Influence of peanut matrix on stability of allergens in gastric‐simulated digesta: 2S albumins are main contributors to the IgE reactivity of short digestion‐resistant peptides

Most food allergens sensitizing via the gastrointestinal tract are stable proteins that are resistant to pepsin digestion, in particular major peanut allergens, Ara h 2 and Ara h 6. Survival of their large fragments is essential for sensitizing capacity. However, the immunoreactive proteins/peptides to which the immune system of the gastrointestinal tract is exposed during digestion of peanut proteins are unknown. Particularly, the IgE reactivity of short digestion‐resistant peptides (SDRPs; <10 kDa) released by gastric digestion under standardized and physiologically relevant in vitro conditions has not been investigated.


| INTRODUCTION
Peanut (Arachis hypogaea) is 1 of the 8 major allergenic foods, called the Big-8, which are mandatorily declared for any processed food products according to US and EU regulations. In addition, a low eliciting dose and high frequency of fatal reactions make peanut 1 of the most potent allergenic foods. 1 The major allergens in peanut are Ara h 1, Ara h 2, Ara h 3 and Ara h 6, based on frequent IgE-binding studies from patient sera. 2 However, based on the potency of the allergic effector activity of the allergens and loss of potency upon their removal, Ara h 2 and Ara h 6 are regarded as the most potent peanut allergens. 3 Food ingestion is the major route of sensitization to food allergens. A great number of allergenic proteins are remarkably resistant to proteolysis in the gastrointestinal tract, and survival of their large fragments is essential for their sensitizing capacity. 4 There is no strict relationship between digestibility and allergenicity, and several major allergens are extremely digestion-labile proteins. However, evaluation of resistance to digestion with pepsin by in vitro simulated digestion remains a central part of allergenicity safety assessment of novel proteins. 4 In vitro methods simulating digestion processes are widely used tools because of their simplicity, low cost, reproducibility and ethical acceptability. 5 Several studies have investigated pepsin digestibility of peanut allergens by in vitro simulated digestion (Table S1), and regardless of the experimental conditions, Ara h 2 and Ara 6 showed higher pepsin resistance compared to Ara h 1 and Ara h 3. 6,7 In most studies, purified peanut allergens or peanut extracts were digested. Only 2 studies have dealt with peanut protein digestibility within their real food matrix. Plundrich et al 8  investigated digestibility of proteins from the whole peanut by complete oral-gastric-intestinal-brush border membrane proteases, but it did not report the gastric digestion products. Therefore, the immunoreactive protein/peptide species to which the immune system of the gastrointestinal tract is exposed during digestion of peanut proteins remains unknown. In particular, the IgE reactivity of short digestion-resistant peptides (SDRPs) released by gastric digestion is yet to be investigated. This fraction also contains important highly immunoreactive peptides. It has been demonstrated that some Ara h 2 peptides of size <3 kDa can cross-link IgE/FceRI complexes and degranulate cells, 10 and that Ara h 1, when digested to small peptide fragments, retains both the sensitizing and the IgE-reacting potential because of the peptide's ability to aggregate. 11 The aim of this study was to comprehensively investigate stability and structures of pepsin-resistant allergens, of their larger fragments and of SDRPs released by pepsin digestion of whole peanut grain under standardized and physiologically relevant gastric conditions. 5 In particular, IgE reactivity of SDRPs released by pepsin digestion was investigated to determine roles of individual peanut allergens digestion products.

| Materials
All details and information about materials and chemicals are available in Methods S1.

| Peanut preparation, simulated oral and gastric in vitro digestion and protein extract isolation
Red skin raw peanuts (Arachis hypogea L.) of the runner variety were obtained from a local grocery. Raw peanuts with skin were milled using a coffee grinder (800 W; Bosh), 3 times, 5 minutes to obtain a particle size <1.5 mm. Ground peanuts were air-dried overnight at room temperature (the amount of evaporated water was 5.4% of total mass).
Oral and gastric in vitro digestions of ground raw peanut were performed according to the previously reported method. 5 The concentrations of electrolytes in stock solutions of simulated salivary fluid (SSF), simulated gastric fluid (SGF) and the final reaction mixtures are presented in Table S2 (more details are available in Supporting Information).
The liquid phase of the digestion mixture (200 lL) was mixed with 200 lL chilled 20% trichloroacetic acid (TCA) in acetone and left overnight at À20°C. After removing the supernatant by centrifugation at 10 000 g for 30 minutes at 4°C, the precipitated proteins were washed 3 times with 1 mL of cold acetone and dried at RT. Protein concentration was determined using BCA assay after re-solubilization of TCA/acetone pellet in 2% SDS. On the other hand, 800 lL of the liquid phase of the digestion was processed without TCA to obtain SDRPs, as described below.     (Table S3). For more details, see Supporting Information.

