INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Despite ongoing advances in the field of asthma therapy, there are few options for specific treatment of the underlying event(s) causing airway inflammation and airway hyperresponsiveness (AHR), two main properties of the disease. To support such initiatives, further insight into the complex pathophysiology of the allergic response in the airways is necessary. Allergen-induced bronchial asthma is characterized by two main features: production of allergen-specific IgE and development of eosinophilic airway inflammation (1). There is evidence of the importance of IgE in the development of bronchial asthma, in that studies have demonstrated a correlation between IgE serum levels and the prevalence (2) or severity (3) of the disease. Murine studies have also supported a role for IgE in allergen-induced airway sensitization. Coyle and colleagues showed that treatment of mice with a nonanaphylactogenic anti-IgE antibody not only neutralizes increased IgE levels after sensitization and airway challenge, but also reduces helper T cell type 2 (Th2) activation and cytokine production, leading to decreased eosinophil airway infiltration compared with control mice (4). IgE has also been shown to enhance antigen focusing (5) and to activate T cell proliferation and T cell cytokine production (6), possibly via the low-affinity receptor for IgE (CD23). Eum and coworkers demonstrated that for the development of AHR in mice, high serum IgE levels were necessary to induce eosinophil trafficking through the airway lumen (7). In addition, antibody against interleukin 4 (IL-4), the main switch factor for the induction of IgE production, prevented eosinophil airway inflammation and development of AHR, suggesting that IL-4 may be essential either via a direct effect on eosinophils or indirectly, through enhancement of IgE production (8).

As well as elevated serum IgE levels, eosinophilic airway inflammation is regularly observed in bronchial asthma (9). Increased numbers of eosinophils are found in the bronchoalveolar lavage fluid (BALF) and in bronchial biopsies, and eosinophil numbers correlate with the severity of the disease (10, 11). The main growth, differentiation, and survival factor for eosinophils is IL-5 (12) and, indeed, IL-5 has been shown to be essential for the induction of allergen-induced eosinophilic airway inflammation in most studies in mice (13) and guinea pigs (14). We previously demonstrated that treatment of high IgE-responder BALB/c mice with anti-IL-5 during airway sensitization completely prevented the development of AHR, despite production of allergen-specific IgE and immediate cutaneous hypersensitivity, suggesting a dissociation between elevated IgE serum levels on the one hand and airway inflammation and AHR on the other (15). Similarly, sensitization and airway challenge of IL-5-deficient mice still resulted in antigen-specific IgE production, but did not increase numbers of eosinophils or airway responsiveness (AR) unless IL-5 production in the lungs was restored by transferring IL-5-transfected vaccinia virus (16).

The aim of the present study was to compare directly the effects of treatment protocols directed against either antigen-specific IgE or IL-5. We chose a model in which systemic sensitization followed by airway challenge with allergen provokes both significantly elevated IgE levels and a sustained eosinophilic airway response. We show that treatment of allergen-sensitized mice with anti-IgE antibody reduces total and allergen-specific IgE serum levels and prevents development of (cutaneous and systemic) anaphylactic reactions, but has little effect on airway inflammation or development of AHR. In contrast, anti-IL-5 treatment of such mice inhibits eosinophilic airway inflammation and development of AHR.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Female BALB/c mice from 8 to 12 wk of age were obtained from Jackson Laboratories (Bar Harbor, ME). The mice were maintained on ovalbumin (OVA)-free diets. All experimental animals used in this study were under a protocol approved by the Institutional Animal Care and Use Committee of the National Jewish Medical and Research Center (Denver, CO).

Sensitization and Airway Challenge

Groups of mice (three or four mice per group per experiment) were sensitized by intraperitoneal injection of 20 µg of OVA (Sigma, St. Louis, MO) emulsified in 2 mg of aluminum hydroxide (AlumInject; Pierce, Rockford, IL) in a total volume of 100 µl on Day 1 and Day 14. Mice were challenged via the airways with OVA (1% in phosphate-buffered saline [PBS]) or PBS for 20 min on Days 28 and 29 by ultrasonic nebulization, and assessed on Day 31 for AR.

