CD4⁺ and CD8⁺ T cells play a central role in a HDM driven model of allergic asthma

Asthma is a chronic inflammatory disease of the airways characterised not only by cellular infiltration into the airways, but also by an accompanying increase in the sensitivity and response to contractile agents (airway hyperresponsiveness-AHR) and to allergen exposure (early and late asthmatic responses) [1, 2]. The global prevalence of asthma is increasing [3, 4], and the symptoms are often inadequately controlled [5]. Pre-clinical models which closely mimic clinical features of asthma are useful to improve our understanding of the mechanisms driving the disease and thereby the ability to identify novel therapeutic targets.

Rodent models of experimental allergic asthma to date have contributed greatly to our current understanding of the pathogenesis of the disease. However, one of the most widely used aeroallergens in model development, ovalbumin (OVA), is rarely implicated in clinical asthma and is often used in more acute challenge protocols due to the development of inhalation tolerance and a diminishing or complete abrogation of the airway inflammatory response following multiple challenges [6‐9]. Conventional rodent OVA-models are thought to require co-administration of an adjuvant to establish successful sensitisation. Therefore OVA may not be the most disease relevant allergen. In contrast, exposure to house dust mite (HDM), the major source of allergen in house dust, has been shown to play an important role in asthma and also in other allergic diseases such as dermatitis and rhinitis [10‐13]. In the clinic it has been shown that sensitisation to HDM is a strong predictor for asthma and there is a correlation between the level of HDM exposure and sensitisation [10, 11]. Because of its clear relevance to human disease, HDM has in recent years become the allergen of choice in the development of animal models of asthma often relying on chronic topical exposure without developing tolerance [14‐19]. However, it is possible that in these models part of the inflammatory response is due to innate mechanisms in response to repeat inflammatory insults rather than allergic mechanisms [20, 21]. We propose that to fully understand the disease phenotype it is important to characterise the different stages of the inflammatory response in vivo. Therefore the aim of this study was to initially demonstrate this acute innate response and briefly investigate the mechanisms involved (i.e. Dectin 1/2 and TLR 2/4 receptors have been linked to this response) and then to go on to develop a murine HDM-driven allergic model based around a separate sensitisation and challenge protocol. Furthermore we also investigated the role of key allergic effector cells in features of allergic asthma using this model. We additionally aimed to compare key features of this model to a standard OVA-model and investigate its potential for chronic allergen exposures without the associated tolerance. We suggest the data presented here will help to elucidate the mechanisms driving the inflammation observed after acute and chronic HDM exposure and present targets for possible future therapeutic intervention.

Pre-clinical models closely mimicking the cardinal features of asthma are useful in understanding the mechanisms driving the disease and can help in identifying novel therapeutic targets. Traditional models of allergic asthma involve the use of OVA along with adjuvants such as Alum and have been profiled extensively. Yet, these models have limitations. On the other hand, HDM is increasingly being used as a disease-relevant antigen in rodent models of allergic airway inflammation since it does not require an adjuvant and is not linked to the development of inhalation tolerance [15‐18]. However, previous studies have suggested that part of the airway inflammation observed in response to topical HDM is due to innate rather than allergic mechanisms [20, 21]. Therefore, to fully understand the disease phenotype and extrapolate in vivo data to the clinic, we suggest it is necessary to clearly distinguish between the events underlying sensitisation and challenge, and characterise all stages of the allergic inflammatory response in vivo. In the present study, we describe an in vivo mouse model of HDM-induced allergic asthma with a distinct sensitisation and challenge phase characterised by allergen induced airway inflammation and AHR dependent on CD4⁺ and CD8⁺ T cells, but not B cells or IgE. This inflammation and AHR was sensitive to treatment with systemic steroids showing the model to be relevant to clinical asthma. In addition this model exhibits persistent airway inflammation and AHR following chronic exposure to the allergen.

A single topical exposure to HDM caused robust inflammatory cell recruitment into the mouse airway. Although the prominent cell type was neutrophils we could also detect an increase in eosinophils and lymphocytes (Fig. 1). This can be recapitulated in the rat [20] and thus far the mechanism underlying this innate response remains unclear. HDM extracts have been shown to contain various other contaminants capable of activating inflammatory pathways [33‐35]. Our data exclude a role for HDM innate protease activity, TLR4 and Dectin-1 (Additional file 1: Figure S1) activation in this response despite previous studies implicating these factors in HDM-mediated inflammatory responses [24‐26, 28], but show that there is a potential involvement of TLR2 and Dectin-2 (Additional file 1: Figure S1) in accordance with previous reports [27‐30]. Although this falls outside the scope of the current publication, we suggest that further investigation of this underlying mechanism driving acute HDM-induced inflammation will lead to a better understanding of animal models, and could potentially uncover disease relevant mechanisms involved in the pathogenesis of asthma.

