Abstract
Background and objective: Dysregulation of respiratory mucins, MUC5AC in particular, has been implicated in respiratory disease and MUC5AC expression is up-regulated in response to environmental challenges and inflammatory mediators. The aim of this study was to examine the effect of genetic variation on susceptibility to common respiratory conditions.
Methods: The association of MUC5AC and the closely linked genes MUC2 and MUC5B with respiratory outcomes was tested in the MRC National Survey of Health and Development, a longitudinal birth cohort of men and women born in 1946. Also examined were the functional variants of the genes encoding inflammatory mediators, IL13, IL1B, IL1RN, TNFA and ERBB1, for which there is a likely influence on MUC5AC expression and were explored potential gene-gene interactions with these inflammatory mediators.
Results: Statistically significant associations between the 3'ter MUC5AC simple nucleotide polymorphism (SNP) rs1132440 and various non-independent respiratory outcomes (bronchitis, wheeze, asthma, hay fever) were reported while the adjacent loci show slight (but largely non-statistically significant) differences, presumably reflective of linkage disequilibrium (allelic association) across the region. A novel association between bronchitis and a non-synonymous functional ERBB1 SNP, rs2227983 (aka epidermal growth factor receptor:R497K, R521K) is also reported and evidence presented of interaction between MUC5AC and ERBB1 and between MUC5AC and IL1RN with respect to bronchitis. The ERBB1 result suggests a clear mechanism for a biological interaction in which the allelic variants of epidermal growth factor receptor differentially affect mucin expression.
Conclusions: The MUC5AC association and the interactions with inflammatory mediators suggest that genetically determined differences in MUC5AC expression alter susceptibility to respiratory disease.
SUMMARY AT A GLANCE
This longitudinal cohort study shows occurrence of the common respiratory conditions bronchitis, wheeze, asthma and hay fever to be associated with genetic variation in a mucin gene, MUC5AC. Functional variation in the epidermal growth factor receptor (epidermal growth factor receptor encoded by ERBB1) is also associated with bronchitis and modulates the MUC5AC effect.
Keywords: airway epithelium, asthma, genetics, inflammation, respiratory function test
Introduction
Disturbances to the normal dynamics of the respiratory epithelial layer as a result of allergens, microorganisms and noxious agents that cause inflammation often leads to secretion of large quantities of airway mucus which allows the expulsion of the offending agent. In chronic airway disease, this mechanism is intensely active for lengthy periods, exacerbating disease symptoms, and in severe asthma, some airways become irreversibly occluded.1 Genetic variation affecting any part of these pathways is likely to influence susceptibility and severity of respiratory symptoms.
The major high molecular-weight glycoprotein components of respiratory mucus are the mucins MUC5AC and MUC5B.2 The quantities of MUC5AC and MUC5B are increased in a number of respiratory conditions and experimental models3–7 and expression of their genes, MUC5AC and MUC5B are up-regulated.6,8
Several mediators of inflammation such as the cytokines, interleukin 13 (encoded by the gene IL13), interleukin 1B (encoded by IL1B) and tumour necrosis factor α (encoded by TNFA), which are present at high levels within the asthmatic airways,9–13 are thought to orchestrate respiratory mucus hypersecretion and are known to up-regulate MUC5AC expression as well as secretion,14–17 and MUC5AC is regulated through an epidermal growth factor receptor (EGFR, encoded by the gene known as ERBB1) signalling pathway.7,8,18–22 See Table S1 in the supporting information available online for details of functional evidence.
Previous studies showed that the genetically variable MUC2 tandem repeat (TR) sequence of the main mucin domain had a different size distribution in atopic individuals with and without asthma.23 Although MUC2 expression shows some evidence of up-regulation in inflammatory disease,23 biologically, MUC2 appears a rather unlikely candidate for altering respiratory disease susceptibility because the protein is found at only very low levels in the airways.24 – 25 MUC5AC is located adjacent to MUC2 in a region of strong linkage disequilibrium (allelic association) on chromosome 11p15.5.26 Thus the association seen between MUC2 and asthma could in fact be a consequence of association between the MUC2 TR and a causative allele (variant) in MUC5AC, or possibly MUC5B.
