Background

Asthma is likely to result from the effects of environmental stimuli in genetically susceptible individuals.¹ Asthma frequently develops on a background of atopy. According to the 'hygiene hypothesis',² frequent exposure to microbes and infections early in childhood is postulated to protect against the development of atopy and asthma. Conversely, a 'clean' environment during immune maturation may lead to persistence of a Th2, instead of a Th1, immune phenotype, which predisposes to atopy.

The 'hygiene hypothesis' was initially proposed to account for the reduced prevalence of hay fever with increasing family size.³ Subsequently, reduced allergic sensitisation in children has been observed with factors that increase microbial exposure, including earlier attendance at day care, frequent colds in infancy and living on a farm, especially with early exposure to livestock or unpasteurised milk.^2,4,5 However, the timing and dose of these exposures are critical, as are genetic and environmental cofactors.⁴ While there is some epidemiological evidence for a causal effect of microbial exposure in protecting against allergy and asthma,⁶ the precise mechanisms underlying any postulated protective effect are not yet clearly defined.

Exposure to endotoxin may provide a unifying mechanism through which the 'hygiene hypothesis' acts.⁵ Endotoxin, or lipopolysaccharide (LPS), is an extracellular component of the cell wall of Gram-negative bacteria.⁴ Recent cohort studies in childhood have shown that higher levels of endotoxin in house dust and mattress dust are associated with a lower risk of allergic sensitisation^7,8 and atopic asthma.⁹ Pattern recognition receptors sense endotoxin, and the innate immune system is activated via the endotoxin receptor complex, which includes LPS-binding protein, CD14, MD2 and Toll-like receptor 4 (TLR4).¹⁰ A single-nucleotide polymorphism (SNP) in the TLR4 gene (chromosome 9q32–33), consisting of Asp299Gly (nucleotide substitution 896A>G), has been associated with hyporesponsiveness to inhaled endotoxin challenge, reduced density of TLR4 in airway epithelium, and reduced production of inflammatory cytokines in response to endotoxin.¹¹ Given that variability in the natural history of allergic disease may be influenced by genetic factors in host responses to allergens, the aim of this study was to determine the effect of TLR4 polymorphism on risk and severity of asthma and atopy, in families with asthmatic offspring.

Results

Genotyping results

Genotypes were in Hardy–Weinberg equilibrium. In five families, one or more members of the parent-affected sibling trios could not be genotyped for the TLR4 polymorphism due to technical reasons, so 336 of the original 341 families recruited were included in this study. Similarly, five of the original 184 hyper-normal controls could not be genotyped, leaving 179 controls for inclusion. The frequency of the Gly299 allele was 6%, similar to previous reports in Caucasian populations.^11,12

TDT analysis

The TDT analysis was performed on parent-affected sibling trios, where the sibling was the first affected sibling with asthma, as defined in subjects and methods. The TDT analysis was also performed in trios in which the sibling had both asthma and the presence of atopy, raised total IgE, PC₂₀ <16 mg/ml or PC₂₀ <4 mg/ml. TDT analysis found no association of the TLR4 Gly299 allele with risk of asthma, or asthma with related asthma or atopy phenotypes (Table 1).

Table 1 - Transmission disequilibrium test (TDT) analysis of parent-affected sibling trios.

Full table

Case–control analysis

Three case–control analyses were performed: first affected sibling, second affected sibling or asthmatic parents, vs 'hyper-normal' controls. No statistically significant associations were detected with asthma in any of these analyses (Table 2).

Table 2 - Case–control analyses of asthmatics vs controls.

Full table

Disease severity

In the first affected siblings, the TLR4 Asp/Gly or Gly/Gly genotypes (combined) were associated with a higher mean atopy severity score of 1.8, compared to the mean score of 1.2 in those with the Asp/Asp genotype (P=0.003) (Table 3). Other asthma and atopy phenotypes were not associated with the polymorphism in the first affected siblings. No phenotypes were associated with the polymorphism in the second affected siblings or in asthmatic parents (all P>0.05, data not shown).

Table 3 - Disease severity associations in the first-affected siblings.

