Animal models of allergic reactions are commonly established in guinea pigs, rats, and mice. Guinea pigs have been used for asthma research for several decades owing to the high anatomical and physiological similarities between the airways of humans and guinea pigs. The important similarity is that the acute bronchoconstriction response to allergens is largely mediated by histamine and leukotrienes. Guinea pigs are recommended for preclinical research on allergen-induced airway hyper-responsiveness (11). However, allergen-induced IgG1 is associated with a high degree of early airway smooth muscle contraction in the guinea pig, leading to a higher level of histamine release by the mast cells (12). Thus, guinea pig models are suited for investigating mast cell-dependent reactions. Among rat strains, Sprague–Dawley (SD) rats are typically used for studying passive cutaneous anaphylaxis (PCA) reactions (13). Brown Norway (BN) rats are used for exploring the allergenicity of food proteins (14) and asthma studies because of their high sensitivity (15). Mice have also been used as allergy models, and different species or strains show variability in terms of physiology and immunology. BALB/c mice are the most commonly used mouse strain in antigen challenge models because they develop a Th2-biased immunological response (16). C57BL/6 and A/J mice have been successfully used as experimental models for allergic asthma (17). C57BL/6 mice are commonly used to establish transgenic or knockout models that provide information about specific genes involved in human diseases. In the OVA-induced asthma model, C57BL/6 mice display a higher eosinophil percentage in the bronchoalveolar lavage fluid (BALF) than that of BALB/c mice, whereas BALB/c mice show an increased lung mast cell count and airway hyperresponsiveness (AHR) than that observed in C57BL/6 mice (18). A/J mice exhibit higher levels of AHR and reactivity to methacholine than do BALB/C and C57BL/6 mice (19).

Humanized mice, that is, mice reconstituted with human immune cells, are a novel tool for studying the immune response during type I hypersensitivity. Eschborn M developed a humanized mouse model of allergen-induced IgE-dependent gut inflammation in PBMC-engrafted immunodeficient mice (20). Another humanized mouse model of immunodeficient γc-deficient mice expressing transgenes for human stem cell factor, granulocyte-macrophage colony-stimulating factor, and IL-3 developed mature functional human mast cells. Human mast cells in mice were sensitized with patient-derived IgE monoclonal antibodies specific for peanut (Arachis hypogaea) allergen 2, which induced fatal anaphylaxis on exposure to the peanut allergen (21). These humanized mice are useful for recapitulating human IgE-mediated or mast cell-induced anaphylaxis.

The classical antigens used for Type I allergic reaction models include ovalbumin (OVA) and 2,4-dinitrophenol (DNP). DNP is a low-molecular-weight chemical hapten that is conjugated to carrier proteins such as human serum albumin or bovine serum albumin. Anti-DNP-IgE and anti-DNP-human serum albumin (or anti-DNP-bovine serum albumin) antibodies have been used in PCA or mast cell models (22). Nevertheless, OVA is most commonly used, primarily for practical reasons such as low costs, high purity, and commercial availability of antibodies (OVA-specific IgE/IgG) and transgenic mice (OVA-specific transgenic mice). OVA is an allergenic protein that is commonly used to induce allergic bronchospasm in guinea pigs and PCA, asthma, and food allergies in rodent models. However, immunization with OVA in mice does not always induce high levels of IgE and may require re-exposure to allergens. In fact, OVA-induced allergy requires the use of both OVA and Th2-adjuvant aluminum hydroxide to sensitize the mice (23).

Antigens that cause food allergies are typically protein antigens such as milk, eggs, wheat, fish, shellfish, peanuts, walnuts, and soybeans (24). To better mimic the gastrointestinal lesions observed in human food allergies, animals are often sensitized via the oral route. Due to their oral tolerance, cholera toxin and staphylococcal enterotoxin B (SEB) are generally used as mucosal adjuvants to enhance the induction of antigen-specific IgE. Subcutaneous sensitization has been proposed as an alternative approach for inducing gastrointestinal symptoms of food allergies in animals (25).

The allergen inhalation challenge is a useful clinical model for evaluating allergic airway diseases that are triggered by the inhalation of food allergens or airborne allergens (26). Although OVA is widely used in asthma experiments, it is not a relevant aeroallergen for human asthma. The clinical relevance of aeroallergens such as house dust mites (HDM) has been investigated. Additionally, HDM sensitize animals via the airways and do not require adjuvants (27) similar to that of the human sensitization route.

