Atopic dermatitis (AD) is a chronic relapsing inflammatory skin disorder associated with interactions between environmental and genetic factors. Food allergy is found in 35 to 45% of children with moderate to severe AD.1 Among various food allergens, hen's egg, particularly egg white, is one of the most common causes of food allergy in young children, and egg allergy is correlated with AD severity and the likelihood of additional food allergies.2,3 The risks for sensitization to aeroallergens4 and asthma5 are also increased in children with egg allergy.

Identifying the responsible gene in food allergy and AD would lead to identification of children who require early preventive intervention. Many studies have attempted to determine the genetic loci involved in allergic responses. Genes such as CD14,6,7 FOXP3,8 STAT6,9 SPINK5,10,11 interleukin (IL)-10,12 and KoreaIL-1313 have been studied in food allergy patients, but showed inconsistent results. One candidate gene that may be involved in food allergy is human leukocyte antigen (HLA), because it plays a major role in immune response regulation and is associated with predisposition to a large number of immunologically mediated diseases.

HLA class I molecules are known to influence apoptosis, an important regulatory function in inflammatory responses. Polymorphisms within this region may therefore be related to the duration and severity of inflammation.14 HLA class II molecules have extensive molecular polymorphisms, almost entirely confined to the peptide-binding groove. These polymorphisms determine which antigen-derived peptides are bound and presented to T cells via T cell receptors. In the presence of appropriate co-stimulatory signals, this results in the activation of the T cells recognizing the HLA-peptide complex. Antigen presentation to T cells by HLA molecules is a defining step in the development of antigen-specific immune responses.15 Certain HLA class II alleles (DR, DQ, DP) influence specific IgE responses to airborne pollens15-17 and house dust mite allergens.18,19 With regard to food allergies, HLA-DQ7 has been positively associated with cow's milk allergy.20 HLA-DR and DP molecules have been shown to restrict T cell recognition of peanut allergens,21 and HLA-B^*07 and HLA-DRB1^*11 are associated with nut allergy.

To our knowledge, little has been documented regarding the association between HLA class II polymorphisms and egg allergy in AD. The purpose of this study was to investigate whether HLA-DRB1 polymorphisms contribute to the development of egg allergy in Korean children with AD.

An association between AD and HLA has been reported by several investigators, but few studies have shown a significant association between AD and HLA molecules. Ozawa et al.30 reported that patients with AD, bronchial asthma, and allergic rhinitis showed a significantly increased frequency of both HLA-B12 and B40, although no overall difference in HLA antigen frequencies was seen in patients with AD only. Schultz Larsen and Grunnet31 were also unable to demonstrate a significant association between AD and the frequency of HLA-A, B, or C antigens. The association of HLA class I gene polymorphisms with AD has been reported in Korean patients.32 HLA-A24 allele frequency was increased in AD, but statistical significance was lost after Bonferroni correction. Saeki et al.33 reported that the frequency of HLA-DRB1^*13:02 was increased in patients with severe AD and high serum IgE compared with controls, showing a susceptibility tendency associated with AD. However, this also did not remain significant after Bonferroni correction. Svejgaard et al.34 reported that the frequency of the HLA-DR7 antigen was decreased in patients with AD, but the significance of this deviation disappeared when P values were corrected.

In the present study, HLA-DRB1^*11:01 was significantly increased in Korean children with AD compared with controls (P=0.002, P_c=0.048, OR=7.796). To our knowledge, this is the first report implicating a specific HLA-DRB1 allele in AD susceptibility. The exact mechanism underlying the effect of HLA class II polymorphisms on AD development is not fully understood; however, it seems that HLA class II alleles play key roles in antigen presentation to CD4+ T lymphocytes via T cell receptors and thus influence specific IgE response to several allergens.35 Therefore, these alleles are excellent candidates for determination of specific allergic responses.

