The aim of the present study was to determine the effect of gastric and duodenal digestion on the biochemical, immunochemical and allergenic characteristics of blg protein, one of the principal allergens responsible for IgE-mediated CMA. Many allergens, but not all, are stable to the extremes of pH and proteases encountered in the mammalian stomach and small intestine [19, 22, 39–41] and this has led to much debate about the importance of stability to determining the allergenic potential of food proteins. Many of the studies carried out on stability of allergens to digestion have used conditions, such as substrate: protease ratios which are far from those found in vivo in the gastro-duodenal environment or do not take into account the effect of pre-processing by mastication and swallowing. Another factor which has been neglected is the effect that physiological surfactants, such as gastric PC and duodenal bile salts, may have on stability of proteins to digestion. We used an in vitro system which takes such factors into account and attempts to model the in vivo human process more closely.
As described before, blg was found to be almost completely resistant to breakdown in the gastric compartment, reflecting the resistance of this protein to pepsinolysis [24, 42]. However, the presence of the physiological surfactant PC in the gastric phase of digestion was found to protect the blg from breakdown in the duodenal environment, in marked contrast to the susceptibility of the protein to trypsin and chymotrypsin in absence of PC. Other studies performed with simulated gastric or intestinal fluid have also shown that blg was almost completely digested by pancreatic enzymes [24, 42]. These data indicate that blg is resistant to both gastric and duodenal digestion when physiologically relevant levels of PC are included in the digestion mix. The mechanism underlying the protective effect of PC on blg breakdown in the duodenal environment is not clear but may relate to the ability of blg to bind lipids, ligand binding stabilising proteins  or interactions of a subtle nature with lipids as have been found to play a role in the protective effect of PC on the digestion of another cow's milk allergen, α-lactalbumin . These results indicate the limitations of using the results of the published protocols for resistance to pepsin digestion as indications of potential allergenic activity and emphasise the importance of including physiologically relevant surfactants in the digestion mix for such studies.
Assessing the digestibility of a protein is not by itself a sufficient measure of the allergenic potential of a protein and hence we also investigated the effects of digestion on various measures of the allergenic activity of blg and compared the properties of digested and undigested extracts as elicitors of allergic reactions. Indeed, there are some intriguing indications that the inherent stability of proteins and the characteristics of peptide antigens displayed to antigen presenting cells have an impact in various and sometimes contradictory ways on the type of induced immune responses which has clinical indications [43–46]. Due to the diversity of immune mechanisms leading to food allergy, a variety of procedures for predicting allergenic reactivity of food antigens has been developed.
Using human CMA sera the residual IgE binding capacity of blg after digestion was found to be entirely consistent with the levels of residual native protein found in the digesta. Thus, IgE binding was only reduced following gastro-duodenal digestion in the absence of PC, indicating that under these conditions digestion either modified the IgE epitopes on blg in such a way that they were recognised a much lower affinity than the intact protein, or that digestion has destroyed many of the epitopes.
In contrast, even though gastro-duodenal digestion of blg in the absence of PC resulted in a dramatic loss of intact native protein and IgE reactivity as judged by EAST inhibition there was still sufficient intact blg and/or fragments to activate IgE sensitized basophils from blg allergic patients. Similarly both gastric and gastro-duodenal digestion, in the presence or absence of gastric PC, did not abolish the ability of blg to induce proliferation of lymphocytes from CMA individuals. These findings suggest that following gastric and duodenal digestion blg maintains its IgE and T-cell reactivity. In this study the presence of PC was found to enhance T-cell reactivity in both digestion phases and IgE reactivity only after phase 1 digestion.
The ability of blg digesta to activate basophils and stimulate lymphocyte proliferation (Table 1, Figure 3) was not significantly correlated with each other or the concentration of residual blg in the digesta. However, this analysis did demonstrate that overall digestion enhanced activation of sensitised basophils and proliferation of sensitized lymphocytes without reaching statistical significance. Blg showed reduced skin reactivity following gastric digestion, although the PC vesicles in the digestion mix increased the potency of undigested blg. PC alone did not cause skin reaction in control non CM allergic individuals. The effect of PC was abolished by the addition of bile salts in the gastro-duodenal mix. These observations may result from the PC vesicles presenting the blg in a more potent form in skin testing after phase 1 digestion, which decreased after phase 1+2. Why there is a divergence between the SPT results and the Basotest is not clear. Perhaps SPT is more complex and sensitive.
In comparing the effect of blg digesta on SPT reactivity in relation to the blg concentration, a correlation was found between SPT reactivity and residual intact blg concentration in all digesta (Figure 4) apart from the gastro-duodenal digesta prepared in the absence of PC where intact blg concentrations were very low but much of the SPT reactivity was retained. Also, in phase 1 it was significantly reduced. Interestingly, no such effect of the presence of PC in phase 1+2 was noted. These findings indicate that the presence of PC inhibits blg digestion resulting in increased allergenic reactivity of blg as assessed by skin prick reactivity only. Titration of the various digesta by serial dilutions showed that they were highly allergenic, even at very low concentrations in the SPT analysis. These data are not entirely consistent; this is not in agreement with the in vitro cellular assays, where only a small effect of the PC effect was less clearly found. Perhaps the skin prick test SPT is a more sensitive means of assessing allergenic reactivity, while the in vitro techniques may require further refinement and validation. Furthermore, despite the dramatically reduced blg concentration in phase 1+2 the allergenic reactivity as assessed by basotest, PCNA and SPT was not affected. Of course, considering the small size of our population we cannot exclude the possibility that larger population might have given us the possibility to detect any small differences that perhaps are now overlooked. However, our purpose here was to detect any big differences that could affect the clinical praxis, and which could be shown through our pilot study population. No gross differences were seen, arguing that our results can be translated in the clinical routine.
As assessed by the methods employed in this study in vitro digestion has a heterogeneous effect on the ability of blg to elicit reactions in vivo and in vitro, one of the major milk proteins: it may leave it unaltered, reduce or enhance it. It also suggests that the residual blg surviving simulated gastrointestinal digestion still contains sufficient immunologically active structures (T-cell and B-cell epitopes) to potentially either sensitise an individual or elicit an allergic reaction, and possibly sensitise. Allergenic activity might be attributed to inactivation or destruction of epitope structures, formation of new epitopes, or improved access of previously hidden epitopes. This also highlights that the nature, structure and biological activity of peptides may be completely different when these are produced from a purified protein in a buffer solution (e.g. in a pepsin resistance test) to those produced in a complex biological medium, where interactions with other constituents may occur.