Establishment of an ELISA-based in vitro test system to quantify the major honey bee venom allergen Api m 10 in crude bee venoms and therapy allergen products

Abstract

In Germany, hymenoptera venoms (mainly from bees and wasps) are the leading cause of anaphylaxis in adults.¹ Diagnosed honey bee venom (HBV) allergies are treated by subcutaneous immunotherapy using standardized venom preparations obtained by electrostimulation.² Crude HBV and HBV allergen-immunotherapy (AIT)-products are complex mixtures of non-proteinaceous and proteinaceous components, some of which are allergens.³ Currently, there are no commercial quantification systems available for these allergenic components. Icarapin (Api m 10) is one of the major HBV allergens and was suggested to be underrepresented in HBV AIT-products.⁴ Moreover, its underrepresentation was postulated to correlate with potential therapy failure.⁵

Treating patients with a clinically relevant Api m 10-sensitization requires the usage of AIT-products containing Api m 10. To both analyze and to monitor Api m 10 amounts in complex HBV AIT-products in the future, we developed a quantitative Api m 10-specific sandwich ELISA.

BALB/c mice (n = 5) were repeatedly immunized with increasing amounts of purified recombinant (r) Api m 10 and monitored for antigen-specific humoral immune responses to generate Api m 10-specific monoclonal antibodies (mAbs). Splenocytes isolated from sensitized animals were fused with immortalized P3X63Ag8U.1 myeloma cells (ATCC CRL-1597, ATCC, Manassas, VA, USA) (Figure 1A). All rApi m 10-immunized mice developed antigen-specific IgG₁ with OD_450-630 >1 in indirect ELISA, which was not observed in placebo controls. Animals used for mAb generation are highlighted in green (Figure 1B). By limiting dilution, two rApi m 10-specific clones were obtained and identified as IgG₁ with kappa light chains (Figure 1C). Both mAbs were successfully purified by chromatography (online repository). To analyze binding properties of mAb1&2, surface plasmon resonance (SPR) spectroscopy was used. Both mAbs showed dose-dependent binding to rApi m 10 with affinity constants in the low nanomolar range (mAb1: K_D = 59.0 ± 2.7 nM (Chi² = 0.38), mAb2: K_D = 1.45 ± 0.12 nM (Chi² = 1.05)) (Figure 1D). Furthermore, binding of monoclonals to the 11 described variants of Api m 10⁶ was investigated. So far, only the alternative Api m 10 splice variants 1 and 2 have been detected in natural HBV at protein level.^{4, 7} The other variants (variant 3–11) have so far only been described as chimeric transcripts at RNA levels,⁶ and the evidence that they also exist as proteins is rather indirect. Here, mAb1 showed exclusive binding to variants 1&2, whereas mAb2 bound variants 1–7 (Figure 1E), which was confirmed by dot blot (data not shown). Putative binding sites of both mAbs to Api m 10 (deducted from their individual binding to Api m 10 variants: for mAb1 [blue] sequence: DSDEGSNWNWNTLLRPNFLDGWYQTLQ and/or HMKKVREQ referring to sequences only contained in variants 1&2 and for mAb2 (red) sequence: EDFDNEIPKNQGDVLTA only contained in variants 1 to 7) are indicated in Figure 1F. Consecutive binding of both monoclonals was demonstrated by SPR (Figure 2A), and a sandwich ELISA was established using mAb1 as capture and biotinylated mAb2 as detection antibody (Figure 2B). The ELISA detected 0.8–100 ng/mL rApi m 10 with a sigmoidal standard curve, but none of the other tested recombinant, commercially available HBV allergens (Figure 2C). Furthermore, using the established Api m 10 ELISA, differentiation between Api m 10-containing HBV and Api m 10-free wasp venom was demonstrated (Figure 2D). Recently, Grosch et al. reported the presence of an icarapin-like protein in the venom of Polistes dominula,⁸ and Jakob and colleagues found (so far on protein level unconfirmed) sequences of icarapin-like proteins in the genomes of Apis cerana, Megachile rotundata, Bombus terrestris, Polistes dominula, Solenopsis invicta, Leptinotarsa decemlineata, and Drosophila grimshawi.⁹ Whether icarapin-like proteins from Polistes dominula or other hymenoptera species can be recognized by our sandwich ELISA remains to be determined. Finally, the ELISA was used to quantify Api m 10 amounts in commercial aqueous HBV AIT-products (Figure 2E, Figure S1). Here, product-specific differences in Api m 10 content were observed with the following Api m 10 amounts being detected: product (P)1: 0.25 ng Api m 10/μg total protein, P2: 0.25 ng/μg, P3: 0.04 ng/μg, P4: 0.37 ng/μg (Table 1). Furthermore, with exception of P1, Api m 10 amounts were stable for 14 days when stored at 4°C. In concordance with previously reported findings,^{4, 5} our results indicate lower amounts of Api m 10 (ranging from 0.03 to 0.37 ng Api m 10/μg total protein) in processed HBV AIT-products than in the investigated crude venoms (HBV1: 1.92 ng Api m 10/μg total protein, HBV2: 1.58 ng Api m 10/μg total protein) (Table 1). The Api m 10 amounts in crude HBV-preparations detected by our ELISA are in line with previous observations,⁴ detected by semi-quantitative western blot.

In summary, we established a highly selective, quantitative Api m 10 ELISA using mAbs generated from Api m 10-immunized mice. This ELISA detected Api m 10 as isolated protein, in crude HBV, and in commercially available, complex aqueous HBV AIT-products. Therefore, our ELISA may be an important tool for future characterization of HBV AIT-products.

Alisa Landgraf (AL), Kathrin Elisabeth Paulus-Tremel (KEPT), Meike Arend (MA), Ann-Christine Junker (ACJ), Sascha Döring (SD), Sandra Schmidt (SaS), and Susanne Kaul (SK) performed animal experiments. AL performed immune assays. Sascha Hein (SH) implemented the sandwich ELISA and performed SPR analyses. Michelle Beatrice Wolff (MBW) expressed recombinant proteins supervised by KEPT. AL, Daniel Strecker (DS), SD, KEPT, MA, and ACJ generated monoclonal antibodies. Elke Völker (EV) and AL purified monoclonal antibodies supervised by Thomas Holzhauser (TH). AL, KEPT, SH, TH, SK, Stefan Schülke (StS), and Vera Mahler (VM) interpreted the data. VM conceptualized the study. KEPT, SaS, SK, and VM designed the animal study. SaS and SK supervised animal experiments. KEPT, StS, and VM supervised the study. AL, StS, SH and VM wrote the paper. All authors reviewed the manuscript.

The authors have no conflict of interest to declare.