IL-10 promoter and IL4-Rα gene SNPs are associated with immediate β-lactam allergy in atopic women

Authors


Dr Laurence Guglielmi
INSERM U454
CHU Arnaud de Villeneuve
371 avenue du doyen Gaston Giraud
34295 Montpellier cedex 5
France

Abstract

Background:  Allergic reactions to β-lactam antibiotics represent the most frequent cause of immunological drug reactions.

Objective:  This study evaluates the involvement of genetic susceptibility factors in patients with immediate allergic reactions to β-lactams. We examined 15 single nucleotide polymorphisms (SNP) of genes coding proteins implicated in immunoglobulin (Ig)E synthesis regulation.

Methods:  We performed a case–control study involving 44 patients with immediate β-lactam allergy and 44 control subjects, all matched for sex and atopy. Interleukin (IL)-4, IL-13, IL-4Rα, signal transducer and activator of transcription 6 (STAT6), interferon (IFN)-γR1, IFN-γR2 and FcɛRIβ gene polymorphisms were determined using polymerase chain reaction (PCR) restriction fragment length polymorphism, and IL-21R gene and IL-10 promoter polymorphisms by direct sequencing.

Results:  Our analysis did not reveal differences in the distribution of the 15 SNPs between allergic patients and controls. However, among atopic subjects, we found two distinct significant associations between immediate β-lactam allergy in women and the Ile75Val variant of IL-4Rα gene (P = 0.012, OR = 5.4, CI: 1.16–27.7), and two linked IL-10 promoter gene polymorphisms, −819C>T and −592 C>A (P = 0.023, OR = 17.5, CI: 1.26–533.07). In contrast, we observed no association in allergic male subjects in the atopic population. Interestingly, the IL-4Rα Ile75Val variant could have a paradoxal protective effect in atopic male patients (P = 0.004, OR = 0.07, CI: 0.01–0.66).

Conclusion:  Our findings suggest that polymorphisms in the IL-10 promoter and IL-4Rα genes are genetic factors that favour β-lactam immediate allergies in female patients with atopy.

Abbreviations:
SNP

single nucleotide polymorphism

OR

odd ratio

IL

interleukin

STAT

signal transducer and activator of transcription

β-Lactams antibiotics are the most frequent inducers of adverse drug reactions that are mainly related to specific immunological mechanisms (1). Immunoglobulin (Ig)E-mediated allergic reactions to β-lactams are responsible for immediate anaphylactic reactions (anaphylactic shocks, urticaria, angio-oedema or bronchospasm), considered the most serious for patients.

The molecular mechanisms underlying the production of IgE, as well as IgG and IgA isotypes, require two signals (2). One signal is delivered by cytokines that target the specific C heavy chain gene for switch recombination by causing its germline transcription. The other signal involves the interaction of the CD40 antigen on B cells with the CD40 ligand on T cells. Interleukin (IL)-4 is an essential cytokine for IgE class switch and IL-4 and IL-13 cooperate to increase IgE production. Both of these two TH2 cytokines use the α-chain of the IL-4 receptor (IL-4Rα) as a part of their respective receptor systems, functioning through the activation of signal transducer and activator of transcription 6 (STAT6; 3). Therefore, IL-4Rα has a pivotal role in both IL-4 and IL-13 signalling. On the other hand, interferon (IFN)-γ, a major cytokine from TH1 cells, inhibits the IgE production induced by IL-4 (4) and the TH2 polarization of immune responses. The IFN-γ exerts its various biological effects by binding to its receptor consisting of heterodimeric chains, IFN-γR1 (α-chain) and IFN-γR2 (β-chain).

