Polymorphism in the STAT6 Gene Encodes Risk for Nut Allergy

Home , Aeroallergen, Peanut allergy, Tree nut allergy

MM Amoli1, S Hand2, AH Hajeer3, KP Jones4, S Rolf4, C Sting5, BH Davies5 and WER Ollier1 1The Centre for Integrated Genomic Medical Research, Manchester University Medical School, Stopford Building, Oxford Road, Manchester M13 9PT, UK; 2Prince Charles Hospital, Merthyr Tydﬁl CF47 9DT, Wales UK; 3Department of Pathology and Laboratory Medicine (1122), King Fahad National Guard Hospital, PO Box 22490, Riyadh 11426, Kingdom of Saudi Arabia; 4University of Wales Institute Cardiff, Western Avenue, Cardiff CF5 2YB, Wales UK; 5Llandough Hospital, Penarth CF64 2XX, Wales, UK

Nut allergy is an important and potentially life threatening food allergy with a prevalence of one in 150 children in the UK population. STAT6 (signal transducer and activator of transcription) is an important molecule in the induction and regulation of an allergic response, which maps to chromosome 12q in a region previously linked with total serum IgE concentration and atopy in different populations. We have examined the frequency of a single nucleotide polymorphism (SNP) in the 3ЈUTR region of STAT6 gene in 71 UK Caucasoid patients diagnosed with nut allergy and 45 atopic patients without nut allergy using PCR-RFLP and compared these with 184 UK healthy controls. The STAT6 G allele frequency was signiﬁcantly increased in nut allergy patients compared with blood donor controls (P Ͻ 0.0001, OR = 2.9, 95% CI: 1.7–4.9), which was under a recessive model (GG vs GA+AA, P = 0.0001, OR = 3.2, 95% CI: 1.7–5.8) but not in atopic patients without nut allergy. The G allele was most frequent in the severe cases and GG homozygosity was associated with the increased risk of severe reaction (OR = 3.9, 95% CI: 1.9–8.3). We conclude that STAT6 3ЈUTR polymorphism is associated with susceptibility and severity in nut allergic patients in our population. Genes and Immunity (2002) 3, 220–224. doi:10.1038/sj.gene.6363872

Keywords: STAT6; polymorphism; nut allergy

Introduction pathway used by IL-4 and IL-13 in IgE isotype switching and production of TH2 cytokines.10–12 In STAT6-deficient Nut allergy is a common food allergy and usually has an mice it has been observed that most IL-4-mediated func- early onset in childhood. It is frequently persistent and tions are reduced.10 Furthermore a study by Miyata et al13 1 life threatening due to the risk of anaphylactic shock. in STAT6-deficient mice demonstrated an absence of Almost 8% of children experience food allergy in the first eosinophilia and lung damage as a consequence of 2 3 years of life. According to population studies in the aeroallergen challenge. STAT6 also regulates the pro- UK the prevalence of peanut and/or tree nut allergy has duction of TH2-type cytokines by both CD4+ and CD8+ 1,3 been estimated as being one in 150 children and is the T cells, further indicating an important role of STAT6 in 4 cause of most fatal and near-fatal allergic reactions. The allergic responses.13 persistent and severe nature of nut allergy and also the It is assumed that IL-13, which has similar biological difficulty in avoiding nuts or nut products in foods activities to IL-4, can also account for the development through accidental ingestion followed by a potentially of an allergic response in the absence of IL-4 in the same fatal reaction attracts attention to this type of allergy as pathway, using STAT6 activation.14 an important clinical problem. The STAT6 gene has been mapped to chromosome TH2 type cytokines especially IL-4 have a major role 12q.15 We have recently identified a single nucleotide 5 in the induction and triggering of the allergic response. polymorphism (SNP) in the 3ЈUTR region of the STAT6 Although environmental factors are likely to play an gene16 and reported allele and genotype frequencies of important role in the development and specificity of an this polymorphism in a healthy UK population. In this allergic response, genetic factors also have been shown study we examined the hypothesis that STAT6 gene poly- 6 to be important. A region on chromosome 12q has been morphism may contribute risk for the development of shown to be linked with total serum IgE concentration nut allergy. and atopy in several different populations.7–9 STAT6 is a central molecule in the signal transduction Results

