HLA Polymorphisms and Food Allergy Predisposition

Abstract

Food allergy (FA) is a growing health problem that affects ∼8% of the children worldwide. Although the prevalence of FA is increasing, the underlying genetic mechanisms responsible for the onset of this immune disorder are not yet clarified. Genetic factors seem to play a leading role in the development of FA, though interaction with environmental factors cannot be excluded. The broader network of genetic loci mediating the risk of this complex disorder remains to be identified. The human leucocyte antigen (HLA) has been associated with various immune disorders, including FA. This review aims to unravel the potential associations between HLA gene functions and the manifestation and outcome of FA disorders. Exploring new aspects of FA development with the perspective to improve our understanding of the multifaceted etiology and the complex biological mechanisms involved in FA is essential.

Introduction

The incidence and prevalence of food allergy (FA) are increasing over the last years. More specifically, the overall prevalence is estimated around 8% in children and 5% in adults, ¹ posing a new public health concern. The prevailing knowledge on the pathophysiology and genetic background of FA is still puzzling; nevertheless this immune disorder is generally considered to be determined by both genetic and environmental factors. ²

Polymorphisms in genetic loci are one of the major risk factors for the predisposition to FA. The genetic mechanisms underlying the development of FA and its interplay with environmental factors, including exposure to sunlight and to environmental pollutants, breastfeeding, and infections, are burning research topics. ^{1 3} Environmental exposures affect FA onset mainly at the postnatal period and they strongly interact with genetic background. ³

Several candidate genes have been experimentally tested in an effort to attain a deeper knowledge of this multifactorial disease. These studies included a sufficient number of genes including STAT6, SPINK5, FOXP3, IL-10, and HLA . ² Among them, human leukocyte antigen (HLA) is highlighted as the most prominent genetic locus, composed of a great number of different alleles and directly associated with the immune stimulation against external antigens. Each of these HLA alleles manifests a different binding-affinity toward antigenic determinants which derive from food allergens, known as epitopes.

In humans, genes of the HLA family are part of the major histocompatibility complex (MHC) locus, located on the short arm of chromosome 6 (6p21) and divided into three groups, classified as class I, II, and III. MHC molecules, encoded from the HLA gene loci, are highly variant heterodimers. Class I and II MHC molecules are expressed on the cell surface and mainly present antigens from the inner and external cell compartments, respectively, to stimulate CD8 ⁺ and CD4 ⁺ T cells. High affinity of the MHC–epitope interaction leads to a stronger response of the immune system. Plurality of the allergen epitope repertoire provokes a range of mild to very strong immune reactions. This feature attracts a strong interest to the HLA system and therefore it is considered as the most intriguing topic for association studies utilized to clarify the aberrant immune responses against peptides derived from food digestion and characterized with the general term of FA.

The first positive association between HLA and FA was reported in 1996 and since then many HLA loci have been implicated. Τhe HLA locus is known to be the most variant genetic system in humans and the affinity of epitope-binding to MHC molecules ultimately plays an important role in determining its immunogenicity. ⁴ A schematic outline of immunological responses to food allergens is presented in Fig. 1 . This review presents relevant studies that have examined the link between HLA alleles predisposition and manifestation of FA symptoms, in an attempt to illustrate the potential crosstalk between genetic background in the HLA gene loci and the risk of this immune disorder. Table 1 presents relevant studies conducted on the subject, results, and population origin.

Fig. 1.

Antigen presenting cells partially digest allergens in cell lysosomes and trigger naïve CD4 ⁺ T cells (Th0) to develop T helper (Th2) cells, through peptide-antigen recognition in the context of major histocompatibility complex (MHC) class II molecules. Augmentation of T cell receptor signaling is enhanced via CD80/CD86 and CD28 co-stimulatory molecules. Differentiation of Th0 to Th2 cells requires specific cytokines. Th2 further secretes cytokines to induce B cells to proliferate and produce allergen-specific immunoglobulin E, which bind to effector cells (e.g., mast cells, not shown) promoting an allergic reaction. IL-4, interleukin-4.

Table 1. Food allergens, HLA alleles, and detection methods.

