|Home | About | Journals | Submit | Contact Us | Français|
Existing asthma models develop tolerance when chronically exposed to the same allergen.
To establish a chronic model that sustains features of asthma long after discontinuation of allergen exposure.
We immunized and exposed mice to a combination of single, double or triple allergens (dust-mite, ragweed, and Aspergillus) intranasally for 8 weeks. Airway hyperreactivity and morphological features of asthma were studied 3 weeks after the allergen exposure. Signaling effects of the allergens were studied on dendritic cells.
Sensitization and repeated exposure to a single allergen induced tolerance. Sensitization to double, and especially triple allergens broke through tolerance and established AHR, eosinophilic inflammation, mast cell and smooth muscle hyperplasia, mucus production and airway remodeling that persisted at least 3 weeks after allergen exposure. Mucosal exposure to triple allergens in the absence of an adjuvant was sufficient to induce chronic airway inflammation. Anti-IL5 and -IL13 antibodies inhibited inflammation and AHR in the acute asthma model but not in the chronic triple allergen model. Multiple allergens produce a synergy in p38 MAPK signaling and maturation of dendritic cells, which provides a heightened T cell co-stimulation at a level that cannot be achieved with a single allergen.
Sensitivity to multiple allergens leads to chronic asthma in mice. Multiple allergens synergize in dendritic cell signaling and T cell stimulation that allows escape from the single allergen-associated tolerance development.
We have developed a model of chronic asthma that allows for the study and treatment of long-lasting features of asthma obviating the need for acute de novo allergen challenges.
Existing mouse models of asthma have provided a wealth of information.1 A major drawback of these models has been the transient nature of the airway pathology and hyperreactivity. In these models airway pathologies peak 24–72 hrs after the allergen challenge and resolve in 1–2 weeks.2, 3 Mice chronically exposed to a single allergen frequently develop tolerance thus mimicking the 80–85% of the human population that is chronically exposed to allergen yet are asymptomatic.4–6 Eosinophilic inflammation and mucus production are modestly present or significantly reduced in many chronic models. Airway hyperreactivity persists in many of these models in substantially attenuated form.7 Development of such tolerance is a major impediment to understanding the biology of chronic asthma. This is very important because the brunt of therapeutic intervention is aimed at controlling the chronic pathologic process of asthma. In an attempt to develop a mouse model of chronic asthma we employed natural allergens that cause human asthma, because they contain non-immunogenic components that have adjuvant-like effects.8–10 Since most allergic asthmatic patients are sensitized to multiple allergens, we sensitized mice with a combination of three allergens—dust mite (D), ragweed (R) and Aspergillus (A). Our results suggest that multiple allergen exposure breaks through tolerance and induces airway pathologies and hyperreactivity that persist long after the allergen exposure.
The majority of the methods used in this paper can be found in the journals Online Repository.
Allergens used include ovalbumin (Sigma) and extracts of dust-mite (D. Farinae), ragweed (A. artemislifolia), and Aspergillus fumigatus (Greer Laboratories, Lenoir, NC). Adjuvant was aluminum and magnesium hydroxide (Imject® alum, Pirece; 1:1 v:v with allergen). Quantities of allergens for subcutaneous (100 μL behind ear) and intranasal allergen (15 μL in saline) challenges are as follows: D. Farinae (5 μg, LPS content 3–35 EU by LAL assay), Ragweed (50 μg, LPS content 5 EU), Aspergillus (5 μg, LPS content 0.1 EU), DRA mix (dust mite 5 μg, ragweed 50 μg and Aspergillus 5 μg) or OVA (60μg). The dose of the allergens was based upon a survey of previous publications indicating successful sensitization and elicitation of allergic inflammation in the lungs.4–10
Female BALB/C mice were immunized at 12–15 weeks of age twice one week apart with various combinations of dust mite, ragweed, and/or Aspergillus in alum as described above. The week following the 2nd immunization intranasal challenges were given twice a week for 8 weeks with the immunizing allergens. A control group of mice were immunized with saline in alum and intranasally exposed to saline. Another group of mice were chronically exposed to a single or multiple (DRA) allergens intranasally twice a week for 8 weeks without the preceding subcutaneous immunization in alum. The mice were rested for 3 weeks after the 8-week exposure period before analyses. For anti-cytokine therapy the antibody was given 2 weeks after the last (eighth) week of allergen challenge and assessed on day 21 post-allergen challenge. A time-line of manipulations and interventions in the acute and chronic protocol is shown in Figure 1A.
