A diet containing immunogenic proteins is extremely important for the maintenance of the immune system homeostasis [19]. But what makes a dietary protein a food allergen? This has not yet been fully established. The knowledge about the factors that determine the allergenicity of food proteins is still incomplete. In theory any protein contained in foods could induce allergy, however, as stated before 14 are responsible for over 90% of all allergies.

In order to perform in vivo or in vitro diagnostic tests, standardized antigens are fundamental. Early characterization of peanut allergen utilized serum from allergic patients to prepare affinity columns. The study of the eluted proteins permitted the molecular characterization of the major allergens [20]. Since the 1980’s several molecular biology strategies were developed and are currently being used to produce purified proteins both from food in naturaas well as recombinant proteins. Purified antigens in diagnostic allergy tests are not necessarily more effective than complex protein solutions [21]. Albeit all efforts there are still inconsistencies between manufacturers of food allergens which may turn the tests unreliable in the clinical setting [22].

We prepared peanut extracts using four different buffers ranging in pH from 6.8 to 10.0. Quantitively, the highest yield was obtained with the alkaline borate buffer with the addition of β-mercaptoethanol, a disulfide bond reducing agent. Regarding the other three buffers we observed that neutral SB extracted more peanut protein than both alkaline BB and acid TB. We did not find in the literature explanation for this fact.

The Brazilian Table of Food Composition (TACO) research group obtained each food from the five Brazilian macro regions, then analyzed its micro and macro elements [23]. The food table refers to the average value of each component contained in the respective foods. It is important to note that the analytical method used to quantify may overestimate or underestimate the protein content of the foodstuff. Protein quantification in the TACO study was performed by the Kjeldahl method that determines the total nitrogen contents which was then converted to protein concentration [24-26]. Our results using the Lowry [11] protein quantification method agree with the TACO results [23]. We can argue that the use of β-mercaptoethanol may have better exposed reactive sites of the protein content than the other three buffers.

Having obtained the protein extracts we determined the immunogenicity of each extract by performing a routine sensitization protocol with alum, an adjuvant typically used to induce TH2 responses [27] in contrast to Freud adjuvant that is typically used to stimulate TH1 responses [21]. The two mouse strains used - C57BL/6 and Balb/c - were chosen due to their Th1 and Th2 profiles, respectively[28]. To study the intrinsic immunogenicity of allergens and the influence of the food matrix, van Wijk, Nierkens (29) compared the in vivo(subcutaneous) response to purified peanut allergens and to CPE similar to the extracts used in this work. These authors inform that none of the purified allergens induced significant immune activation while, CPE induced an increased immune response to the individual allergens. In accordance with the authors, one must argue that purified proteins injected in the subcutaneous tissue without an inducer of inflammation may be immunogenic in the long run but, most probably, not after a single inoculation. For example, as seem in the clinical scenario, with the use of bovine or porcine insulin by diabetic patients that frequently develop resistance to the daily injected insulin [30].

Although defatting the milled seeds is a common procedure in protein purification, we did not include this step prior to the extraction procedures, taking in account that the ingested food is, in general, not in a defatted condition. However, during the extraction process (incubation and centrifugation) most of the fat was removed as it comes out of phase. The amount of reminiscent fat may have influenced the immunogenicity of each extract. A frequently used immunopotentiator is incomplete Freund’s adjuvant, a mineral oil, in which antigen solutions are emulsified [31]. The possible influence of the different buffers on the separation of the fat contained in the seeds was not calculated.

Both cooked and raw food undergo harsh physical and chemical reactions before it can be absorbed by the gut. The majority must be broken down to the elemental components for absorption, however, to continue immunogenic and induce allergy some of the molecules must resist processing. Only to site a few, potential influencers of allergenicity are: food matrix [29], protein stability during processing [32], and the immunological status of the entry port as has been previously demonstrated by our group [33].