Many allelic variants of cytokine and cytokine receptor genes have been reported, resulting of single nucleotide polymorphisms (SNPs). In this study, we analyse cytokine gene SNPs that have been associated previously with asthma, atopy or high IgE serum level in the context of β-lactam immediate reactions: IL-4 −589 C>T and −34 C>T (5). Interleukin-13 −1512 C>T, −1112 C>T and Arg130Gln (6), IL-4Rα Ile75Val and Gln576Arg (7), IFN-γR1 Val14Met and IFN-γR2 Gln64Arg (8). The STAT6 polymorphism located in intron 18 (C>T) and associated in one study (P = 0.007; 9) but not in another (10) with high serum IgE phenotypes, was also studied.

Another good candidate to investigate is the gene coding for IL-10. This cytokine displays pleiotropic effects in immunoregulation and inflammation. Interleukin-10 inhibits both TH1 production of IFN-γ and IL-2, and IL-4 and IL-5 expression by TH2 lymphocytes (11). Interleukin-10 is also important in suppressing the immune response by inhibiting proinflammatory cells. It has been reported that three SNPs in the IL-10 gene promoter, namely IL-10 −1082 G>A, −819 C>T and −592 C>A, may influence IL-10 production levels in vitro (12, 13).

Interleukin-21 is a cytokine that has been shown, in different experimental mouse models, to modulate the production of IgE (14, 15). In human as well, IL-21 induces the IL-4-driven IgE synthesis by mitogen-stimulated peripheral blood mononuclear cell (PBMC; 16). Moreover, an association was found between a SNP in the putative promoter region and the presence of elevated IgE serum levels, suggesting a regulatory function of IL-21 in the production of this isotype. We were therefore interested in looking at a SNP in the IL-21R gene, −83 T>C, which has been associated with an elevated serum total IgE level (>100 IU/ml) in Caucasian women but not in men (17).

Recently, Qiao et al. reported, in a Chinese population, an association between the Glu237Gly variant of the FcɛRIβ gene and allergic reactions to benzylpenicillanyl (BPA), phenoxomethylpenicilloyl (PVO) and ampiciolloyl (APO) antigens and with an increase of BPA-, PVO- or APO-IgE antibodies in patients with positive specific IgE (18). They also showed that the frequency of the IL-4Rα 576Gln allele was significantly increased in patients with penicillin allergy when compared with control subjects and that this allele was strongly associated with the expression of some penicillin-specific IgE (19).

The aim of this study was to determine whether one or more of the 15 polymorphisms listed above contribute to any susceptibility to immediate β-lactam allergy in a Caucasian population. We report herein that the frequencies of IL-10-819CT/TT and IL-10-592CA/AA genotypes or IL-4Rα 75Ile allele are associated with immediate β-lactam allergy in atopic women only.

Methods

Subjects

Allergic and control subjects were recruited from the outpatient department for drug allergy at the University Hospital of Montpellier (France). Between September 1996 and December 2004, we included all the patients who had consulted with a clinical history suggestive of a β-lactam allergy. We excluded patients having experienced severe life-threatening skin reactions or drug-induced autoimmune disease. Every patient underwent diagnostic procedures including a standardized questionnaire and recordings of skin tests to β-lactams and provocation tests, according to internationally validated procedures as previously described (1). Briefly, skin tests were performed with the major and minor determinants of penicillin PPL and MDM (Allergopharma, Merck, Darmstadt, Germany), penicillin G, amoxicillin, ampicillin and any other β-lactam suspected from the patient's history when the injectable form was available. About 15–20 min following skin prick tests, a wheal larger than 3 mm accompanied by erythema was considered positive in the absence of response to saline control. After prior testing with a negative control reagent, intradermal reactions to potential allergens were considered positive when a change in the size of the initial wheal of 3 mm greater in diameter, associated with a flare, was observed within 20–30 min. A late reading in patients with an unknown chronology or a nonimmediate reaction was made after 24–48 h. Drug provocation tests with the suspected β-lactams were carried out only when skin tests were negative and under strict hospital surveillance. In addition, every patient underwent skin prick tests to common aeroallergens. Subjects presenting one or more positive skin prick test are included in the atopic population.

