BALB/c wild type (WT), BALB/c IL-3^-/-, and BALB/c DO11.10 TCR Tg mice were purchased from the Jackson Laboratory (Bar Harbor, ME). IL-3^-/- mice and IL-3Rβ^-/- mice, deficient in both β_c and β_IL-3 and G4 knock-in mice expressing GFP under the IL-4 promoter were previously described (20-22). IL-25R (IL-17RB)^-/- mice were provided from Dr. Wenjun Ouyang (Genentech). Mcpt8^DTR mice were previously reported (18). All experimental procedures were conducted according to the guidelines of the Institutional Animal Care and Use Committee.

Mice were injected into the ear pinna with 100 µg chicken ovalbumin protein (Sigma, St. Louis, MO) plus 50 µg papain (Sigma) or 500 dead Nb parasites in a 10 µl volume. Seven days later, draining cervical LN was harvested and ex vivo restimulated with OVA_323-339 peptide for 24 hours. IL-4 production was measured by ELISA or ELISPOT assay as described below. In some experiments, 250 µg anti-TSLP mAb (R and D system, Minneapolis, MN) was injected at days 0 and 3 post injection. For basophil depletion, mcpt8^DTR or control mice were intravenously injected with diphtheria toxin (750 ng/20 g body weight; Sigma) prior to immunization.

Draining cervical LN cells and blood were examined for basophils. In brief, cells were stained with anti-FcγR (clone 93) and anti-CD45 (30-F11). Cells were ex vivo stimulated with PMA (CalBiochem, San Diego, CA) and Ionomycin (CalBiochem) for 4 hours. Monensin was added in the culture for the last 2 hours of stimulation. Cells were stained for intracellular IL-4 expression. Samples were acquired using a FACSCalibur or a LSRII (Becton Dickinson, Franklin Lakes, NJ) cytometer and analyzed using FlowJo (Treestar, Ashland, OR). All antibodies used were purchased from eBioscience (San Diego, CA).

To measure T cell cytokine production harvested cells were stimulated with 10 µg OVA peptide for 24 hours. IL-4 production in the culture supernatant was measured by ELISA. In some experiment, IL-4 producing cells were determined by IL-4 ELISPOT assay. Abs used for both assays were purified anti-IL-4 (11B11) and biotinylated anti-IL-4 (BVD6-24G2). These Abs were purchased from eBioscience.

Statistical significance was determined by the Student's t-test using the Prism software (GraphPad Software, La Jolla, CA). p<0.05 was considered to indicate a significant difference.

In response to Th2-inducing allergen challenges or to parasite infections, circulating basophils are transiently recruited into the draining LN, where they are believed to play a key role in supporting the development of antigen specific Th2 type CD4 T cell responses (2). We previously reported that LN recruitment of circulating basophils after helminth infection was completely abolished in IL-3- or IL-3Rβ-deficient mice, indicating an IL-3-dependent process (8). First, we examined whether IL-3 also mediates basophil recruitment after immunization of papain, a prototype allergen shown to recruit basophils to the draining LN and induce an antigen specific Th2 immunity when coimmunized with the antigen (23). Groups of WT and IL-3-deficient mice were immunized with OVA plus papain and basophil recruitment in the draining LN was subsequently examined. Consistent with a previous report (23), injection of OVA protein plus papain induced a transient recruitment of FcγR^high CD45^intermediate basophils into the draining cervical LN of WT mice (Fig. 1). By contrast, such recruitment was completely absent in IL-3-deficient mice (Fig. 1). Therefore, papain-induced basophil LN recruitment also requires IL-3.

Helminth-induced Th2 immunity normally develops in the absence of IL-3 or IL-3Rβ (i.e., in the absence of basophil recruitment into the draining LN) (8). In case of papain-induced Th2 immunity, however, the question of whether basophils are critical for the resulting Th2 immune responses remains unresolved. Since basophil LN recruitment was found to be absent in IL-3-deficient mice immunized with papain, we therefore examined OVA specific CD4 T cell responses in IL-3- or IL-3Rβ-deficient mice after immunization with OVA protein plus papain. First, OVA-specific DO11.10 TCR Tg CD4 T cells that are WT, IL-3-, or IL-3Rβ-deficient were adoptively transferred into WT recipients, and the recipients were subsequently immunized with OVA plus papain. As shown in Fig. 2A, comparable numbers of KJ1.26⁺ CD4 T cells were found in the draining LN following immunization regardless of the phenotypes of CD4 T cells. When proliferation was examined by CFSE dilution, the pattern of CFSE dilution was indistinguishable between WT, IL-3^-/-, and IL-3Rβ^-/- DO11.10 CD4 T cells (Fig. 2B). Therefore, the production of IL-3 or IL-3 responsiveness of CD4 T cells plays little or no role in Ag-induced T cell proliferation/expansion in vivo.