This apparent innate inflammatory response to HDM could confound the understanding of mechanisms underlying asthma-like features in animal models and complicate extrapolation of results to the clinic. We therefore developed a HDM-driven murine model in which airway inflammation and AHR only occurred in animals which had been previously sensitised to HDM. Using direct comparison studies we demonstrated that Alum during the sensitisation stage was not required for the allergic airway inflammation and AHR observed in this HDM model (Figs. 4 and 5). This is a great advantage as it avoids the use of an exogenous adjuvant and more closely resembles the allergenic cocktail to which patients are exposed. HDM has been demonstrated to possess innate adjuvant capacity and as such the fact that an exogenous adjuvant was not required for successful sensitisation or for the development of allergic asthma following allergen exposure, is not surprising. Interestingly, we also show that an adjuvant was not required for the production of OVA-specific IgE or the induction of airway inflammation and AHR in an OVA-dependent murine model (Additional file 4: Figure S4 and Additional file 5: Figure S5). This is surprising when one considers current dogma and that the majority of published models suggest that the OVA model requires the use of a systemic adjuvant to induce allergic sensitisation. Despite being widely criticised for not being clinically relevant and although adjuvant-free OVA models have been described [36‐38] adjuvant dependent OVA-models continue to be used in the field. It has been noted that OVA is contaminated with lipopolysaccharide which might account for some of its auto-adjuvant properties [39].

Our results also highlight the importance of choosing the appropriate route for allergen delivery. Although several murine models utilising systemic sensitisation have been described [40, 41], the majority of research groups that utilise HDM-driven murine models currently favour topical sensitisation rather than systemic routes [14‐19]. However, in contrast to these published models, this new HDM model uses systemic allergen sensitisation with a full HDM extract with all its associated contaminants without the need for an adjuvant. It is currently accepted that asthmatics become sensitised to HDM and other aeroallergens through the airways. However, several other mechanisms can lead to sensitisation and the development of allergic airways disease. Infants may have some features of allergy at birth through prenatal in-utero sensitisation [42‐44]. In addition, in those atopic patients whereby atopic dermatitis is developed early on, followed by allergic rhinitis and subsequently atopic asthma later in life systemic sensitisation seems more likely, rather than sensitisation through airway exposure [45‐47]. The fact that not every person exposed to HDM becomes sensitised and develops allergic asthma suggests that an animal model where allergen exposure only elicits a response in previously sensitised animals seems preferable. In contrast to reports that Balb/c mice can be primed to respond to HDM challenges via intranasal sensitisation [48] in our hands in this C57BL/6 murine model this did not occur (Fig. 2 and Additional file 3: Figure S3) and thus supports our use of systemic sensitisation during model development.

Due to the difficulty developing appropriate models capable of recapitulating key features of chronic asthma, the mechanisms underlying the chronicity of the disease have been far less investigated when compared to acute, immediate responses. The common allergen, OVA, frequently used in model development has been reported by many groups following chronic exposure protocols to result in inhalation tolerance and the abrogation of the inflammatory airway response [6‐9]. It has been suggested that intermittent allergen exposures over prolonged periods of time do not result in the development of tolerance [31, 32]. Our data show that the administration of HDM over the course of 5 weeks in the absence of exogenous adjuvants does not lead to tolerance but a robust inflammatory response in HDM sensitised animals characterised by persistent airway eosinophilic, neutrophilic and lymphocytic airway infiltration accompanied by AHR to 5-HT (Fig. 6). It is plausible that the robust allergic response in our model is attributable to the non-continuous allergen exposure protocol. These results parallel data collected in those models employing chronic topical HDM challenges [18‐23] and highlight the feasibility of using HDM to model chronic disease symptoms and the potential of the model to be used to assess the effect of new therapeutics dosed prophylactically and therapeutically in established ‘disease’ conditions.