Correspondence: Professor WER Ollier, The Centre for Integrated Gen- Patients characteristics omic Medical Research, Manchester University, Medical School, Stopford Building, Oxford Road, Manchester M13 9PT, UK. Table 1 shows characteristics of the nut allergic cohort E-mail: BillȰfs1.ser.man.ac.uk and the atopic volunteers. No such data are available for Received 10 December 2001; accepted 13 February 2002 the blood donor controls. The average age of the atopic STAT6 and nut allergy MM Amoli et al 221 Table 1 Characteristics of patients with nut allergy and atopic con- Table 3 STAT6 2964 G/A 3ЈUTR polymorphism allele and geno- trol groups type frequencies in nut allergy patients compared with atopic patients without nut allergy and healthy controls Controls Nut allergic Atopic patients Controls Nut allergy Atopic patients (n = 71) (n = 45) patients (without nut allergy) = = = Age (yrs) at the onset of atopy 3.2 ± 5.95 9.97 ± 6.78 (n 184) (n 71) (n 45) (mean ± s.d.) Age (yrs) at the onset of nut 9.7 ± 12.1 – GG 88 (48%)a,b 53 (75%)a 26 (58%)b allergy GA 78 (42%) 17 (24%) 16 (35%) Gender AA 18 (10%) 1 (1%) 3 (7%) Female 36 (51%) 30 (67%) G 254 (69%)c,d 123 (87%)c 68 (76%)d Male 35 (49%) 15 (33%) A 114 (31%) 19 (13%) 22 (24%) Family history of atopy n = 70 n = 42 Yes 52 (74%) 29 (69%) aP Ͻ 0.0001; bP = 0.4; cP Ͻ 0.0001 (OR 2.9, 95% CI: 1.7–4.9); dP = 0.2 No 18 (26%) 13 (31%) (OR 1.5, 95% CI: 0.7–2.8). Nut allergy severity Severe 46 (65%) – Moderate 18 (25%) – Mild 7 (10%) – mate of the true underlying relationship. The frequency of the G allele (76%) and GG homozygotes (58%) in atopic patients without nut allergy were increased compared with blood donor controls (69% and 48% controls was significantly higher than that of nut allergy respectively) but these differences did not achieve a level cases because of ethical issues relating to recruitment and of statistical significance. bleeding of children as age-matched controls. Most of the nut allergy patients had a history of severe symptoms of STAT6 genotype frequencies in severe nut allergy nut allergy. The percentage of patients who were skin cases prick test (SPT) and radio allergo-sorbent test (RAST) The nut allergy patients were divided to severe, moder- positive to different allergens are summarised in Table 2. ate and mild cases according to their reported reactions after the ingestion of nuts. The majority of cases in our STAT6 genotype frequencies in nut allergy patients cohort (65%) had previously experienced a severe reac- and control groups tion. The G allele was most frequent in the severe group The STAT6 G allele frequency was significantly increased (Table 4) and GG homozygosity was associated with the in the nut allergy patients compared with blood donor highest risk for a severe reaction (OR = 3.9, 95% CI: 1.9– controls (87% vs 69%, P Ͻ 0.0001, OR = 2.9, 95% CI: 1.7– 8.3). The STAT6 GG genotype was less associated with 4.9) (Table 3). STAT6 GG homozygosity was also signifi- mild/moderate reactions (P = 0.05, OR = 2.3, 95% CI: cantly increased in these patients compared with controls 1.0–5.5). using a recessive model (GG vs GA+AA, P = 0.0001, OR = 3.2, 95% CI: 1.7–5.8). Unfortunately no information Total and specific serum IgE level association with was available regarding the atopic status of these controls STAT6 polymorphism in different populations and previous studies suggest that the prevalence of There was no significant difference in mean total or spe- allergy to environmental allergens may be considerable.17 cific IgE level in patients with the GG genotype compared This suggest that using such controls would make it more with to those with other genotypes. When nut allergy and difficult to demonstrate a genetic association with atopy atopic patients were combined a small increase in median and any association found is likely to be an underesti- total serum IgE level was observed in patients carrying

