Hymenoptera stings may cause life-threatening and sometimes fatal IgE-mediated anaphylactic reactions with the major threat emanating from the yellow jacket V. vulgaris and the honeybee A. mellifera. Although venom immunotherapy is highly effective, an adequate diagnosis and identification of the culprit venom is hampered by the use of crude venoms for measurement of specific IgE levels. Thereby, the main problem arises from serologic double-positivity for A. mellifera and V. vulgaris venom of up to 50% of patients that have IgE against hymenoptera venoms [1]. Apart from true double-sensitisation this phenomenon is largely attributed to molecular cross-reactivity, either based on the presence of cross-reactive epitopes in homologues proteins of both venoms such as the hyaluronidases and dipeptidylpeptidases, or the presence of so-called cross-reactive carbohydrate determinants (CCD) which account for 70-80% of cross-reactive patients found within the double-positive cohort [2].

Since common diagnostic approaches are not capable of circumventing or differentiating these cross-reactivities considerable interest has emerged in strategies enabling an improved diagnosis. Fail-safe identification of the culprit venom is imperative for design of appropriate therapeutic intervention with either one or both venoms and thus key to any improvement.

The most promising approach for the development of reliable diagnostics as well as safer and more efficacious patient-tailored treatment modalities relies on the use of defined recombinant allergens [3]. For honeybee venom (HBV), phospholipase A2 (Api m 1) has emerged as surrogate marker, however, for YJV access to native proteins is limited and only a minor number of recombinant allergens are available [4, 5].

The three yellow jacket allergens thought primarily responsible for IgE-mediated allergic reactions include phospholipase A₁ (Ves v 1), hyaluronidase (Ves v 2), and antigen 5 (Ves v 5) [6]. To some extent this estimation is prompted by the concentration of the particular proteins in the venom. However, the relevance of Ves v 2 as allergen has been disregarded [7], while the novel high molecular weight compound Ves v 3 recently was reported to exhibit a pronounced allergenic potential [8]. Both Ves v 2 and Ves v 3 are glycoproteins prone to CCD reactivity with homologues in HBV.

By contrast, Ves v 1 and Ves v 5 are non-glycosylated and unique candidates for diagnosis of YJV allergy. While expression of Ves v 5 could be demonstrated in various eu- and prokaryotic hosts [9], only insoluble protein was obtained in scarce attempts of Ves v 1 expression [10] rendering a reliable assessment of IgE reactivities on the basis of such a protein questionable.

In this study, we report the successful expression of both, Ves v 1 and Ves v 5, in insect cells and their subsequent biochemical and immunological characterisation. The unanticipatedly pronounced prevalence of IgE reactivity in YJV-sensitised patients to rVes v 1 emphasises the need for two unique recombinant major allergens from YJV, especially in terms of double-positivity.

Standard diagnostic approaches in hymenoptera venom allergy, but also in plant associated allergies are often hampered by multiple IgE reactivities affecting the interpretation of ambiguous results and the correct choice of the proper venom for immunotherapy, a prerequisite for efficient therapy [1].

Causative for this phenomenon is IgE binding either to peptide epitopes of closely related or homologous proteins or to conserved carbohydrate structures of related or otherwise unrelated glycoproteins. Obviously, the more common incidence in hymenoptera venom allergy is the latter one, the molecular basis of which could be attributed to α-1,3-core-fucose and, exclusively in plants [14], β-1,2-xylose. These residues are absent in mammalian glycosylation and, therefore, constitute a highly immunogenic epitope in men. In contrast to protein-directed cross-reactivity, the carbohydrate-directed reactivity in food and hymenoptera allergy is mainly believed to be clinically irrelevant, but diagnostically cumbersome [15, 16].

Since identification of the culprit venom is strongly affected by such double positivities the choice of the proper venom for immunotherapy is often difficult. A sophisticated method to identify and circumvent such reactivities are inhibition tests based on mutual inhibition of IgE by venom of the particular species [17]. However, these tests are not widely used in standard diagnosis and the obtained data are difficult to interpret. Hence, the more advanced and promising option relies on the use of unique recombinant major allergens which are representative for the respective venom and fulfil all criteria regarding high prevalence and low cross-reactivity of both types.

