Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Allergic asthma is characterized by airway hyperresponsiveness (1), allergen-specific immunoglobulin (Ig)E in serum (2), and infiltration of inflammatory cells into the airways (3). Allergen-specific CD4⁺ T-helper (Th) 2 cells are involved in the induction and effector phase of asthma. These CD4⁺ T cells not only produce interleukin (IL)-4, which stimulates IgE production and Th2 development (4), but also produce IL-5, which stimulates eosinophil maturation, attraction, and survival (7, 8). The important role of the allergen-specific T cell in the immunologic and pathologic processes of allergic asthma makes them an interesting target cell for therapy.

Recently it was reported that incubation with high doses of immunodominant T-cell epitopes can inhibit clonal expansion, cytokine production, and/or B-cell help by human and murine T-cell clones in vitro (9). In addition, it has been shown in several models of T cell-mediated autoimmune diseases that the use of immunodominant T-cell epitopes in a vaccination strategy could prevent disease induction by modulating T-cell responses (12). More importantly, Briner and colleagues (15) and Hoyne and coworkers (16) demonstrated that administration of an immunodominant T-cell epitope in primed mice reduced T-cell reactivity not only to the epitope but also to the entire protein upon challenge with the protein. Because most studies so far have been performed in autoimmune or Th1 models, little is known about the possibilities of epitope treatment on Th2-mediated immune responses. The use of allergen-derived immunodominant epitopes as treatment for allergic disorders is of extreme interest; it not only may downregulate the allergen-specific T- and B-cell responses, but also can make immunotherapy safer because it does not bear the risk of crosslinking IgE on mast cells.

To study the effects and mechanisms of immunotherapy with allergen and an allergen-derived immunodominant epitope, we used a murine model of allergic asthma in which immunologic and pathologic features are displayed similar to those observed in patients with allergic asthma (17). In this model, ovalbumin (OVA)-sensitized and -challenged Balb/c mice display high serum levels of OVA-specific IgE, airway hyperresponsiveness in vivo, infiltration of eosinophils in the airways and bronchoalveolar lavage fluid (BALF), and OVA-specific CD4⁺ Th2 cells (17, 18). We recently reported that immunotherapy by subcutaneous administration of increasing doses of OVA for 8 wk (semi-rush protocol used for human immunotherapy), after sensitization but before challenge, inhibited antigen- induced airway hyperresponsiveness, eosinophilia, and OVA-specific Th2 cytokine production (19). Moreover, we demonstrated that using three separate high doses of OVA resulted in similar effects as the semi-rush protocol. In the present study we used the murine model to study and compare the effects of immunotherapy with entire OVA and the immunodominant epitope OVA_323-339, which is reported to be responsible for 25 to 35% of the T-cell response (20) and 50% of the IgE response to entire OVA (21) in Balb/c mice.

Our data show that in sensitized animals, immunotherapy with suboptimal doses of entire OVA results in downregulation of airway hyperresponsiveness and a decreased influx of eosinophils in the BALF, concomitant with decreased production of Th2-related cytokines. In contrast, immunotherapy with the immunodominant epitope OVA_323-339 aggravates airway hyperresponsiveness and eosinophil infiltration in the BALF.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Sensitization and Challenge

Animal care and use were in accordance with the guidelines of the Dutch Committee of Animal Experiments. Specified pathogen-free male Balb/c mice (6 to 8 wk) were obtained from the Central Animal Laboratory, Utrecht, The Netherlands. Mice were housed in Macrolon cages and provided with OVA-free food and water ad libitum. Active sensitization was performed by seven intraperitoneal injections of 10 µg ovalbumin in 0.5 ml pyrogen-free saline on alternate days. Immunotherapy was performed 14 and 17 d after the last sensitization by injecting a subcutaneous, suboptimal, 150 µg/200 µl dose of OVA, OVA_323-339, or pyrogen-free saline (vehicle). Two days after the last immunotherapy treatment, mice were exposed to either OVA (2 mg/ml in pyrogen-free saline) or saline aerosols on eight consecutive days. To study the immunodominance of OVA_323-339 in the model, mice were sensitized as described above and were challenged 17 d after the last sensitization. Aerosol was generated with an ultrasonic nebulizer (particle size 3 to 5 µm; DeVilbiss) connected to a Plexiglas exposure chamber. Exposure was performed for 5 min with a maximal group of six animals.

