Editor—Atopy is characterised by raised IgE levels and it underlies the clinical disorder bronchial asthma.1-3 IL-4 and IL-13 play a key role in the development of asthma acting through the IL-4 receptor (IL-4R) and IL-13R, respectively.1-3 Since these two receptors share the IL-4Rα chain, this subunit is the crucial component required for IL-4 and IL-13 signalling.1-3 Recently, three coding variants in the human IL4R gene have been identified and these showed genetic and functional association with atopic asthma in different ethnic groups. The Ile50 variant was associated with atopic asthma in a Japanese population4 5and Arg551 or Pro478 were associated with IgE levels in American6 and German populations,7respectively, but perhaps through different mechanisms. Ile50 specifically and strongly upregulates cellular IgE synthesis through stat6 when tested by cDNA transfection into human and murine B cell lines.4 5 Functional assay of Arg551Gln IL-4Rα showed impaired binding of the negative regulatory molecule, protein tyrosine phosphatase, SHP-1, and increased expression of CD23 was found in peripheral blood mononuclear cells after challenge with human IL-4.6 However, this finding was not replicated by another group.8 In contrast, Pro478 is in tight linkage disequilibrium with Arg551 in a German population and may change the structure of the receptor, leading to altered phosphorylation patterns of signal molecules and hence lower IgE levels.7 These findings emphasise the genetic heterogeneity of atopy and question whether the molecular interaction between variants of IL-4Rα and its associated signal transduction molecules, such as stat6, may be important.1-3

The Stats are a family of transcription factors evolutionarily conserved from Drosophila to humans.9 To date seven different STAT genes have been identified in humans.9 Stat6 activation correlates with functional responses induced by IL-4 and IL-13.1-3 Mice deficient in Stat6 lack IgE production and Th2 inflammatory reactions.10 11 Stat6 is therefore an essential molecule for IL-4 and IL-13 signal transduction. Moreover, it is assigned to chromosome 12q13-14, the site of genetic linkage to atopic asthma.

The aim of this study was to test whether variants of theSTAT6 gene relate to atopic asthma in both British (n=300) and Japanese (n=400) populations.

For the Japanese study, 100 adults from Osaka were randomly selected as controls from clients of a commercial medical examination company, with adjustment for the sex and age distribution of this area.12 13 We recruited 400 patients, 100 in each group, with atopic asthma, intrinsic (non-atopic) asthma, atopic rhinitis, and atopic eczema from Kawai Clinic, Wakayama Red Cross Medical Centre, and Habikino Hospital, respectively. We selected patients with one principal atopic clinical disorder at the time of examination.

For the British study, 150 subjects were randomly recruited from an obstetric outpatient clinic in Oxfordshire. One hundred and fifty subjects with physician diagnosed atopic asthma were selected from the Oxford Chest Unit; both controls and patients were white.

All the asthmatic subjects had specialist physician diagnosed asthma with (1) recurrent breathlessness and chest tightness requiring continuing treatment, (2) physician documented wheeze, and (3) documented labile airflow obstruction with variability in serial peak expiratory flow rates >30%. Marked asthma was designated as chronic rather than episodic asthma, needing chronic therapy with steroids. The diagnosis of rhinitis was based on chronic nasal obstruction or congestion, rhinorrhoea, and sneezing, not caused by chronic infection. The diagnosis of atopic eczema was made on the basis of the morphological appearance of active skin disease, the distribution of skin lesions, the clinical course of the disorder, and a family history of atopy. All cases of eczema met the diagnostic criteria of Hanafin and Rajka.14 There were no heavy smokers (>20 cigarettes per day) among these subjects.