Food allergen	Study (y) (Ref)	Type of study	Age	Population of the study	HLA genotypes, haplotypes, SNPs-associated positively with allergic patients	Origin	Detection methods	Sample size
Peanut	Donovan et al 1996 5	Family based Case–control	21, 22, 26, 27, 28 y, 49, 54 y	4 peanut allergic, 1 allergic to cat fur, 1 allergic to f Dermatophagoides pteronyssinus , 1 control (mother)	Haplotype Paternal HLA-DR4 haplotype	Australian	HLA-DR typing was performed by PCR with SSO (low-resolution typing)	5
	Madore et al 2013 12	Case–control	∼4 y	311 children peanut allergic, 226 controls	4 digit HLA allele HLA -DQB1*06:03P 2 digit HLA allele HLA-DQB1*02	Caucasian	Allele SEQR HLA-sequence-based typing packs by Abbott molecular diagnostics (high-resolution typing)	537
	Hong et al 2015 13	Observational study GWAS	Children (0–12 y), adults (>21 y)	2,759 US participants (1,315 children, 1,444 parents) 2,197 participants of European ancestry (316 peanuts allergic, 144 nonallergic non sensitized controls, 1,737 controls of uncertain phenotypes	SNPs rs9275596 (HLA-DQB1/HLA-DQA2) rs7192 (HLA-DRA)	European Non-European	Human Omni1-Quad Bead Chip, SHAPEIT and imputation usingIMPUTE2, Infinium HumanMethylation450 BeadChips (high-resolution typing)	2,694 1,283 children, 1,411 parents
	Martino et al 2017 14	Observational study GWAS	∼1 y	73 infants peanut allergic, 148 controls	Amino acid positions 37 and 71 in HLA-DRB1 gene 2 digit HLA alleles	Caucasian Asian	Illumina Human Omni 2.5–8 SNP array, MassARRAY platform and IPLEX chemistry (high-resolution typing)	121
	Shreffler et al 2006 16	Case–control	6.5–8 y	73 peanut allergic subjects, 75 siblings tolerant to peanut allergy (atopic and nonatopic)	HLA-DRB1*03*08 HLA-DQB1*04 4 digit HLA alleles HLADQB1*03:02, not associated 4 digit HLA allele	69 families Caucasian4 families Asian	Low-resolution genotyping	148
	Dreskin 2010 17	Case–control	3–69 y	53 peanut allergic subjects, 59 peanut tolerant siblings Control group of bone marrow donor (7,870)	HLA-DRB1*08:03 2 digit HLA allele HLA -DRB1*08 possible association	European	High-resolution typing	112
Milk	Marenholz et al 2017 15	GWAS	∼2 y	497 allergic children 2,682 controls	SNP rs9273440 (HLA-DQB1)	European	Haplotype Reference Consortium data, Illumina's HumanOmniExpressExome-8 v1.2, HumanOmniExpress-12 v1.0, or HumanOmni1-Quad v1(high-resolution typing)	3,179
	Marenholz et al 2017 15	GWAS	∼2 y	497 allergic children 2,682 controls	No association	European		3,179
	Camponeschi et al 1997 18	Case–control	1 mo–9 y	35 children with CMA 37 controls	HLA serotype HLA-DQ7	Italian	RFLP, PCR (low-resolution typing)	72
	Savilahti et al 2010 1 9	Case–control	8–9 y	87 children with CMA 76 controls	HLA haplotype HLA-(DR15)-HLA-DQB1*06:02 2 digit HLA alleles Hazelnut	Finnish	Low-resolution, full-house genotyping	163
Hazelnut , carrot	Boehncke et al, 1998 22	Case–control	Adults	120 patients with pollen allergy, 80 patients with pollen associated food allergy 4251 healthy controls	HLA-DRB1*01 Carrot HLA-DRB1*12 4 digit HLA alleles Hazelnut HLA-DQA1*01:01 HLA-DQB1*05:01 2 digit HLA alleles	Caucasian	PCR (low-resolution typing)	4,451
Nut	Hand et al 2004 21	Case–control	3–56 y	84 nut allergic 82 atopic controls 1,798 HLA typed random blood donors as control population	HLA-B*07† HLA-DRB1*11 HLA haplotypes Shrimp	81 Caucasian and3 mixed race (1 Afro-Caribbean/Caucasian, 1 Afro-Caribbean/Arab and 1 Asian/Caucasian)	PCR (using specific primers) (low-resolution typing)	1,984
Shrimp, peach	Khor et al 2018 26	Observational study GWAS	Adults >20 y	1,1011 Japanese women	HLA-DRB1*04:05-HLA-DQB1*04:01 Peach HLA-DRB1*09:01-HLA-DQB1*03:03 SNPs Shrimp rs28359884 (HLA-DR/DQ) Peach rs74995702 (HLA-DR/DQ)	Asian	PCR-SSOP (high-resolution typing)	1,1071
Egg	Park et al, 2012 23	Case–control	1,5–4 y	185 children (96 AD patients with egg allergy and 89 patients without egg allergy) 109 normal control	4 digit HLA allele HLA-DRB1*11:01	Asian	PCR- SSO and PCR-SSCP (high-resolution typing) Haplotype Reference Consortium data, Illumina's HumanOmniExpressExome-8 v1.2, HumanOmniExpress-12 v1.0, or HumanOmni1-Quad v1 (high-resolution typing)	294
	Marenholz et al 2017 15	GWAS	∼2 y	497 allergic Children 2,682 controls	No association	European		3,197
Hydrolyzed wheat protein	Noguchi et al 2019 25	GWAS	Adults > 20 y Mainly women aged 40 y	452 food allergic, 2,700 control group	Amino acid positions 34 for HLA-DQa1 13 and 26 for HLA-DQb1 SNP rs9271588	Japanese	Illumina Human Omni ExpressBeadChip, Taqman genotyping assay, PCR-SSO (high-resolution typing)	3,652