Determination of OVA- and OVA_323-339-Specific Immunoglobulins

Serum samples were taken 24 h after the last challenge and OVA-specific IgG1, IgG2a, and IgE were determined by enzyme-linked immunosorbent assay (ELISA). To determine serum levels of OVA-specific IgE, microtiter plates (Nunc A/S, Roskilde, Denmark) were coated overnight at 4°C with 2 µg/ml recombinant human FcR1-IgG fusion protein diluted in phosphate-buffered saline (PBS). After blocking with 0.5% bovine serum albumin in ELISA buffer (2 nM ethylenediaminetetraacetic acid, 136.9 nM NaCl, 50 nM Tris, and 0.05% Tween-20), serum samples or duplicate dilution series of an OVA-specific IgE reference (10,000 U/ml) were incubated for 2 h. Reference standards were obtained by intraperitoneal immunization of mice with OVA, as described previously (17). After washing, wells were incubated for 1 h with 10 µg/ml of OVA in ELISA buffer. As second antibody, horseradish peroxidase (HRP)-conjugated goat anti-OVA antibody was used. Incubation was continued for 1 h, followed by washing procedures. Color development was performed by incubation with 10 mM o-phenylenediamine dihydrochloride (OPD) substrate solution and stopped by adding H₂SO₄ (4 M). Optical density was read at 492 nm by use of a Titretek Multiscan (Flow Labs, Irvine, UK). The detection limit of the ELISA was 0.5 U/ml. OVA-specific IgG1 and IgG2a ELISAs were performed as described elsewhere (19). The detection limits of the ELISA were 0.005 U/ml for IgG1 and 0.05 U/ml for IgG2a.

To determine the OVA_323-339-specific antibody titer, serum samples were preincubated with OVA_323-339 (1 mg/ml, 1 h) before the samples were added to the ELISA plates. As a positive control, serum samples were incubated with OVA (1 mg/ml). Preincubated and nonincubated samples were developed on the same ELISA plates.

Airway Responsiveness In Vivo

Airway responsiveness was measured in vivo 24 h after the last aerosol exposure by means of a modified plethysmograph, as described by Corry and associates (22). In short, mice were anesthetized by intraperitoneal injection of urethane (2 g/kg), and placed on a heated blanket (30°C). The trachea was then cannulated and a small polyethylene catheter was placed in the jugular vein for intravenous administrations. Spontaneous breathing of the animals was suppressed by intravenous injection of tubocurarine chloride (3.3 mg/kg). When the breathing stopped, the tracheal cannula was attached to a ventilator (C. F. Palmer, London, UK). The inflation volume of the ventilator was 0.8 ml, of which the mouse inhaled approximately 0.15 ml per breath with a rate of 200 breaths/min. Under these conditions, mice maintain physiologic arterial blood gas parameters (data not shown). Changes in resistance were measured by use of a plethysmograph coupled to a pressure transducer (M45; Validyne Engineering Corp., Northridge, CA). By use of a pulmonary mechanics analyzer (Model 6; Buxco Corp., Sharon, CT), lung resistance (RL) was measured by quantitating PT, ¹ (where PT = change in tracheal pressure and = change in flow) at points of equal volume (70% tidal volume). Changes in tracheal pressure were measured using a pressure transducer connected to the tracheal ventilation cannula, and changes in flow were measured by use of a pressure transducer connected to the plethysmograph (pressure changes were calibrated to changes in volume over the physiologic range studied). At time intervals of at least 4 min and after the response had returned to baseline level, doses of methacholine ranging from 40 to 640 µg/kg were administered via the jugular catheter. Concentrations of methacholine were prepared in saline and kept on ice for the duration of the experiment. For each dose of methacholine the increase in airway resistance was measured at its peak and expressed in cm H₂O/ml/s. At least six mice were evaluated per experimental group.

