"Allergolds": Gold Nanocluster-Based Bioconjugates of Food Allergens with Reduced Immunoglobulin E Binding

Abstract

Allergen-specific immunotherapy represents the only method of achieving a lasting reduction in the severity of allergic symptoms. However, the need to expose patients to the allergens to which they are sensitized carries risks. One solution is to use denatured allergens whereby the structure of allergenic proteins is disrupted, preventing their recognition by immunoglobulin E (IgE) antibodies and thus reducing the risk of adverse reactions. Denaturation is often carried out by using chemical cross-linking to generate allergoids. Gold nanoclusters (AuNCs) are emerging as versatile tools in biotechnology due in part to their ability to conjugate a wide range of biological molecules. Previous works have described the formation of AuNC using egg allergens such as Gal d 4 (lysozyme), Gal d 2 (ovalbumin), and whole egg whites. In all cases, AuNC bioconjugation disrupted the protein structure, allowing for their use in biosensing applications. In this work, we hypothesize that these AuNC-allergen bioconjugates could be used to generate "Allergolds", chemically altered versions of allergenic proteins analogous to traditional allergoid formulations. Using spectroscopic techniques, we confirm that the formation of AuNC bioconjugates of the chicken egg Gal d 4 and Gal d 2 disrupts protein structure when generated from both purified protein and whole egg whites. This structural perturbation was found to be resilient to a range of chemical conditions and successfully disrupted recognition by human IgE. These results establish Allergolds as a potential tool for generating systematically denatured allergens from both purified proteins and biological extracts.

Figure 1

Formation of AuNC-Gal 4 nanoclusters. (A) Representative fluorescence spectra of AuNC-Gal d 4 (0.215 mM) showing the characteristic emission peak at 700 nm (λ_ex:360 nm). A control spectrum obtained from an equivalent concentration of unfunctionalized Gal d 4 is shown for comparison. Representative fluorescent spectra displayed here and elsewhere throughout the work are taken from three representative trials unless otherwise specified. Note that the three individual unfunctionalized Gal d 4 spectra cannot be distinguished due to overlap. (B) Representative image of Gal d 4 sample under visible light showing the characteristic orange color. (C) Average DLS size distribution per volume plot of prepared AuNC-Gal d 4 and Gal d 4, further confirming bioconjugate formation. (D) Representative TEM image of AuNC-Gal d 4. Additional images are available in Figures S1–S4. (E) Comparative size distribution analysis of the AuNC core size.

Figure 2

Conjugation with AuNC perturbs the Gal d 4 structure. (A) Representative tryptophan fluorescence spectra (λ_ex: 280 nm) of equivalent concentrations of Gal d 4, AuNC-Gal d 4, and the NaOH-treated negative control showing a loss of fluorescence intensity upon AuNC bioconjugation. (B) ¹H NMR spectra of Gal d 4 and AuNC-Gal d 4 showing the amide e ¹H region. Spectra were normalized for relative protein concentrations. (C) Representative CD spectra for equivalent concentrations of Gal d 4 and AuNC-Gal d 4, showing a loss of α-helical structure, represented by the peak intensity at 220 and 210 nm. Representative CD spectra are taken from at least three representative trials. Note that the individual Gal d 4 spectra cannot be distinguished due to overlap. (D) Quantification of secondary structure content from the CD spectra shown in panel (C) where orange bars correspond to the AuNC-Gal d 4 and blue bars correspond to the native Gal d 4. (E) CD spectra obtained for Gal d 4 (0.2 mg/mL) at various pH’s. (F) Representative size-exclusion chromatography elution profile for AuNC-Gal d 4. A single peak is observed corresponding to the AuNC-Gal d 4 bioconjugate at 0.38 column volume (CV) or 9.1 min. The representative fluorescence spectrum taken of peak fractions confirms the presence of AuNC bioconjugate with characteristic emission at 700 nm (inset). Small peaks at >0.7 CV (denoted by *) represent leftover reagent from the bioconjugation process. No significant fluorescence was observed for these fractions.

