Mouse models of allergic asthma: acute and chronic allergen challenge

Human allergic asthma is defined as a chronic inflammatory disorder of the airways and is characterised by airway inflammation, persistent airways hyperresponsiveness (AHR) and intermittent, reversible airways obstruction (GINA, 2006; Bousquet et al., 2000). In addition, structural changes in the airway including subepithelial and airway wall fibrosis, goblet cell hyperplasia/metaplasia, smooth muscle thickening and increased vascularity are observed (Bousquet et al., 2000; Fish, 1999). These changes are termed ‘airway remodelling’ and may be the result of repeated exposure to the allergen, which causes repeated or continuing inflammation in the airways (Zosky and Sly, 2007). Chronic inflammation and structural changes are thought to have functional consequences that contribute to asthma symptoms.

The exact cellular and biochemical processes underlying chronic inflammation and airway remodelling are poorly understood. Although the best approach to investigate these processes, and to identify crucial pathways and potential novel targets for drug therapy, is to perform studies in human asthmatics, the required mechanistic studies are not acceptable owing to ethical reasons. Animal models provide an alternative for investigating disease mechanisms and progression. Because asthma is a complex multifactorial disease, it is unlikely that a single animal model of asthma that replicates all of the morphological and functional features of the chronic human disease will ever be developed. However, we can use animals to model specific features of the disease, and much of our current understanding of disease processes in asthma, and in particular the response to allergens, comes from studies in laboratory animals such as guinea pigs, rats and mice. The mouse is the most widely used species, mainly because of the availability of transgenic animals and because of the wide array of specific reagents that are available for analysis of the cellular and mediator response. This Commentary will, therefore, focus on the development of allergen challenge models in the mouse.

Mice do not spontaneously develop asthma; so, in order to investigate the processes underlying this disease, an artificial asthmatic-like reaction has to be induced in the airways. Mouse models of the acute allergic response to inhaled allergens have been widely used to elucidate the mechanisms underlying the immunologic and inflammatory responses in asthma, and for the identification and investigation of novel targets for controlling allergic inflammation.

A variety of different acute allergen challenge models have been developed in mice and a number of sensitisation and challenge protocols have been employed. Some of these are summarised in Table 1.

The nature of the acute inflammatory model may be influenced by the choice of mouse strain, the allergen, and the sensitisation and challenge protocol (Zosky and Sly 2007; Kumar et al., 2008). The most commonly used strain of mouse for antigen challenge models is BALB/c as they develop a good T helper cell 2 (Th2)-biased immunological response (Boyce and Austen, 2005). However, other strains (C57BL/6 and A/J) have been used successfully in allergen challenge studies (Kumar et al., 2008). Ovalbumin (OVA) derived from chicken egg is a frequently used allergen that induces a robust, allergic pulmonary inflammation in laboratory rodents. A review of OVA challenge models has recently been published by Kumar et al. (Kumar et al., 2008). OVA, however, is seldom implicated in human asthma, and other groups have used alternative allergens that may have greater clinical relevance, for example house dust mite (HDM) and cockroach extracts (Johnson et al., 2004; Sarpong et al., 2003).

Although many different sensitisation and challenge protocols have been used, the basic model is consistent. Acute sensitisation protocols usually require multiple systemic administration of the allergen in the presence of an adjuvant. Adjuvants such as aluminium hydroxide (AlOH₃) are known to promote the development of the Th2 phenotype by the immune system when it is exposed to an antigen. Adjuvant-free protocols have also been described (Blyth et al., 1996), but these usually require a greater number of exposures to achieve suitable sensitisation. Sensitisation solely via the airways has also been attempted using both OVA and HDM. With OVA models, success has been limited and only modest pulmonary inflammation and mild AHR have been observed. This might be because tolerance can develop when model protein antigens are delivered via inhalation without systemic sensitisation. Inhaled delivery of HDM has been more successful, possibly because of the intrinsic enzymatic activity of this allergen. After the sensitisation period (usually 14–21 days), the animal is challenged with the allergen via the airway, usually over a period of several days. Allergen may be inhaled as a nebulised formulation (aerosol), or administered by intratracheal (i.t.) or intranasal (i.n.) instillation of an aqueous formulation.

The acute challenge mouse models reproduce many key features of clinical asthma, for example elevated levels of IgE, airway inflammation, goblet cell hyperplasia, epithelial hypertrophy, AHR to specific stimuli and, in some models, early- and late-phase bronchoconstriction in response to allergen challenge. However, there are also some major differences. In acute challenge models, the pattern and distribution of pulmonary inflammation is different from that observed in individuals with asthma (Kumar and Foster, 2002). For example, bronchoalveolar lavage and histology studies indicate that the influx of inflammatory cells is dominated by eosinophils. Because of the short-term nature of the acute models, many of the lesions observed in chronic human asthma, such as chronic inflammation of the airway wall, and airway remodelling changes are absent. Furthermore, many of the key features appear to be short-lived and, in some models, airways inflammation and AHR have been shown to resolve within a few weeks after the final allergen challenge (McMillan and Lloyd, 2004).