Bronchoalveolar Lavage

After completion of the dose-response curve to methacholine, the animals were lavaged five times through the tracheal cannula with 1-ml aliquots of pyrogen-free saline warmed to 37°C. The bronchoalveolar lavage (BAL) cells were washed with cold PBS (400 × g, 4°C, 5 min) and the pellet was resuspended in 150 µl cold PBS. A Burker- Türk chamber was used to count the total number of BAL cells. For differential BAL cell counts, cytospin preparations were made and stained with Diff-Quik (Merz & Dade, Düdingen, Switzerland). Cells were identified and differentiated into mononuclear cells, lymphocytes, neutrophils, and eosinophils by standard morphology.

Proliferation of Lung-Draining Lymph Node Cells

At 24 h after the last aerosol, lung-draining lymph nodes (LN) were collected from the thorax and single-cell suspensions were made. Cells (2 × 10⁵ cells/well in 96-well plates) were cultured in RPMI supplemented with 10% heat-inactivated fetal calf serum, 2 nM L-glutamine, 100 EU penicillin, 100 µg/ml streptomycin, and 50 µM -mercaptoethanol in the presence of OVA (10 µg/ml), OVA_323-339 (10 µg/ml), medium, or hen egg lysozyme (Hel) (10 µg/ml) (18). As a positive control, cells were cultured in the presence of immobilized CD3 (clone 17A2, 50 µg/ml). After 72 h of culture, cells were pulsed for 16 h with 0.3 µCi of [³H]thymidine and harvested, and proliferation was determined.

Cytokine Production of Draining LN Cells

Lung-draining LN cells were cultured for 120 h in the absence or presence of OVA (10 µg/ml) (18). As a positive control, cells were cultured in the presence of immobilized CD3. Levels of IL-4, IL-5, and interferon (IFN)- in supernatant were determined by capture ELISA (PharMingen, San Diego, CA). The detection limits of the ELISAs were 16 pg/ml for IL-4 and IL-5, and 160 pg/ml for IFN-.

Materials

OVA (chicken egg albumin crude grade V) and Hel were purchased from Sigma Chemical Company (St. Louis, MO). OVA_323-339 was synthesized by standard solid-phase g-fluorenylmethyloxy carbonyl (Fmoc) chemistry, analyzed and purified by reverse phase high-performance liquid chromatography, and checked by fast atom bombardment mass spectrometry. Recombinant human FcR1-IgG and HRP-conjugated goat anti-OVA antibody were generously provided by Dr. P. M. Jardieu, Genentech, Inc. (South San Francisco, CA). Urethane and methacholine (acetyl--methylcholine) were purchased from Janssen Chimica (Beerse, Belgium), tubocurarine chloride from Nogepha (Alkmaar, The Netherlands), and Tween-20 from Merck (Darmstadt, Germany).

Statistics

Unless stated otherwise, data are expressed as means ± standard error of the mean (SEM) and evaluated using a two-way analysis of variance (ANOVA), followed by Student's t test for comparison between two groups. A probability value of P < 0.05 was considered statistically significant. For cell types with a very low number in control animals (i.e., neutrophils and eosinophils), a Poisson distribution was assumed.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