Figure 3

AuNC bioconjugation irreversibly disrupts the protein structure/function. (A) Catalytic activity of AuNC-Gal d 4 and unfunctionalized Gal d 4. The activity was assessed using an equivalent amount of protein and presented as relative (normalized) values. Three independent batches of AuNC-Gal d 4 were analyzed in triplicate (for a total of 9 trials). The average of each individual batch is shown in circles along with the age of each batch at the time of analysis. Three trials were run on unfunctionalized Gal d 4. The overall averages and standard deviations for the two conditions are shown in the bar graph. (B) Gal d 4 activity following thermal denaturation at 90 °C at different heat exposure times: 0, 15, 30, and 60 min showing both the incomplete denaturation of Gal d 4 and the rapid recovery of catalytic activity upon returning to room temperature. All Gal d 4 activity samples were allowed to cool for 10 min after heat exposure, and their values were normalized against the control sample of equivalent concentration of unconjugated Gal d 4. (C) Average thermal denaturation curves for Gal d 4 (blue) and AuNC-Gal d 4 (orange), as assessed using tryptophan fluorescence (λ_ex: 280 nm, λ_em: 340 nm). Forward (25 → 95 °C) and reverse (95 → 25 °C) melting curves are shown in solid and dashed lines, respectively. (D) A close-up view of the individual AuNC-Gal d 4 melting curves from three separate trials.

Figure 4

Allergolds display reduced antibody binding. (A) Average relative IgG binding of three different batches of prepared AuNC-Gal d 4 vs a Gal d 4 control determined by ELISA. Each AuNC-Gal d 4 batch was analyzed three times (for a total of 9 trials). The average of each individual batch is shown in circles, along with the age of each batch at the time of analysis is shown. Note that the values from batch 2 and batch 3 overlap. Four trials were run on the unfunctionalized Gal d 4, and the individual results are shown in circles. The overall averages and standard deviations for the two conditions are shown in the bar graph. (B) Relative IgG binding of prepared AuNC-HEW vs a HEW control. Average and standard deviations for all trials are shown. Individual results for AuNC-HEW overlap and are omitted for clarity. All IgG ELISA were carried out using 3000 ng/mL analyte. (C) Competitive inhibition-ELISA assay examining the binding capability of unfunctionalized and AuNC-Gal d 4 against human IgE’s from allergic patients. This assay measures the ability of the analyte to compete with a fixed concentration of Gal d 4 for IgE binding (Materials and Methods). (D) The ability of HEW and AuNC-HEW to bind IgE. Data from the individual repeats (minimum 3) are shown in circles.

Figure 5

Impact of Allergold formation on Gal d 2 structure and morphology. (A) Average fluorescence emission spectra (λ_ex: 365 nm) of prepared AuNC-Gal d 2 and aqueous solution of 0.180 mM Gal d 2. (B) Representative AuNC-Gal 2 under visible light. (C) Average fluorescence emission spectra (λ_ex: 280 nm) of prepared AuNC-Gal d 2 and aqueous solution of 0.180 mM Gal d 2. (D) Average CD spectra of prepared AuNC-Gal d 2 and aqueous Gal d 2 solution. (E) Average DLS size distribution per volume plot of prepared AuNC-Gal d 2 and aqueous solution of 0.180 mM Gal d 2.

Figure 6

Comparison of known Gal d 4 IgG and IgE epitopes. X-ray crustal structure of Gal d 4 (PDB: 1dpx) showing dominant (A) IgE binding epitopes in blue and (B) IgG epitopes in green. In both images, cysteine residues are shown in orange, while residues common to both IgE and IgG (AA 31–34 and AA 84–91) are shown in red. (C) IgE and IgG epitopes are displayed on the Gal d 4 sequence.