Abbreviations: AD, atopic dermatitis; CMA, cow’s milk allergy; GWAS, genome wide association study; HLA-DR4, human leukocyte antigen DR4; PCR, polymerase chain reaction; RFLP, restriction fragment length polymorphism; SNP, single-nucleotide polymorphism; SSO, sequence-specific oligonucleotide; SSCP, single-strand conformation polymorphism.

Original research studies published in English, between 1996 and 2018, were retrieved from PubMed database. Combinations of keywords relating to “FA,” “HLA,” and “MHC class I and II” were used to collect relevant studies. Collectively 15 studies were selected and discussed here (4 genome wide association studies [GWAS], 10 case–control studies, and 1 bioinformatics approach).

HLA and Food Allergens

Patients enrolled in the following studies were considered as food allergic on the assumption of specific symptoms displayed: (1) a convincing clinical history of immunoglobulin E (IgE)-mediated allergic reaction with objective symptoms from the skin (urticaria, angioedema), the respiratory system (wheezing), and the gastrointestinal tract (vomiting) or (2) positive skin prick tests/and or positive specific IgEs. In cases of conflicting clinical history concerning allergic reactions, subjects would undergo an oral food challenge under strict medical monitoring. Control group consisted of subjects with no history of atopy. ^{5 6 7 8}

Peanut

Peanut is one of the most common food allergens and on account of this it is the most studied allergen. One of the first studies attempting to relate specific allergen responsiveness to HLA included a family with four out of the five siblings demonstrating allergic reaction to various allergens including peanut. The purpose of the study was to discriminate whether peanut allergy could be attributed to T-cell receptor chains a- and b- (TCR-a, TCR-b), or HLA haplotypes. The results suggested that peanut allergy segregates only with a paternal HLA allele, while TCR-a and TCR-b alleles are not associated with the allergic phenotype. ⁹

In 1998, another investigation revealed the role that HLA class II alleles have in the development of peanut allergy. A familial study searching for resemblance of genetic predisposition to FA was conducted and included family members from 37 families, who were allergic to peanut. Also, nonatopic individuals were involved in the study. The overall population of this study included 161 individuals, representing the study group and encompassing 50 peanut-allergic, 34 nonpeanut-allergic siblings both atopic and nonatopic, and 77 nonpeanut-allergic parents. Two-hundred ninety-three unrelated controls participated in this same study. Participants were typed by polymerase chain reaction (PCR) amplification combined with sequence-specific oligonucleotide probing SSOP for the following HLA class II alleles, HLA-DPB1, HLA-DRB1, and HLA-DQB1. Four of the studied genotypes, specifically DQB1*04, DPB1*03:01, DRB1*08, and DRB1*08/12 tyr16, were significantly overrepresented, while two genotypes (HLA-DPB1*02:01, HLA-DPB1* 01:01) were less frequent in the study group compared with control group. When peanut-allergic individuals were compared with (family unrelated) controls only three of the above associations remained significant. Particularly, DQB1*04, DRB1*08, and DRB1*08/12 tyr16 genotypes were more frequent in the peanut-allergic individuals than in control group. Finally, genotypic comparisons between peanut-allergic and nonpeanut-allergic siblings revealed a weak association of DRB1*03 allele that became insignificant after statistical correction for multiple comparisons. At this particular time (1998), these results, although perplexed, supported the concept that at least some of the HLA class II polymorphisms favored the susceptibility to peanut allergy. ¹⁰ Although this study comprised a small study group, it was established as the first evidence that HLA genes affect the development of FA. ⁹