In the present study, we selected all patients with an immediate reaction and randomly chose 44 with β-lactam demonstrated allergy and matched them for sex and atopy to 44 controls, defined as patients exposed to β-lactams who had no positive skin tests, or positive provocation test for β-lactams (Table 1). All selected women were aged <45 years when they experienced their initial allergic reaction, to take into account a possible influence of hormonal factors.

Table 1.   Characteristics of allergic and control group
 Allergics (n = 44)Controls (n = 44)
Women2222
 Atopics/nonatopics11/1111/11
 Age (years)24.2 ± 10.530 ± 10.7
Men2222
 Atopics/nonatopics11/1111/11
 Age (years)34.1 ± 19.534.3 ± 15.6

The ethics committee of Montpellier Hospitals authorized this study and all subjects gave their written informed consent.

Genotyping

DNA was isolated from peripheral blood by standard phenol-chloroform extraction followed by ethanol precipitation. All polymorphisms, except those of IL-10 and IL-21R genes, were analysed by polymerase chain reaction (PCR), restriction fragment length polymorphism (RFLP). Briefly, PCR fragments including the polymorphism were amplified using specific primers detailed in Table 2. Polymerase chain reaction (PCR) were carried out in a total volume of 50 μl containing 100 ng DNA, 0.5 μM of each primer, 0.2 mM of each dNTP, 1.5 mM MgCl2, 5 μl 10X buffer and 1 U Taq DNA polymerase (Invitrogen, Cergy Pontoise, France). Samples were denatured at 94°C for 10 min followed by 32 cycles at 94°C for 45 s, hybridization temperature for 30 s, 72°C for 60 s and a final 10 min extension at 72°C. Fifteen microlitres of each PCR product were digested with a specific restriction enzyme (Table 2). Digested fragments were visualized after agarose gel electrophoresis.

Table 2.   Primer sequences, PCR condition and restriction enzyme PCR–RFLP
PolymorphismsPrimers (5′→3′)Annealing temperature (°C)Restriction enzymeReference
  1. STAT, signal transducer and activator of transcription; PCR, polymerase chain reaction; RFLP, restriction fragment length polymorphism; IFN, interferon.

STAT6 (in18C>T)ACCTCCCTTCCTCCCTAACC
TAGCCCATCTGCACAGACC
55PstI 
IL-13
 Arg130GlnCTTCCGTGAGGACTGAATGAGACGGTC
GCAAATAATGATGCTTTCGAAGTTTCAGTGGA
55NlaIV(6)
 −1512C>ACAACCGCCGCGCCAGCGCCTTCTC
CCGCTACTTGGCCGTGTGACCGC
60BstUI(6)
 −1112C>TGGAATCCAGCATGCCTTGTGAGG
GTCGCCTTTTCCTGCTCTTCCCGC
54BstUI(6)
IL-4RA
 Ile75ValGGAAGAGTCTGATGCGGTTCC
CAGCCCACAGGTCCAGTGTATAG
55MslI(27)
 Gln576ArgGCCCCCACCAGTGGCTACC
GAGGTCTTGGAAAGGCTTATAC
55HpaII(19)
IL-4
 −589C>TACCTAAACTTGGGAGAACATGGT
TGGGGAAAGATAGAGTAATA
51AvaII(19)
 −34C>TATTTTCCCTCGGTTTCAGCAA
TTGTGTCCGTGGACAAAGTTG
55MnlI 
IFNGR1 (Val14Met)GACGGAAGTGACGTAAGG
CCATCTCAGCCCTGGTCA
55TspRI(8)
IFNGR2 (Gln64Arg)CAGCTGCCCGCTCCTCAG
GGCTTACTATTTAAACTGGACT
55HinfI(8)
FcɛRIβ (Glu237Gly)TATCAGGTTCCAGAGGATCGTGTT
GATTCTTATAAATCAATGGGAGGAAAC
55XmnI(18)