OVA_323-339 Is an Immunodominant T-Cell Epitope of OVA

Lung-draining LN cells from OVA-sensitized animals were isolated 24 h after the last OVA or saline challenge. LN cells obtained from animals that received saline challenge did not proliferate after in vitro stimulation with OVA or OVA_323-339, but did proliferate after stimulation with immobilized CD3 (Figure 1). LN cells obtained from animals that received OVA challenge proliferated after stimulation with OVA as well as with OVA_323-339 (Figure 1). However, the proliferative response to OVA_323-339 was 30% lower than the proliferation to entire OVA. Stimulation with higher doses of OVA or OVA_323-339 did not change this difference in proliferation (data not shown), indicating that maximal responses were reached.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Proliferative response of lung-draining LN cells from OVA- and saline-challenged animals to OVA (10 µg/ml), OVA323-339 (10 µg/ml), Hel (10 µg/ml), and immobilized CD3. LN cells were cultured for 72 h, after which the proliferative response was measured by [3H]thymidine incorporation. Data are expressed as means ± SEM (n = 4 per group). *P < 0.01 compared with medium values (ANOVA).

Although LN cells obtained from animals that received saline challenge produced IFN-, IL-4, and IL-5 after CD3 stimulation, no cytokines were detected after stimulation with OVA or OVA_323-339 (data not shown). LN cells from animals that received OVA challenge produced IL-4 and IL-5 after OVA stimulation (Figure 2). Interestingly, OVA_323-339 stimulation induced LN cells to produce IL-5 but not IL-4 (Figure 2). In these cultures, IFN- could be detected only after incubation with CD3 (data not shown).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Cytokine production by lung-draining LN cells from OVA-challenged animals to OVA (10 µg/ml), OVA323-339 (10 µg/ml), Hel (10 µg/ml), and immobilized CD3. LN cells were cultured for 120 h, and IL-4 and IL-5 production was determined by ELISA. LN cells from animals that were challenged with saline did not produce IL-4 or IL-5 (data not shown). Data are expressed as means ± SEM (n = 4 per group). *P < 0.001 compared with medium values (ANOVA).

OVA_323-339 Contains a T-Cell Epitope and a B-Cell Epitope

Serum samples from OVA-challenged animals were tested for the presence of OVA- and OVA_323-339-specific antibodies. To determine the OVA_323-339 antibody titer, serum was preincubated with OVA_323-339. Preincubation of the serum with OVA_323-339 resulted in 50% blocking of OVA-specific IgG1 and IgE, suggesting that OVA-specific IgG1 and IgE consisted of at least 50% OVA_323-339-specific antibodies (Table 1). As a control, serum was incubated with entire OVA before samples were tested. This resulted in a 90% decrease of OVA-specific IgG1 and IgE antibodies compared with nonincubated samples (Table 1).

To study the effect of immunotherapy in an ongoing immune response, three different treatment groups were examined, namely control therapy (vehicle), OVA therapy, and OVA_323-339 peptide therapy. Within each treatment group, mice were challenged with either OVA or saline (Table 2).

Serum Levels of OVA-Specific IgG2a or IgE Are Not Affected by Immunotherapy, Whereas IgG1 Is Increased After OVA Therapy

Levels of OVA-specific IgG1, IgG2a, and IgE were determined 24 h after the last OVA or saline challenge.

In vehicle-treated animals, OVA-specific IgG1 serum levels were significantly increased after OVA challenge, compared with saline-challenged animals (Figure 3A). Remarkably, after OVA immunotherapy, OVA-specific IgG1 serum levels were significantly increased in both saline and OVA-challenged animals, compared with vehicle-treated animals. Further, no difference in OVA-specific IgG1 serum level was observed between the OVA- and saline-challenged animals. In OVA_323-339-treated animals, OVA-specific IgG1 serum levels were increased after OVA challenge, compared with saline challenge. This rise in OVA-specific IgG1 levels was similar to the rise that was observed in vehicle-treated animals.

View larger version (18K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Figure 3.   Serum levels of OVA-specific IgG1 (A), IgG2a (B), and IgE (C) in units per milliliter (U/ml), as determined by ELISA, 24 h after the last OVA (solid bars) or saline (hatched bars) challenge. Animals were treated with vehicle (Veh), OVA, or OVA323-339. Values are expressed as means ± SEM (n = 8 per group). # P < 0.05 and ## P < 0.005 compared with saline-challenged animals of the same treatment group; *P < 0.01 and **P < 0.001 compared with vehicle-treated, saline-challenged animals (ANOVA).

