Aplicación de la tecnología microarray al diagnóstico de la alergia a los alimentos

Abstract

Background: Food allergies are a growing global health problem, that are estimated to affect 3-7% of children and 2% of adults. The foods most frequently involved in allergic reactions are cow’s milk, eggs, fruits and tree nuts. Currently there is no an active treatment that can prevent or completely relieve symptoms, and they are usually focused on avoidance of the food. The tools available are useful to assist in the diagnosis of IgE-mediated food allergy, but it is necessary to incorporate new techniques to facilitate management of patients and to predict the course of the disease. Peptide microarray technology has been proposed as a useful tool for the diagnosis and prognosis of food allergy. Objectives: The goal of this work is to validate the applicability of peptide microarray technology on food allergic patients to the main allergens of cow’s milk: αs1-casein (αs1-cas), αs2-casein (αs2-cas), β-casein (βcas), κ-casein (κ-cas), and β-lactoglobulin (β-lac) (COALE study), egg: ovalbumin (OVA) and ovomucoid (OM) (OVALE study), peanut (Ara h 9) and peach (Pru p 3) (LTP study). In addition we assess peptide microarray technology utility for diagnosis and prognosis of allergy to referred foods. Methods: COALE was a longitudinal study that included a cohort of infants with allergy to cow milk assessed at 6 follow-up visits. OVALE was also a longitudinal study that included a cohort of infants with allergy to egg evaluated at 5 follow-up visits. Finally, LTP was a cross-sectional study carried out on a cohort of peach allergic patients sensitized to peanut. In all studies relevant clinical data were collected and allergy diagnosis was confirmed by prick test, in vitro levels of specific-allergen antibody, and oral food challenges in every follow-up visit. Microarray immunoassay was performed using sera from allergic patients. A library of 20-mer overlapping aa (3-offset) was printed on nanosurface microarray slides, corrresponding to the primary sequences of: αs1-cas, αs2-cas, β-cas, κ-cas, β-lac of milk in COALE study, OVA and OM of egg in OVALE study, and Ara h 9 and Pru p 3 in LTP study. Biostatistics and bioinformatics methods were applied for data analysis. Results: In the COALE study no differences were observed in prick test and specific IgE and IgG4 levels at baseline. Microarray immunoassay was performed in a subgroup of 118 patients. The population of study showed similar IgE-recognition patterns at baseline and no significant differences were observed at the beginning. IgE binding intensity and IgE diversity peptides were significantly higher, both at visit 2 and 3 in allergic than tolerant subjects. 6 significant peptides were identified at visit 2 and 17 at visit 3. Only 10 peptides of the latter were selected through ROC analysis as biomarkers: 4 of αs1-cas (αs1-casp1, αs1-casp16, αs1-casp17 and αs1-casp18), 4 of β-lac (β-lacp5, β-lacp6, β-lacp7 and β-lacp38) and 2 of κ-cas (κ-casp30 and κ-casp31). Additionally, a decision tree model in machine learning combining αs1-casp16 and κ-casp30 biomarkers was identified for maximizing diagnostic power. In the OVALE study, no differences were observed in prick test and specific IgE and IgG4 levels at baseline. Microarray immunoassay was performed in a subgroup of 42 patients. IgE binding intensity and IgE diversity peptides showed no significant differences at any follow-up visits or at baseline. However individual analysis of peptides found 10 significant peptides identified at visit 3 as biomarkers: 25-27, 73-75, 91-92, and 118-119 peptides of OVA protein. Additionally, a decision tree model in machine learning combining peptides 74 and 75 of OVA protein was identified. IgG4 recognition did not show a predictive value for tolerance. Finally, the LTP study included 48 peach allergic-patients, of whom 37 were peanut-allergic, and 11 were peanut-tolerant. Unexpectedly, peanut-tolerant patients showed greater binding intensity and number of peptides, both IgE and IgG4, than allergic patients. Interestingly, peanut-tolerant patients also presented significantly high levels of IgG4 antibodies against Ara h 9 peptide 4, which could partly explain the tolerance of these patients through their IgE epitope blocking activity. Despite the high homology between Ara h 9 and Pru p 3 proteins, the structural differences between peptide 4 of both proteins agree with the possible blocking role of Ara h 9 peptide 4 and not of Pru p 3 peptide 4. Conclusions: Using peptide microarray technology, differences in IgE recognition profile between allergic and tolerant patients have been established. Besides, a group of diagnosis peptide biomarkers has been found in both COALE and OVALE studies. In LTP study, IgE and IgG4 binding profile between peanut-allergic and peanut-tolerant patients have been identified in Ara h 9 and Pru p 3. Ara h 9 peptide 4 could have a predictive value of tolerance in peanut allergy. Peptide microarrays could be very useful in the diagnostic management of the reported allergies.