Table 2 Percentage of nut allergy patients with a positive skin prack test (SPT) and radio allergo-sorbent test (RAST) to nuts and other allergens. Mean scores have been given for SPT and RAST tested for each antigen

SPT No. % Mean ± s.d. RAST No. % Mean ± s.d.

Peanut 71 79 4.6 ± 3.3 Peanut 69 71 2.6 ± 1.9 Brazil nut 71 58 3.2 ± 4.1 Hazelnut 69 48 1.1 ± 1.2 Walnut 71 54 2 ± 2.6 Brazil nut 69 48 1.2 ± 1.4 Hazelnut 71 49 1.9 ± 2.1 Almond 69 32 0.5 ± 0.8 Pecan nut 71 78 2 ± 2.1 Pecan nut 68 32 0.8 ± 1.3 Cashew nut 71 39 1.6 ± 2.5 Cashew nut 69 34 0.7 ± 1.1 Almond 71 65 2.4 ± 1.9 Walnut 68 40 0.7 ± 1.2 Mite 69 91 4.3 ± 2.3 Grass 70 86 4.8 ± 2.8 Cat 69 69 3.3 ± 2.5 Apple 57 19 0.4 ± 0.9 Soy 57 42 0.9 ± 1.3 Birch 59 37 1.1 ± 1.9

Scores have been given based on measurement of 2 mm greater than the positive control for SPT and 2 units above the standard unit for RAST test.