In all three treatment groups, OVA-specific IgG2a serum levels were not increased after OVA challenge, compared with saline challenge (Figure 3B). Furthermore, OVA-specific IgG2a levels were not affected by OVA or by OVA_323-339 immunotherapy. In all three treatment groups, OVA challenge induced a significant increase of OVA-specific IgE serum levels, compared with the respective saline-challenged groups (Figure 3C). More importantly, both OVA and OVA_323-339 immunotherapy did not change the level of OVA-specific IgE in OVA-challenged animals compared with the vehicle-treated group. However, after OVA immunotherapy, saline-challenged animals showed significantly increased OVA-specific IgE levels compared with vehicle- or OVA_323-339-treated saline-challenged animals.

Airway Responsiveness In Vivo Is Reduced after OVA Therapy and Increased after Peptide Therapy

In all three treatment groups, OVA challenge induced a significant (P < 0.001) increase in the responsiveness to methacholine compared with the respective saline-challenged controls. Vehicle immunotherapy before OVA challenge resulted in a significantly increased hyperresponsiveness to methacholine at doses of 480, 560, and 640 µg/kg compared with their saline-challenged controls (Figure 4A). After OVA immunotherapy, hyperresponsiveness to methacholine was observed at doses of 560 and 640 µg/kg compared with their saline-challenged controls (Figure 4B). The OVA-induced hyperresponsiveness to methacholine was significantly reduced compared with the vehicle-treated animals at doses of 480, 560, and 640 µg/kg. After OVA_323-339 immunotherapy, hyperresponsiveness to methacholine was observed at doses ranging from 160 to 640 µg/kg compared with their saline-challenged controls (Figure 4C). In contrast to OVA immunotherapy, OVA_323-339 immunotherapy induced a significant increase in the hyperresponsiveness to methacholine at doses of 160 and 320 µg/kg compared with vehicle-treated OVA-challenged animals.

View larger version (18K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Figure 4.   Airway responsiveness in vivo. Increase in airway resistance after intravenous administration of methacholine in mice challenged with OVA (solid bars) or saline (hatched bars) after treatment with vehicle (A), OVA (B), or OVA323-339 (C). Data are expressed as means ± SEM (n = 6 per group). # P < 0.05 and ## P < 0.001 compared with saline-challenged animals of the same treatment group; *P < 0.05; **P < 0.001 and ***P < 0.0001 compared with vehicle-treated, OVA-challenged animals (ANOVA).

The effective dose that provokes 50% of the maximal effect (ED₅₀) of the methacholine dose-response curve in OVA-challenged animals was significantly increased after OVA immunotherapy (vehicle 399 ± 14 µg/kg versus OVA immunotherapy 518 ± 48 µg/ml), and significantly decreased in OVA_323-339-treated animals (238 ± 32 µg/ml) (Figure 5).

View larger version (12K): [in this window] [in a new window]   Figure 5.   Methacholine ED50 in OVA-challenged animals after subcutaneous immunotherapy with vehicle (Veh), OVA, or OVA323-339. Data are expressed as means ± SEM (n = 6 per group). *P < 0.05 and **P < 0.01 compared with vehicle-treated, OVA-challenged animals (ANOVA).

Inflammatory Cells in the BALF Are Decreased after OVA Therapy and Increased after Peptide Therapy

After OVA challenge, all three treatment groups showed a significant increase in total numbers of cells (P < 0.02) and eosinophils (P < 0.001) in BALF compared with their saline-challenged controls (Table 3). Interestingly, compared with vehicle treatment, immunotherapy with OVA resulted in a significantly reduced eosinophil infiltration in OVA-challenged animals whereas neutrophil, lymphocyte, and mononuclear cell infiltration was not affected (Table 3, Figure 6). Remarkably, OVA_323-339 immunotherapy resulted in a significant increase in eosinophils (P < 0.05) and neutrophils (P < 0.05) but not in mononuclear cells and lymphocytes, compared with the vehicle-treated OVA-challenged animals (Table 3, Figure 6).