Full table

As an illustration of the clinical relevance of the observed difference in atopy severity score, we characterised the relationship of components of the score to the numbers of skin-prick test and specific IgE responses that an individual has. The score ranged between 0 and 4.5, with nonatopic individuals having a mean of 0.02 (s.d. 0.04) and atopic individuals having a mean of 1.7 (s.d. 1.0). The highest scores >4 (very severe atopy) were observed in 12 subjects. These individuals had four to six positive skin-prick tests (out of six) and six to seven positive specific IgE tests (out of seven). Subjects with a score of 1.8 had, on average, a greater number of positive skin-prick tests (two to five) and positive specific IgE tests (two to six), than those with a score of 1.2 (one to three positive skin-prick tests, and two to four positive specific IgE tests).

Discussion

This study demonstrates, for the first time, an association, between polymorphism in the TLR4 gene and the severity of atopy, in asthmatics. First asthmatic siblings carrying the Gly299 allele, which confers hyporesponsiveness to endotoxin, had increased severity of atopy. This finding supports the notion that reduced exposure to microbes enhances the risk of allergic sensitisation, at least in an asthmatic population.

According to the 'hygiene hypothesis', the putative mechanism for increased atopy prevalence is persistence of a Th2 phenotype, due to reduced endotoxin sensing and signalling. The endotoxin receptor complex on immune effector cells senses endotoxin, and then signals via TLR4 to activate the immune system, evoking a Th1 cytokine and cellular response. During maturation of the immune system in early life, genetically determined reduction of TLR4 activation and signalling could lead to reduced Th1 differentiation, thereby favouring persistence of a Th2 phenotype. The TLR4 Asp299Gly polymorphism in the extracellular domain of the receptor results in reduced TLR4 function, as demonstrated by the lower density of the receptor in airway epithelium and reduced downstream Th1 cytokine production in response to endotoxin.¹¹ Moreover, the clinical importance of reduced TLR4 function due to this polymorphism is demonstrated by associations with increased risk of bacterial infection,¹³ Gram-negative infection¹⁴ and septic shock.¹⁵ Thus our findings of increased severity of atopy in those carrying the TLR4 Gly299 allele, presumably via reduced endotoxin sensing, is consistent with, and adds to, previous studies showing a reduced response of the innate immune system as a result of this polymorphism.

We found a positive association of TLR4 polymorphism with atopy severity in asthmatics. To overcome the limitations of using a single marker of atopy, we developed the atopy severity score using principal component analysis. This score was comprised of measures of magnitude of allergic response (wheal size, specific IgE level) and its range (number of wheals, number of specific IgE responses).¹⁶ The mean atopy severity score of 1.8 in first asthmatic siblings with Asp/Gly or Gly/Gly genotypes was higher (indicating more severe atopy) than the mean of 1.2 in Asp/Asp individuals. This difference was clinically important, with a score of 1.8 reflecting a greater range of positive skin-prick tests and positive specific IgE tests than a score of 1.2. Given the positive findings of the present study, this composite score may therefore be more useful than univariate measures of the severity of atopy in genotype–phenotype correlations.

In contrast, no association was found between genotype and total serum IgE, which was in concordance with previous reports.^12,17 When we initially derived the atopy severity score by principal component analysis, there were five components with total serum IgE making up the fifth component. It formed the least contribution (smallest coefficient) to the first principal component, and was subsequently removed, which improved the overall explained variance of the score (S Rorke, personal communication). In a component plot of the five variables, total serum IgE was separated out from the variables of positive skin-prick test and specific IgE (probably indicating different regulating genetic factors for example, HLA vs host response genes) and was also the least correlated with the first principal component. Thus, total serum IgE may not be an important factor in determining the overall atopy severity and therefore may play a less important role in atopic asthma.

The association with severity of atopy was found in the analysis of the first asthmatic siblings, and not in the second asthmatic siblings or asthmatic parents. It is known that birth order affects the risk of atopy, such that children with a greater number of older siblings have a reduced prevalence of atopy.³ It is conceivable that birth order and other factors may dominate over genetic effects on atopy in children born later in the pedigree, whereas in children born earlier in the pedigree (which would most likely include the first asthmatic sibling) the genetic effect is detectable.