Mice were immunized with ovalbumin or Aspergillus in alum as above twice one week apart. The week following the second immunization animals were challenged intranasally with the immunizing allergens five days in a row (OVA model) or twice a week for 2 weeks (Aspergillus model). This slight difference in the Aspergillus model was done in order to be consistent with the chronic model where allergen exposure is done twice a week. Airway pathologies and hyperreactivity were assessed at 72 hr after this acute exposure. In another set of experiments in the OVA model symptoms were left to resolve for two weeks before intranasal anti-cytokine therapy (50 μg anti-IL-5 or anti-IL-13) was administered and a secondary (recall) allergen challenge was given 1 hour later and symptoms assessed 24 hours after this challenge.
Inflammation was quantified using Metamorph image acquisition and analysis software on H&E stained lung sections (5 μm) at 200x magnification. Airway inflammation (AI) and airway epithelial hypertrophy (AEH) were measured as the area of inflammatory infiltrates and area of epithelium per μm of basement membrane, respectively using a minimum of seven airways per mouse and 3–5 mice per group.
We studied the effect of chronic exposure to a single allergen—Aspergillus fumigatus extract and compared it with that after acute exposure. We also compared the effect of acute exposure to Aspergillus with that to ovalbumin, since the latter is a widely used experimental allergen in mouse studies. We found that Aspergillus immunization followed by a short series of intranasal challenge (acute model) caused significant airway inflammation, epithelial hypertrophy and airway hyperreactivity to methacholine and these effects were comparable to that with ovalbumin (Figure 1B, upper panel and 1C). However, chronic (twice a week for 8 weeks) exposure to Aspergillus followed by a 3-week rest period led to a remarkable reduction in airway inflammation, AHR (Figure 1B, middle panel and 1C), and epithelial hypertrophy (online repository Figure E1A). Airway inflammation reduced from a 12-fold increase over the saline control in the acute model to a 3.4-fold increase over the saline control in the chronic model (Figure 1C). AHR (Figure 1D, compare acute A+ with A+) and epithelial hypertrophy (Figure E1A) showed a near-complete resolution. The possibility was examined that the mice in the chronic model had a higher level of inflammation immediately after the last allergen exposure and simply failed to maintain it over the next 21 days. To this objective we studied airway inflammation 24h after the last allergen exposure in the chronic Aspergillus model. The degree of inflammation at 24h was no different from that observed on day 21 (Figure E1B and E1E) suggesting that these mice developed tolerance in regards to inflammation.