Passive anaphylactic models are broadly categorized into local cutaneous and systemic reactions. The principle of PCA is the same as that of anti-allergen IgE sensitization and allergen re-exposure in mast cells. The model is generated by administering a subcutaneous injection of mouse IgE-rich antiserum or anti-allergen-IgE monoclonal antibodies (28) at the back or dorsal portion of either ears in rats (29) or paw of mice (30). After 24–48 h, Evans blue dye containing the antigen is administered via the tail vein. As the local vascular permeability increases owing to the allergic reaction, the degree of the reaction is assessed by measuring the diameter of the local blue skin spots or observing the extent of blue skin staining. However, this parameter is non-quantitative and may not reflect a dose-response relationship for evaluating drug efficacy; hence, some researchers have collected local skin samples and quantitatively determined Evans blue absorbance using formamide (22) or acetone–saline extraction (28). Allergen-dependent paw swelling is confirmed by measuring changes in hind paw width using digital calipers (31).

In the passive systemic anaphylaxis (PSA) model, sensitized animals are injected with anti-allergen IgE monoclonal antibodies at the tail vein, followed by administration of the antigen via the tail vein after 24 h, which causes a decline in body temperature and an increase in symptom scores with respect to antigen-specific IgE-dependent PSA reactions (32). In our previous study, mice were intravenously sensitized with anti-shrimp tropomyosin monoclonal IgE. After 24 h, the mice were challenged intravenously with shrimp tropomyosin. After 30 min, rectal temperature decreased by 1.6℃ (33). The degree of PSA was determined by measuring histamine, leukotrienes, and prostaglandins levels in the blood, which were collected through cardiac puncture 5 min after allergen challenge (34).

The ASA is commonly used to evaluate drug safety in guinea pigs. The drug is administered via multi-point injection into the muscles three times every alternate day. On days 14 and 21, an intravenous drug challenge is administered to evaluate the degree of systemic reactions (35). Additionally, ASA is elicited by injecting a protein antigen into the tail vein of mice immunized with the antigen and adjuvant for to 2–4 weeks. Similar systemic symptoms, including shock, develops during ASA therapy. Antigens and adjuvants induce IgG1/2 and IgE in ASA mice. IgG-class antibodies recognize and bind to Fc gamma receptors (FcγRs) that are expressed on mast cells, neutrophils, monocytes and macrophages in humans (36). In mouse models, antigens induce basophils (37), macrophages, and neutrophils (38) to release PAF by activating FcγRs, accounting for IgG-induced anaphylaxis. A higher mortality rate is observed during ASA testing than during PSA testing. Moreover, clinical surrogate measurements such as rectal temperature measurements and behavior scales were used similar to that used for PCA. The central temperature variation (ΔT) observed during the 30 min following the challenge was classified according to the standard described by Jonsson F et al. (39) as Grade 1: no shock, 1°C > ΔT ≥ −1°C, no mortality; Grade 2: mild shock, −1°C > ΔT ≥ −4°C, no mortality; Grade 3: severe shock, −4°C > ΔT, and/or mortality.