Few studies have considered immunological responses to food. Camponeschi et al.20 found a relationship between HLA-DQ7 and cow's milk allergy; however, as the control group was not atopic, it is unclear whether the association was related to atopy or milk allergy. Boehncke et al.36 evaluated patients with pollen allergy and those with pollen and food allergies compared with healthy controls. Although they found some associations with specific alleles, the associations with food allergies such as carrot and peanut were explained primarily by susceptibility to pollen allergy. With regard to peanut allergy, HLA-DR and DP molecules restrict T cell recognition of peanut allergens.21 Nevertheless, no association was reported in a cohort of sibling pairs.37 HLA-B^*07 and HLA-DRB1^*11 were investigated as predisposing factors in nut allergy, but there was no statistical significance.22 The synergistic effect of anti-HLA class II monoclonal antibodies on proliferative responses of T cells to ovalbumin in egg-sensitized patients was investigated.38 Proliferation of T cells from AD patients with a low immune reaction was restored by anti-HLA-DQ monoclonal antibodies, and the proliferation of T cells from other patients was inhibited by anti-HLA-DP monoclonal antibodies. These results suggest that a component of the T cell repertoire that is reactive with ovalbumin in egg-sensitized patients may be restricted by HLA-DP molecules.

In the present study, we found no statistically significant association between HLA-DRB1 and egg allergy. HLA-DRB1^*15:01 and HLA-DRB1^*08:02 were regarded as susceptibility (P=0.036) and protective factors (P=0.021), respectively, for the development of egg allergy in AD. However, neither remained significant after P values were corrected. As many egg allergy patients in our study were also sensitized to other foods, we selected the patients who were allergic to egg alone (n=22) and compared the HLA-DRB1 phenotype frequency between these patients and normal controls. Again, we failed to find alleles significantly associated with egg allergy (data not shown).

Thus far, nucleotide sequence polymorphisms have been identified in at least 762 distinct HLA-DRB1 alleles.29 However, nucleotide differences do not account well for the observed variations in the associations with disease susceptibility or resistance. Differences in amino acid sequences are more relevant, because a single substitution at a critical residue of the HLA binding groove can alter the binding of antigenic peptides and totally change the nature of the immune response. HLA-DRB1 proteins contain subregions (pockets) in the DR binding groove which exert a major influence on peptide binding and its subsequent recognition by T cells.28 In particular, the polymorphic residues 70, 71, and 74 of the β chain located in pocket 4 play a critical role in the recognition of the HLA-DR/peptide complex by the CD4+ helper T cell.39 HLA-DRB1 alleles are categorized into seven DR restrictive supertypes (A, De, Dr, E, Q, R, and a) on the basis of physicochemical characteristics, especially charge, of the polymorphic residues. This novel categorization of DR alleles on the basis of function allows the prediction of specific sequence motifs within pocket 4 of the HLA-DR peptide-binding groove that are associated with autoimmune and infectious disease susceptibility.27 According to a Spanish study, 'Dr' was associated with atopic disease.40 In our study, AD patients with egg allergy exhibited a significantly higher frequency of functional group 'a' compared with AD patients without egg allergy or normal controls (P=0.033). A trend toward an association between egg allergy development in AD and functional group 'a' was observed, but significance was lost after Bonferroni correction. Further studies are needed to elucidate the association between HLA-DR functional groups and egg allergy.

Although no association between HLA-DRB1 polymorphisms and egg allergy was found in this study, several points should be taken into account. First, the study population was small. Given the number of HLA-DR alleles, it is difficult to obtain significant results using a small number of patients and controls owing to statistical limitations. Second, the diagnosis of egg allergy was not confirmed by oral food challenge. Controversy exists regarding the accuracy of serum-specific IgE levels in diagnosing food allergies.41 Third, multiple epitopes from various egg allergens may be bound by a number of discrete HLA class II molecules during processing and presentation, and it is possible that each allergen is associated with a specific HLA class II allele. The most important allergenic egg proteins are ovomucoid, ovalbumin, ovotransferrin, and lysozyme.42 Although ovalbumin is the most abundant protein in egg white, ovomucoid has been shown to be the dominant egg allergen.43 We did not confirm the presence of specific IgE against each egg allergenic protein in our patients. Finally, children who had outgrown an egg allergy might have been included in the AD without egg allergy group at enrollment. The selection of persistent egg allergy may assist in identifying alleles significantly associated with egg allergy.

In conclusion, the HLA-DRBI polymorphism was not associated with egg allergy. However, HLA-DRB1^*11:01 was associated with AD in Korean children.