Genes and Immunity STAT6 and nut allergy MM Amoli et al 222 Table 4 Distribution of STAT6 2964 G/A 3ЈUTR polymorphism allergy there is little data relating clinical severity and IgE allele and genotype frequencies in nut allergy patients stratified level.19 In study by Clarke et al20 serum levels of IgE to by severity a 15 kDa protein has been shown to be significantly higher in peanut allergy patients experiencing severe Controls Nut allergy patients symptoms than those classified as having a moderate or mild reaction. In our study there was no obvious associ- Severe Mild and ation between the levels of total or specific serum IgE and moderate (2N = 368) (2N = 92) (2N = 50) STAT6 GG genotype in nut allergy patients. However when the data relating to total serum IgE levels were G 254 (69%) 82 (89%) 41 (82%) pooled for nut allergy and atopic patients, a small but A 114 (31%) 10 (11%) 9 (18%) not significant increase in level was observed in the GG GG 88 (48%)a,b 36 (78%)a 17 (68%)b homozygotes. Given the normal variability that exists Others 96 (52%) 10 (22%) 8 (32%) day to day in atopic patients it is likely that a much larger + (GA AA) group of patients will need to be examined before a categorical conclusion can be drawn. a = b = P 0.0002 (OR 3.9, 95% CI: 1.9–8.3); P 0.05 (OR 2.3, 95% CI: Although the effect of STAT6 polymorphism appears 1.0–5.5). to be highly significant (P Ͻ 0.0001, for the allele frequency), it is unlikely to account for the major genetic the GG genotype compared to patients carrying either component to this condition. However this association is GA or AA genotype but this did not reach a level of sig- compatible with nut allergy having a complex aetiology nificance using the Mann-Whitney test (262 vs 196 units where both environmental and oligogenic factors contrib- respectively, P = 0.4). ute. It is likely that the aetiology of nut allergy has several discrete components to its development and that genetic Discussion factors could contribute to each stage. An obvious component comprises the specificity of the initiating In this study we have examined whether genetic poly- response to the allergens. Dorion and Leung21 have pro- ␤ morphism in the STAT6 gene is associated with nut posed that restricted use of the V 2 T cell receptor (TCR) allergy. Our results suggest that presence of the STAT6 may account, for the severity of the immune response to 3ЈUTR G allele is significantly associated with increased peanut allergens. Moffatt et al22 have also described an risk of nut allergy. association between T-cell receptor ␣/␦ complex and spe- The GG homozygous genotype conveys the highest cific IgE responses to the dust mite allergen but no link- risk and this suggests that the G allele may act in an addi- age was found with the ␤ chain. tive or recessive way. As yet it is not known whether the Linkage and association studies of atopy and allergy STAT6 polymorphism described is of functional conse- have already identified a number of chromosomal quence and whether it directly effects gene expression. areas.6,23 Genes, which are important in the specificity Previous studies have demonstrated that polymorphisms and recognition of allergens, represent good candidates within the 3ЈUTR region of genes can influence mRNA for susceptibility factors. This is supported by markers stability and in consequence effect the level of protein within the HLA region on chromosome 6 and within the production.18 It is not possible yet to conclude that the TCR gene complex on chromosome 14q11, which show STAT6 polymorphisms described is directly responsible significant linkage with atopy.6,23 Previous studies have for the association of this gene with nut allergy. The provided evidence for association between susceptibility association of STAT6 with nut allergy could be due to to allergy (with or without asthma-related symptoms), linkage disequilibrium of this SNP with another poly- and HLA-DR4 and/or the DR7 allele.24 Howell et al25 morphism in the coding region or promoter region of the have also demonstrated significant increase in HLA- STAT6 gene or another gene in close physical proximity. DRB1*08, and the DQB1*04 allele in peanut allergy Although a significant association was observed patients. between STAT6 polymorphism and nut allergy it is A region in chromosome 5, which contains a cluster of unlikely that this represents a specific association with cytokines, IL-4, IL-13, IL-9 and GM-CSF, has also been this condition. Indeed although non-significant increase considered as being an important locus in regulating the in the STAT6 G allele frequency was observed in the allergic response. A divergent association between SNP’s atopic patients without nut allergy, this was at an inter- in promoter region of the genes within this region has mediate level between the nut allergy group and blood been observed in different atopic patient populations.26 donor controls. This is likely to be explained by the A marker has been found on chromosome 11q13 which STAT6 G allele being associated with severity of allergy. has been strongly linked with atopy and total serum IgE A large proportion of the nut allergy cases had either a level and shows a maternal effect. The association history of severe reaction or anaphylaxis. In contrast between the Fc⑀R1 polymorphisms in this region with atopic patients without nut allergy had been identified atopy was also divergent in patients from different eth- from healthy volunteers who exhibited asthma, eczema nic groups.27 or hay fever. Consequently the majority of these did not Environmental factors are another major component in exhibit severe atopic reaction. the development of atopy in different populations. A STAT6 maps to a region on chromosome 12q, which large increase in the prevalence of atopic disease in the has been previously linked with total serum IgE concen- western countries is striking and a rural lifestyle has been tration and atopy in several different populations.7–9 IgE shown to be protective both in Europe and Africa.28 It is antibodies are largely responsible for triggering type 1 assumed that the differences in the nature of childhood immediate-hypersensitivity reactions. In studies of food infection or bacterial flora of the bowel are responsible