Since recipient cells are fully capable of producing IL-3 in the experiments described above, we next transferred WT DO11.10 CD4 T cells into different types of recipients (WT, IL-3^-/-, and IL-3Rβ^-/-) to test whether IL-3 production or IL-3Rβ expression by the recipient cells plays any role in this process. As shown in Fig. 2C, DO11.10 CD4 T cells efficiently expanded in WT and IL-3^-/- recipients. Interestingly, in IL-3Rβ^-/- recipients the expansion was significantly reduced, which may be attributed to the defects in the expansion and maturation of antigen presenting cells in IL-3Rβ^-/- mice (20). Nevertheless, CFSE dilution profiles of Ag specific T cells in all tested recipients was not significantly different (Fig. 2D), indicating that IL-3 production or IL-3Rβ expression of the endogenous recipient cells does not play a role in activating T cells, although IL-3Rβ expression of the recipients may contribute to the expansion of Ag specific T cells.

We then examined whether the inactivation of IL-3 or IL-3Rβ on the recipient derived cells including basophils affects papain-induced Th2 immune responses. Draining LN cells from immunized recipients were restimulated ex vivo with OVA peptide and IL-4 secretion was measured by ELISA. As shown in Fig. 2E, OVA-specific IL-4 secretion was similar regardless of the IL-3 (or IL-3Rβ) deficiency in the recipient cells. Likewise, IL-3^-/- and IL-3Rβ^-/- DO11.10 CD4 T cells transferred into WT recipients and immunized with OVA plus papain also secreted similar levels of IL-4 after OVA restimulation (data not shown). Therefore, basophils (or IL-3) seem to be dispensable in papain-induced Th2 immunity.

We recently reported that injection of nonviable Nb larvae similarly induces basophil LN recruitment and robust Th2 responses and that basophil depletion using anti-FcεRI (MAR-1) mAb did not alter the responses (24). Similar to papain immunization or live helminth infection, dead Nb injection failed to recruit basophils into the draining LN in IL-3^-/- mice (Fig. 3A). Interestingly, the magnitude of basophil recruitment by dead parasites was substantially greater than that induced by papain (Fig. 1 and 3A). Dead parasite immunization also induced stronger CD4 T cell IL-4 responses when compared to that induced by OVA/papain immunization (Fig. 3B and 3C). No IL-4 production was observed without restimulation or with irrelevant peptide stimulation (data not shown). Likewise, the lack of IL-3 from either transferred DO11.10 CD4 T cells and/or endogenous cells did not affect the generation of OVA-specific IL-4-producing T cells (data not shown). Therefore, these results suggest that in vivo Th2 immunity can arise without IL-3 (i.e., basophils).

To ensure the dispensable role for basophils we utilized Mcpt8^DTR mice in which basophils can be selectively depleted by DT administration (18). WT and Mcpt8^DTR mice were transferred with DO11.10 CD4 T cells, injected with DT, and then immunized with OVA plus dead Nb. Basophil depletion was achieved by DT injection as measured by blood basophils (Fig. 4A). Consistent with earlier findings, basophil depletion by DT did not affect the development of IL-4-producing Ag specific CD4 T cell responses (Fig. 4B). Therefore, basophils are not critical for in vivo Th2 immunity.