Despite these apparent shortcomings, acute allergen challenge models have been successfully used to investigate disease processes, in particular the relationship between cells and inflammatory mediators, and how they orchestrate the inflammatory processes in the lung. Evidence supporting the hypothesis that asthma is a Th2-mediated disease, the role of the T cell in the allergic response, and the role of the eosinophil and its involvement in the development of AHR came from studies in the acute challenge models.

A wide range of cellular and molecular targets have been identified, and subsequently evaluated, in the acute challenge models. These include anti-cytokine approaches, specific mediator antagonists, monoclonal antibodies and specific enzyme inhibitors. With a number of these targets, efficacy has been demonstrated in the acute challenge models, but not in clinical trials in patients. Anti-IL-5 therapy is one of the most frequently discussed examples. IL-5 appears to have a key role in allergen-induced inflammatory responses and the development of AHR in mice. IL-5 knockout mice were protected both from acute allergic inflammation, AHR (Foster et al., 1996) and chronic airway remodelling (Cho, et al., 2004). However, IL-5 antagonists have been evaluated in both mild acute and chronic severe asthma, and have failed to show clinical benefit (Leckie et al., 2000; Kips et al., 2003). There are other examples of drugs that have shown efficacy in acute challenge models but not in clinical studies; these include VLA4 antagonists (Norris et al., 2005), PAF antagonists (Hozawa et al., 1995) and IL-4 antagonists (Riffo-Vasquez and Spina, 2002; Borish et al., 1999). One possible explanation for the contrasting data from the mouse and human studies is that in mouse acute challenge models it is only possible to assess the affect on the development of allergic responses; whereas with clinical studies, potential suppression of established inflammatory disease is investigated.

Thus, although the acute challenge models have been useful for investigating the process underlying acute airway inflammation and AHR, there are concerns relating to the suitability of acute challenge models for both the investigation of disease processes associated with chronic asthma, and assessing potential novel treatments that may slow, or reverse, the changes of airway inflammation and remodelling, and its effect on lung function.

In order to address some of the issues associated with acute challenge models, several research groups have investigated chronic allergen exposure in mice. The key aims were to reproduce more of the features of clinical asthma, such as airway remodelling and persistent AHR, and to enable novel therapies to be evaluated in a therapeutic setting rather than a prophylactic setting, i.e. evaluating the effect of a novel drug in an established pulmonary inflammation setting. A number of different chronic challenge models have been developed by increasing the number of allergen challenges, some examples are shown in Table 2. Kumar et al. and Lloyd have discussed some of these models in recent review articles (Kumar et al., 2008; Lloyd, 2007).

Chronic allergen challenge models involve repeated exposure of the airways to low levels of allergen for periods of up to 12 weeks. Different allergens have been employed and co-administration of an adjuvant is not always required. A number of investigators, for example Temelkovski et al., Fernandez-Rodriguez et al. and Wegman have used OVA to establish a chronic exposure (Temelkovski et al., 1998; Fernandez-Rodriguez et al., 2008; Wegmann, 2008). Others have used environmentally relevant allergens such as HDM extract or grass pollen (Johnson et al., 2004; Kim et al., 2006). One anticipated problem with chronic allergen exposure, in particular with model protein antigens such as OVA, was that long-term challenge might lead to the development of tolerance and downregulation of inflammation and AHR (Jungsuwadee et al., 2004; Kumar et al., 2008). Both the strain of mouse and route of allergen administration are factors that may influence the induction of tolerance (McMillan and Lloyd, 2004; Shinagawa and Kojima, 2003). Kumar et al. suggested that tolerance might be related to the high mass concentrations of the aerosolised allergen used, which could overwhelm the clearance mechanisms (Kumar et al., 2008). This group successfully developed a model in which sensitised animals were exposed to inhalation of controlled mass concentrations of aerosolised antigen to minimise parenchymal inflammation and tolerance.

Chronic allergen exposure in mice has been shown to reproduce some of the hallmarks of human asthma including allergen-dependent sensitisation, a Th2-dependent allergic inflammation characterised by eosinophilic influx into the airway mucosa, and AHR. In addition, in some models, there is evidence of airway remodelling with goblet cell hyperplasia, epithelial hypertrophy, and either subepithelial or peribronchiolar fibrosis. Importantly, some of the key features of the chronic allergen exposure models have been shown to persist after the final challenge. After cessation of allergen challenge, some features of airway remodelling have been shown to persist in a number of models (Johnson et al., 2004; Macmillan and Lloyd, 2004; Kumar et al., 2004a). In these models, the persistence of AHR and lung inflammation varies depending on the exposure protocols employed (Lloyd, 2007).