View larger version (13K): [in this window] [in a new window]   Figure 6.   Total numbers of eosinophils recovered 24 h after the last OVA challenge in BALF of animals treated with vehicle (Veh), OVA, or OVA323-339. Results are expressed as means ± SEM (n = 8 per group). *P < 0.05 compared with vehicle-treated, OVA-challenged animals (ANOVA).

Cytokine Production by Draining LN Cells Is Reduced after OVA Therapy but Unaffected by Peptide Therapy

Cultures of lung-draining LN cells isolated from saline-challenged animals produced no detectable levels of IL-4, IL-5, or IFN- after in vitro stimulation with OVA (data not shown), whereas LN cultures from OVA-challenged animals showed substantial levels of both IL-4 and IL-5 (Figure 7). Although no IFN- production by draining LN cells could be detected after in vitro stimulation with OVA, high levels of IFN- were found after incubation with immobilized CD3 in either of the treatment groups (data not shown). After OVA immunotherapy, cytokine levels produced by isolated draining LN cells stimulated in vitro with OVA were significantly reduced for IL-4 (P < 0.01) and IL-5 (P < 0.05) compared with the vehicle-treated group (Figure 7). In contrast, immunotherapy with OVA_323-339 did not alter the IL-4 and IL-5 production in vitro compared with the vehicle-treated animals (Figure 7).

View larger version (16K): [in this window] [in a new window]   Figure 7.   Cytokine production by draining LN cells after in vitro stimulation with OVA (10 µg/ml). LN cells from animals that were challenged with saline did not produce IL-4 or IL-5 (data not shown). LN cells isolated from OVA-challenged animals, treated with vehicle (Veh), OVA, or OVA323-339, produced detectable levels of IL-4 (solid bars) and IL-5 (hatched bars). Data are expressed as means ± SEM (n = 6 per group). *P < 0.05 and **P < 0.01 compared with vehicle-treated, OVA-challenged animals (ANOVA).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

To study the effects of peptide and OVA immunotherapy, we first determined whether the OVA_323-339 peptide is an immunodominant epitope in our murine model. In agreement with literature (20, 21), our data show that after sensitization and challenge with OVA 50% of the OVA-specific IgG1 and IgE was directed against OVA_323-339. We also demonstrated that OVA_323-339 induced proliferation and cytokine production in LN cell cultures from OVA-challenged animals. Remarkably, OVA_323-339 induced high levels only of IL-5, but not IL-4, indicating that the epitope is recognized by a subpopulation of OVA-specific T cells. This subpopulation may play an important role in the pathologic and immunologic features in our model because IL-5 is an important inducer of eosinophil maturation, attraction, and survival. In the present study we have shown that immunotherapy using entire OVA results in downregulation of airway hyperresponsiveness, decreased eosinophil infiltration, and decreased production of IL-4 and IL-5 by draining LN cells. Interestingly, OVA_323-339 immunotherapy resulted in opposite effects. Although in vitro IL-4 and IL-5 production were not influenced, both airway hyperresponsiveness and eosinophil infiltration were significantly increased.

In all experimental groups, OVA challenge induced a high serum level of OVA-specific IgE, which was not affected by OVA or OVA_323-339 immunotherapy. Although the serum titer was measured only 24 h after the last challenge, this finding indicates that the production of OVA-specific IgE and the development of airway hyperresponsiveness and eosinophilia are differently regulated. In agreement herewith, we found that OVA immunotherapy increased OVA-specific IgE in saline-challenged animals, whereas airway hyperresponsiveness and eosinophilia were absent. Recent studies using B cell-deficient (23) or IgE knockout mice (24) demonstrated that airway hyperresponsiveness and lung eosinophilia can occur even in the absence of antigen-specific IgE.

