Grantee Research Project Results

The objective of the research project was to identify the role of the epithelial cytokine thymic stromal lymphopoietin (TSLP) in the development of tolerance or allergy to foods. Epithelial cells of the gastrointestinal tract, lung and skin express TSLP. Our experimental approach was to use mice genetically deficient in the receptor for TSLP, as well as wild-type controls, to determine if TSLP was required for the development of immune tolerance or gastrointestinal or systemic anaphylaxis to food. Our next goal was to determine the mechanism by which TSLP could exert its effects on the mucosal immune system. These experiments were performed using purified T cells and dendritic cells that were either wild-type or genetically deficient for the TSLP receptor.

 

A modified objective of our research project was to determine the role of TSLP at extra-intestinal sites in the development of allergic sensitization to foods. Specifically, we examined the role of TSLP in sensitization to food allergens through cutaneous exposure, a route of exposure that is thought to be highly relevant to development of allergic sensitization in humans.

In aim 1 of our studies, we examined the necessity of the TSLP receptor in the development of oral tolerance or allergy to a model food allergen, ovalbumin (OVA). Using a standard oral tolerance procedure (low-dose OVA feeding for 5 days followed by systemic immunization), we observed that mice lacking the TSLP receptor had no deficiency in the development of immune tolerance to food allergens delivered through the gastrointestinal tract. We therefore concluded that TSLP is not required for oral tolerance. We then used two models of food allergy. One model of gastrointestinal manifestations of food allergy is generated by systemic immunization with OVA, followed by repeated oral challenge to induce an acute diarrheal response associated with intestinal inflammation. We observed that gastrointestinal manifestations of food allergy were dependent on TSLP receptor, as quantifiable measures of symptoms and inflammation were significantly reduced in TSLP receptor-deficient mice compared to wild-type mice. We therefore concluded that TSLP is required for gastrointestinal manifestations of food allergy. We next tested a model of systemic anaphylaxis in response to OVA challenge. Mice were orally sensitized by feeding OVA together with the mucosal adjuvant cholera toxin (CT). Mice lacking the TSLP receptor were able to generate an allergen-specific IgE response and a Th2-skewed cytokine response as well as respond to allergen challenge with anaphylaxis to a similar degree as wild-type mice. Therefore TSLP is not required for primary allergic sensitization via the gastrointestinal tract.

 

In aim 2, we determine the mechanism by which TSLP played a role in gastrointestinal manifestations of food allergy. We found that transferring mesenteric lymph node cells from wild-type mice with allergic diarrhea to TSLP receptor knockout mice could restore disease in these mice, including development of allergen-specific IgE, Th2-skewed inflammation in the intestine, mast cell counts in the intestine, and most importantly, symptoms. When we co-cultured dendritic cells and T cells together with exogenous TSLP, we found that the main target of TSLP was the T cell. From these results we conclude that TSLP plays a role in the development of gastrointestinal manifestations of food allergy by amplifying a local Th2 response in the intestine. The results of Aim 1 and Aim 2 have been peer-reviewed and published, and the manuscript (Blazquez et al, Gastroenterology, 2010) is appended to this report.

 

 

 

 

 

 

 

 

 

 