Since most allergic subjects are sensitized to multiple allergens, we reasoned that sensitization to multiple allergens might alter the outcome of chronic exposure. To this goal we sensitized and exposed mice for 8 weeks to a combination of 2 and 3 common allergens—dust mite (D), ragweed (R) and Aspergillus (A) extracts. The combination of allergens, especially, the three allergen (DRA) combination induced severe airway inflammation (12.4μm2 area/μm of basement membrane) (Figure 1B, middle panel and 1C), epithelial hypertrophy (Figure E1A) and AHR (Figure 1D, compare DRA+ with A+), which persisted for 3 weeks after the last allergen exposure. Examination of a smaller group of mice 30 days after the last allergen exposure exhibited a level of airway inflammation that was similar to that observed at 3 weeks (Figure E1C and E1E). The combination of two allergens (DR, DA and RA) produced an intermediate phenotype with DA and RA (both contain the Aspergillus extract) being more potent than DR (Figure E1D and E1E). The chronic model showed a higher level of airway resistance as opposed to tissue resistance and increased hysteresivity when analyzed by a forced oscillation method (Figure E2).11 If tolerance is defined by the loss of inflammation from the acute to chronic phase to a level similar to saline-treated control mice, 93% of DRA mice (N=14) broke tolerance whereas no single allergen mouse and 67% of two-allergen mice (only mice with Aspergillus extract combinations) broke tolerance. We asked whether mucosal sensitization to DRA without the preceding subcutaneous immunization in alum would induce sustained airway inflammation. Mice chronically exposed to DRA without alum developed airway inflammation (Figure 1B, bottom panel and 1C) and AHR (Figure 1D, compare DRA- with R-) that persisted for 3 weeks. Mice exposed to ragweed alone failed to develop inflammation and AHR. To confirm that the mice had systemic sensitization to the allergens following mucosal exposure we studied splenic T cell proliferation. Spleen T cells from DRA but not ragweed mice showed strong antigen-driven proliferation (Figure E3).
DRA mice showed a 30 fold increase over the chronic Aspergillus (A) model in toluidine blue-positive intra-epithelial mast cells (Figure 2A). Airway inflammation was associated with a significant increase in MBP+ eosinophils (Figure 2B), goblet cell hyperplasia (fluorescent PAS staining for mucus glycoproteins, Figure 2C), peribronchial smooth muscle mass (alpha-smooth muscle actin-positive cells, Figure 2D), and collagen deposition (Sirius red stain, Figure 2E) when compared to the single allergen (A) model and saline control. Peribronchial inflammation in DRA was characterized by the presence of CD45+ leukocytes, CD4 and CD8 T cells (Figure E4).
We measured cytokine and chemokine levels in lung homogenates on day 21 after the last allergen exposure in chronic models. Eotaxin, IL-5, IL-10, IL-13, and IL-17 were significantly elevated in the lungs of DRA mice compared to saline and Aspergillus (Figure 3A). RANTES and TGF-β were significantly higher in DRA versus saline but not Aspergillus. The other cytokines and chemokines did not show any difference among the study groups. IL-4 was detectable in low quantities and did not show any difference between the study groups (Figure E5). All sensitized animals had elevated levels of Aspergillus-specific IgE and IgG1 antibodies but differences were not significant (Figure 3B). Th2 cells, eosinophils and basophils selectively utilize VCAM-1 (vascular cellular adhesion molecule 1) for adhesion to endothelium and entry into the tissue.12, 13 DRA mice had endothelial VCAM-1 expression 25 times higher per μm of basement membrane than that expressed in chronically challenged single allergen immunized and control mice (Figure 3C).
Anti-IL5 and anti-IL13 antibodies are highly effective in the mouse model of asthma.14, 15 Human studies with an anti-IL5 antibody failed to demonstrate clinical usefulness.16 A recent human trial with a mutant protein that blocks the activation of IL-4Rα, a receptor common to both IL-4 and IL-13, showed selective inhibition of the late-phase allergic reaction without affecting airway hyperreactivity, FEV1, PEF, symptoms score, IgE and the eosinophil count.17 We examined the effect of neutralizing antibodies against IL-5 and IL-13 in the acute (OVA) and chronic asthma models. We chose the acute OVA and not the acute Aspergillus model because most previous studies with anti-cytokine antibodies used the former model. Antibody was administered 1 hour before the allergen challenge in the acute OVA model. In human chronic asthma this pretreatment approach is unrealistic since it is difficult to predict the timing of the next asthma attack. To mimic an intervention in human asthma we administered antibody 2 weeks after the last allergen challenge (1 week before sacrifice) in the chronic asthma model. As reported previously, both anti-IL5 and anti-IL13 antibodies prevented airway inflammation & AEH (Figure 4A&B) and AHR (Figure 5C) in the acute asthma model.14, 15 However, both antibodies were totally ineffective in reducing airway inflammation and hyperreactivity in the chronic asthma model (Figure 4A–C). Thus, our chronic model is better suited to predict therapeutic outcome in response to anti-cytokine therapy.