Genes and Immunity STAT6 and nut allergy MM Amoli et al 223 for these patterns.28 The specificity of antigen also seems References to be important in the development and also severity of clinical reaction. Several proteins in the nut kernel have 1 Sicherer SH, Sampson HA. Peanut and tree nut allergy. Curr Opin Pediatr 2000; 12: 567–573. been identified as being allergenic. These include the 2 Bock SA. Prospective appraisal of complaints of adverse reac- major storage proteins arachin and conarachin, peanut tions to foods in children during the first 3 years of life. Pediatrics lectin and phospholipase D, the acidic glycoprotein, pea- 1987; 79: 683–688. nut 1 and three major allergens, which have been charac- 3 Emmett SE, Angus FJ, Fry JS, Lee PN. Perceived prevalence of terised and sequenced as Ara h1, 2 and 3.20 peanut allergy in Great Britain and its association with other Current in vitro tests cannot distinguish between atopic conditions and with peanut allergy. Allergy 1999; 54: patients experiencing mild or severe nut allergic reac- 380–385. tions, or predict the likely event of a likely further severe 4 Sampson HA, Mendelson L, Rosen JP. Fatal and near-fatal ana- reaction. Given that the STAT6 homozygous GG geno- phylactic reactions to food in children and adolescents. N Engl 327 type is associated with a more severe phenotype it may JMed1992; : 380–384. 5 Walley AJ, Cookson WO. Investigation of an interleukin-4 pro- be possible that such information, when combined with moter polymorphism for associations with asthma and atopy. J other genetic and laboratory data, could help predict the Med Genet 1996; 33: 689–692. risk of subsequent severe reactions. Such an approach 6 Barnes KC, Marsh DG. The genetics and complexity of allergy could be of clinical use. and asthma. Immunol Today 1998; 19: 325–332. 7 Barnes KC Neely JD, Duffy DL et al. Linkage of asthma and total serum IgE concentration to markers on chromosome 12q: Methods evidence from Afro-Caribbean and Caucasian populations. Gen- omics 1996; 37:41–50. Population studied 8 Nickel R, Wahn U, Hizawa N, Maestri N, Duffy DL, Barnes KC. Seventy-one unrelated Caucasians nut allergic patients Evidence for linkage of chromosome 12q15-q24.1 markers to were recruited for this study from the Allergy Clinic at high total serum IgE concentrations in children of the German Llandough Hospital, Cardiff. All patients who had had Multicenter Allergy Study. Genomics 1997; 46: 159–162. 9 Thomas NS, Wilkinson J, Holgate ST. The candidate region positive SPT to peanut, cashew, brazil nut, pecan, approach to the genetics of asthma and allergy. Am J Respir Crit almond, hazelnut or walnut since 1993 were asked to Care Med 1997; 156: S144–S51. attend. Out of 83 patients recruited, a total of 12 were 10 Takeda K, Tanaka T, Shi W et al. Essential role of Stat6 in IL-4 excluded: five because they were related to nut allergic signalling. Nature 1996; 380: 627–630. patients already in the study and seven because they 11 Kaplan MH, Schindler U, Smiley ST, Grusby MJ. Stat6 is gave a poor history of nut allergy despite being SPT posi- required for mediating responses to IL-4 and for development tive. The remaining 71 (mean age: 10 ± 12.1 years at of Th2 cells. Immunity 1996; 4: 313–319. onset) were analysed in the study. Forty-five healthy vol- 12 Takeda K, Kamanaka M, Tanaka T, Kishimoto T, Akira S. unteers from hospital and university staff with a history Impaired IL-13-mediated functions of macrophages in STAT6- deficient mice. J Immunol 1996; 157: 3220–3222. of asthma, eczema or hay fever were recruited as atopic 13 Miyata S, Matsuyama T, Kodama T et al. STAT6 deficiency in a controls. All were SPT negative to nuts and SPT positive mouse model of allergen-induced airways inflammation abol- for one of the common allergens (house dust mite, cat ishes eosinophilia but induces infiltration of CD8+ T cells. Clin dander, grass pollen). A total of 184 UK Caucasoid blood Exp Allergy 1999; 29: 114–123. donors were also available for comparison. No infor- 14 Zurawski G, de Vries JE. Interleukin 13, an interleukin 4-like mation regarding atopic status was available on this cytokine that acts on monocytes and B cells, but not on T cells. group. Ethical approval was obtained for this study and Immunol Today 1994; 15:19–26. all cases and controls gave informed consent. 