Specific IgE against 15 airborne antigens was detected by solid phase immunoassay, MAST (Hitachi, Tokyo, Japan), and we used previously described criteria for a positive titre. A high concentration of total serum IgE (Pharmacia CAP system, Uppsala, Sweden) was taken to be greater than the mean +2 SD.13 14 Atopy, defined as IgE responsiveness, was diagnosed as the presence of a high concentration of total serum IgE, a positive specific IgE titre (>0.35 kU/l) against one of 15 highly purified aeroallergens, or a combination of these two features.13 14

DNA samples were extracted using a commercial kit (IsoQuick, Microprobe Corporation, Garden Grove, USA). cDNA was extracted using commercial kits (Pharmacia, Sweden). To detect variants, cDNA for the Stat6 coding region was amplified with primers 5′GAA GAC AGC AGA GGG GTT GC and 5′CTG TGG GGG TAGT AGA AGA G.15 Sequencing was conducted with the big-dye system (ABI, UK) and images were visualised in the commercialised POP-6 gel using the automated sequencer (ABI Prism 310 Genetic Analyser). Primers were 5′TGT CTC TGC CCC TGG TGG TC or 5′GGT CCC CAC CTC AGC CAT G.15

To conduct the genetic association study, PCR including 1.5 mmol/l of magnesium chloride and genomic DNA in a 30 μl mixture was performed in a Perkin Elmer Cetus thermal cycler. Primers were 5′GGG AAG TTC AGG CTC TGA GAC CC and 5′AGC TCT GTA TGT GTG TGT G.15 The underlined sequence was exchanged to incorporate the restriction site. The amplification products were digested withAciI.

Contingency table analysis, odds ratios, 95% confidence intervals, and significance values were estimated by computerised exact methods (SPSS program version 8). Odds ratio (OR) was calculated between high and low risk genotypes and between high and low risk alleles. If the number in the column was less than 10, Fisher's exact test was used. Probability values from this test were corrected for multiple comparisons by multiplying the p value by the number of comparisons that were made (Bonferroni correction).

Mutation scanning enabled us to identify a G2964A variant in a 3′ untranslated region of STAT6.5As shown in tables 1 and 2, the allele frequencies forSTAT6 in the two control populations were quite different, p(G)=0.76, p(A)=0.24 in the British population and p(G)=0.33 and p(A)=0.67 in the Japanese population (χ²=82.2, df=2, p<0.000001). There was no significant association between atopic asthma and thisSTAT6 variant in the British population; no association was seen with any atopy or asthma phenotype. In the Japanese population, mild atopic asthma characterised by positive antigen specific IgE response (ASE) (<0.75 kU/l) or high total IgE (mean + between 1 SD and 2 SD) showed a strong association with this variant (OR=3.19 (95% CI: 1.40-7.08), p=0.0043, Pc=0.016). Neither adult nor childhood asthma with a marked atopy phenotype (ASE>3.5 kU/l and total IgE >mean +2 SD) was associated with this variant ofSTAT6. Again, no atopy phenotypes were associated with this variant.

Recent genetic and functional studies highlight the importance of variants of IL-4Rα in the development of atopy; Ile50 upregulates cellular IgE synthesis and is associated with atopic asthma in a Japanese population,4 5 while Pro478 alters the binding affinity of IRS-1/2, and is associated with lower IgE levels.7 No association was found between atopic asthma and these variants in a British population.5 These findings emphasise the genetic heterogeneity of atopic disorders, even within one locus in different ethnic groups; they raise the question of interaction between variants of different components of the IL-4/IL-13 signalling pathway. BCL6 might repress stat6 activated transcription16 17 and a variant ofBCL6 has been associated with marked atopy.18 Interactions among signal transduction molecules of IL-4Rα might explain the complex relationships between genetic variants and phenotypes in atopy and asthma.1-3

The gene encoding humanSTAT6, spanning 19 kb on 12q13-14, consists of 23 exons.19 Genome wide searches for atopic asthma showed a strong linkage to 12q14-24.20-22 Our present data, showing association between variants ofSTAT6 and asthma with mild atopy, support the candidacy of STAT6 as an “asthma” locus on 12q in a Japanese population. Also, the marked difference in genotype frequencies between the British and Japanese populations may explain the “genetic heterogeneity” among and within ethnic groups.1-3 Since G2964A variant is located in 3′UTR of the gene, the functional role of this variant remains unknown; it may be in linkage disequilibrium with so far unidentified but functional variants in the regulating or coding parts ofSTAT6 or variants of the immediately adjacent genes. Further studies are needed to clarify the role of theSTAT6 variants in predisposing to asthma among different ethnic groups.