Skin prick tests and serum allergen-specific IgE levels are routinely used for clinical diagnosis of food allergies. Alternatively, basophil activation test is performed, where flow cytometry is used to measure the expression of both CD63 and CD203c on the surface of basophils following stimulation with food allergens (40). In animal models, the sensitization routes may be intragastric, epicutaneous, subcutaneous, intraperitoneal, or inhalational. However, oral sensitization is the primary route in food allergy models as it is similar to the natural oral route in humans. As an adjuvant, cholera toxin disrupts the intestinal barrier and promotes mucosal immune reactions. The combination of whey protein (15–20 mg) or β-lactoglobulin (1 mg) (to represent cow milk allergen) and 10 μg of cholera toxin is often administered for 5 weeks for sensitization to generate a mouse model of cow milk allergy. Notably, a whey-free diet was provided to the mice for at least 1 week prior to initiating the sensitization procedure (41, 42). Shrimp could trigger severe food allergies with tropomyosin as the major cross-reactive allergen to other shellfish. The quantity of shrimp tropomyosin required for sensitization varies from 30 to 200 μg (43, 44). Peanut is another common food allergen. In addition to using peanut allergen with cholera toxin for sensitization, Joseph J. Dolence developed a mouse model for inhalation-based peanut allergen sensitization. A large number of follicular helper T cells have been detected in the mediastinal lymph nodes of allergen-sensitized mice, and the IL-1 pathway is involved in the Tfh response to peanut allergen (45). Although OVA-specific IgE levels were elevated in mice that were orally immunized with OVA–cholera toxin combination, high serum IL-4 levels, increased mast cell counts in the jejunum, and anaphylaxis were not detected following oral challenge. Upon epicutaneous sensitization with OVA, mice exhibited IgE-mediated mast cell expansion, intestinal allergy, and anaphylaxis, as evidenced by decreased core temperature and increased serum mMCP-1 levels (46). Keiko Kameda developed a food allergy mouse model using 1 mg of chicken egg ovomucoid or cow milk casein in epicutaneous sensitization process without adjuvants (47), indicating that the skin may be an efficient route of sensitization to food antigens. Thus, the protocol design for a mouse food allergy model including the allergen, adjuvant, route, dose, timing of sensitization and challenge exposures, endpoint selected, and outcome of the study is described in Table 1.

Human allergic asthma and rhinitis are triggered by aeroallergens and foodborne allergens. Most animal models of asthma and rhinitis are based on initial sensitization to antigens, followed by a local challenge. None of the currently used animal models emulate all the characteristics of allergic airway diseases. Guinea pigs are commonly used to induce immediate mast cell-dependent hypersensitivity reactions. Bronchial circulation, number of mast cells and mucus glands, and airway neural control in the lungs of guinea pig lungs are significantly similar to those described in humans. Notably, the immediate hypersensitivity reaction in the lung mimics that observed in humans and involves the activation of H1 and cysLT1 receptors (53). However, large doses of antigens induce severe immediate hypersensitivity and favor IgG production, whereas low doses favor mixed IgE and IgG production (54). Moreover, few protocols have been suggested for airway remodeling in guinea pigs. In rats, strains such as Wistar, SD, Fisher, and Lewis do not always develop an allergic response to IgE production, whereas Brown Norway (BN) rats have a high IgE response (55). Airway smooth muscle contraction elicited by allergens in rats is relatively weak and primarily mediated by serotonin, which differs between guinea pigs and humans (54).

BALB/c mice are commonly used to generate mouse models to evaluate Type 2 eosinophilic inflammation or airway remodeling. T2 asthma responds to steroid treatment. Briefly, 10–100 μg of OVA is used to elicit systemic sensitization through intraperitoneal administration for 2–3 weeks, primarily in combination with aluminum hydroxide. In the challenge phase, the animal is exposed to 1%–5% aerosolized OVA inhalation or OVA intranasal instillation (3–7 days for acute asthma or 4–6 weeks for airway remodeling). Large doses of OVA (500 μg) (56) or long-term aerosolized OVA inhalation challenge (57, 58) induce more severe subepithelial fibrosis, whereas low dose (5 μg) challenges favor bronchial epithelial thickening and mucus accumulation in the epithelium (56). Similarly, in our recent studies, shrimp tropomyosin was used as an alternative to OVA, and it elicited high IgE production, AHR, eosinophilic inflammation, Th2 response, basophil and M2 macrophage activation, and airway remodeling in mice (23, 59, 60). Complete Freund's adjuvant (CFA) favors IgG2a activation, indicating a Th1 type immune response (61). OVA/CFA-sensitized mice were developed to induce a non-T2 asthma phenotype, which is characterized by airway neutrophil-dominant inflammation and glucocorticoid insensitivity (62, 63). However, most foodborne allergens, including OVA and shrimp tropomyosin, do not induce airway inflammation or AHR in humans.