Tolerance to a single allergen due to chronic exposure has previously been shown to be mediated by natural regulatory T cells and tolerizing dendritic cells.5, 6, 18–21 Interestingly, we observed increased FoxP3+ cells in the lungs from DRA mice as compared to A mice (Figure E6). This is in agreement with the increased concentrations of IL-10 and TGF-β in the lung tissue from DRA mice. The results suggest that the lack of tolerance in DRA mice is not due to the absence of FoxP3+ regulatory T cells. Next we examined the effect of single vs. multiple allergens on dendritic cells. We generated CD11c+ dendritic cells by culturing bone marrow cells in GM-CSF for 7 days.22,23 At the end of the culture period more than 72% of the cells were CD11c+ and a majority of these cells expressed surface MHC class II (Figure E7A, left panel). We observed a dose dependent increase in ERK1/2 signaling in these dendritic cells when incubated with a single allergen. Figure E7B shows pERK1/2 induction by the Aspergillus extract. Incubation with a single dose (3 μg/ml) of dust mite, ragweed and Aspergillus and at the equivalent dose of DRA (1 μg of each allergen, total 3 μg/ml) induced a similar level of pERK1/2 in dendritic cells (Figure 5A, left panel). The effect on p38 signaling was strikingly different. Dust mite, ragweed and Aspergillus at the dose of 3 μg/ml induced p38 phosphorylation in 22 ± 3%, 28 ± 5% and 25 ± 3% cells. DRA at the same total dose (i. e, 1/3 the dose of each individual allergen) induced p38 phosphorylation in 88 ± 7% cells, suggesting a synergistic effect (Figure 5A, right panel, also Figure E7C). We measured the LPS content and protease activity of the allergen extracts. Dust mites and Aspergillus had the highest and lowest levels of LPS, respectively (see Allergens and Adjuvant in the Method Section). Despite the 30 to 50-fold difference in the LPS content between Aspergillus and dust mites, the difference in p38 and ERK1/2 phosphorylation among these allergens was negligible. The Aspergillus extract showed the highest level of protease activity among the three allergens (Figure E8), but this activity also did not make any difference in dendritic cell MAPK signaling.
Next we examined the biological relevance of this increased p38 signaling by DRA. The expression level (mean fluorescence intensity) of MHC class II was significantly increased in triple allergen (DRA)- as compared to single allergen-treated cells (Figure 5B, left panel). Unlike MHC II, CD40 was expressed at a low level on dendritic cells and this level changed very little after allergen stimulation. Individual allergens increased the number of cells expressing CD40 (Figure 5B, right panel) as reported previously.24 However, DRA induced significantly greater number of CD40+ cells. The DRA-induced increase in MHC class II antigen and CD40 was significantly inhibited in cells that were pretreated with the p38 MAPK inhibitor SB202190 (Figure 5C). These results suggest that multiple allergens induce a higher level of dendritic cell signaling, which leads to increased expression of co-stimulatory molecules.
Next we examined the TCR-mediated proliferative response of CD4 T cells in the presence and absence of CD11c+ dendritic cells from the chronic DRA, A and saline control mice. Anti CD3 stimulation did not induce significant differences in proliferation of T cells from DRA and A mice but DRA T cells showed increased proliferation as compared to A and saline T cells when stimulated with anti-CD3 + anti CD28 antibodies (Figure 5D). Similarly, when dendritic cells were used in lieu of anti CD28, DCs from DRA mice provided the highest level of co-stimulation (Figure 5E). In contrast, dendritic cells from A mice significantly inhibited DRA T cell proliferation. DCs from DRA mice showed a tendency to augment proliferation of T cells from saline and A mice but the difference did not reach statistical significance.