There are limitations associated with the chronic models. In some cases, features peculiar to the mouse are observed that are not seen in human asthma. For example, in the mouse, inflammation is not restricted to the conducting airways, whereas it is in humans, and both lung parenchymal and vascular inflammation/remodelling are observed (Wenzel and Holgate, 2006). Large increases in airway smooth muscle, which is a characteristic feature of chronic asthma (Carroll et al., 1993), either does not appear to occur (Wenzel and Holgate, 2006) or, if it does, only after high level challenge (Shinagawa and Kojima, 2003; McMillan and Lloyd, 2004). Furthermore, there is little or no recruitment of mast cells into the airway wall or epithelium (Boyce and Austen, 2005), which may reflect the paucity of mast cells in the airways of mice (Kumar and Foster, 2002).

Despite these issues, using repetitive allergen provocation in mice has been shown to mimic important features of the human disease and would appear to represent an improvement over the acute challenge models.

Chronic allergen exposure in mice appears to now be the model of choice for studying the role of specific cell types and inflammatory cytokines and mediators in the processes involved in chronic inflammation and, in particular, some of the structural changes to the airways. For example, Humbles et al. have used a chronic challenge model to investigate the role of the eosinophil in airway remodelling (Humbles et al., 2004), and Kumar et al. and McMillan et al. have investigated the role of TGF-β in the development of airway remodelling in their chronic challenge model (Kumar et al., 2004a; McMillan et al., 2005). In addition, Yang et al. have used chronic models to investigate the role of IL-13 in regulating chronic inflammatory changes to the airway (Yang et al., 2005). These, and future, studies will add to our understanding of the mechanisms of chronic inflammation and airway remodelling and their contribution to asthma symptoms.

Chronic models may also provide a more suitable system for the preclinical evaluation of novel therapeutic agents. Unlike acute challenge models, the chronic challenge model enables investigations to be carried out in a setting of established airway inflammation and AHR and, therefore, more closely reflects most of the experimental studies undertaken in the clinic. A number of groups are using chronic models to profile established and novel pharmacological treatments for asthma. These studies might provide an insight into the predictive nature of the chronic models with regard to clinical outcome. A summary of some of the compounds that have been evaluated in mouse acute and chronic allergen challenge models and in human asthma trials is provided in Table 3. A number of anti-inflammatory agents are being profiled in chronic models, and reports of efficacy with corticosteroids (Macmillan et al., 2005), leukotriene receptor antagonists (Henderson et al., 2002) and phosphodiesterase 4 (PDE4) inhibitors (Herbert, et al., 2008) are emerging. Preliminary data suggest that certain chronic models might reproduce sensitivity to drug therapy in the clinic. However, further studies using a wider range of molecules, including examples that have demonstrated positive and negative outcomes in the clinic, are required. These studies will enable investigators to judge whether the chronic models might better reflect clinical outcome and represent an improvement over the acute challenge models.

Mouse allergen challenge models are a basic, and frequently used, tool for asthma research. Because of the complexity and diverse nature of the disease, it is unlikely that a mouse model that is truly representative of clinical asthma will be developed. Efforts are focusing on modelling specific disease phenotypes rather than trying to reproduce all features of asthma in a single model. Acute allergen challenge models have been extensively used to investigate pulmonary inflammation and AHR, but their use is limited, in particular when relating findings to chronic asthma. Chronic allergen challenge models appear to reproduce some of the features of chronic asthma and might enable key questions, relating to both the pathogenesis of asthma and potential approaches to novel therapy, to be addressed. Both acute and chronic allergen challenge models have limitations that need to be taken into account when extrapolating findings from the animal model to the human disease.

It is important that mouse models are valid and reflect clinical asthma as closely as possible. Studies are ongoing to refine and improve existing models in order to enhance their utility. For example, asthma exacerbations account for the majority of asthma-related hospital visits and this is becoming a focus of research. Mouse models are being developed to model asthma exacerbations, and investigators are using acute and chronic allergen challenge models as a background for superimposing additional, higher intensity allergen challenge (Siegle et al., 2006; Ito et al., 2008) or viral infection (Bartlett et al., 2008).

A combination of improved animal models, which more closely reflect asthma, and use of relevant human systems will broaden our knowledge of the disease and help identify and evaluate new therapeutic targets.

Deposited in PMC for immediate release.