In the present study, no association was observed between TLR4 polymorphism and the risk of asthma, asthma with atopy, or asthma with BHR. This lack of association was found in the TDT analysis of the trios, and also the comparison of each of the three asthmatic groups with controls. This was in agreement with a North American study which examined TLR4 polymorphisms in asthma.¹² In their study, the TLR4 gene was screened and a total of 29 SNPs were discovered. Five of the common polymorphisms, including the Asp299Gly SNP, were genotyped in 589 families from North America, and replicated in a second cohort of 167 families from Quebec. Using a family-based association test, the authors found no evidence of association with asthma or asthma-related phenotypes (FEV₁ % predicted, PC₂₀ methacholine).¹² A study of 334 German subjects also showed that there was no difference in the overall risk of asthma between wild-type and variant TLR4 genotypes.¹⁷ However, whereas subjects carrying wild-type genotypes had increasing risk of asthma with greater endotoxin exposure, there was no such effect in subjects with variant genotypes.¹⁷ Taken together, these results would appear to exclude a role for genetic variation in TLR4 in the overall susceptibility to asthma, but support the role of gene–environmental interaction in the development of allergic disease.

It has been estimated that one-third of asthmatic symptoms can be attributed to atopy.¹⁸ Thus, in addition to atopy, it is likely that other disease processes such as airway remodelling contribute significantly to the development of asthma.¹⁹ Hence, genes that are likely to affect airway remodelling (eg ADAM33)²⁰ have been linked to asthma. Our findings highlight the possible link between reduced endotoxin exposure and allergy, but does not support a link with asthma, a disease which is likely to be further along the multistep process of disease development.

Potential limitations of this study should be addressed. As families in this cohort were recruited on the basis of asthma, there were too few atopic individuals without asthma to study the effect of the polymorphism on atopy alone. Several related phenotypes of relevance to asthma were analysed in this study, and a number of comparisons were made. We did not employ Bonferroni correction for the number of comparisons as this adjustment is too conservative when phenotypes are not independent. However, at least for the disease severity analysis, even when adjusting the uncorrected P-value of 0.003 for the number of comparisons (15), the corrected P-value of 0.045 remains statistically significant. Finally, we cannot exclude that the association observed is due to another causal allele in linkage disequilibrium with the SNP studied here. Overall, we feel that the findings presented here meet epidemiological criteria of plausibility and adds genetic data in support of the notion that infection may protect against atopy.⁶

In summary, this study confirms the previously observed lack of association of TLR4 polymorphisms with asthma. In contrast, the findings suggest that genetically determined hyporesponsiveness to endotoxin may increase atopy severity in asthmatic populations. Future studies should examine the role of TLR4 polymorphisms in the development and severity of atopy in individuals without asthma, in order to generalise these results and explore the importance of endotoxin exposure and predisposition to allergy.

Subjects and methods

Subjects

Affected sib-pair families

In all, 341 Caucasian affected sib-pair families were collected in the UK, as previously described.^16,20 The definition of asthma used consisted of a physician's diagnosis of asthma, together with current asthma medication use. Atopy was defined as either a positive skin-prick test (>3 mm) or a raised specific IgE (>0.35 IU) to one or more common allergens.¹⁶ Bronchial hyper-responsiveness (BHR) to methacholine, total serum IgE and specific serum IgE were measured,^16,20 and severity scores for atopy and asthma were generated.^16,21

Hyper-normal controls

In total 184 Caucasian 'hyper-normal' controls were recruited in the UK. These individuals had no diagnosis of asthma, and no family history of asthma.

Genotyping

A tetra-primer amplification refractory mutation system (ARMS)-polymerase chain reaction (PCR) assay²² was used to genotype the TLR4 Asp299Gly (896A>G) polymorphism. The following primers were used: (small letters in primer indicate mismatch) (See Table 4)

Table 4.

Full table

Genomic DNA (20ng) was amplified in a total reaction volume of 10 l containing dNTP 0.2 mM, MgCl₂ 4 mM, primers (outer upper 0.1 M, outer lower 0.1 M, inner upper 2 M, inner lower 2 M), AmpliTaq Gold DNA polymerase (0.025 U/l) and buffer (Applied Biosystems, Warrington, UK). The PCR cycling conditions were 95°C 5 min; then nine cycles of 94°C 30 s, X°C 30 s (where X is initially 72°C, decreasing 1°C per cycle to 64°C), 72°C 30 s; then 31 cycles of 94°C 30 s, 64°C 30 s and 72°C 30 s; and finally 72°C 10 min, on a PTC-225 DNA Engine Tetrad (MJ Research Inc, Waltham, MA, USA).