Treatment Protocol

The anti-IgE antibody termed 1-5 is a nonanaphylactogenic rat anti-mouse antibody that does not activate mast cells or basophils. It recognizes an epitope within the FcRI-binding region of murine IgE and is IgE isotype specific (17, 18). Anti-IgE was administered intraperitoneally (200 µg in 200 µl of PBS) on Day 27, and 2 h before each airway challenge on Days 28 and 29. Control animals received similar amounts of rat IgG antibody (Sigma), following the same protocol. One set of experiments was performed using a higher concentration (800 µg) of anti-IgE antibody three times before airway challenge.

Anti-IL-5 (TRFK-5; kindly provided by R. Coffman, DNAX Institute, Palo Alto, CA) was administered intranasally (50 µg in 30 µl of PBS) into lightly anesthesized animals (tribromo) ethanol [Avertin], 2%, administered intraperitoneally) on Day 27 and 2 h before each airway challenge on Days 28 and 29. Control animals received similar amounts of rat IgG antibody (Sigma), following the same protocol. In preliminary experiments, the intranasal route of anti-IL-5 antibody delivery was shown to be superior to intravenous or intraperitoneal injections (200 µg in 200 µl of PBS) in preventing airway inflammation and AHR (data not shown).

Measurement of Anti-OVA Antibody and Total Immunoglobulin Levels

Anti-OVA immunoglobulin serum levels were measured by enzyme-linked immunosorbent assay (ELISA) as previously described (19). The antibody titers of the samples were related to pooled standards that were generated in the laboratory. Total IgE levels were determined using the same method as previously described (19). Total immunoglobulin levels were calculated by comparison with known mouse IgE standards (PharMingen, San Diego, CA). The limit of detection was 100 pg/ml for IgE.

To determine the neutralizing effect of anti-IgE on IgE serum concentrations, antibody was added to serum samples from sensitized animals at increasing concentrations and the ELISA was performed as described above.

Stimulation of Mast Cells and Kinase Assay for Extracellular Regulated Kinase (ERK2)

Mast cells were obtained by culturing femoral bone marrow cells in IL-3-saturated medium for 4 wk as described (20). The purity of mast cells was > 98% as assessed by toluidine blue staining. Mast cells (3 × 10⁶/ml) were passively sensitized with anti-OVA IgE (500 ng/ml) for 2 h, then washed extensively and incubated in the presence of OVA (10 µg/ml) or PBS for 5 min. An in vitro kinase assay for ERK2 was carried out as previously described (20).

Systemic and Cutaneous Anaphylaxis

To determine the effect of anti-IgE treatment on the development of anaphylaxis, mice were passively sensitized by intravenous injections of anti-OVA IgE antibody (2 µg) on Days 1 and 2 (19). Anti-IgE (300 µg) or rat IgG control antibody was injected intraperitoneally on Day 0, and 2 h before passive sensitization with anti-OVA IgE on Days 1 and 2. Twenty-four hours later, OVA was injected intravenously (500 µg in 200 µl) or intradermally (10 µg in 20 µl) and the animals were monitored for development of acute symptoms or wheal responses for up to 30 min. Identical results were obtained using anti-DNP IgE (20 µg, intravenous) and DNP-HSA (200 µg, intravenous) (Sigma) instead of the combination of anti-OVA IgE and OVA.

In a different set of experiments, mice (actively) sensitized to OVA (using the sensitization protocol described above) were injected with anti-IgE antibody (200 µg, intraperitoneal) on Days 27, 28, and 29 of the protocol. Twenty-four hours after the last antibody injection, OVA was injected intravenously (500 µg in 200 µl) or intradermally (10 µg in 20 µl).