As a related aim 3, we examined the impact of TSLP on the development of allergic sensitization through the skin. Cutaneous exposure to food allergens, such as peanut, has been proposed to be a significant exposure route leading to sensitization in human disease. To our surprise, we found that the allergen OVA did not require any exogenous adjuvant to sensitize mice when it was applied topically to the skin in the absence of any damage to the skin (such as through tape-stripping) or with any need for occlusive bandages. Topical exposure once a week for a period of 6 weeks was sufficient to generate a robust IgE response and anaphylaxis when the mice were orally or systemically rechallenged. This sensitization occurred in mice lacking the TSLP receptor and also in mice lacking a functional TLR4 receptor, indicating that sensitization was not dependent on endotoxin contamination of the allergen. To determine if the skin could function as a sensor of allergenicity, we compared a panel of strong and weak allergens. We found that the purified milk allergen α-lactalbumin was unable to sensitize mice without adjuvant. To compare crude allergen extracts, we compared peanut to soy and found that mice could be sensitized to peanut without adjuvant, but could not be sensitized to soy through the skin. Comparing purified allergens of peanut, the stronger allergen Ara h 2 could sensitize mice without adjuvant, whereas the weaker allergen Ara h 1 was a poor inducer of sensitization and anaphylaxis. These results indicate that the skin may be an important sensor of allergenicity. Exposure through the skin is not sufficient to predict allergenicity. The use of a crude rice extract resulted in high levels of allergic sensitization and anaphylaxis in mice, despite the fact that rice is not a significant food allergen in humans. We speculate that the susceptibility of rice protein to digestion may prevent its allergenicity in humans. We propose that a combination of skin exposure to assess potential allergenicity, together with digestion assays to determine resistance of food allergens to proteolytic digestion, be tested to determine their predictive value in determining allergenicity of novel food proteins. The results of Aim 3 were presented at the 2012 meeting of the American Academy of Allergy, Asthma, and Immunology (poster appended to this report).

 

 

 

1. The in vivo model systems used in the first two aims of this project are well established. The models of gastrointestinal manifestations of food allergy, food-induced anaphylaxis and oral tolerance were all performed with the antigen ovalbumin, which provided reproducible results between experimental replicates. All studies were performed in at least two independent experiments, which provided sufficient power to observe significant differences between groups. Outcome measures were (1) onset of diarrhea symptoms; (2) anaphylaxis as measured by drop in body temperature; and (3) suppression of ear swelling induced by antigen injection in the ear. These outcome measures were reliable and simple to measure or observe without the need for expensive specialized equipment beyond a rectal thermometer (obtained from World Precision Instruments) and an ear caliper (also from World Precision Instruments). Secondary outcomes included measurement of allergen-induced cytokine secretion from spleen on lymph node cultures. These were performed using commercial ELISA kits (all from eBioscience), and were without technical difficulties. Allergen-specific IgE was measured using a monoclonal anti-IgE capture antibody and DIG-labeled OVA to detect OVA-specific IgE antibodies. We have found this to be a more sensitive method for detecting allergen-specific IgE in the absence of competition from allergen-specific IgG. We measured cytokine expression in the gastrointestinal tissues by real-time PCR. Isolation of RNA from intestinal tissues was an initial technical challenge that we overcame by combining a traditional approach of homogenization in Trizol (Life Technologies) together with a RNA cleanup step using RNeasy mini-kits (Qiagen). To assess the impact of TSLP on antigen presentation, we performed conventional antigen presentation assays using transgenic OVA-specific T cells isolated from DO11.10 mice. These are well-established assays used in the literature and provide reliable data on T cell priming.

 

2. The model used for epicutaneous exposure was adapted from that described by other investigators (Strid 2005; Birmingham 2007; Oyoshi 2010). We used a depilatory cream (Veet^TM) to remove the abdominal hair of mice. No further preparation of the skin was performed (no tape stripping). Antigen was applied topically while mice were anesthetized, without occlusive bandages. This approach was technically simple and led to reliable sensitization to a distinct subset of antigens (ovalbumin, peanut proteins) but not others (milk allergens, soy). Sensitization was tested using serum antibody levels and functionally by induction of anaphylaxis. These experiments, together with the published work of Gangur and colleagues (Birmingham 2007; Gonipeta 2009; Parvataneni 2009; Gonipeta 2010), support the hypothesis that the skin functions as a sensor of allergenicity and may be an optimal system for testing food allergenicity. Future studies will determine if we can use an in vitro system modeling the skin response to food allergens, which would make this a more cost-effective method for testing allergenicity. The use of in vivo models makes this a relatively expensive method for testing allergenicity, but provides an exciting proof of principal showing that within the skin there is a mechanism of recognition of food allergens with relevance to human health.