PCR products were resolved by micro-array diagonal gel electrophoresis (MADGE),²² stained with Vistra Green (Amersham Biosciences, Little Chalfont, Buckinghamshire, UK) and visualised by a Fluoroimager 595 (Molecular Dynamics, Sunnyvale, CA, USA). Allele-specific PCR products (Asp 147 bp, Gly 292 bp) and a constant product (385 bp) were visualised. Two researchers independently scored the genotypes using the Phoretix 1D gel analysis software (Nonlinear Dynamics, Newcastle upon Tyne, UK). Representative genotypes were confirmed by dideoxy dye terminator cycle sequencing (BigDye Terminator Version 3.0, Applied Biosystems) on an ABI PRISM 377 DNA Sequencer (Applied Biosystems).

This study protocol was approved by the Southampton and South West Hampshire joint ethics committee.

Statistical analysis

TDT analysis

Association of the TLR4 polymorphisms with asthma and atopy was tested in the trios of mother, father and first-affected sibling with available genotype, as previously described.¹⁶ The transmission disequibrium test (TDT) was performed in STATA (Version 6.0, Stata Corporation, TX, USA) using the programme written by David Clayton (MRC Biostatistics Unit, Cambridge, UK). Any TDT test with a P-value of <0.05 was considered significant. Dichotomous variables analysed using TDT were: (i) asthma positive on questionnaire ('Have you ever had asthma?', 'Was this confirmed by a doctor?' and 'Have you used any medicines to treat asthma, or any breathing problems, at any time in the last 12 months?'); (ii) asthma positive on questionnaire, with atopy (defined by raised specific IgE and/or positive skin-prick test); (iii) asthma positive on questionnaire, with raised total serum IgE (age corrected); (iv) asthma and PC₂₀ methacholine <4 mg/ml (severe asthmatics); (v) asthma and PC₂₀ methacholine 16 mg/ml.

Case–control analysis

Data were analysed in three separate groups of cases, and participants within each group were considered to be independent. The group of cases were: first affected sibling, second affected sibling, and parents with a diagnosis of asthma. Hardy–Weinberg equilibrium was confirmed using ² with 1 degree of freedom. Genotype frequencies for each group of cases were compared to the 'hyper-normal' control population, and analysed using ² in SPSS Version 11 (SPSS Inc, Chicago, IL, USA). The less common Asp/Gly and Gly/Gly genotypes (combined) were compared with the more common Asp/Asp genotype, since the Gly/Gly genotype was very rare, and heterozygotes have previously been shown to have functional differences in endotoxin responsiveness when compared to wildtypes.¹¹

Disease severity

Data were analysed in three separate groups of cases, and participants within each group were considered to be independent. The groups of cases were: first affected sibling, second affected sibling, and parents with a diagnosis of asthma. t-Tests and nonparametric tests (SPSS Version 11) were performed as appropriate on the following variables: (i) total serum IgE (age corrected and log₁₀ transformed to improve normality); (ii) FEV₁% predicted; (iii) slope of FEV₁ response to methacholine (transformed to [1/(least squares (LS) slope+30)]1000 to improve normality, avoid negative values and overcome 'censored' PC₂₀ values); (iv) atopy severity score; (v) asthma severity score. A P-value of <0.05 was considered significant in all analyses.

Notes

Contributions of authors

ST Holgate, JB Clough and TP Keith designed the original study in which the subjects were recruited. JW Holloway and IA Yang initiated this study of the TLR4 polymorphism. IA Yang and JA Cakebread performed the genotyping. SJ Barton and S Rorke undertook the statistical analysis. IA Yang and JW Holloway drafted the report. All authors approved the final version of this report.

Contributions of the funding support

Genome Therapeutics Corporation and Schering-Plough were responsible for the design of the original study and data collection. The funding bodies had no role in the genotyping, data analysis or writing of the report.