The IL-10 and IL-21R SNPs were determined by sequencing. A 521-bp PCR fragment including the IL-21R −83T>C polymorphism was amplified using the primers: 5′-GCC TGC TGC ATC TAG TGT C-3′ (upstream) and 5′-CCG TGC TTC ATG AGA AAG A-3′ (downstream). A 588-bp PCR fragment including all three IL-10 SNPs was amplified using the primers: 5′-ATC CAA GAC AAC ACT ACT AA-3′ (upstream) and 5′-TAA ATA TCC TCA AAG TTC C-3′ (downstream). The fragments were sequenced using the ABI Prism Big Dye Terminator DNA Sequencing Kit (Applied Biosystems, Courtaboeuf, France) according to the manufacturer's instructions, and analysed on an ABI Prism 310 Genetic Analyser (Perkin-Elmer, Courtaboeuf, France). The IL-21R and IL-10 sequencing primers are: 5′-CGT GCT TCA TGA GAA AGA TG-3′ and 5′-CTC AAA GTT CCC AAG CAG CCC T-3′ respectively.

Statistical analysis

The chi-squared test or Fisher's exact test were used to test for significant association between β-lactam allergy or atopy, and alleles or genotypes distribution. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. P-values below 0.05 were considered statistically significant.

Results

Our population consisted of 44 allergic patients and 44 control subjects. Among each group, half subjects were women, half were men (Table 1). We first looked for an association with β-lactam immediate allergy, then with atopy. Further analysis, combining allergy, atopy and gender, were performed to reveal a statistical association.

Cytokine gene polymorphisms in patients with immediate β-lactam allergy and in control subjects

We first compared the prevalence of 15 SNPs in candidate genes in the demonstrated immediate β-lactam allergic and the control population. The genotype distributions and allele frequencies are shown in Table 3. We found no significant genotype or allele association with immediate allergic reactions for all the SNPs studied. Furthermore, none of the polymorphisms was associated with gender (data not shown).

Table 3.   Genotype distribution and allele frequency of the 15 SNPs in 44 patients with immediate β-lactam allergy and 44 control subjects
PolymorphismsGenotypeGenotype frequency (%)P-valueAlleleAllele frequencyP-value
AllergicsControlsAllergicsControls
  1. *Genotype of all control subjects could not be determined (n < 44).

  2. STAT, signal transducer and activator of transcription; IFN, interferon; IL, interleukin; SNP, single nucleotide polymorphism.