Cell Preparation and Culture

Peribronchial lymph nodes (PBLNs) were harvested, and mononuclear cells (MNCs) were purified by passing the tissue through stainless steel mesh, followed by density gradient centrifugation (Organon Teknika, Durham, NC), and resuspended in RPMI 1640 medium (GIBCO, Grand Island, NY) containing 10% fetal calf serum (FCS), penicillin (100 U/ml), streptomycin (100 g/ml), 5 mM glutamine, and 50 M 2-mercaptoethanol (2-ME). MNCs were plated in 96-well round- bottom plates at 200,000 cells per well and cultured in the presence or absence of OVA and mitogen at 37° C. Cell-free supernatants were harvested and stored at 20° C.

Cytokine Production

Cytokine levels in the BAL fluid and the supernatants after 48 h of incubation were measured by ELISA as described (15). Briefly, ELISA plates were coated with purified anti-cytokine antibodies (all reagents from PharMingen) and blocked with 10% FCS-PBS. Samples and dilution rows of purified cytokines as standards were incubated at 4° C overnight. Biotinylated anti-cytokine antibodies followed by avidin-peroxidase and 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonate) (ABTS) substrate were used for detection. The limit of detection was 4 pg/ml for IL-4 and IL-5.

Bronchoalveolar Lavage

BALF was obtained as described (15). Briefly, lungs were washed via the trachea with 1 ml of Hanks' balanced salt solution (HBSS), and cells were counted and resuspended in HBSS. Slides were stained with Leukostat (Fisher Diagnostics, Pittsburgh, PA), and cell differentiation percentages were determined by counting at least 300 cells under a light microscope.

Immunohistochemistry

After perfusion via the right ventricle, lungs were inflated through the tracheas with 2 ml of saline and fixed in 10% formalin. Major basic protein (MBP) in lung sections was localized by immunohistochemistry using rabbit anti-mouse MBP (kindly provided by G. Gleich [Major Clinic, Rochester, NY] and J. Lee [Scottsdale, AZ]) as described (15). Slides were examined in a blinded fashion with a Zeiss (Oberkochen, Germany) microscope equipped with a fluorescein filter system. Numbers of eosinophils in the submucosal tissue around central airways were evaluated using IPLab2 software (Signal Analytics, Vienna, VA) for the Macintosh, counting four different sections per animal (15).

Determination of Airway Responsiveness

AR was measured in unrestrained animals, using barometric plethysmography (Buxco, Troy, NY) as previously described (21). Briefly, the whole body plethysmograph is calibrated by injecting 150 µl of air into the main chamber. Flow-derived parameters are calculated by measurements of pressure differences between the main chamber of the plethysmograph, containing the animal, and a reference chamber. Mice were challenged with aerosolized PBS or methacholine (MCh) in increasing concentrations (3 to 50 mg/ml) through an inlet of the main chamber of the plethysmograph for 3 min and readings were taken and averaged for 3 min after each nebulization. AR is expressed as the fold increase in P_enh (enhanced pause).

Statistical Analysis

Analysis of variance was used to determine the levels of difference between all groups. Pairs of groups were compared by Student's t test. Comparisons for all pairs were performed by Tukey-Kramer honest significant difference (HSD) test for airway responsiveness and histology data. The p values for significance were set to 0.05. Values for all measurements are expressed as the mean ± standard deviation, except values for P_enh, which are presented as the mean ± standard error of the mean (SEM). MCh doses for increases in P_enh are reported as the geometric mean and the lower and upper levels of the 95% confidence interval.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Anti-IgE Antibody Decreases Total and OVA-specific IgE Levels

We studied the effects of anti-IgE and anti-IL-5 treatment in actively sensitized animals. BALB/c mice were sensitized by intraperitoneal injection of OVA emulsified in alum on Day 1 and Day 14. Airway challenge with OVA was performed on Days 28 and 29. Serum levels of OVA-specific and total immunoglobulins were measured 2 d after the last airway challenge, on Day 31. Sensitization and challenge with OVA resulted in significantly increased serum levels of anti-OVA IgE and IgG₁ and of total IgE when compared with mice challenged alone (Table 1). Sensitization did not significantly alter total IgG serum levels.