Here we explore the possible association between variants in these MUC genes (see Table S2 in the supporting information available online) and various respiratory- and allergy-related outcomes in the 1946 British birth cohort. We also test for gene–gene interactions with functional simple nucleotide polymorphism (SNP) within the genes IL13, IL1B, TNFA and IL1RN which encodes the interleukin 1 receptor agonist and interacts with IL1B (see Table S1 in the supporting information available online), and ERBB1 with respect to disease outcome.
Methods
Study participants
The MRC National Survey of Health and Development is a socially stratified sample of 5362 of all British births during 1 week of March 1946. The data collections from which we have collated information were carried out at age 43(1989) and 53 years (1999) when research nurses visited study members in their own homes and asked a series of health and lifestyle questions. At age 53, 2989 of the cohort members were interviewed. Contact was not attempted for the 1979 individuals who had previously refused to take part, were living abroad, were untraced since the previous contact at 43 years or had already died. The responding sample at age 53 is in most respects representative of the national population of a similar age27 and considered to be representative of a European population since the study began before mass immigration into the United Kingdom. Blood and buccal samples were collected from consenting participants at age 53 (ethical approval reference MREC no. 98/2/121).
Outcome variables
Table 1 shows a description of all outcome variables and measures. The outcome variables indicating whether individuals had ever had asthma or hay fever were as described previously.28 Forced expiratory volume in 1 s and forced vital capacity, were recorded at each visit using a Micromedical turbine electric spirometer (Cardinal Health UK 232 Ltd, Basingstoke, UK).
Table 1.
Outcome variables used; questions asked and measured in 1989 or 1999
| Questions asked | Responses | Recoded variable as analysed in this study |
|---|---|---|
| Have you ever had asthma (1946–1989)? | No | No or yes (once/recurring) combined with question from 1999 below |
| Yes, once | ||
| Yes, recurring | ||
| In the last 10 years (1989–1999), did you have asthma? | No | Ever asthma to 1999 |
| Yes | ||
| Have you ever had hay fever (1946–1989) | No | No or yes (once/ recurring) combined with question from 1999 below |
| Yes, once | ||
| Yes, recurring | ||
| In the last ten years (1989–1999), did you have hay fever? | No | Ever hay fever to 1999 |
| Yes | ||
| Have you ever had bronchitis (1946–1989)? | No | Ever bronchitis to 1989 |
| Yes, once | ||
| Yes, recurring | ||
| During the past 3 years (1996–1999), did you have any chest illness, such as bronchitis or pneumonia, which kept you off work or indoors for a week or more? | Yes | Bronchitis 96–99 |
| No | ||
| Does your chest ever sound wheezy or whistling? Do you get this most days (or nights)? (Asked in 1989) | No, or not most days or nights | Wheeze 89 |
| Yes, most days or nights | ||
| Does your chest ever sound wheezy or whistling? Do you get this most days (or nights)? (Asked in 1999) | No, or not most days or nights | Wheeze 99 |
| Yes, most days or nights | ||
| Has this baby ever had a lower tract respiratory infection (i.e. bronchitis, bronchopneumonia or pneumonia)? (asked of mother when cohort member 24 months) | Yes | LRTI |
| No | ||
| Max FEV1 reading in 1989† | See methods | FEV89 |
| Max FEV1 reading in 1999† | See methods | FEV99 |
| Difference in Max FEV1 over 10 years (1989–1999).† | DELTA | |
| FEV1/FVC ratio | FEV1/FVC89 and 99 |
Three measures were recorded at age 43 and two at age 53. Only participants where at least two measures were taken, and deemed satisfactory, were included. The maximum readings were selected for analysis. Variables used shown in bold.
FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; LRTI, lower respiratory tract infection.
Confounders
Potential confounders were chosen because they were previously reported to be significantly associated with one or more of the outcome variables, or considered to be of direct biological relevance.29–30 These were smoking status, childhood social class, own social class, gender and region of birth, as well as height for lung function measurements (see Table 2 for demographic details of the key outcome and confounder variables).
Table 2.