HLA-DQB1 gene has been associated with asthma, ¹¹ but it also exhibits a clinical correlation with FA. The association between HLA-DQB1 and peanut allergy has been evaluated in a study with hundreds of young participants: 311 children with peanut allergy and 332 children as the nonallergic-control group. Among studied HLA-DQB1 alleles, the frequency of DQB1*06:03P was increased, whereas frequencies of DQB1*05, DQB1*05:01P, DQB1*02, and DQB1*03:02P were decreased in the peanut-allergic group compared with the controls. After adjustments for asthma status and sex only DQB1*02 and DQB1*06:03P showed significant differences. The results propose that DQB1*02 shows a potential protective role to allergy, whereas DQB1*06:03P acts as a critical risk factor specifically for peanut allergy. ¹² These results suggested that HLA class II genes predispose to atopy including not only FA but asthma as well.

Hong et al conducted in 2015 the first GWAS of FA. This study included specific food allergens such as egg, peanut, and milk. The study group consisted of 2,759 individuals, of whom 1,315 were children and 1,444 were parents. Τhe researchers identified genetic variants in the HLA-DR, HLA-DQ gene-loci areas that were significantly associated with peanut allergy in children of European ancestry, tagged by two single-nucleotide polymorphism (SNPs), rs9275596, and rs7192. These associations were replicated in an independent sample of European ancestry (88 peanut-allergic, 128 controls). Furthermore, differential deoxyribonucleic acid methylation of the HLA-DQB1 and HLA-DRB1 genes was obtained to partially mediate the identified SNP-peanut allergy associations. The results indicated that the HLA-DR, HLA-DQ gene region definitely set an important genetic risk for peanut allergy. ¹³

Later on, Martino et al conducted a GWAS in which genetic variants were detected and associated with peanut allergy. The study included 73 infants with peanut allergy and 148 controls. A total of 3.8 million SNPs were tested. An association with amino acid positions 37 and 71 of the HLA-DRB1 gene was found significant during this analysis. The same association with this polymorphism was previously reported in the GWAS described by Hong et al. These findings are supportive for the role of specific polymorphisms within the HLA class II region to the risk of IgE-mediated peanut allergy development, more precisely those relating to the nucleotide triplets encoding HLA-DRB1 amino acids 37 and 71. ¹⁴ Subjects with peanut-allergic siblings acquire more than 10-fold greater risk to present peanut allergy compared with the general population.

Another GWAS on FA by Marenholz et al in 2017 stratified the results from the three most common FA: egg, peanut, and cow milk allergy. A replication study was performed. HLA region was among the five loci that showed genome wide significance to the allergic phenotypes studied. More particular, the SNP rs9273440 located in the 3ʹ-untranslated region (UTR) of HLA-DQB1 was strongly associated with peanut allergy, while no association was reported with egg and cow milk allergy. ¹⁵

As already mentioned, Howell et al observed a weak association of HLA class II alleles with peanut allergy in sibling pairs, which eventually became insignificant after statistical corrections. Two subsequent studies attempted to elucidate these results; however, both studies failed to replicate any previously reported associations between HLA allelic polymorphisms and peanut allergy. The first, by Shreffler et al, study investigated 25 different HLA alleles (7DQ and 18 DR alleles) in a population of 73 peanut-allergic children and their 75 peanut-tolerant siblings, with no FA history. Low-resolution HLA-typing revealed no significant HLA class II differences between. Comparison with unrelated control group was not performed. ¹⁶

