbumin (OVA) and then challenged intranasally with OVA or saline (6). We analyzed airway infiltration by inflammatory cells (7), cytokine release in the bronchoalveolar lavage (BAL) fluid (8), bronchopulmonary hyperreactivity to inhaled methacholine (9), and OVA-specific IgE and IgG1 titers in the serum (10).

Repeated intranasal OVA challenges in immunized γδ+/+mice resulted in a significant increase in the number of eosinophils and of CD4+and CD8+ T lymphocytes infiltrating the bronchial tissue (Fig. 1, A, D, and G). Eosinophils and T cells were located in the bronchial submucosa and around the blood vessels (Fig. 1, B, E, and H). Antigen-induced eosinophilia also occurred in the blood, BAL fluid, and bone marrow (11, 12). Unlike γδ+/+ mice, OVA-challenged γδ−/− mice showed only a moderate increase in the number of eosinophils in bronchial tissue, BAL fluid, blood, and bone marrow; no significant changes in T cell counts in lung tissue were observed (Fig. 1) (12).

Figure 1

Inflammatory cell distribution in the bronchial tissue of antigen-challenged γδ+/+ and γδ−/− mice. Statistical data from six tissue sections per mouse and six mice per group are plotted for eosinophils (A), CD4+ T cells (D), and CD8+ T cells (G). Cryostat lung sections from γδ+/+ (B, E, and H) or γδ−/− (C, F, andI) mice were processed for immunohistochemical analysis (7). Original magnifications, ×200 (B and C) and ×100 (E, F, H, and I). *P < 0.05; †P< 0.05 [one-way analysis of variance (ANOVA) followed by Student's t test for unpaired values] compared with saline- or OVA-challenged γδ+/+ mice, respectively. BL, bronchial lumen; V, vessel.

Airway eosinophilia in OVA-challenged γδ+/+ mice paralleled IL-5, but not interferon-γ (IFN-γ), production in the BAL fluid (Fig. 2), a finding consistent with selective induction of a TH2 response in the airways. In contrast, γδ−/− mice failed to release IL-5 in response to intranasal administration of OVA (Fig. 2A) and the amounts of IFN-γ remained very low in both saline- and OVA-challenged animals (Fig. 2B). High concentrations of IL-4 were detected in the BAL fluid of saline-challenged γδ+/+and γδ−/− mice (445.6 ± 98.9 and 280.8 ± 42.9 pg/ml, respectively). These quantities decreased slightly after intranasal OVA challenge in both γδ+/+ and γδ−/− mice (210.6 ± 15.1 and 189.4 ± 28.5 pg/ml, respectively), which suggests that IL-4 release in the lung is not enhanced under these experimental conditions.

Figure 2

Concentrations of IL-5 (A) and IFN-γ (B) in the BAL fluid of OVA-immunized γδ+/+ and γδ−/− mice. Seventy-two hours after the final intranasal saline or OVA challenge, cell-free supernatants from BAL fluids were harvested and assayed for IL-5 and IFN-γ production by enzyme-immunometric assay and ELISA, respectively (8). Results are means ± SEM (vertical bars) of five or six mice per group. *P < 0.05; †P< 0.05 (one-way ANOVA followed by Student'st test for unpaired values) compared with saline- or OVA-challenged γδ+/+ mice, respectively. Dotted line in (B) indicates the sensitivity of the assay.

Despite the differences in OVA-induced airway eosinophilia and IL-5 release in the BAL fluid, γδ+/+ and γδ−/− mice developed similar bronchial hyperreactivity to methacholine (13). This may be due to the comparable concentrations of serum IgE observed in both types of mice after intranasal OVA challenge (see below). The dissociation between eosinophil infiltration into the airways and bronchial hyperreactivity has been observed in other models (14).

To determine whether impaired pulmonary inflammation observed in γδ−/− mice resulted from an inefficient antigen priming at the periphery or from a defective response in the lung, we analyzed serum titers of OVA-specific IgE and IgG1 in OVA-immunized γδ+/+ and γδ−/− mice that were challenged intranasally with either saline or OVA (Fig.3). Repeated intraperitoneal injections of OVA into γδ+/+ mice resulted in high titers of OVA-specific IgG1 and in production of low, but detectable, concentrations of OVA-specific IgE (Fig. 3, saline). This same protocol of immunization elicited only 1/100th of the production of OVA-specific IgG1 and undetectable concentrations of IgE in γδ−/−compared with wild-type mice (Fig. 3, saline). Intranasal OVA challenge boosted the specific response to comparable levels in both γδ+/+ and γδ−/− mice (Fig. 3, OVA). Decreased specific IgE and IgG1 titers in OVA-immunized saline-challenged γδ−/− mice were not compensated by an increase in the number of OVA-specific antibodies of other isotypes (15). Thus, the peripheral immune response to soluble OVA that is generated as a consequence of the multiple intraperitoneal injections of antigen was impaired in γδ−/− mice. This may be the basis of the reduction in pulmonary inflammation after intranasal antigen challenge. The possibility that residual 129/Sv background genes present in the backcrossed γδ−/−mice could account for the differences presented here was minimized by analysis of OVA-induced peripheral and pulmonary responses in BALB/c, 129/Sv, and (BALB/c × 129/Sv) F1 animals, which were found to be comparable (16). The possibility still exists, however, that a recessive 129/Sv gene interacts with a BALB/c gene to generate the observed phenotype. Experiments conducted with TCRγ−/− mice could further clarify this issue. However, the organization of the TCRγ locus in the mouse, with four different Cγ regions spanning several megabases in the chromosome, prevents the production of these mice by currently available techniques.