IL-4 (−589)CC31 (74.4)32 (72.7)0.81C0.840.860.67
CT12 (27.3)12 (27.3)T0.160.14
TT1 (2.3)0 (0)    
IL-4 (−34)CC32 (72.7)35 (79.5)0.45C0.850.900.36
CT11 (25.0)9 (20.5)T0.150.10
TT1 (2.3)0 (0)    
IL-13 (Arg130Gln)GG24 (54.4)29 (65.9)0.27G0.750.820.27
GA18 (41.0)14 (31.8)A0.250.18
AA2 (13.6)1 (2.3)    
IL-13 (−1512)CC20 (45.4)26 (59.1)0.20C0.660.740.25
CA18 (41.0)13 (29.5)A0.340.26
AA6 (13.6)5 (11.4)    
IL-13 (−1112*)CC22 (50.0)28 (65.1)0.15C0.720.800.18
CT19 (43.2)13 (30.2)T0.280.20
TT3 (3.8)2 (4.7)    
IL-4RA (Ile75Val)GG11 (25)16 (36.4)0.24G0.550.600.44
GA26 (59.1)21 (47.7)A0.450.40
AA7 (15.9)7 (15.9)    
IL-4RA (Gln576Arg)*AA29 (65.9)33 (78.6)0.19A0.830.870.47
AG15 (34.1)7 (16.6)G0.170.13
GG0 (0)2 (4.8)    
IL-10 (−1082*)GG10 (22.7)5 (12.2)0.20G0.450.330.09
GA20 (45.4)17 (41.5)A0.550.67
AA14 (31.9)19 (46.3)    
IL-10 (−819)CC22 (50.0)24 (54.5)0.67C0.680.720.62
CT16 (36.4)15 (34.1)T0.320.28
TT6 (13.6)5 (11.4)    
IL-10 (−592)CC22 (50.0)24 (54.5)0.67C0.680.720.62
CA16 (36.4)15 (34.1)A0.320.28
AA6 (13.6)5 (11.4)    
IL-21R (−83)TT22 (50.0)23 (52.3)0.83T0.700.740.46
CT18 (41.0)19 (43.2)C0.300.26
CC4 (9.0)2 (4.5)    
STAT6 (in 18)CC36 (81.8)31 (70.4)0.21C0.910.840.17
CT8 (18.2)12 (27.3)T0.090.16
TT0 (0)1 (2.3)    
FcɛRIβ (Glu237Gly)AA40 (91.0)39 (88.6)1A0.950.940.73
AG4 (9.0)5 (11.4)G0.050.06
GG0 (0)0 (0)    
IFNGR1 (Val14Met)AA8 (18.2)10 (22.8)0.59A0.350.460.12
AG15 (34.1)21 (47.7)G0.650.54
GG21 (47.7)13 (29.5)    
IFNGR2 (Gln64Arg)AA36 (81.8)33 (75.0)0.43A0.900.870.63
AG7 (15.9)11 (25.0)G0.100.13
GG1 (2.3)0 (0)    

We then investigated the haplotypes formed by the combinations of the three IL-10 promoter gene SNPs (−1082 G>A, −819 C>T and −592 C>A) already demonstrated by others (12) and confirmed in our study. As expected, only three haplotypes were found: GCC, ACC and ATA. The fourth haplotype, GTA, has been reported yet only in a Chinese population (20). No significant difference in the haplotype frequencies between allergic patients and controls could be demonstrated (Table 4).

Table 4.   Haplotype frequencies of interleukin (IL)-10 promoter gene in β-lactam allergic patients and controls
Haplotype (IL-10)Allergics (n = 44)Controls (n = 41)
GCC45.532.9
ACC22.736.6
ATA31.830.5

Cytokine gene polymorphisms in atopic patients and in nonatopic subjects

Most of the SNPs investigated herein were previously associated with atopy in different studies. To extend our observations, we delineated the atopic population within both allergic and control groups. Patients with one or more positive skin prick tests to common aeroallergens were considered atopic. Finally, 44 atopic subject genotypes were compared with those of 44 nonatopic subjects. Among the 15 SNPs analysed, no associations were found between allele/genotype distributions and atopy (data not shown). Meanwhile, by this criterion and considering results by sex, we found that the frequency of the IL-4Rα 75Ile allele is higher in men's nonatopic group when compared with atopic men (95.4%vs 4.5%; P = 0.004) when subjects are homozygous or heterozygous for the polymorphism (GA + AA genotypes). The 75Ile allele seems therefore to have a protective effect (OR = 0.07, CI: 0.01–0.66; Table 5). This frequency difference was not observed in atopic women (Table 5).

Table 5.   Genotype distribution and allele frequency of the IL-4Rα Ile75Val polymorphism in 22 atopic men and 22 nonatopic men
IL-4RαAtopicsControlsP-value
  1. *The P-value for the genotype frequency is calculated for IL-4RαGG vs IL-4RαGA/CA (OR: 0.07, 95% CI: 0.01–0.66).

  2. IL-4R, interleukin-4 receptor; OR, odds ratio; CI, confidence interval.

Genotype frequency (%)
 GG9 (40.9)1 (4.6)0.004*
 GA9 (40.9)16 (72.7) 
 AA4 (18.2)5 (22.7) 
Allele frequency
 75Val0.610.410.054
 75Ile0.390.59 

Polymorphisms in allergic atopic patients

Because atopy is a genetic tendency to develop the classic allergic diseases, we next compared, within the atopic population, the prevalence of all SNPs between the immediate β-lactam allergic group (n = 22) and the control group, consisting of nonallergic but atopic patients (n = 22). We observed no significant differences between determined genotypes in allergic atopic patients and nonallergic atopic patients (data not shown).