Demographic details and frequencies of outcome variables in the test population
| All n (%) | Men n (%) | Women n (%) | |
|---|---|---|---|
| Gender | |||
| Male | 1441 (49.4) | NA | NA |
| Female | 1447 (50.6) | NA | NA |
| Own social class | |||
| Non manual | 1756 (68.1) | 824 (63.1) | 932 (73.3) |
| Manual | 821 (31.9) | 481 (36.9) | 340 (26.7) |
| Father's social class | |||
| Non manual | 1139 (42.5) | 558 (42.0) | 581 (43.0) |
| Manual | 1539 (57.5) | 770 (43.0) | 769 (57.0) |
| Region | |||
| Scotland | 320 (11.0) | 160 (11.1) | 160 (10.8) |
| Northern England | 721 (24.7) | 388 (26.9) | 333 (22.6) |
| Central England | 1187 (40.7) | 572 (39.7) | 615 (41.6) |
| London/Southeast | 690 (23.7) | 321 (22.3) | 369 (25.0) |
| Ever smoking 1989 | |||
| No | 813 (29.7) | 347 (25.9) | 466 (33.7) |
| Yes | 1924 (70.3) | 993 (74.1) | 931 (66.6) |
| Ever asthma 1999 | |||
| No | 2464 (89.9) | 1210 (90.2) | 1254 (89.6) |
| Yes | 277 (10.1) | 132 (9.8) | 145 (10.4) |
| Ever hay fever 1999 | |||
| No | 2081 (76.1) | 1013 (75.6) | 1068 (76.5) |
| Yes | 655 (23.9) | 327 (24.4) | 328 (23.5) |
| Ever bronchitis 1989 | |||
| No | 2180 (79.6) | 1082 (80.7) | 1098 (78.5) |
| Yes | 560 (20.4) | 259 (19.3) | 301 (21.5) |
| Wheeze 1989 | |||
| No | 2493 (91.3) | 1189 (92.4) | 1304 (95.8) |
| Yes | 155 (5.9) | 98 (7.6) | 57 (4.2) |
| Wheeze 1999 | |||
| No | 2664 (91.3) | 1303 (90.5) | 1361 (92.2) |
| Yes | 253 (8.7) | 137 (9.5) | 116 (7.9) |
| LRTI | |||
| No | 2038 (75.3) | 995 (74.3) | 1043 (76.2) |
| Yes | 670 (24.7) | 344 (25.7) | 326 (23.8) |
The table sample includes all those that had data for the MUC genes and at least one of the outcomes.
LRTI, lower respiratory tract infection; NA, not applicable.
Genotyping
Details of DNA extraction, genotyping and validation as well as choice of SNP are given in the online supporting information (Text and Tables S1, S2 in the supporting information available online).
Data analyses
LDmax (http://www.sph.umich.edu/csg/abecasis/GOLD/docs/ldmax.html) was used to calculate pairwise measures of linkage disequilibrium. All further statistical analyses were performed using SPSS or STATA software. For each categorical outcome, contingency tables were constructed to compare the distribution of genotypes or alleles between the ‘affected’ and ‘unaffected’ groups with respect to disease variables. Multiple logistic regression models were then used to adjust the important associations for potential confounders. For these analyses, each of the SNP markers was coded by genotype (co-dominant model, with alleles grouped where necessary—see supplementary information). MUC5AC TR has two common length alleles and several rare ones which were considered as genotypes made up of three alleles L, (long) S (short) and R (rare). For binary (yes/no) outcomes, because the MUC2 TR data are recorded as a continuous variable, we compared the MUC2 TR allele size distributions (using a Mann–Whitney test) between the two groups, as done for our previously published study in which allele length was associated with asthma.23 Regression analysis was carried out to relate MUC2 allele length to lung function, using the combined MUC2 allele lengths for each individual and categorized into four gender-specific quartiles. Finally, potential interactions between each of the inflammatory loci and MUC5AC rs1132440 were explored and assessed using the likelihood ratio test. To display these differences in distribution in a simple manner graphically, we combined heterozygotes and homozygotes for the minor allele in each case.
Results
11p15.5 mucin gene variants typed in the National Survey of Health and Development cohort
Details and allele frequencies for polymorphisms within MUC2, MUC5AC and MUC5B are shown in Table S2 in the supporting information available online. No significant deviation from Hardy–Weinberg equilibrium was observed for any polymorphism. The MUC2 TR allele lengths ranged from 3.21 to 11.64 kb. As previously reported,23 26 there was a major mode between 7 and 8 kb with a minor mode of around 4 to 4.5 kb.
Linkage disequilibrium within the 11p15.5 MUC gene complex
As reported previously for other markers in this region,26 all of the MUC5AC and MUC5B markers are significantly associated with one or more of the others (see Table S3 in the supporting information available online).