| ImmunoCAP inhibition
IgE binding of the SDRPs fraction of digested peanut was determined using ImmunoCAP inhibition (ImmunoCAP System; Phadia/ Thermo Fisher Scientific). Seven undiluted individual sera (200 lL; patients #1-7 Table S2) were pre-incubated with 200 lL peptides prior to the measurement for specific IgE on solid surface for the following: peanut (f13), Ara h 1 (f422), Ara h 2 (f423) and Ara h 3 (f424). Applied peptides were released from about 3.3 mg of milled peanut, for example released from about 800 lg of peanut proteins extracted to liquid phase during digestion. The inhibition of IgE binding was expressed as percentage based on non-inhibited serum, using the following formula:

| Immunoblotting
The samples (120 lg for 2D immunoblots on 7 cm IPG strips) were loaded and resolved on 14% gel. Proteins were transferred onto nitrocellulose membranes and incubated overnight at 4°C with 1:10 diluted serum pool from patients sensitized to peanut.
The serum pool consisted of sera of the first 8 peanut-sensitized patients presented in Table S3 (range of total peanut-specific IgE: 11-415 kU A /L; median 109 kU A /L). For more details, see Supporting Information.

| De novo modelling and molecular graphics
The sequences of Ara h 1 and Ara h 3 were obtained from UniProt   Table S4).

| Ara h 1 and Ara h 3 allergens' cascade pattern of pepsin proteolysis
It was possible to draw pepsin proteolytic pattern for Ara h 1 and Ara h 3 due to its high content within peanut proteome, by determining the peptides sequence coverage for each of their isoforms within spots (Table S4, Figure 2A that were the most intense in DPS contained basic peptides from acidic Ara h 3 subunit, due to lack of its acidic residues at C-terminus ( Figure 2B). Beside 2S albumins, in CPS we detected Ara h 8 in spot 20 and Ara h 10 in spot 18 ( Figure 1, Table S4), but not in 2D gel of DPS.

| Ara h 2 and Ara h 6 remained almost intact during pepsin digestion
Intact Ara h 2 and 6 were identified in all the peanut proteome preparations, SPE, CPS and DPS (Figures 1 and 3, Table S4). Mass spectrometry identification of Ara h 2 and Ara h 6 spots was acquired from the CPS and DPS gels as shown in Figure 3. IgE-binding potency of proteins extracted from CPS and DPS were compared in inhibition ELISA test ( Figure 4A) with a pool of sera from ten peanut-sensitized patients (Table S3). As a reference   (Tables S5 and S6).
We searched the IEDB database to match SDRPs released during peanut digestion (Tables S5 and S6) Figure 5A,B). We also detected 2 non-epitope peptides from Ara h 8 (Table S5). Analysing SDRPs after reduction/alkylation and digestion by trypsin, 2 peptides of Ara h 2, both being part of continuous epitopes, were found (  Figure S5). In vitro digestion studies with purified proteins reported dramatically higher protein digestibility than possible under physiological conditions. In addition, it is well known that pure proteins in solution can have different sensitivities to proteolysis compared to proteins adsorbed at oil-in-water interface owing to changes in the protein structure. 18,24 Therefore, the proteins regarded as highly digestible, as estimated by pepsin digestion in their purified form, could be markedly resistant to proteolysis within a complex food matrix. Regions with identified peptides of Ara h 1, Ara h 2 and Ara h 3 found in the short digestion-resistant peptides (SDRPs) of peanut digested by pepsin. A, 3D structure of Ara h 1; non-core N-terminal is in yellow, non-core C-terminal is in green, and the core is in grey. Intact peptides (middle) and peptides found after reduction, alkylation and trypsin digestion of SDRPs of peanut digested by pepsin (right). Regions with peptides matched with peanut continuous epitopes (IEDB database search) are in red, and non-matching peptides are in blue. B, 3D structure of Ara h 3; the basic subunit is in orange, and acidic subunit is in light blue. Intact peptides (middle) and peptides found after reduction, alkylation and trypsin digestion of short digestion-resistant peptides (SDRPs) of peanut digested by pepsin (right). Regions with peptides matched with peanut continuous epitopes (IEDB database search) are in red, and non-matching peptides are in blue. The crystal structures of peanut major allergen Ara h 1 (PDB entry 3SMH), Ara h 2 (PDB entry 3OB4) and Ara h 3 (PDB entry 3C3V). C, 3D structure of Ara h 2; flexible loop is shown in violet colour, C-terminus flanking region is shown in green colour, and N terminus flanking region is shown in yellow colour. Peptides found after reduction, alkylation and trypsin digestion of SDRPs of peanut digested by pepsin are shown in red colour, and they are matched with peanut continuous epitopes (IEDB database search) end of the sequence (S522-A530, Figure 2B), occupying <2% of the C-terminal domain. 25 This explains why the acidic subunits are much more prone to pepsin proteolysis than the more compact basic subunit and also why natural Ara h 3 processing results in higher diversity in the mass of acidic subunits , as compared to the basic subunits (about 23 kDa).