Figure 3

OVA-specific IgE and IgG1 titers in the sera of immunized γδ+/+ and γδ−/− mice. All animals were immunized with OVA and challenged with either saline or OVA (6). Blood was harvested from post vena cava 72 hours after the final intranasal saline or OVA challenge and sera were tested for the presence of IgE (A) and IgG1 (B) by ELISA (10). Results are means ± SD (vertical bars) of five animals per group. One experiment representative of three is shown.

OVA-induced pulmonary responses largely depend on early production of IL-4, because airway eosinophilia and local IL-5 release are prevented by administration of antibodies to IL-4 during the peritoneal immunization period (17). Because some γδ T cells secrete IL-4 (3, 4), we postulated that the impaired immune response and allergic airway inflammation observed in OVA-immunized γδ−/− mice could result from a lack of early IL-4 production. To verify this hypothesis, we injected γδ−/− mice intraperitoneally during the immunization period with a complex of active IL-4 and a monoclonal antibody (mAb) to IL-4 to increase half-life and prolong IL-4 activity in vivo (18). Seven days after the last OVA immunization, IL-4-treated γδ−/− mice had OVA-specific serum IgE (Fig. 4A) and IgG1 (Fig. 4B) concentrations similar to those of γδ+/+ mice (compare Fig. 4, A and B, with Fig. 3). IL-4 administration also rendered γδ−/− mice prone to respond to OVA stimulation with eosinophil accumulation and IL-5 release into the BAL fluid (Fig. 4, C and D). Thus, γδ T cells were essential for initial IL-4 production, early IgE and IgG1 synthesis, and development of a TH2 response in the airways.

Figure 4

In vivo IL-4 administration restored TH2 responses in OVA-immunized γδ−/−mice. Mice were untreated or were injected intraperitoneally with a mixture of 5 μg of recombinant murine IL-4 and 50 μg of rat mAb to murine IL-4 (11B11) (18) at the time of the first OVA immunization (day 0) and on days 6 and 12. Blood was collected from the retroorbital plexus on day 21 (7 days after the last intraperitoneal OVA immunization) and from post vena cava 72 hours after the final intranasal antigen challenge (day 49) and serum concentrations of OVA-specific IgE (A) and IgG1 (B) were quantitated (10). The same animals were used for eosinophil (C) and IL-5 (D) determination in BAL fluid [legend to Fig. 2; see (8,11)]. Results are means ± SEM (vertical bars) of four or five mice per group. *P < 0.05 compared with IL-4-untreated mice (one-way ANOVA followed by Student'st test for unpaired values).

Mast cells, CD8+ TCRαβ cells and NK1.1+ TCRαβ T cells also secrete IL-4 (19), which suggests that all these cell types have the capacity to initiate a TH2 response in vivo. However, mast cell-deficient mice have reduced eosinophil accumulation in the BAL fluid but no changes in OVA-specific IgE and IgG1 (20), indicating that factors originating from mast cells participate in the onset of pulmonary eosinophilia but not in development of the humoral response. The analysis of OVA-driven specific IgE response, BAL eosinophilia, and the expression of TH2-type cytokines in bronchial lymph nodes of β2-microglobulin (β2M)-deficient mice, which lack both CD8+ TCRαβ and NK1.1+TCRαβ T cells, showed that these cells are not required for TH2 cell-mediated in vivo pulmonary allergic reactions (21). Together, these observations suggest that the source of IL-4 early in this response may be the γδ T cells themselves. IL-4-producing γδ T cells are present in mutant mice lacking β2M, because they are selected independently from β2M-associated class I molecule expression (4). By releasing IL-4, such TCRγδ cells may actively participate in the initiation of TH2 immune responses, as shown for IgE and IgG1 production in αβ T cell-deficient mice (22). Further characterization of the subset of peripheral γδ T cells directly involved in IL-4 production will represent an important step for understanding and modulating the development of TH2 responses, particularly in the context of allergic diseases such as bronchial asthma.

  • * Present address: Novartis Horsham Research Centre, Horsham, UK.

  • These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: mpretol{at}


Stay Connected to Science

Navigate This Article