Statistically significant association was also found between the MUC2 TR allele length distribution and the MUC5AC TR genotypes, LL and LS being associated with shorter MUC2 alleles, (P < 0.001, Kruskal-Wallis test) as found previously using family inferred haplotypes.26 The MUC2 TR and MUC5AC rs1132440 showed a similar trend, although this was not statistically significant (P = 0.084). Thus a general pattern of association can be seen to extend from MUC2 to MUC5AC.
Tests of association between mucin genetic variants and respiratory variables
Each mucin genetic variable (see Table S2 in the supporting information available online) was analysed for association with each of the respiratory outcomes detailed in Table.1
MUC5AC rs1132440 genotype counts showed statistically significant association with hay fever, bronchitis and wheeze at 43 years (3 × 2 contingency tables chi-square P-values 0.001 to 0.02) and were marginally associated with asthma (P = 0.06, Table S4 in the supporting information available online). Curiously, for all outcomes, there was an increase in heterozygote frequency and a decrease in the rarer GG homozygote in affected individuals (Fig. S1 in the supporting information available online). The change in heterozygote frequency had the effect of causing a statistically significant deviation from Hardy–Weinberg equilibrium in both the yes and no groups for hay fever (P = 0.01). Allele count differences (2 × 2 contingency tables) were only statistically significant for bronchitis and wheeze (P = 0.026 and 0.027 respectively).
The MUC5AC TR genotype variable showed marginally significant association with hay fever (P = 0.044) but was not significant with any other outcomes. Because the MUC5AC TR dataset is somewhat smaller than the SNP dataset (Table S2 in the supporting information available online) because of the requirement for high-quality blood DNA for the Southern blot analysis, tests for association of MUC5AC rs 1132440 with all outcomes were also performed on the smaller dataset. Significance remained for bronchitis, wheeze and hay fever, suggesting that associations with MUC5AC rs 1132440 are stronger than those with MUC5AC TR. For the MUC5B SNP data, a significant association was observed between the exon 2 SNP (rs 2672785) and wheeze at 43 years (P = 0.022). There was a trend towards longer MUC2 TR alleles in the asthma and wheeze groups, but this was not statistically significant.
The measures of lung function (forced expiratory volume in 1 s, forced vital capacity, Δ forced expiratory volume and forced expiratory volume in 1 s/ forced vital capacity), adjusted for gender and height showed just one significant association, namely heterozygotes for MUC5B rs2672785 showed slightly (1%) but significantly reduced forced expiratory volume in 1s/forced vital capacity in 1989 (P = 0.028). This remained significant after full adjustment for the other confounders.
Because MUC5AC rs1132440 showed both stronger association and association with more respiratory outcomes than any of the other loci, all further analyses were conducted using only this locus.
Adjusting for confounders and identifying risk genotypes
In an adjusted model, all previously identified associations remain significant (Table 3).
Table 3.
Association between MUC5AC rs1132440 genotypes and respiratory outcomes
| Outcome | Genotypes (CC versus) | Before adjustment | After adjustment | n | ||
|---|---|---|---|---|---|---|
| OR (95% CI) | P-value | OR (95% CI) | P-value | |||
| Ever bronchitis 1989† | GC | 1.042 (0.83–1.30) | 0.717 | 1.033 (0.83–1.29) | 0.779 | 2361 |
| GG | 0.689 (0.50–0.94) | 0.019 | 0.683 (0.50–0.93) | 0.017 | ||
| Wheeze 1989 ‡ | GC | 0.938 (0.64–1.37) | 0.739 | 0.898 (0.61–1.32) | 0.582 | 2285 |
| GG | 0.382 (0.20–0.74) | 0.005 | 0.372 (0.19–0.73) | 0.004 | ||
| Ever asthma 1999 | GC | 0.975 (0.73–1.31) | 0.865 | 0.979 (0.73–1.31) | 0.887 | 2362 |
| GG | 0.614 (0.40–0.94) | 0.025 | 0.615 (0.40–0.94) | 0.026 | ||
| Ever hay fever 1999 | GC | 1.220 (0.99–1.51) | 0.068 | 1.242 (1.00–1.54) | 0.049 | 2357 |
| GG | 0.930 (0.70–1.23) | 0.614 | 0.949 (0.71–1.26) | 0.721 | ||
Outcomes are described in Table 1. Logistic regression OR and P-values are shown both before and after adjusting for the possible confounders; smoking status, region of birth, father's social class, own social class, gender. Significant associations are shown in bold and OR 95% confidence intervals are in parentheses.