In general, our data agree with those of the literature [34, 35] regarding the ionic strength of the buffer used in the extraction of peanut proteins. However, different from these authors the use of the reducing agent increased the extraction efficiency in our model. The difference between the two protocols consists in the buffer and the reducing agents used. We used an acid while they used an alkaline tris buffer and we used βME and they used SDS and other reducing agents. As proposed by these authors the CPE is immunogenic as it is a complex mixture of proteins, carbohydrates, and fatty acids. The latter be responsible for the adjuvant effect due to their inflammatory potential. [29].

As can be seen in the gel electrophoresis the profile of protein bands is influenced by the extraction buffer used. Although similar molecular weight proteins are extracted by each of the extraction buffers, none of them successfully extracted all proteins from peanuts. In our experiment, all protein extracts are immunogenic. All animals produced significantly more antibodies to the Ag-alum solution than their sham counterparts. Irrespective of which protein extract was inoculated or adsorbed to the Elisa plate positive reactivity was observed. However, the intensity of serum reactivity varied. C57BL/6 mice sensitized with one of the four peanut extracts did not show significant differences when the plates were adsorbed with CPE- BB2ßME or CPE-TB/HCl. These results suggest a dominance in the immunogenicity of the protein profile that was obtained by these two buffers. On the other hand, the same sera show distinct reactivity profiles when the plates are coated with CPE-BB or CPE-SB. Our hypothesis is that the immunogenicity of each CPE differs and induces distinct antibody profiles. This hypothesis has not yet been tested. We are currently performing further experiments to elucidate this fact.

Each CPE induced different intragroup variability. C57BL/6 mice sensitized with CPE-SB presented more homogeneous reactivities when compared to those sensitized with CPE-TB/HCl which presented quite heterogeneous reactivities. Some individuals with high levels and others with low levels of anti-peanut antibodies. This pattern of response was less prevalent in Balb/c mice, being visualized only in animals inoculated with CPE-2βME and tested with plates adsorbed with the homologous extract. In populations with a heterogeneous genetic profile, one would expect a pattern of intragroup discrepancies however the experiments were undertaken with inbred mouse strains. We found no similar findings in the literature and our working hypothesis is that the animals’ microbiota may be influencing their responses, since the animals used were neither SPF nor germ-free.

The mean reactivity of sera from the C57BL/6 and Balb/c mice sensitized with one of the four peanut extracts varied according to the extract used for adsorption on the plates. The mean reactivity titers of sera plated on CPE-2βME were lower than the remaining plates suggesting that either its adhesion is less effective or that 2βME decreases the immunogenic capacity of this extract as it dissociates disulfide bonds. The latter is supported by the fact that immunizations with this extract also yields lower antibody titers on plates covered with the extracts obtained with other buffers (CPE-BB, CPE-SB and CPE-TB / HCl).

The immunoblotting revealed that the pattern of sera reactivity to the corresponding extract used to sensitize also varies according to the CPE used. Astuti et al found that peanut allergen induced individual specificity patterns, suggesting that it is better to use CPE rather than purified proteins as a tool for diagnosing allergies, since immunodominance may vary due to food processing, absorption and physiology therefore, each individual may be allergic to different allergens of any food [36].

We did not observe differences in the comparison of the mean reactivities of sera from C57BL/6 mice sensitized with one of the four peanut extracts and tested on an ELISA plate adsorbed with the homologous extract. In this lineage although the average reactivity for the peanut was not significantly affected it is possible to observe the great intragroup dispersion with some animals responding more than others. This observation is not the result of possible inter-plate variations because all the sera from a same group were evaluated on the same plate suggesting the interference of biological factors in the response to the peanut by C57BL/6 mice.

This scenario is different for the Balb/c mice sensitized with the peanut extracts. The form of extraction clearly alters the reactivity. CPE-TB/HCl showed the lowest titers. In this lineage only those inoculated with CPE-BB2βME showed a large dispersion in anti-peanut polyisotypic IgG titers. Although CPE-TB/HCl yielded the lowest protein concentration, this fact does not justify the low antibody titers since all the animals were inoculated with 100 μg of antigen independent of the extract, suggesting that genetic factors of the lineage associated with the differentiated immunogenicity of the extracts influenced the observed immune response.