Treatment of sensitized mice with anti-IgE antibody (200 µg) 2 h before each airway challenge with antigen reduced IgE serum levels to 25% of the levels in sensitized and challenged mice that received rat IgG (Table 1). Increasing the antibody dose (to 800 µg per injection) did not further reduce IgE serum levels. We analyzed the capacity of anti-IgE to interfere with IgE detection and function. Serum concentrations of total and OVA-specific IgE from sensitized, challenged mice were established by ELISA as described above. Anti-IgE was added in increasing concentrations (1-1,000 µg/ml) to the samples. We observed a concentration-dependent neutralization of total and specific IgE, which resulted in a 75% reduction at an anti-IgE antibody concentration of 100 µg/ml. Despite the reduction in total and allergen-specific IgE, OVA-specific IgG₁ antibody production was not reduced in anti-IgE-treated mice. Treatment of mice with anti-IL-5 antibody was without effect on immunoglobulin production after sensitization and challenge with OVA (Table 1).

Anti-IgE Antibody Blocks IgE and Prevents Mast Cell Activation In Vitro

One of the main actions of IgE is the activation of mast cells on cross-linking with antigen, resulting in release of proinflammatory mediators, which in turn lead to immediate hypersensitivity (22). We analyzed the effects of preincubation of bone marrow cultured mast cells with anti-IgE antibody on ERK activation, an early event following antigen cross-linking of sensitized mast cells. As shown in Figure 1, passive sensitization of mast cells with anti-OVA IgE followed by incubation with OVA resulted in a significant activation of ERK2. In contrast, preincubation of mast cells with anti-IgE at the time of passive sensitization with anti-OVA IgE prevented kinase activation after incubation with OVA (Figure 1), demonstrating the activity of the IgE antibody in preventing mast cell activation.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Effect of anti-IgE on FcRI-medicated activation of ERK2 in bone marrow-derived mast cells. Bone marrow-derived mast cells (3 × 106/ml) were incubated with anti-OVA IgE (500 ng/ml) in the presence of anti-IgE (200 µg/ml) or rat IgG (control) antibody for 2 h. After extensive washing, the cells were then incubated with OVA (10 µg/ml) or PBS (control) for 5 min and ERK2 activities in the cells were measured. MBP, Myelin basic protein. A representative autoradiograph from three independent experiments is shown.

Anti-IgE Antibody Prevents Cutaneous and Systemic Anaphylaxis In Vivo

IgE-mediated mast cell activation after intradermal injection of antigen results in immediate-type hypersensitivity. To study the effects of anti-IgE antibody treatment on the development of anaphylactic reactions, we pretreated mice before passive sensitization with anti-OVA IgE with anti-IgE antibody. Intravenous delivery of antigen resulted in anaphylactic responses (slowing of breathing, tachycardia, weakness) in all control animals receiving anti-OVA IgE and rat IgG antibody before OVA (four of four animals within 15 min). These systemic anaphylactic reactions were completely inhibited by anti-IgE antibody treatment before and during passive sensitization. Similarly, anti-IgE treatment of OVA-sensitized mice before intradermal or intravenous injection of allergen (OVA) completely inhibited systemic anaphylactic reactions (four of four) and the development of cutaneous hypersensitivity responses (zero of four positive responders). Further, development of immediate cutaneous hypersensitivity responses in sensitized and challenged mice after intradermal injection of allergen was completely inhibited (zero of four positive responders) in anti-IgE-treated mice compared with mice receiving control IgG (four of four positive responders). Moreover, significantly diminished cutaneous hypersensitivity responses (smaller wheals, fewer positive responders) were observed in the anti-IgE-treated mice (data not shown).