The second, by Dreskin, study focused on previously reported elevated IgG levels in peanut-allergic subjects compared with controls and examined the relationship between IgG levels produced and performance of capture and processing of allergens via HLA class II. HLA class II alleles, HLA-DRB1, DPB1, and DQB1 alleles, were determined by high-resolution methodology, though no significant differences between peanut-allergic children and their nonpeanut-allergic siblings were obtained. In contrast, both antipeanut IgG and IgE levels showed significant differences between discordant sibling pairs with identical HLA class II alleles. HLA DRB1*0803 allele was overrepresented in peanut-allergic subjects as in nonallergic siblings tested in a large control group of bone marrow donors of European ancestry indicating that it could be an emerging factor of potential risk for peanut allergy in families of European origin. ¹⁷

Μilk

A research team from Italy studied the relationship between HLA antigens and children with cow milk allergy. The study consisted of 72 children, 37 of them with cow milk allergy and 35 nonallergic controls. Total IgEs were also counted. Results of the study clearly denoted that children with cow milk allergy were positively associated with the elevated presence of HLA-DQ7 antigens. ¹⁸

Cow milk allergy proteins have also been studied regarding their association with genes known to interfere in immunological reactions. A study focused on the filaggrin protein tyrosine phosphatase nonreceptor type 22 ( PTPN22 ) and HLA class II genes in an effort to discriminate whether the immunological response to food allergens is influenced by these genes. The population participating in the study, consisted of 163 children 8- to 9-year-old; among them 87 were allergic to cow milk and 76 were healthy controls. Data derived from this study indicated that HLA class II haplotype HLA-(DR15)-DQB1*06:02 has a positive association with IgG responses to cow milk peptide-antigens, β-lactoglobulin, and α-casein, among children with cow milk allergy. Furthermore, another HLA class II haplotype, the HLA-(DR1/10)-DQB1*05:01 was not associated with IgG responses to ovalbumin and to β-lactoglobulin. These findings strongly defined HLA class II haplotypes as positively associated features with cow milk allergy, whereas filaggrin mutations were rather implicated to the atopic dermatitis (AD) development than aberrant immune responses against food antigens. ¹⁹

Two GWAS were conducted, by Hong et al and Marenholz et al, to examine the association between the three most common food allergens (milk, egg, and peanut) and HLA alleles. The studies showed that milk unlike peanut had no correlation with the HLA alleles examined. ^{13 15}

It should be noted that the association between FA and HLA genes has also been analyzed using bioinformatics. Dimitrov and Doytchinova composed an algorithm that imitated the way that the human body treats food allergens. These food allergens were initially degraded by digestive enzymes and then their affinity and binding ability to DQ and/or DR molecules was examined. The algorithm was adjusted to 13 egg and milk allergens. The results showed that the peptides that generated both from egg and milk allergens preferentially bind to DQ7 (DQA1*05: 01/DQB1*03:01), DQ8 (DQA1*03: 01/DQB1*03:02) and DRB1 *01:01. Moreover, the peptides, derived from egg allergens, bind also to DQ4 (DQA1*04: 01/DQB1*04: 02). Peptides generated from milk allergens are unidentifiable from DRB1*03: 01, DRB1*04:04, DRB1*12:01 and DRB1*15: 01 and the same applies to the peptides generated from egg allergens with DRB1*03: 01, DRB1*04: 04, and DRB1*12: 01. It is worth mentioning that the alleles that bind to allergen peptides can be considered as susceptible to the particular allergy, whereas the nonbinding alleles have a protective role. This study failed to reveal any protective DQs against egg allergy. ²⁰ Based on these findings, it can be assumed that HLA class II and many of respective haplotypes, such as the above-mentioned DQ and DR interact with the IgE epitopes and they both have an important impact on the development of FA.

Nuts

Apart from peanut and milk, nuts have also been studied as usual food allergens. A study was conducted in UK with distinct study groups, 84 individuals allergic to nuts, 82 nonallergic to nuts, though with an atopic background and 1798 HLA typed blood donor controls. All participants were typed by PCR using sequence-specific primers for the HLA class I (HLA-A, HLA-B) and class II (HLA-DRB1, HLA-DQB1) loci. The study indicated that two genotypes, HLA-DRB1*11 and HLA-B*07, were prevalent significantly in the nut-allergic individuals compared with the atopic control group, but reverse results were obtained when the nut-allergic group was compared with the blood donor controls. All frequencies comparisons lost significance after statistical corrections. The authors suggested that the HLA-DRB1*11 and HLA-B*07 could be correlated specifically to nut allergy, due to their low frequency detection in the atopic control group. ²¹ These results provided more evidence that, irrespective of the polymorphisms detected in the HLA system, both HLA class I and II alleles are involved in the development of nut allergy.