Separate analysis for each gender revealed that a significant association of the IL-10 promoter polymorphisms −819 C>T and −592 C>A, which are two linked SNPs, was only detected in atopic women (P = 0.023, OR = 17.5, CI: 1.26–533.07, n = 11) but not in atopic men (P = 0.076, n = 11) when IL-10-819CT/TT and IL-10-592CA/AA genotypes are compared with IL-10-819CC and IL-10-592CC genotypes respectively (Table 6).

Table 6.   Allele frequency and genotype distribution of the IL-4RA Ile75Val, IL-10 −819C>T and IL-10 −592C>A polymorphisms in 11 β-lactam allergic atopic women and 11 control atopic women
PolymorphismsGenotypeGenotype frequency (%)P-valueAlleleAllele frequencyP-value
AllergicsControlsAllergicsControls
  1. *OR: 5.4, 95% CI: 1.16–27.17.

  2. **The P-value for the genotype frequency is calculated for IL-10 −819CC vs IL-10 −819CT/TT (or IL-10 −592CC vs IL-10 −819CA/AA; OR: 17.50, 95% CI: 1.26–533.07).

  3. IL, interleukin; OR, odds ratio; CI, confidence interval.

IL-4Rα (Ile75Val)GG2 (18.2)7 (63.6)0.08G0.450.820.012*
GA6 (54.5)4 (36.4)A0.550.18
AA3 (27.3)0 (0)    
IL-10 (−819)CC1 (9.1)7 (63.6)0.023**C0.410.680.069
CT7 (63.6)1 (9.1)T0.590.32
TT3 (27.3)3 (27.3)    
IL-10 (−592)CC1 (9.1)7 (63.6)0.023**C0.410.680.069
CA7 (63.6)1 (9.1)A0.590.32
AA3 (27.3)3 (27.3)    

Additionally, the frequency of IL-4Rα 75Ile allele (A allele) is much higher in immediate β-lactam allergic atopic women when compared with control atopic women (55%vs 18%, P = 0.012, OR = 5.4, CI: 1.16–27.7; Table 6). No such difference was observed when comparing the group of immediate β-lactam allergic and atopic men with the control atopic men's group (32%vs 45%, P = 0.352).

Conclusions

In order to delineate the involvement of cytokine and cytokine receptor gene polymorphisms in immediate β-lactam allergy, we analysed 15 SNPs of candidate genes that have been all associated by previous studies either to upregulated serum IgE levels, or to asthma and atopy. Here, we show that the polymorphisms of IL-13, IL-4, IFN-γR1, IFN-γR2, IL-21, FcɛRIβ, STAT6 genes are not linked in this drug allergy. Furthermore, atopy in general is not a risk factor for drug hypersensitivities and allergies (21–23), but atopic patients may; however, suffer from more severe reactions. Our genetic findings seem therefore in keeping with the difference between drug allergic and atopic patient populations suggested by available epidemiological data.

Interestingly, we also found no association of the studied SNPs with atopy in our overall group of patients. It has already been pinpointed that published data are controversial as many studies have been unable to reproduce previous findings (24). The inconsistency of these results may be attributable to differences in studied populations, such as ethnicity and differences in definition of phenotype. In this study, we favoured a highly stringent clinical selection of patients in order to ensure a high homogeneity of the studied group. The consequence is a limited size of the population considered herein.