Whether basophils function as professional antigen presenting cells in the draining LN to induce Th2 immunity has been the matter of debate. Since a stronger Th2 immunity develops after immunization of dead Nb than of papain, we exploited this model system to compare cell types responsible for antigen presentation. Mice were immunized with OVA/papain or OVA/Nb, and MHC II expression of recruited basophils (CD45^intermediate FcγR^high, population II) and conventional antigen presenting cells (CD45^high FcγR^high, population I) was compared. As shown in Fig. 5A, MHC II expression of FcγR bearing antigen presenting cells (I) was very high; >90% of them were MHC II^bright. Moreover, no obvious difference was noticed between papain or dead Nb immunization. By contrast, MHC II^bright population was dramatically lower (~10%) in basophils (Fig. 5A) and most of the basophils did not express MHC II, similar to negative control CD45^neg LN cells (population III). Therefore, basophils are not the major MHC II-expressing cells in the draining LN.

It was demonstrated that endocytosis of soluble proteins by basophils is crucial for antigen presentation (3). In fact, ~15% of basophils are MHC II^bright in OVA/papain immunized mice, thus they may still be involved in antigen presentation to naïve T cells and in Th2 immunity. Alternatively, FcεRI-expressing 'inflammatory DC' subsets may function as key Th2-inducing antigen presenting cells (7,9). To address this possibility, mice were immunized with allophycocyanin (APC)-conjugated OVA plus papain or dead Nb. Fluorescent APC uptake of CD11c⁺ (I), FcεRI⁺ CD11c⁺ (II), and FcεRI⁺ CD11c^- (III) cells was next analyzed. We found that APC-OVA uptake was predominantly found in FcεRI⁺ CD11c⁺ cells (Fig. 5B). By contrast, FcεRI⁺ CD11c^- cells (III), which primarily include basophils, were not OVA-APC⁺ (Fig. 5B), strongly indicating that basophils are unable to uptake injected antigens. Importantly, OVA-APC uptake by FcεRI⁺ CD11c⁺ cells was dramatically enhanced when dead Nb was used instead of papain (Fig. 5B), which may result in a stronger Th2 immunity (Fig. 3B and 3C). OVA-APC uptake of conventional CD11c⁺ DC was substantially lower, although it was higher upon dead Nb injection (from 1.6% to 6.9%). Overall, these results provide strong evidence that dead Nb- and possibly papain-induced Th2 immunity is likely to be induced by antigen presentation of 'inflammatory DC' subsets that express FcεRI rather than basophils.

To further investigate a mechanism(s) of how IL-3- and basophil-independent Th2 immunity develops we first tested the contribution of IL-4. WT DO11.10 CD4 T cells were transferred into WT and IL-4-deficient G4/G4 recipients. The recipients were then immunized with OVA plus dead Nb, and examined the development of OVA specific Th2 immune responses. Surprisingly, IL-4 production of OVA specific CD4 T cells was completely abolished in IL-4-deficient recipients, while robust Th2 immune responses were observed in WT recipients (Fig. 6A). Given that naïve CD4 T cells are capable of producing IL-4 sufficient for Th2 differentiation (25), these results strongly suggest that IL-4 produced by recipient non-T cells but not by Ag specific CD4 T cells in response to Nb/OVA immunization is required for the development of Th2 immune responses. Recently, innate type 2 cells such as nuocytes, MPP^type2, and NBNT cells have been implicated to play crucial roles in developing in vivo Th2 type immune responses particularly in the gut associated tissues (13,14,26). IL-25 and TSLP are candidate cytokines that activate those innate type 2 cells to promote Th2 immunity. In order to examine the contribution of innate type 2 immune components, we first used mice deficient in IL-25R (IL-17RB)^-/- as recipients of DO11.10 CD4 T cells. The recipients were immunized with OVA plus dead Nb, and subsequent OVA-specific IL-4 responses were determined. As shown in Fig. 6B, the lack of IL-25R by recipient cells played no roles in generating IL-4-producing cells, indicating that IL-25-responsive innate Th2 cells may not be involved in this process. In addition, a potential role of TSLP was examined by treating WT recipients of DO11.10 CD4 T cells with neutralizing anti-TSLP mAb after immunization of OVA/Nb as previously reported (23). TSLP neutralization did not affect the generation of OVA-specific IL-4 responses (Fig. 6C). Therefore, neither IL-25 nor TSLP plays a role in the generation of IL-4-producing CD4 T cells in vivo and that IL-4 produced by recipient derived cells plays an essential role in promoting antigen specific Th2 immunity.