Anti-IL-5 or Anti-IgE Does Not Alter Th2 Cytokine Production

Th2 cytokine production by T cells may be a key regulatory event in the induction of airway inflammation and AHR (23). To evaluate the importance of IgE in the induction of Th2 responses after sensitization, we measured the production of OVA-induced IL-4 and IL-5 levels in cultured PBLN T cells after 48 h. Sensitization and challenge significantly enhanced antigen-specific IL-4 and IL-5 production by PBLN T cells compared with nonsensitized animals (Figure 2). Neither anti-IgE nor anti-IL-5 treatment significantly altered Th2 cytokine production by PBLN cells from sensitized and challenged mice (Figure 2).

View larger version (25K): [in this window] [in a new window]   Figure 2.   IL-4 and IL-5 production by peribronchial lymph node T cells after sensitization and airway challenge with allergen. BALB/c mice were sensitized by intraperitoneal injection of OVA in alum on Days 1 and 14 followed by airway challenges with OVA on Days 28 and 29 (ipNeb, n = 14). Control animals were nonsensitized mice challenged with OVA via the airways on Days 28 and 29 (Neb, n = 12). Groups of sensitized mice were treated with rat IgG (ipNeb, n = 14), anti-IgE (200 µg, intraperitoneally: ipNeb--IgE, n = 16), or anti-IL-5 antibody (50 µg, intranasally; ipNeb--IL-5, n = 14) on Day 27 and 2 h prior to each airway challenge with allergen. Two days after the last challenge, PBLN cells were harvested and cultured (2 × 105 cells per well) in the absence or presence of OVA (50 mg/ml) for 48 h. Cytokine levels in supernatants were determined by ELISA. Expressed are the means ± SD (pg/ml) from three independent experiments. Controls (Neb) had protein cytokine levels of < 25 (pg/ml).

Because T cells might not be the only source of cytokines, we measured concentrations of IL-5 in the BAL fluid. Allergen sensitization and airway challenge increased IL-5 concentrations in the BAL fluid significantly (< 10 pg/ml in nonsensitized versus 144 ± 35 pg/ml in sensitized and challenged mice, p < 0.01); anti-IgE antibody treatment of sensitized mice before and during airway challenge did not significantly alter IL-5 production in the BAL fluid (122 ± 27 pg/ml, p > 0.05 versus ipNeb).

Anti-IL-5 But Not Anti-IgE Antibody Prevents Eosinophil Airway Infiltration

Eosinophilic airway inflammation is a pivotal event in allergen-induced airway sensitization. We assessed the changes after allergen sensitization and challenge in cells from BALF and by in situ immunocytochemistry of lung tissue. After sensitization and airway challenge with allergen, the absolute numbers of leukocytes, lymphocytes, and neutrophils were significantly increased in the BALF compared with nonsensitized mice. Moreover, numbers of eosinophils were significantly increased in OVA-sensitized/challenged mice by more than 200-fold (Figure 3).

View larger version (24K): [in this window] [in a new window]   Figure 3.   Eosinophil infiltration of BALF after sensitization and airway challenge with allergen is inhibited by anti-IL-5 antibody treatment. Groups and protocols are as described in Figure 2. Two days after the last airway challenge, BALF cells were harvested and relative numbers of different cell types were determined by light microscopy. Expressed are the means ± SD of the percentage of the different cell types from three independent experiments. *p < 0.01 versus Neb; p < 0.01 versus ipNeb.

To localize the eosinophils in the lung tissue of sensitized, challenged mice, immunohistochemistry with anti-MBP antibody was performed on formalin-fixed lung sections (Figure 4). The numbers of MBP-positive cells were measured in the peribronchial tissue using computer-assisted analysis and compared for the different groups. In control mice (receiving rat IgG), sensitization followed by airway challenge resulted in significant increases in the number of peribronchial eosinophils compared with nonsensitized controls (Figure 5).