|
Almost completely preserved cores of these 2 allergens explain the similarity in the secondary structures of digested and undigested samples observed by CD spectrometry.
In IgE ELISA inhibition, the same order of magnitude of IC 50 values obtained for control and digested sample suggests that partly digested peanut allergens mainly retained their allergenic potential.
Pepsin proteolysis only slightly reduced the IgE-binding potential of peanut proteins extracted during digestion. These results are in agreement with data shown in 2D immunoblots.
We also examined the IgE reactivity of SDRPs (<10 kDa) from the gastric-simulated digesta by an inhibition study on peanut allergens on ImmunoCAP and demonstrated that 2S albumin SDRPs are more potent than Ara h 1 and Ara h 3 SDRPs. This peptide fraction could be the result of extensive proteolysis of peanut proteins, thus representing the digestion-resistant peptides of the smallest molecular mass, or they could be truncated parts of larger DRPs, excised following proteolytic attack on exposed loops of peanut's proteins.
Although Ara h 1 and Ara h 3 are the most abundant storage proteins in peanuts, patients with peanut allergies recognize Ara h 2 and Ara h 6 more frequently and with greater intensity. 20 IgE binding to Ara h 2 and Ara h 6 is primarily dependent on the discontinuous epitopes, 26,27 and it is likely that CMCEALQQIMENQ peptide that we have identified in SDRP fraction, besides being a part of continuous Ara h 2 and Ara h 6 epitopes, could be a part of a potent discontinuous epitope by creating S-S crosslinking adducts.
Schocker et al 27 investigated the transfer of Ara h 2 in human breast milk and identified 5 tryptic fragments of Ara h 2, of which 2 (ANLRPCEQHLMQK and CMCEALQQIMENQSDR) were same as the peptides that we identified in this study in SDRP fraction of gastric digesta. These peptides could be part of the same discontinuous epitope held together by several disulphide bridges.
The results of our study demonstrated that after gastric digestion of whole peanut matrix, high molecular mass digestion-resistant fragments of Ara h 1, 2, 3 and 6, with mostly retained core structure bearing discontinuous epitopes, survive. On the other hand, large proportion of generated SDRPs are part of continuous epitopes resulting in high allergenic potential of SDRPs. Therefore, it can be expected that in vivo these species become exposed to the intestinal immune system and transported to circulation. Increased intestinal permeability (as found under different physiological and pathological conditions) or disruption of tight junctions may enable transport of peanut protein fragments across the intestinal epithelium. Therefore, in peanut allergic individuals, in both intestine and circulation, they can induce allergic reaction. In intestine, these fragments can induce intestinal anaphylaxis via mast cell-dependent IgE-FceRI-IL-13 pathway 28 and histamine-depended mesenteric lymph node and lamina propria DC accumulation. 29 In conclusion, we demonstrated that the most potent allergens, Ara h 2 and Ara h 6, remained mostly intact to proteolysis by pepsin, and SDRPs originating from Ara h 2 were most potent in inhibiting IgE binding, suggesting that not only intact Ara h 2 but also its SDRPs are of clinical relevance. Ara h 1 N-and C-terminal parts and acidic forms of Ara h 3 were most susceptible to proteolysis by pepsin.
However, Ara h 1 exhibited sequential digestion into a series of DRPs with preserved allergenic capacity. Thus, the major peanut allergens and SDRPs play important roles in allergic reactions to peanut.
Compared to the findings of in vitro digestion studies performed on purified proteins under similar conditions, digestibility of proteins when they are within real food matrix was dramatically lower.