References

1. Holgate ST. The epidemic of allergy and asthma. Nature 1999; 402 (Suppl): B2–B4. | Article | PubMed | ISI | ChemPort |
2. Strachan DP. Family size, infection and atopy: the first decade of the 'hygiene hypothesis'. Thorax 2000; 55 (Suppl 1): S2–S10. | Article | PubMed |
3. Strachan DP. Hay fever, hygiene, and household size. BMJ 1989; 299: 1259–1260. | PubMed | ChemPort |
4. Liu AH. Endotoxin exposure in allergy and asthma: reconciling a paradox. J Allergy Clin Immunol 2002; 109: 379–392. | Article | PubMed | ISI | ChemPort |
5. von Mutius E. Environmental factors influencing the development and progression of pediatric asthma. J Allergy Clin Immunol 2002; 109: S525–S532. | PubMed |
6. Liu AH, Murphy JR. Hygiene hypothesis: fact or fiction? J Allergy Clin Immunol 2003; 111: 471–478. | Article | PubMed | ISI |
7. Gereda JE, Leung DY, Thatayatikom A et al. Relation between house-dust endotoxin exposure, type 1 T-cell development, and allergen sensitisation in infants at high risk of asthma. Lancet 2000; 355: 1680–1683. | Article | PubMed | ISI | ChemPort |
8. Gehring U, Bischof W, Fahlbusch B, Wichmann HE, Heinrich J. House dust endotoxin and allergic sensitization in children. Am J Respir Crit Care Med 2002; 166: 939–944. | Article | PubMed | ISI |
9. Braun-Fahrlander C, Riedler J, Herz U et al. Environmental exposure to endotoxin and its relation to asthma in school-age children. N Engl J Med 2002; 347: 869–877. | Article | PubMed | ISI |
10. Beutler B. Innate immune responses to microbial poisons: discovery and function of the Toll-Like receptors. Annu Rev Pharmacol Toxicol 2003; 43: 609–628. | PubMed |
11. Arbour NC, Lorenz E, Schutte BC et al. TLR4 mutations are associated with endotoxin hyporesponsiveness in humans. Nat Genet 2000; 25: 187–191. | Article | PubMed | ISI | ChemPort |
12. Raby BA, Klimecki WT, Laprise C et al. Polymorphisms in toll-like receptor 4 are not associated with asthma or atopy-related phenotypes. Am J Respir Crit Care Med 2002; 166: 1449–1456. | Article | PubMed | ISI |
13. Kiechl S, Lorenz E, Reindl M et al. Toll-like receptor 4 polymorphisms and atherogenesis. N Engl J Med 2002; 347: 185–192. | Article | PubMed | ISI | ChemPort |
14. Agnese DM, Calvano JE, Hahm SJ et al. Human toll-like receptor 4 mutations but not CD14 polymorphisms are associated with an increased risk of gram-negative infections. J Infect Dis 2002; 186: 1522–1525. | Article | PubMed | ISI | ChemPort |
15. Lorenz E, Mira JP, Frees KL, Schwartz DA. Relevance of mutations in the TLR4 receptor in patients with Gram-negative septic shock. Arch Intern Med 2002; 162: 1028–1032. | Article | PubMed | ISI | ChemPort |
16. Sayers I, Barton S, Rorke S et al. Allelic association and functional studies of promoter polymorphism in the leukotriene C4 synthase gene (LTC4S) in asthma. Thorax 2003; 58: 417–424. | Article | PubMed | ISI | ChemPort |
17. Werner M, Topp R, Wimmer K et al. TLR4 gene variants modify endotoxin effects on asthma. J Allergy Clin Immunol 2003; 112: 323–330. | Article | PubMed | ISI | ChemPort |
18. Pearce N, Pekkanen J, Beasley R. How much asthma is really attributable to atopy? Thorax 1999; 54: 268–272. | PubMed |
19. Davies DE, Wicks J, Powell RM, Puddicombe SM, Holgate ST. Airway remodeling in asthma: new insights. J Allergy Clin Immunol 2003; 111: 215–225. | Article | PubMed | ISI | ChemPort |
20. Van Eerdewegh P, Little RD, Dupuis J et al. Association of the ADAM33 gene with asthma and bronchial hyperresponsiveness. Nature 2002; 418: 426–430. | Article | PubMed | ISI | ChemPort |
21. Holloway JW, Rorke S, Beghé B, Holgate ST. Genetic influences on asthma severity In: Szefler SJ, Leung DYM (eds). Severe Asthma: Pathogenesis and Clinical Management. Marcel Dekker: New York, 2001 pp 19–38.
22. Ye S, Dhillon S, Ke X, Collins AR, Day IN. An efficient procedure for genotyping single nucleotide polymorphisms. Nucleic Acids Res 2001; 29: E88–E98. | Article | PubMed | ChemPort |

Acknowledgements

We thank the patients and families who participated in this study. The collection of families was funded by Genome Therapeutics Corporation and Schering-Plough. IAY was supported by an Allen+Hanburys/Thoracic Society of Australia and New Zealand Respiratory Research Fellowship. JWH was supported by a Medical Research Council (MRC) Research Training Fellowship.