View larger version (56K): [in this window] [in a new window]   View larger version (64K): [in this window] [in a new window]   Figure 4.   Localization of eosinophils in peribronchial tissue after airway sensitization with OVA. Groups and protocols are as described in Figure 2. Sections of peribronchial tissue were stained with rabbit anti-mouse MBP antibodies. Shown are representative lung sections treated with anti-IgE (A) and anti-IL-5 (B).

View larger version (13K): [in this window] [in a new window]   Figure 5.   Eosinophil infiltration of peribronchial regions after sensitization and airway challenge with allergen is inhibited by anti-IL-5 antibody treatment. Sections of peribronchial tissue were stained with rabbit anti-mouse MBP antibodies. Four sections around similar-sized central airways were chosen randomly and eosinophils in the connecting tissue were counted in a blinded fashion. Expressed are the mean (± SD) numbers of eosinophils per square millimeter of lung tissue from three independent experiments. *p < 0.05 versus Neb; #p < 0.05 versus ipNeb.

To study the role of IgE and IL-5 in the induction of eosinophilic airway inflammation, sensitized mice were treated intraperitoneally with anti-IgE (200 µg) or intranasally with anti- IL-5 antibody (50 µg) once before and prior to each of the two airway challenges with allergen. Intranasal delivery of anti-IL-5 was chosen because it was shown in preliminary experiments to be even more effective in preventing inflammation and AHR than were other forms of delivery (intraperitoneal, intravenous) and less was required (50 versus 150 µg) to induce the same degree of inhibition. Treatment of sensitized mice with 50 µg of anti-IL-5 antibody before airway challenge with allergen completely prevented the increases in eosinophil numbers in the lungs and prevented peribronchial eosinophil infiltration (Figure 5). In contrast, anti-IgE treatment (including 200 µg of anti-IgE administered intranasally at the same time points) failed to affect eosinophilic inflammation significantly. Increasing the concentration of anti-IgE antibody to 800 µg per injection did not affect the outcome (data not shown).

Anti-IL-5 But Not Anti-IgE Antibody Prevents Development of AHR

Development of AHR is associated with increased IgE serum levels and eosinophilic airway inflammation. To assess the effects of anti-IgE and anti-IL-5 treatment on the development of AHR after systemic sensitization and airway challenge with OVA, we measured in vivo AR to aerosolized MCh by barometric plethysmography. This method reliably detects AHR after sensitization and airway challenge in mice and the measurements of AR are closely correlated to invasive measurements of intrapleural pressure and in vivo lung resistance to MCh (24). Systemic sensitization and airway challenge with allergen significantly increased airway reactivity (Figure 6) in mice receiving rat IgG. The concentration of MCh necessary to increase AR by 100% decreased from 17.5 mg (upper and lower limits, 11 and 25 mg) in nonsensitized controls to 5.1 mg (upper and lower limits, 4.2 and 7.1 mg) in sensitized, challenged mice (p < 0.01). Anti-IgE treatment had little effect on the development of AHR, and increasing the anti-IgE antibody concentration did not alter this result. Sensitized mice receiving anti-IL-5 antibody before allergen airway challenge failed to develop increased AR; the dose-response curve after challenge with aerosolized MCh was similar to that of nonsensitized controls (Figure 6). These data confirm that IL-5, in contrast to IgE, plays a key role in the induction of eosinophil airway inflammation and development of AHR after allergic sensitization and airway challenge.

