- Top of page
- Materials and methods
Background: Hazelnuts are a common cause of food allergic reactions. Most hazelnut allergic individuals in central and northern Europe are sensitized to Cor a 1, a member of the PR-10 protein family, while the lipid transfer protein Cor a 8 acts as a major allergen in the south of Europe. Other allergens, including profilin and seed storage proteins, may be important in subgroups of patients. Reliable detection of specific IgE in the clinical diagnosis of food allergy requires allergen reagents with a sufficient representation of all relevant allergen components. Some reported observations suggest that natural hazelnut extract may not be fully adequate in this respect.
Methods: The capacity of immobilized natural hazelnut extract to bind Cor a 1-, Cor a 2- and Cor a 8-specific IgE and IgG antibodies was investigated by serum adsorption and extract dilution experiments and by the use of allergen specific rabbit antisera. All measurements were performed with the ImmunoCAP assay platform.
Results: The experimental results revealed an incomplete capacity of immobilized hazelnut extract to capture IgE antibodies directed to the major allergen Cor a 1. Spiking of hazelnut extract with recombinant Cor a 1.04 prior to solid phase coupling gave rise to significantly enhanced IgE antibody binding from Cor a 1 reactive sera. The spiking did not negatively affect the measurement of IgE to extract components other than Cor a 1.
Conclusion: A hazelnut allergen reagent with enhanced IgE detection capacity can be generated by supplementing the natural food extract with recombinant Cor a 1.04.
Hazelnuts are one of the most common causes of food allergic reactions in Europe. Symptoms upon ingestion are most often restricted to the oral cavity but may also be systemic and in rare cases life-threatening or even fatal (1, 2).
Several hazelnut allergens have been identified and characterized (3). Hazelnut allergic individuals from northern and central Europe are almost invariably sensitized to Cor a 1, a 17 kDa member of the PR-10 protein family of allergens related to the major birch pollen allergen Bet v 1 (4–8). In contrast, hazelnut allergic individuals from the Mediterranean region predominantly react to Cor a 8, a 9 kDa nonspecific lipid transfer protein related to peach and cherry allergens Pru p 3 and Pru av 3, respectively, and rarely to Cor a 1 (7). Profilin, designated Cor a 2 in hazelnut (4), appears to be of minor importance in hazelnut allergy regardless of geographical region (7, 9, 10).
Further hazelnut allergens reported include different seed storage proteins: a 2S albumin (7), the glycinin-like protein Cor a 9 (1) and the vicilin-like protein Cor a 11 (11). Reactivity to these allergens would arise independently of pollen sensitization and may be important in subgroups of patients but their clinical importance in wider populations of hazelnut allergic subjects remains to be established. The same is true for oleosin, an oil body associated protein which was recently reported as a hazelnut allergen (12).
Sensitization to Cor a 1 is believed to be caused mainly by cross-reaction to Bet v 1 following birch pollen sensitization. In this type of hazelnut allergy, symptoms upon ingestion are usually mild and restricted to the lips and oral cavity, referred to as the oral allergy syndrome (2, 7, 13–16). Local oral reactions to hazelnut are also common in individuals from the Mediterranean area, where Cor a 8 is a main culprit allergen, but individuals from this region more frequently also experience severe reactions to hazelnut, involving distal organs (7, 17, 18). Possible reasons for this asymmetry in reaction pattern include dose and stability of the allergen to which the patient reacts, affinity of IgE antibody binding to the food allergen, influence of pollen sensitization pattern and dietary habits.
Assessment of sensitization by allergen-specific IgE testing or skin prick testing (SPT) is a primary tool in routine clinical diagnosis of food allergy. To reliably detect sensitization, it is critical that the allergen reagent applied contains an adequate amount of all relevant allergen components. However, preparation of food allergen extracts with high and consistent quality is a difficult task. Several studies have shown great variation in allergen content between different manufacturers’ extracts and, in some cases, severe shortage or lack of important components (19–25) which is likely to affect the diagnostic outcome. For hazelnut, several clinical studies have been reported and given quite different results regarding diagnostic sensitivity of testing with natural hazelnut extract (2, 26–28). While acceptable sensitivity was observed in some studies, a significant lack in sensitivity in other studies prompted us to undertake the work reported here.
In this study, we investigated the capacity of immobilized natural hazelnut extract to bind antibodies specific for Cor a 1, Cor a 2 and Cor a 8, and explored the possibility of enhancing its diagnostic sensitivity by supplementing the extract with recombinant Cor a 1.04.
- Top of page
- Materials and methods
Clinical diagnosis of food allergy relies in part on allergen extracts which allow reliable assessment of sensitization. However, the composition of natural allergen extracts is affected by a multitude of factors that are difficult to control, including differences and variation in abundance of individual allergens in the raw material, extraction efficiency and postextraction stability and susceptibility to protein modification activities. Such factors are known to affect the quality of certain food extracts, particularly those prepared from foods comprising fully hydrated and metabolically active tissues, passing through natural stages of maturation and ripening (30, 31).