15 Leek JP, Hamlin PJ, Bell SM, Lench NJ. Assignment of the STAT6 gene (STAT6) to human chromosome band 12q13 by in situ hybridization. Cytogenet Cell Genet 1997; 79: 208–209. Symptom severity 16 Amoli M, Ollier WE, Hajeer AH. A novel PCR-RFLP assay for Nut allergy symptoms were graded using a modification the detection of a polymorphism in the 3Ј of STAT6 gene. Genes 20 of the system of Clarke et al. These were as follows: mild Immun 2000; 1: 349–350. (vomiting, abdominal pain, irritability, pruritus and/or 17 Huss K, Naumann PL, Mason PJ et al. Asthma severity, atopic urticaria), moderate (facial oedema including lip and status, allergen exposure and quality of life in elderly persons. mouth swelling), severe (laryngeal oedema, cyanosis, Ann Allergy Asthma Immunol 2001; 86: 524–530. wheeze and collapse or anaphylaxis). 18 Jacob CO, Lee SK, Strassmann G. Mutational analysis of TNF- alpha gene reveals a regulatory role for the 3Ј-untranslated Genotyping of the STAT6 polymorphism region in the genetic predisposition to lupus-like autoimmune disease. J Immunol 1996; 156: 3043–3050. The genotyping of the samples were carried out using 19 Hourihane JO, Dean TP, Warner JO. Peanut allergic subjects’ 16 the method has been described previously. peripheral blood mononuclear cell proliferative responses to crude peanut protein. Clin Exp Allergy 1998; 28: 163–168. Statistical analysis 20 Clarke MC, Kilburn SA, Hourihane JO, Dean KR, Warner JO, Continuous data were described as median, mean, and Dean TP. Serological characteristics of peanut allergy. Clin Exp standard deviation (mean ± s.d.), and categorical vari- Allergy 1998; 28: 1251–1257. ables as percentages. Strength of association was esti- 21 Dorion BJ, Leung DY. Selective expansion of T cells expressing mated using odds ratios (OR) and by calculating 95% V beta 2 in peanut allergy. Pediatr Allergy Immunol 1995; 6:95–97. 22 Moffatt MF, Hill MR, Cornelis F et al. Genetic linkage of T-cell confidence intervals (CI). Levels of significance were receptor alpha/delta complex to specific IgE responses. Lancet determined by chi-square analysis. The Mann-Whitney 1994; 343: 1597–1600. test was used to assess phenotypic differences between 23 Holgate ST. Asthma genetics: waiting to exhale. Nat Genet 1997; carriers of different genotypes. Analyses were carried out 15: 227–229. using the STATAV6 software package and SPSS v9. 24 Senechal H, Geny S, Desvaux FX et al. Genetics and specific

Genes and Immunity STAT6 and nut allergy MM Amoli et al 224 immune response in allergy to birch pollen and food: evidence 26 Rosenwasser LJ, Klemm DJ, Dresback JK et al. Promoter poly- of a strong, positive association between atopy and the HLA morphisms in the chromosome 5 gene cluster in asthma and class II allele HLA-DR7. J Allergy Clin Immunol 1999; 104: 395– atopy. Clin Exp Allergy 1995; 25 (Suppl 2): 74–78 401. 27 Shirakawa T, Li A, Dubowitz M et al. Association between atopy 25 Howell WM, Turner SJ, Hourihane JO, Dean TP, Warner JO. and variants of the beta subunit of the high-afﬁnity immuno- HLA class II DRB1, DQB1 and DPB1 genotypic associations with globulin E receptor. Nat Genet 1994; 7: 125–129. peanut allergy: evidence from a family-based and case-control 28 Moffatt MF, Cookson WO. Genetics of asthma and inﬂam- study. Clin Exp Allergy 1998; 28: 156–162. mation: the status. Curr Opin Immunol 1999; 11: 606–609.

Genes and Immunity