It is worth noting that HLA class II genotypes also show an association with pollen allergy and pollen-associated FA. In a sample study composed of 120 patients with pollen allergy, 80 patients with pollen-associated FA exhibited strong association with HLADRB1∗08 genotype in comparison with controls. Concerning the hazelnut allergy, HLA-DRΒ1*01, HLA-DQA1*01:01, and HLA-DRB1*05:01 alleles were significantly decreased in comparison to controls. Furthermore, HLA-DRB1*12 allele showed significant association with carrot allergy. Surprisingly, no such associations were observed when peanut-allergy cohorts were compared with grass-pollen allergic. Similarly, in birch-associated, hazelnut, and carrot allergy frequencies of haplotypes HLA- DRΒ1*01, -DQA1*01:01, -DRB1*05:01 and HLA-DRB1*12, respectively, no statistical significance was reached. These data suggested that this HLA allele could potentially serve as a marker for atopy rather than being specific for FA. ²²

Egg

Furthermore, HLA-DRB1 polymorphisms association with egg allergy in Korean children with AD were examined by Park et al. The study included 96 AD patients with egg allergy, 89 AD patients without egg allergy, and 109 nonallergic controls. No statistically significant association was found between HLA-DRB1 polymorphisms and egg allergy after genotypic comparison among AD patients with egg allergy with either nonallergic controls or with AD patients without egg allergy. However, the study limitations were the small sample size, the enrollment of egg allergic individuals who had outgrown their egg allergy and the fact that only the specific IgE levels for egg white (and not ovomucoid, ovalbumin, ovotransferrin, lysozyme) were used as a diagnostic tool. ²³ ( Fig. 2 ) In addition, egg is one of the food allergens that has been examined with bioinformatics. In particular, the researchers created an algorithm adjusted to 13 milk and egg allergens. This algorithm had the ability to treat food allergens in the same way as the human body. The algorithm was adjusted to 13 egg and milk allergens. The results showed that the peptides that generated both from egg and milk allergens preferentially bind to DQ7 (DQA1*05: 01/DQB1*03: 01), DQ8 (DQA1*03: 01/DQB1*03:02), and DRB1 *01:01. Moreover, the peptides that derived from egg allergens bind also to DQ4 (DQA1*04: 01/DQB1*04: 02). ²⁰

Fig. 2.

Genomic regional association plot on chromosome 6, for polymorphic sites that reached genome-wide significance [log10(p-value) ≥3.2]. X-axis shows the chromosomal positions of major histocompatibility complex class II genetic loci and y-axis represents the −log10 ( p -value) for single nucleotide polymorphisms and/or polymorphic amino acid positions. Each color indicates association with different type of food allergy. The plot derived from data of referred genome wide association study.

Notwithstanding, egg was examined in two GWAS about its correlation with HLA alleles. The researchers found no association between this food allergen and the HLA alleles. ^{13 15}

Wheat

In another GWAS that was performed in Japan, the researchers examined the potential relationship between wheat allergy, specifically on the hydrolyzed wheat protein (HWP) and HLA alleles. HWP is a commonly used substance in food as well as in cosmetics, such as soap and it has been implicated in severe allergic reactions after ingestion or skin contact. Four-hundred fifty-two allergic subjects, especially women and 2,700 controls, were included in this study. The allergic patients were of Japanese origin and presented allergic symptoms toward a specific kind of HWP, Glupearl 19S. It is worth mentioning that these women unlike women of European origin developed permanent intolerance for wheat products after exerting allergic symptoms. ²⁴ A total of 6.6 million SNPs were genotyped with the Illumina Human Omni Express BeadChip. Taqman genotyping was used in the replication analysis. The results showed a positive association between HWP allergy and HLA alleles class II, HLA-DQa1 amino acid position 34, HLA-DQb1 amino acid positions 13 and 26, respectively, that are foundin the P4 binding pockets on the HLA-DQ molecule and the HWP and rs9271588. In addition, a replication research was further performed of 45 HWP allergic subjects and 326 controls. Results of the study remained statistically significant after the replication analysis confirming the substantial involvement of the HLA system to wheat allergy. ²⁵