Meanwhile, when combining all criterions of allergy, atopy and gender, we found that the risk of developing an immediate reaction to β-lactams is most significantly associated with IL-10 promoter and IL-4Rα gene polymorphisms when it is linked with atopy and female gender. The results showed two independent associations between immediate β-lactam allergy in atopic women and IL-10 −592/−819 genotype and the IL-4Rα 75Ile allele. Another gender discrepancy was observed concerning IL-4Rα genotype: in atopic men, the 75Ile allele could have a protective effect when patients are either homozygous or heterozygous for the polymorphism. That depending on the gender, a given polymorphism could be associated either with increased risk or relative protection, is a puzzling finding not previously described in the literature on gene polymorphisms and atopy.

We could not confirm the associations made by Qiao et al.’s study of IL-4Rα 576Gln polymorphism and of the Glu237Gly variant of FcɛRIβ gene with penicillin allergy. Indeed, this analysis was performed on a likely homogenous Chinese population and concerned allergy to particular major and minor penicillin antigens without distinguishing immediate and nonimmediate reactions (18, 19).

The IL-4 receptor-α chain is a subunit that is common to the IL-4 and IL-13 receptor complexes and therefore has a key role in allergic disease by promoting IgE production. In this study, we analysed two naturally occurring variants of the IL-4Rα chain that are a consequence of SNPs in the coding sequence of the IL-4Rα gene resulting in single amino acid changes: Ile75Val and Gln576Arg. These two polymorphisms have been linked to atopic disease (7), suggesting that alterations of IL-4 and IL-13 signaling may be associated with allergic disease. Studies examining their functional consequences, have shown that the Gln576Arg polymorphism had no effect on IL-4 signal transduction in murine cell lines (25, 26). Although the substitution Ile75Val is located in an extracellular domain, two independent groups have demonstrated that there is no alteration in IL-4-binding affinity for its receptor (25, 27, 28). Studies performed in cell lines indicate that the 75Ile variant mediates a modest increase of STAT6 activation, proliferation and transcription activity of the Iɛ-promoter in response to IL-4 stimulation (25, 28). Additional study has suggested that the pertinent susceptibility trait is, in fact, a haplotype in which 75Val is linked to 576Arg (27).

Interleukin-10 is an anti-inflammatory cytokine that has been shown to regulate both cellular and humoral immunity. It regulates T-cell functions by suppressing the expression of TH1 cytokines by TH1 cells (29, 30). It has immunoregulatory effects by downregulating class II major histocompatibility complex (MHC) expression and inhibits the production of proinflammatory cytokines by monocytes (31). These inhibiting effects are not found in B-cell population where IL-10 enhances lymphocyte survival, proliferation, differentiation and antibody production (11).

The effect of the three SNPs analysed in this study, on IL-10 production by lymphocytes in vitro has been assessed (12). The SNP −1082 G>A was associated with varying levels of cytokine production after stimulation with concanavalin A (con A). The GCC haplotype has been associated with high IL-10 production in PBMC cultures (12). Another study using luciferase assay found compatible results (13).

Our study shows that atopic women homozygous or heterozygous for the IL-10 −819T/−592A allele present a higher risk of developing an immediate IgE-dependent allergy to β-lactams. These two alleles are always associated with the IL-10 −1082A allele that is responsible for a low expression of the cytokine. Indeed, IL-10 modulates many cells and effector functions associated with allergic diseases, and several studies provide evidence for an inverse correlation between IL-10 levels and the severity of allergic and asthmatic disease [reviewed in Hawrylowicz and O'Garra (32)].

The gender-restricted association of IL-10 promoter and IL-4Rα SNPs with β-lactam allergy in atopic patients highlights the importance of considering sex difference as an important cofactor. Interestingly, women are shown to have higher incidence of adverse drug reactions than men (approximately 65% of the cases; 21, 33, 34).

Acknowledgments

The authors thank the nurses for their cooperation and Monique Rongier for collecting patient samples. The authors also acknowledge Haoi-Bich Co Minh for her assistance. This study was supported by the Paul Hamel Institute and the French Society of Allergology and Clinical Immunology (SFAIC).

Ancillary