For HBV, phospholipase A2 (Api m 1) is considered an ideal surrogate marker as it is structurally unrelated to the phospholipase A2 in vespid venom and shows a high prevalence of sIgE recognition [18]. Nevertheless, due to its nature as glycoprotein cross-reactivity on CCD level is possible, even if this seems to be reduced in a natural conformation as indicated by patient sera that showed IgE reactivity to MUXF and glycoproteins of insect origin (data not shown) but not to Api m 1. No other proteins in HBV fit to the diagnostic needs due to presence of multiple glycosylation sites and confirmed CCD reactivities. However, recombinant approaches might offer opportunities for establishment of improved allergen molecules devoid of glycosylation.

In contrast, YJV contains two non-glycosylated major allergens without cross-reactive homologues in other species, Ves v 1 and Ves v 5, both showing high IgE prevalence, as shown for proteins purified from venom [19]. Allergens like Ves v 2 a and b and Ves v 3 do not meet those criteria as they are glycoproteins and in case of the Ves v 2 isoforms show only minimal IgE prevalence apart from carbohydrate based IgE binding. In contrast, the HBV hyaluronidase Api m 2 appears to be a true, but minor allergen recognised by approx. 30% of sensitised patients only [7, 8]. Furthermore, the dipeptidylpeptidases from HBV and YJV have been shown to be cross-reactive due to their high sequence homology [8].

The phospholipase A2 from HBV can be purified from venom that is easily obtained by electrostimulation, nevertheless, recombinant production could be shown in bacteria. Although refolded from insoluble aggregates enzymatic as well as biological activity of the recombinant phospholipase in terms of effector cell activation proved to be comparable to the native protein [20].

In contrast to Api m 1 and Ves v 5 [9, 21] and to the best of our knowledge, no expression of functional phospholipase A1 from YJV has been reported so far [10]. This is even more important since the purification from natural sources is hampered by the significantly higher effort to obtain substantial amounts of YJV.

However, the approach used in this study, production of hymenoptera venom allergens in a nearly autologous system, yielded Ves v 1 for the first time as a soluble and enzymatically active molecule. Interestingly, expression of the active enzyme appears not to be detrimental for the host cells, as shown with the inactive variant of Ves v 1. This finding might be contributed to the lytic nature of the baculovirus mediated overexpression, resulting in apoptosis 48-72 h after infection. Biophysical and biochemical measurements of both recombinant Ves v 1 and Ves v 5 were entirely in accordance with activity and native folding. Under the aegis of insect cell-mediated folding and expression the allergenic characteristics were compatible even with activation of human basophils, an ex vivo assay considered to reflect the pathophysiological situation.

By the use of such defined rVes v 1 and rVes v 5 molecules a true sensitisation to YJV for ≥ 80% of the patients independently of their sensitisation could be confirmed, while a true double-sensitisation was verified in only 10% of the double-positive patients as indicated by additional sIgE reactivity to nApi m 1. Thereby, the prevalence of CCD reactive patients was approx. 60% in the double-positive cohort which is in accordance with the literature [1] and can be assumed the only reason for cross-reactivity. In depth follow up studies with increased patient numbers in the future can provide detailed information about the prevalence of IgE reactivity as well as clinical relevance of Ves v 1 and Ves v 5. Additionally, the potential need for further vespid allergens and implementation of the allergens into laboratory systems for quantitative sIgE measurements to enable fail-free detection of all YJV sensitised individuals should be thoroughly evaluated.

In summary, the use of the defined recombinant major allergens Ves v 1 and Ves v 5 provides a significant improvement for the identification of the culprit venom which is indispensable for the choice of the appropriate immunotherapeutic strategy. Thereby, the need of Ves v 1 for detection of true sensitisation to YJV could be established. Implementation of rVes v 1 to routine diagnosis thus can allow for assessing its true IgE prevalence and clinical relevance beyond estimations from immunoblot studies. Furthermore, component-resolved diagostics using recombinant allergens such as Ves v 1 and Ves v 5 may provide new insights into the role and relevance of particular venom compounds during sensitisation and hyposensitization.