Long-term immunotherapy in humans (25) and mice (19) has been reported to induce a rise in allergen-specific IgG, which has been suggested to compete with IgE for allergen binding, thereby blocking IgE-mediated events (25). Although this increase in allergen-specific IgG is not always correlated with clinical effectiveness (26), and the "blocking antibody theory" is still controversial, we also found an increase in OVA-specific IgG1 serum level after OVA immunotherapy.

Immunotherapy in atopic patients has been suggested to induce allergen-specific Th1 cells that suppress the effector function of allergen-specific Th2 cells and alter allergen-specific antibody production (27, 28). Although reduced Th2-type cytokine production by allergen-specific T cells after immunotherapy has been demonstrated in most studies (29), increased Th1-type cytokines are not always concomitant (27, 31). Remarkably, Akdis and colleagues (32) reported both a decreased Th2 and Th1 cytokine production by peripheral blood mononuclear cells in vitro after bee venom therapy, suggesting the induction of allergen-specific T-cell unresponsiveness rather than the induction of an allergen-specific Th1-type population or exclusive downregulation of the allergen-specific Th2 response. The present study shows that immunotherapy with entire OVA resulted in decreased airway symptoms that were accompanied by an OVA-specific downregulation of the production of IL-4 and IL-5 by draining LN cells in vitro. Although we cannot fully exclude the induction of OVA-specific Th1-type cells, the facts that no IFN- production by lung-draining LN cells could be detected after OVA stimulation in vitro, and that cell cultures from spleen and LNs draining the site of immunotherapy (brachial and axillary LNs [33]) did not produce detectable levels of IFN- after stimulation with OVA 5, 9, and 12 d after immunotherapy (Janssen and colleagues, unpublished data), argues against the induction of Th1-type cells as being the main regulatory mechanism.

Because it has been reported that in vitro stimulation of allergen-specific T cells with immunodominant epitopes resulted in T-cell unresponsiveness, interest has focused on the development of immunotherapy using immunodominant epitopes of the allergen. In the first human trial, it was reported that immunotherapy using dominant T-cell epitopes of Fel d I ameliorated allergic symptoms during exposure to the entire allergen in allergic patients (34). Although the use of a T-cell epitope was expected to circumvent the risks of side effects, the peptide immunotherapy provoked clinical side effects in 65% of the patients directly after injection with the peptide (34). In the present study peptide treatment resulted in increased airway hyperresponsiveness and eosinophilia, whereas we observed no increase in allergen-specific production of Th2-related cytokines by splenocytes, mesenteric LN cells, or cells that drain the site of immunotherapy (Janssen and colleagues, unpublished data). Also, no increase in allergen-specific cytokine production by lung-draining LN cells was observed. However, the significant increase of eosinophils in the BALF suggests that IL-5 production in the lungs was augmented. Other OVA_323-339 immunotherapy strategies, including the use of higher doses of peptide and increased time between immunotherapy and challenge, resulted in similar effects (Janssen and colleagues, unpublished data).

Our findings show that immunotherapy using OVA or OVA_323-339 peptide affects the ongoing Th2 response by different mechanisms. Although OVA_323-339 is the immunodominant T-cell epitope of OVA, we cannot exclude the possibility that other T-cell epitopes of OVA are involved in the reduction of airway symptoms and cytokine production. Alternatively, the biodistribution of the protein and peptide may differ after subcutaneous administration, which may result in T-cell activation at different locations. Moreover, because OVA_323-339 contains a B- as well as a T-cell epitope, the presence of this B-cell epitope may influence the uptake of the peptide by an antigen-presenting cell (APC) (35). If peptide and protein are presented by different APCs, and with different costimulation, this may greatly affect the outcome of the T-cell response (36).

In summary, we have shown that immunotherapy with the entire protein had beneficial effects, whereas peptide immunotherapy resulted in deterioration of airway hyperresponsiveness and eosinophilia. Knowledge about biodistribution and the antigen presentation during protein immunotherapy may help in understanding the mechanism of successful immunotherapy, and could facilitate the development of safe peptide immunotherapy.