Other Food Allergens

More recently, Khor et al conducted a GWAS, in which 11,011 Japanese women were involved. These women had an allergic response to 27 food allergens (fruits, legumes/grains, nuts, vegetables/mushrooms, eggs/dairy, meats, fish, shellfish product food groups). The peach and shrimp allergy specific loci were identified in the HLA-DR/DQ gene region tagged by rs28359884 and rs74995702, respectively. After HLA imputation, the most strongly associated haplotype was HLA-DRB1*09:01-HLA-DQB1*03:03 for peach allergy and HLA-DRB1*04:01-HLA-DQB1*04:05 for shrimp allergy. The findings of this study suggested that additional mechanisms, beyond the genetic predisposition for the HLA polymorphic alleles, are required for an allergic response. Moreover, gene expression modifications of related HLA genes (HLA class II in particular) influenced by genetic variants could play an important role in the etiology of FA. ²⁶

HLA Alleles Associated with Other Immune Disorders

Hitherto, the class II MHC region has been significantly implicated in allergy, which is composed of the HLA-DP, HLA-DQ, and HLA-DR loci, each one subdivided into hundreds of alleles. Specifically, class II molecules present peptides derived from extracellular proteins that have been partially digested through lysosomes into smaller, discrete molecules, and processed for presentation to T cells ( Fig. 1 ). Class II molecules have been associated with several types of autoimmune and other disease with hereditary predisposition, such as diabetes type 1, hemochromatosis, celiac disease, hypothyroidism, multiple sclerosis, rheumatoid arthritis, and systemic lupus erythematosus. ^{27 28 29 30 31 32 33 34 35 36} Class II HLA allele DRB1*15:01 is reported to conduct a dominant genetic risk for multiple sclerosis. Also, HLA class II genes have been recognized with variable disease susceptibility for Crohn's disease, ulcerative colitis, and celiac disorder. ^{29 30 31} The celiac disease is further associated with an increased risk of autoimmune diseases including Hashimoto's thyroiditis, Graves' disease, and type 1 diabetes. ^{29 30 34} Τype 1 diabetes has been strongly associated with heterozygosity for certain HLA class II alleles, the HLA-DR3, and/or HLA-DR4. The vast majority of patients with type 1 diabetes are homozygous for DQB1 allele encoding the β-chain that forms a dimer with an α-chain to constitute the class II DQ protein. Aspartic acid at position 57 of the DQ β-chain is closely related to resistance to type 1 diabetes, whereas other amino acids at this position, such as alanine, valine, or serine, confer susceptibility. Other loci and alleles in the MHC, however, are also important for the onset of diabetes type 1 diseas. ^{33 34 35} HLA-DRB1 haplotypes are tightly linked with both rheumatoid arthritis and systemic lupus erythematosus susceptibility and severity. ^{35 36}

Conclusion

In this review, we discussed the association between FA and HLA system. HLA system is known to be the most polymorphic genetic loci in humans, and it is also influenced by the ancestry of the population involved. Specifically, HLA class I and II and their multiple haplotypes contribute to a yet undefined level in the development of FA. It is worth mentioning that a considerable proportion of individuals who carry specific HLA class II risk alleles do not present FA. This is probably related to genome epigenetic modifications or protective functions οf other, still unconnected to allergy, genetic loci. Different methylation patterns have already been found in the HLA-DRB1 and HLA-DQB1 genes. So, it is worthwhile to further focus and elucidate any epigenetic mechanisms involved. Moreover, a variety of food allergens not previously studied extensively, such as fish, shellfish, soy, and wheat, is a strong incentive for further studies in the direction of their possible linkage with the HLA polymorphic allele. Furthermore, populations of different origin must be included in future studies as the majority of the already completed are referred to the Caucasian population only. New insights in genetics and molecular screening technologies are expected to shed light on the multifaceted etiology involved in various pathways of FA and, ultimately leading to a more effective prevention or even treatment of FA.