HDM sensitisation induced the production of total and specific IgE (Fig. 2 and Additional file 2: Figure S2). Decades of research have implicated IgE-mediated responses in asthma and IgE is now a key target for asthma therapy. However, some have argued that increases in measurable IgE could be merely a result of asthma i.e. a marker of inflammation [49]. Additionally, studies utilising IgE directed therapy have shown mixed results [50]. Thus, IgE independent mechanisms have been suggested to also contribute to allergic disease [13, 51‐53]. Through the use of KO mice deficient in IgE and functional B cells, our findings suggest that, at least in this model, IgE and B cells are not required for the generation of immune responses in animals systemically sensitised with and topically exposed to HDM (Figs. 10 and 11). This is in accordance with the suggestion that HDM can contribute to allergic disease independent of IgE, but is in contrast with a recent publication showing that mice, deficient in B cells, repeatedly exposed to an allergenic cocktail of HDM, Alternaria and Aspergillus exhibited reduced airway inflammation, lung pathology and AHR [54]. However, it is likely that the different allergens administered can act through different pathways some of which might indeed be independent of B cells and IgE production [55‐58]. Furthermore, existing experimental model data, in contrast to these observations, suggested that the role of B cells in allergen induced immune responses may depend on the route of allergen delivery and on the type of allergen used [54, 58]. This could account for the differences with our model and the divergent data regarding the role of B cells in the regulation of immune responses to various allergens. Moreover, we show that the airway inflammatory response and associated AHR in this model is T cell dependent as CD8⁺ and CD4⁺ T cell KO mice failed to develop a response distinguishable from WT saline exposed control mice (Figs. 8 and 9). In the clinic elevated numbers of lymphocytes correlate with eosinophilia and asthma severity and this lymphocytosis has been shown to include both CD4⁺ and CD8⁺ T cells [59‐61]. The CD4⁺ T cell is thought to be the dominant active T cell subtype in clinical allergic asthma as well as in animal models of asthma [57, 60, 62‐64]. In contrast, the role of CD8⁺ T cells both in human asthma and animal models of allergic disease remains controversial, having been reported to be both protective or deleterious [65]. Supporting our data however, an increase in airway CD8⁺ T cells following allergen challenge predicts annual FEV₁ decline in asthmatic patients [66, 67] and increased numbers can be found in patients following a fatal asthma attack [68, 69].

It is clear from published experimental data that CD4⁺ and CD8⁺ T cells can contribute to AHR development and allergic inflammation. However, the interpretation of their role and importance during sensitisation and/or challenge across various different models can vary largely. It is well established that CD4⁺ T cells are required during sensitisation [57, 70, 71]. Reconstitution of CD4⁺ T cells alone in RAG^-/- mice, deficient in both T-and B-cells, prior to sensitisation was sufficient for Aspergillus fumigatus-induced AHR and airway inflammation [57]. Interestingly, while CD8⁺ T cells were not required, this observation was also dependent on IL-4. Additionally, CD4⁺ T cell depletion before allergen challenge resulted in reduced AHR and airway eosinophilia [62] suggesting they are indeed crucial for the development of allergen induced airway disease. CD8⁺ T cells can differentiate, under the right conditions, in separate subsets [72‐75]. It has been argued that these different subsets within different models could be responsible for the contradictory reports. Specific depletion of CD8⁺ T cells in sensitised rats enhanced the airway inflammation and related late asthmatic responses [76] and in addition augmented airway remodelling and mucus production. [77, 78]. Similarly, certain CD8⁺ T cell subsets have been shown to prevent OVA-induced inflammation in mice [79, 80]. Others have shown that CD4⁺, but not CD8⁺ T cells are required for allergic airway responses [57, 81]. In contrast, a body of work by Gelfand and colleagues [70, 71, 82, 83] using mainly a systemic sensitisation and topical challenge OVA mouse model of allergic asthma corroborates our results in this HDM model and highlights a role for both CD8⁺ and CD4⁺ T cells in AHR and airway inflammation. Using a variety of adoptive cell transfer studies and KO mice, they demonstrated the activation of allergen primed CD8⁺ T cells and IL-13 production within the airway [71, 83] critical for AHR and airway inflammation in this model. Additionally, their data suggest that CD4⁺ T cells and IL-4 are crucial during sensitisation for CD8⁺ T cell activation [70]. More recent publications also highlighted a possible role for IL-13 producing CD8⁺ T cell populations in augmented allergen driven inflammatory responses even in the absence of prior sensitisation [84]. It would be interesting to see whether similar pathways underlie the AHR and airway inflammation in our HDM model and merits further investigation. It is likely that a close interaction between CD4⁺ and CD8⁺ T cells is key to driving the allergic responses within our HDM model.

In summary, we have developed a murine model of allergic asthma characterised by allergic airway inflammation and AHR which is dependent on prior allergen sensitisation without the need for an exogenous adjuvant. We show that this inflammation and AHR is dependent on CD4⁺ and CD8⁺ T cells and is persistent following multiple challenges. Our data reinforce the notion that T cells play a pivotal role in orchestrating pathological and physiological changes observed in human asthma. This newly developed model holds advantage over models that utilise topical administration in isolation in that sensitisation and challenge phases can be easily distinguished which allows separate interrogation. Additionally, with the possibility of repeat challenges this model could be useful in studies exploring therapeutic interventions in established chronic disease.