Allergenicity in food allergy : influence of food processing and immunomodulation by lactic acid bacteria

Allergic diseases such as allergic rhinitis, allergic asthma, atopic eczema and food allergy have become an increasing health problem world-wide, affecting between 20-30% of the total population. Peanut allergy (prevalence ~1%) is a common and persistent food allergy accounting for severe allergic reactions. Peanuts are often consumed after thermal processing (e.g. boiling, roasting) which can alter the protein structure and change its immunoreactivity and allergenicity. In vitro diagnostic testing, however, is generally performed using the native, unprocessed protein and more knowledge on the effect of processing on allergens is necessary to improve these diagnostic procedures. In addition, rationally designed processing could also lead to reduction of the allergen content in certain products and therefore be an effective food technological approach in allergy management. Another approach in allergy management is the use of immunomodulating foods, such as probiotics. There are indications that probiotics, e.g. specific lactic acid bacteria, could be beneficial for many conditions, including different clinical expressions of allergy. Chapter 1 gives an overview of several aspects of allergy with a focus on food allergy. Firstly the basic mechanism and the involved immune cells are discussed, after which the prevalence of food allergy in the context of the EuroPrevall project is described. Different food allergens are discussed with an emphasis on the allergens from peanut and different methods are described to assess the potential allergenicity of proteins under widely used processing conditions, including heating and the Maillard reaction. Lastly, different methods to prevent or treat allergies are discussed with a special emphasis on immunomodulation by lactic acid bacteria. This introduction chapter is concluded with the research aim and thesis outline. Section 1: Influence of processing on allergenicity of proteins In our first study, described in Chapter 2, Ara h 2/6 was purified from raw peanut and heated in solution (boiling) in the presence or absence of glucose. Ara h 2 and 6 were also purified from roasted peanuts for comparison. Structural changes, the capacity to induce cell proliferation and cytokine production, and IgE-binding and IgE cross-linking capacity were evaluated. Although no effect of processing on T-cell reactivity was observed, heat-induced denaturation reduced the IgE-binding and cross-linking capacity. Interestingly, the soluble fraction of the Ara h 2/6 isolated from roasted peanuts retained the conformation and allergenic activity of the native protein. In Chapter 3 similar methods were used to assess the effect of heating and glycation on Ara h 1. Heating in solution, irrespective of their level of glycation, resulted in formation of aggregates having reduced IgE-binding and cross-linking capacity, while T-cell reactivity was retained. The soluble fraction of Ara h 1 isolated from roasted peanuts appeared to be highly denatured, formed more globular and smaller aggregates, and showed no evidence of glycation. However, these smaller aggregates retained IgE-binding capacity, unlike the aggregates formed after heating and glycating purified Ara h 1. These results could account for observed differences between boiled and roasted peanuts and suggest that other modifications than the Maillard reaction affect the allergenicity of Ara h 1. As peanuts are often consumed after roasting, the wet-thermal processing procedures, employed in the two previous described studies, were related to the effect of thermal treatment and Maillard reaction under low moisture conditions, which is described in Chapter 4. The extensive heating at low moisture resulted in hydrolysis of both Ara h 1 and Ara h 2/6. However, in contrast to Ara h 2/6, soluble Ara h 1 formed large aggregates. Thermally treated Ara h 2/6 had both a lower IgE-binding and degranulation capacity compared to the native form, and the presence of glucose during heating partly counteracted both the decrease in IgE-binding and degranulation capacity. The IgE-binding capacity of Ara h 1 was also decreased; however, the basophil degranulation capacity increased significantly. This demonstrates the importance of including degranulation assays in addition to IgE-binding assays, when assessing allergenic potency of allergens. In addition, we here propose a role for large aggregates in the increased IgE-cross-linking capacity of individual allergens. Chapter 5 describes the effect of glycation on the immunoreactivity and basophil degranulation capacity of Cor a 11, the 7S globulin from hazelnut (and thus a homologue of Ara h 1). Three processing methods (heating at low moisture content at 37, 60 and 145°C) resulted in proteins with increasing degrees of glycation. Glycation at 37°C did not influence the specific IgG or IgE binding, while both were decreased after heating at 60°C and 145°C. However, heating at 145°C in the absence or presence of glucose resulting in the formation of aggregated structures, increased the basophil degranulation capacity of Cor a 11 using sera high in Cor a 11 specific IgE, but not when using sera from peanut allergic patients low in Cor a 11 specific IgE. Therefore, this study besides showing the importance of the use of a combination of tests also indicated the importance of using well-characterized sera as a source of IgE. In Chapter 6 we focused on the clinical features of all our clinically well-defined peanut allergic patients of which immune cells and sera were used for the previously described studies. In addition, soy allergic patients were included and an extensive IgE profile was determined for all patients. Gly m 4 (Bet v 1 homologue from soy) sensitization was suggested to be an important indicator of severe soy allergy in the soy allergic patients, while in peanut allergic patients sensitization to allergens from soy and pea extract nor Gly m 5 and 6 was found to have a good diagnostic specificity. This is likely due to the presence of clinically non-relevant cross-reactivity between peanut-specific IgE and homologues soy and pea components. Section 2: Immunomodulation by Lactobacillus strains In the first in vitro study, described in Chapter 7, initially 51 Lactobacillus strains were screened of which 8 were selected and tested for their immunomodulating effects on peripheral blood mononuclear cells (PBMC) of healthy donors. All tested Lactobacillus strains were capable of inducing the production of IL-1β, IL-10, IFN-γ and TNF-α. Clear strain-specific effects were observed with L. plantarum strains showing significantly higher induction capacity of IFN-γ, IL-12 and TNF-α compared with L. acidophilus strains. We therefore concluded that especially L. plantarum strains are promising candidates in IgE-mediated allergy by their stimulation potential of the T-cell response toward a putative Th1 response. As healthy subjects, in contrast to allergic individuals, are assumed to finely regulate the Th1/Th2 balance by inducing sufficient Treg cell activity, immunomodulatory effects of six selected Lactobacillus strains were investigated on PBMC of pollen-allergic patients in Chapter 8. All strains could modulate PBMC to induce innate cytokine production and in addition, all strains had the ability to repress IL-13 production. Again a differential effect on IFN-γ and IL-12 induction was observed. In addition, one strain could extensively suppress proliferation induced by anti-CD3/anti-CD28 stimulation. Specific strains that were able to suppress the Th2 cytokine induction and induce Th1 cytokines might be beneficial for allergic patients. Effects found in vitro cannot directly be extrapolated to in vivo and therefore, in Chapter 9, we performed an in vivo screening including five Lactobacillus strains. Blood samples were collected before and after a 4-week intervention with probiotics from all 62 birch-pollen-allergic patients included. Four strains caused a decrease in birch-pollen-specific IgE and for one specific strain this coincided with significant decreases in IL-5 and IL-13 and an increase in IL-10 production by anti-CD3/anti-CD28 stimulated PBMC cultures and might therefore have the potential to alleviate seasonal allergy symptoms. The last chapter, Chapter 10, gives an overview of the most important results of this thesis and discusses the research limitations and future research perspectives. We hypothesize the role of protein aggregation in allergenicity and we elaborate on the importance of a proper stepwise approach to realize selection of a proper lactic acid strain for in vivo human testing. In conclusion, this thesis showed that processing effects can have profound and specific effects on the structure and the allergenicity of relevant allergens. However, to test putative effects on allergenicity, IgE-binding tests only are not sufficient and mediator release assays are important to include, particularly when testing aggregated proteins. These results might have consequences for the proper diagnosis of food allergy in daily practice. Finally, as effects of lactic acid bacteria are strain specific, a proper pre-selection of candidate strains is important to choose the most promising strains for clinical testing. In our in vivo screening, one strain, L. plantarum CBS125632, was found to be promising because of its desired immunomodulatory activity to test in a follow-up trial to reduce symptoms of birch-pollen allergy.