View larger version (14K): [in this window] [in a new window]   Figure 6.   Development of airway hyperresponsiveness after sensitization and airway challenge with allergen is inhibited by anti-IL-5 antibody treatment. Groups and protocols are as described in Figure 2. Airway responsiveness to aerosolized MCh was measured in unrestrained, conscious mice and PENH values were determined. Expressed are the means ± SEM of the increase in Penh values compared with Penh values after PBS nebulization from three independent experiments. *p < 0.05 versus Neb. #p < 0.05 versus ipNeb on anti-IgE. Differences between ipNeb and anti-IgE were not significant.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Current experimental directions in the treatment of allergic asthma include administration of nonanaphylactogenic IgE and anti-IL-5 or modifiers of IL-5 interactions with receptors on eosinophils. We studied the effects of anti-IgE and anti-IL-5 treatment in a murine model of allergen-induced airway inflammation and AHR to better understand prospects for future therapy. In this model, sensitized and challenged mice develop increased levels of allergen-specific IgE, immediate cutaneous reactivity, eosinophilic infiltration of the airways, and increased responsiveness to MCh (24). Increased levels of IgE are seen in most atopic diseases and have been implicated in the development of allergic symptoms (2, 3). IgE-mediated mast cell activation is the main mechanism for immediate-type hyperresponsiveness (25), and the absence of mast cells prevents the appearance of anaphylactic reactions after allergen provocation (26). We show here that the anti-IgE antibody is highly effective: blockade of IgE-mast cell interaction via the high affinity IgE receptor by the nonanaphylactogenic anti-IgE antibody, which binds to the main Fc receptor epitope of the IgE molecule, inhibits mast cell activation. In our in vitro experiments, we show that mast cell kinase activation, after passive sensitization with allergen-specific IgE plus allergen provocation, is completely inhibited by preincubation of the mast cells with the anti-IgE antibody before adding the allergen. Intravenous or intradermal injection of allergen into sensitized animals induces systemic and cutaneous anaphylactic reactions. These reactions were significantly decreased or eliminated after anti-IgE antibody treatment before allergen provocation. Treatment with anti-IgE also reduced detectable levels of IgE in sensitized mice in vitro and in vivo.

However, this reduction in IgE and IgE-mediated activity did not appear to interfere with T cell function or the induction of eosinophilic airway inflammation after airway challenge of sensitized animals. This contrasts with the results of Coyle and coworkers, who describe a similar model of sensitization and intranasal allergen challenge but in which anti-OVA IgE inhibited eosinophilic accumulation in the airways (4). In this model, they proposed that enhancement of T cell activation by B cells via the low-affinity IgE receptor (CD23) was blocked by the anti-IgE antibody. This resulted in less IL-4 and IL-5 production, leading to a significantly reduced eosinophilic inflammation in the airways. Besides the differences in the sensitization and allergen challenge protocol, we cannot explain the discrepancy in results at the present time. We have shown previously that in mice genetically lacking CD23, systemic sensitization followed by airway challenge as described in this article leads to normal (even enhanced) T cell activation and eosinophilic airway inflammation (27). This implies that neither T cell activation nor eosinophil infiltration after sensitization and airway challenge is dependent on B-T cell interaction via CD23. Further, one could assume that OVA-specific IgE generated during sensitization has become bound to FcRI⁺ cells and is therefore not inhibited by the anti-IgE. On the other hand, anti-IgE competes with FcRI for IgE (17, 18) and therefore might reduce IgE on mast cells over a longer period of time. Our data indicate that, indeed, anti-IgE treatment of actively sensitized mice 3 d before intravenous or intradermal allergen challenge was able to prevent systemic and cutaneous anaphylactic reactions. Importantly, results from our own and other experiments have shown that neither B cells/B cell products nor mast cells are necessary or important for airway inflammation or AHR in this model. We have demonstrated that in mice genetically lacking B cells (µMt^-/-), normal T cell activation and eosinophilic inflammation are observed (28). Further, IgE-deficient mice were shown to have normal airway inflammation and airway responsiveness after systemic sensitization and airway provocation (29). Similarly, development of AHR after systemic sensitization and airway challenge in B cell-deficient mice is independent of immunoglobulin production or IgE-mediated mass cell activation (30). An addition to these results is the finding that mast cell-deficient mice respond similarly to control mice, i.e., they show similar eosinophilic airway inflammation and increased AR after this mode of sensitization and challenge (31). Taken together, these data indicate the independence of T cell activation, T cell cytokine production, eosinophil infiltration and development of AHR from B cells, B cell-derived IgE, or IgE-mediated mast cell activation.