In the case of hazelnut, questions have been raised regarding the content of Cor a 1 in commercially available extracts used for SPT (19, 20). Further, only 12 of 31 (39%) Dutch hazelnut allergic patients were reported to test positive with hazelnut ImmunoCAP (28), even if this in part was because of the use of a cut-off level of 0.7 kUA/L. Eighteen of the 31 (58%) patients in that study showed specific IgE levels above the conventional cut-off of 0.35 kUA/L. In contrast, a sensitivity of 75% (n=67) of the same test was reported from a multicentre DBPCFC study on hazelnut allergy (2), despite the fact that the 0.7 kUA/L cut-off level was also applied in this study.
In the work reported here, three approaches were taken to assess the relative antibody binding activity of Cor a 1, Cor a 2 and Cor a 8 in immobilized hazelnut extract: (i) analysis of differential IgE antibody adsorption by immobilized extract; (ii) analysis of allergen-specific rabbit IgG antibody binding to immobilized extract and (iii) monitoring of differential loss of specific IgE binding activities as a result of serial dilution of extract prior to immobilization.
Taken together, the experimental results indicated a scarcity of Cor a 1 and an abundance of Cor a 8 in natural hazelnut extract. Although the reason for the low level of Cor a 1 activity was not addressed in this work, the situation resembles that described for certain fruits, where rapid loss of PR-10 allergen activity upon extraction has been attributed to degrading or modifying enzyme activities or the release of endogenous phenolic compounds (25, 32–34). However, it is also possible that the observed shortage of Cor a 1 activity is in part related to an abundance of major storage proteins, competing with Cor a 1 for solid phase attachment sites. Yet another possible reason may be that Cor a 1 is inefficiently coupled to the solid phase or that its antibody binding function is negatively affected by that procedure. An indication that this may be the case comes from the assessment of allergen content in liquid hazelnut extract which showed a relatively high concentration of Cor a 1 in comparison with Cor a 2 and Cor a 8. Even if these measurements were performed using a comparative IgE inhibition approach rather than a validated quantitative assay, the results suggest that the differential allergen activity of immobilized extract observed is not simply a function of allergen concentration in the extract used.
It should be noted that the concentration estimates made in our study do not necessarily reflect allergen concentrations in the actual food as the solubility of a protein depends on its isoelectric point and hydrophobicity as well as the pH and salt conditions in the extraction buffer. PR-10 proteins, such as Cor a 1, and LTPs, such as Cor a 8, have very different biophysical properties, and factors of this nature may in part explain the unexpected differences in concentration estimates of Cor a 1 and Cor a 8 in liquid extract obtained in this study.
The scarcity of Cor a 1 activity in immobilized hazelnut extract made it of interest to examine whether this deficiency could be rectified by addition of recombinant Cor a 1.04 to the extract prior to solid phase coupling. Even though this would seem a straightforward adjustment to the extract composition, it could not be taken for granted that the recombinant allergen would remain intact and attach efficiently to the solid phase in the presence of competing extract substances or that it would not itself outcompete endogenous allergens to some extent. Evaluation of an initial experiment, where increasing amounts of rCor a 1.04 were simply added to the extract prior to coupling to the solid phase, showed that progressively increased binding of Cor a 1 reactive IgE was achieved, but not to a level comparable with that obtained with rCor a 1.04 alone. To examine whether this might be a result of competition for attachment sites, rCor a 1.04 was added to different dilutions of hazelnut extract before immobilization to the solid phase. Evaluation was then performed using sera that were predominantly reactive to either Cor a 1, Cor a 2 or Cor a 8, as well as sera reactive to hazelnut components other than those three allergens, potentially including storage proteins. Satisfyingly, an almost quantitative binding of Cor a 1 reactive IgE antibodies was achieved at a 25-fold extract dilution and above, while very limited loss of binding of Cor a 2- and Cor a 8-reactive antibodies occurred.
Evaluation of an experimentally enhanced hazelnut reagent, using sera of subjects with diagnosed hazelnut allergy, demonstrated both an improved sensitivity in detection of hazelnut sensitization and a dramatic increase in binding capacity for Cor a 1-reactive IgE antibodies. Among the 50 central European subjects, all eight that tested negative with the unmodified extract tested positive with the enhanced reagent, equivalent to a rise in sensitivity from 84% to 100% in the particular population studied. With respect to IgE quantitation, a median increase of 6.8 times with the enhanced reagent was observed. Considering the recent developments regarding the use of specific IgE quantitation for prediction of clinical reactivity to foods (35), we believe that this aspect of improved assay performance may also be of clinical importance.
For the Spanish subjects, all of whom were Cor a 1 negative, the unmodified and enhanced reagent compositions gave results that agreed within less than 20% in all cases. Thus, no negative effect of either the extract dilution or the rCor a 1.04 addition on the detection of hazelnut specific IgE antibodies in these sera was observed.
In conclusion, the present work demonstrates that a hazelnut allergen reagent with significantly improved diagnostic sensitivity and quantitative performance can be generated by supplementing natural extract with recombinant Cor a 1.04. This application exemplifies how recombinant allergens may be utilized to important clinical advantage already at a stage where not all relevant components in an allergen source are known or available in a pure form.