We have developed a mouse model of chronic asthma with the following important new features: 1). Resistance to tolerance despite repeated allergen exposure: 2). Persistence of eosinophilic inflammation, AHR and other features of asthma beyond the first 3 weeks after exposure; 3). Mucosal sensitization to allergens and induction of chronic inflammation without the need for adjuvants; and 4). Resistance to anti-IL-5 therapy, thus mimicking aspects of human chronic asthma. In most previous studies pathologic features of experimental asthma were investigated within 24–72 hr after the allergen exposure. In a few studies where the pathologic features were observed longer than 72 hr, near-complete resolution took place within one week after the allergen exposure.25 In our model the Aspergillus extract was of pivotal importance. A combination of allergens that contained Aspergillus induced sustained inflammation. Nonetheless, Aspergillus alone was not sufficient to produce the phenotype. Previously the effect of a combination of two allergens (recombinant cockroach and dust mite proteins rBla g2 & rDer f1) was studied in the acute asthma model at 72 hr. These mice did not develop AHR.26 Dust mite was previously shown to facilitate sensitization to inhaled ovalbumin,27 which is similar to the effect of Aspergillus in our model. It should be noted that there chronic models of asthma where chronic exposure to a single allergen leads to persistent albeit low level inflammation and AHR.28,29
One shortcoming of mouse models is that they do not have sustained airway hyperreactivity. Although methacholine produces meaningful total lung resistance, the response from the parenchyma dominates while airway resistance is largely absent.30 The DRA model distinguishes itself from many other models by the persistence of airway resistance. One of the remarkable features of our model is the presence of intraepithelial mast cells. The accumulation of intraepithelial mast cells in tracheal sections and peribronchial regions has been reported previously.31 Intraepithelial mast cells are ideally situated to encounter inhaled allergens and promote sustained inflammation, and airway hyper-reactivity.32
A common practice in preclinical drug development is to pre-treat animals with a therapeutic agent before an allergen challenge and then study the outcome. This simple approach has limitations in human asthma. Our model has an advantage in this regard because in its chronic stage it can be used to test therapeutic agents without performing a de novo allergen challenge. The result from the anti-IL5 antibody experiment suggests that our asthma model is a better predictor of outcome in human asthma. Our results are in agreement with a recent report by Kumar et al, who showed a lack of efficacy of an anti-IL5 antibody.33 An anti-IL13 antibody inhibited airway reactivity only at the highest dose (50 mg/ml) of methacholine but had no effect at lower doses.
The mechanism of persistence of inflammation and airway hyperreactivity in DRA mice is likely related to the effect of multiple allergens on dendritic cells. DRA stimulates p38 MAPK phosphorylation in more dendritic cells than a single allergen at the same protein concentration. This heightened p38 MAPK signaling increases MHC class II and CD40 expression on dendritic cells, which is in agreement with previous reports.34,35 We believe that the increased dendritic cell signaling and activation contribute to the heightened co-stimulatory activity of DRA dendritic cells. As a result T cell activation remains unabated, and the airway inflammation persists.
The mechanism by which multiple allergens induce a synergy in p38 but not ERK1/2 signaling is unknown. This effect is unlikely to come from the protein load or the protein-derived antigenic epitopes alone as we have used the same amount of the protein. It is possible that the synergy results from the adjuvant-like activities of the allergens. Many allergens contain proteases and NADPH oxidase. In one study intranasal ovalbumin failed to induce airway inflammation. However, when administered along with the proteases from Aspergillus fumigatus, it produced significant airway inflammation.10 Pollen-derived NADPH oxidase has been shown to be essential for induction of airway inflammation by allergens.8 NADPH oxidase generates reactive oxygen species, which induces non-specific inflammation in a p38-dependent manner. These studies suggest additional non-antigenic functions of allergens. We recognize the allergen extracts we used are mixtures containing multiple proteins and that the effects seen in the model may be due to unknown components of the allergens.
This work was supported by NIH grants RO1 AI059719 and AI68088, PPG HL 36577 and N01 HHSN272200700048C
“This article has an online data supplement, which is accessible from this issue’s table of content online at the Journal of Allergy & Clinical Immunology website”
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.