In light of these data it is not surprising that sensitized mice treated with an effective IgE antibody before allergen airway challenge developed increases in AR, similar to control mice. These results after systemic sensitization and challenge are in stark contrast with our findings in a model of airway sensitization using a 10-d nebulization protocol with allergen (in the absence of adjuvant). This approach leads to increases in allergen-specific IgE but only modest eosinophilic airway inflammation. Under these conditions, the combination of both allergen-specific IgE and IL-5-mediated eosinophilia is required for the development of AHR (28, 32). The role of IgE in this development of AHR in approach is unclear, but may be explained by a function of IgE in eosinophil trafficking (7). Thus, it appears that when substantial eosinophilic airway inflammation can be induced, the requirement for IgE in the development of AHR is lost.

The present results support a critical role for IL-5-mediated eosinophilic inflammation in the induction of AHR. Treatment of sensitized mice with anti-IL-5 antibody before allergen challenge prevented eosinophilic airway infiltration and inhibited the development of AHR, despite elevated serum levels of allergen-specific IgE. This confirms data from previous studies in guinea pigs (14, 33), mice (15), and monkeys (34) using anti-IL-5 antibody. Whereas treatment with anti-IL-5 blocked eosinophil airway accumulation and prevented development of AHR, there was no interference with T cell (proliferative responses, cytokine production), B cell (immunoglobulin production), or mast cell (cutaneous hypersensitivity) function. Despite many contrasting studies, at least two studies investigating the effects of anti-IL-5 antibody on the development of AHR in a model of systemic sensitization followed by airway challenge failed to detect an effect on the development of AHR (8, 35). It is difficult to account for these differences. It is possible that the mode of administration, the amount of antibody used, or the differing times to allergen challenge of the airways may account for the different outcomes of the experiments. In the present experiments, we used intranasal delivery of the IL-5 antibody because we found it superior compared with systemic administration in inhibiting eosinophil accumulation and development of AHR. Only one-third (50 µg) of the amount of anti-IL-5 antibody was necessary (after intranasal administration) to achieve a similar degree of inhibition of eosinophilic inflammation and development of AHR as seen after systemic antibody treatment (150 µg). This resembles our experience with treatment of sensitized mice with IFN-, in which local (nebulization) but not systemic (intraperitoneal) delivery was effective in reducing allergen-specific IgE production, eosinophilic inflammation, and AHR (36). Our study suggests that local anti-IL-5 antibody treatment may work through different mechanisms: first, it reduces eosinophil differerentiation, maturation, and accumulation in the airways. In addition to the reduction in lung eosinophils, we also observed a significant reduction of bone marrow and peripheral blood eosinophilia in animals treated with anti-IL-5 antibody compared with sensitized and challenged control animals (data not shown), suggesting systemic effects of the intranasally delivered antibody. Second, anti-IL-5 may inhibit local activation and interfere with the anti-apoptotic effects of IL-5 on eosinophils (37), thus reducing the survival time of the eosinophils in the inflamed tissue. Both mechanisms could together reduce eosinophil numbers in the airways, leading to substantially reduced mediator release.

Comparison of the two treatment protocols suggests that anti-IgE may be an effective approach when immediate-type allergic reactions and mast cells are involved in disease pathogenesis. Indeed, clinical trials suggest that anti-IgE treatment may be successful therapy in the treatment of allergic rhinitis and conjunctivitis (38). In contrast, our data indicate that anti-IgE may not be as effective in preventing allergic symptoms when a (chronic) inflammatory (eosinophilic) process is involved. Under these circumstances, interference with the induction of Th2 cytokine production (IL-4, IL-5) and/or the accumulation of the primary effector cells (eosinophils) at the site of the allergen-induced inflammation should prove more effective. IL-5 antibody treatment is only one approach to successfully prevent airway eosinophilic inflammation and AHR in allergen-induced bronchial asthma.