Worldwide, Dermatophagoides pteronyssinus (Der p) and Dermatophagoides farina (Der f) are the most common HDM allergens inhaled by humans. Blomia tropicalis is a mite species prevalent in both tropical and subtropical regions (64). Unlike OVA, HDM sensitization and challenge were achieved in mice without the use of adjuvants. Murine eosinophilic asthma is induced by intranasal exposure to Blomia tropicalis allergen without an adjuvant (65). Three different asthma phenotypes, namely, eosinophilic, mixed, and neutrophilic were induced in mice via different doses and routes of HDM (a mixture of Der p and Der f allergen extracts) sensitization and challenge. Mice were sensitized with three intraperitoneal injections of low-dose HDM extract and aluminum hydroxide and challenged with low-dose HDM for eosinophilic airway inflammation. To induce the mixed or neutrophilic phenotype, mice were intranasally sensitized and challenged with low or high doses of HDM, respectively (66). Wang et al. compared asthma induction in mice treated with Der p or Der f extracts. When compared with the Der f-challenged group, Der p-exposed mice exhibited higher BALF neutrophil counts and increased levels of IL-17A and MCPT-1 in lung homogenates, whereas the mice challenged with Der f exhibited a higher expression of MMP-9 and MMP-12 (67). Der p1 and Der f1 are the major allergen proteins in HDM. Recombinant purified Der p1 or Der f1 has been used to develop a capture IgE-ELISA to detect allergen-specific IgE responses in mite-allergic subjects (68). In addition, Der p1 or Der f1 challenged mice exhibited a similar phenotype of airway eosinophilic inflammation, as observed in mice induced with HDM. These recombinant proteins are often used to investigate the molecular mechanisms underlying allergen sensitization. CD163 is a Der p1-binding protein. Der p1 increases CCL24 secretion from bone marrow-derived macrophages in Cd163^−/− mice when compared to WT mice (69). Der p 2 drives airway Th2 inflammation via TLR4/Der p 2 interaction. Der p 2 and low-dose lipopolysaccharide (LPS) promote airway inflammation in WT mice, but not in TLR4^−/− mice (70).

LPS is a known risk factor for asthma exacerbation in humans. A combination of allergens and LPS was used to generate a severe asthma model. However, successfully establishing a severe asthma phenotype considerably depends on the allergen (OVA or HDM), dose of LPS (100 ng, 1 μg, or 10 μg), the starting time of LPS administration (in the sensitization or challenge phase), and LPS administration duration. A single dose of LPS (100 ng) in HDM-treated mice induces the accumulation of NET-releasing CXCR4^hi neutrophils in the lungs and promotes HDM-induced type 2 allergic airway inflammation (71). In OVA-induced asthma mice, LPS (1 μg) was intraperitonially injected before each OVA/alum sensitization. LPS treatment significantly alleviates Th2 allergic airway inflammation after OVA challenge (72). In our recent experiment, LPS (250 ng) was intranasally instilled to mice in the shrimp tropomyosin/alum sensitization phase, which reduced the eosinophil count in the BALF. A modified LPS/shrimp tropomyosin-induced severe asthma model was used in this study. Mice were intraperitonially injected with 20 μg of shrimp tropomyosin mixed with aluminum hydroxide on days 0, 7, 14, and 21. The combination of shrimp tropomyosin (20 μg) and LPS (1 μg) was intranasally instilled to mice on days 7, 14, and 21. In the challenge phase, the mice were intratracheally instilled with 50 μg of shrimp tropomyosin once on day 28 and exposed to an intranasal challenge with shrimp tropomyosin from days 29–31. Shrimp tropomyosin combined with LPS induced significant increase in eosinophil and neutrophil counts in the BALF, excessive mucus secretion, increase in IL-17A and IL-1β levels, and decrease in E-cadherin level, which are partially improved by dexamethasone treatment (unpublished data). Additionally, LPS has been used in the OVA challenge phase for determining the neutrophilic phenotype, which is characterized by increased neutrophil count in the BALF (>30% of the total cell count) (73, 74). However, LPS/allergen-induced peribronchial inflammation has often been complicated by the development of alveolitis (75), which differ from the airway lesions in asthmatic humans. Protocol design for mouse asthma model including allergen, adjuvant, route, dose, timing of sensitization and challenge exposures, endpoint selected, and outcomes of the study are described in Table 2. The experimental protocol, including OVA-induced eosinophilic asthma model (76) and airway remodeling model (57), OVA/LPS (73), or HDM-induced mix phenotype model (83), and OVA/CFA-induced neutrophilic asthma model (63), is shown in Figure 1.