Identification of HLA-DR and -DQ alleles conferring susceptibility to pollen allergy and pollen associated food allergy

Authors


Dr Boehncke Department of Dermatology, Frankfurt University School of Medicine, Theodor-Stern-Kai 7, D — 60590 Frankfurt/Main, Germany.

Abstract

Background

Allergenic crossreactivity of pollen and foods due to the antigeneic similarity of oligopeptides is a well established clinical phenomenon.

Objective

To determine the immunopathological relevance of antigen presentation, we analysed the HLA class-II genotype of patients with either pollen allergy or pollen associated food allergy.

Methods

One hundred and twenty patients with pollen allergy and 80 patients with pollen associated food allergy were evaluated by skin- prick tests, RAST, and HLA class-II genotyping. The control population comprised 4251 healthy blood and bone marrow donors.

Results

Monovalent pollen allergy was observed in 57% (n = 68) of patients with pollinosis (57x grass pollen, 11x birch pollen), but only in 15% (n = 12) of patients with food allergy (9x grass pollen, 3x birch pollen). Hazelnut (71%), almond (65%), walnut (44%) and apple (41%) were the most common food allergens and frequently associated with birch pollen allergy. Grass pollen allergy was associated with an increased frequency of HLA-DQB1*0301 (RR = 2.3; EF = 0.4; = 0.0016) when compared with the control population. HLA-DRB1*08 confered a sixfold higher risk for peanut allergy (EF = 0.3; = 0.0013) and -DRB1*12 a 13-fold higher risk for carrot allergy (EF = 0.3; < 0.000001). The differences on allele frequencies detected among patients with food allergies diminished or turned statistically insignificant when their genotypes were directly compared to those of patients with the corresponding pollen allergies. This was found in the case of birch pollen associated hazel nut allergy for the extended haplotype HLA-DRB1*01, -DQA1*0101, -DQB1*0501 as well as in grass pollen associated peanut allergy for HLA-DRB1*08 (from RR = 6, = 0.0013 to insignificant) and in birch pollen associated carrot allergy for HLA-DRB1*12 (from RR = 13, < 0.000001 to insignificant).

Conclusion

We were able to identify HLA class-II alleles associated with some allergies thus indicating that these alleles might confer susceptibility to the respective allergens. Similarities at the level of the HLA class-II genotype parallel the empirical finding of distinct cross-reactivity patterns thus complementing investigations of IgE specificities. Our observations provide evidence for the major importance of antigen presentation on the manifestation of distinct crossreactivity patterns.

Ancillary