Note that this variable does not exist for 1999.
Wheeze most days and nights; was not statistically significant in 1999.
CI, confidence interval; OR, odds ratio.
The rare homozygote genotypes of rs1132440 appeared to confer protection against bronchitis (odds ratio (95% confidence interval) = 0.689 (0.50–0.94)), wheeze (0.382 (0.20–0.74)) and asthma (odds ratio = 0.614 (0.40–0.94)) as the odds ratios are significantly less than 1 (P-values ≤ 0.025 in all cases) (Table 3). For hay fever, heterozygosity appeared to confer risk (adjusted odds ratio (95% confidence interval) = 1.24 (1.00–1.54); P = 0.049).
The inflammatory markers
Details of the inflammatory response markers tested are given in Table S1.
Significantly different allelic distributions between the affected and unaffected groups were found for: the IL13 promoter SNP (rs1800925) in asthma (P = 0.038); the IL13exonic SNP (rs20541) in asthma (P = 0.0007) *; the ERBB1 SNP (rs2227983) in bronchitis (P = 0.007). For both IL13 SNPs, the rare allele confers risk and is overrepresented in the asthmatic affected group. In contrast, the rare ERBB1 rs2227983 allele is significantly underrepresented in the affected bronchitis group. Logistic regression analysis showed significant associations between genotype and outcome in each of these cases, which remained significant after adjustment for the potential confounders and the association between IL1B and asthma became significant (Table 4).
Table 4.
Association between inflammatory mediator genotypes and respiratory outcomes
| ERBB1 (L and S)* | ERBB1 (rs2227983) | IL13 (rs1800925) | IL13 (rs20541) | IL1B (rs16944) | IL1RN (TR) 2 and X | TNF (rs1800629) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | 1 | 2 | 1 | 2 | 1 | 2 | 1 | 2 | 1 | 2 | 1 | 2 | 1 | 2 |
| Ever bronchitis 1989 | 0.96 (0.75–1.23) | 1.01 (0.76–1.34) | 0.90 (0.72–1.11) | 0.53 (0.32–0.88) | 1.22 (0.98–1.52) | 1.02 (0.57–1.82) | 1.03 (0.82–1.29) | 1.03 (0.55–1.91) | 0.93 (0.75–1.15) | 1.02 (0.72–1.44) | 1.08 (0.87–1.33) | 0.93 (0.64–1.37) | 0.92 (0.73–1.14) | 0.99 (0.57–1.72) |
| Wheeze 1989 | 1.20 (0.77–1.87) | 1.03 (0.60–1.77) | 0.69 (0.46–1.02) | 1.20 (0.62–2.34) | 0.82 (0.54–1.25) | 0.85 (0.30–2.41) | 0.65 (0.41–1.02) | 0.65 (0.20–2.12) | 1.11 (0.76–1.64) | 1.32 (0.74–2.38) | 0.99 (0.67–1.47) | 1.75 (0.99–3.08) | 0.76 (0.50–1.16) | 1.81 (0.83–3.92) |
| Ever asthma 1999 | 1.04 (0.75–1.45) | 1.01 (0.69–1.49) | 1.13 (0.92–1.38) | 1.13 (0.76–1.66) | 1.45 (1.08–1.93) | 1.02 (0.46–2.26) | 1.47 (1.11–1.96) | 1.46 (0.68–3.14) | 1.20 (0.90–1.60) | 1.54 (1.00–2.37) | 1.22 (0.92–1.63) | 1.54 (0.98–2.41) | 0.87 (0.64–1.17) | 1.39 (0.74–2.63) |
| Ever hay fever 1999 | 1.10 (0.87–1.39) | 1.02 (0.78–1.34) | 0.90 (0.67–1.19) | 1.15 (0.68–1.95) | 1.24 (1.00–1.53) | 0.96 (0.55–1.67) | 1.13 (0.91–1.39) | 1.22 (0.69–2.17) | 1.12 (0.92–1.37) | 0.98 (0.70–1.38) | 1.04 (0.85–1.27) | 1.14 (0.81–1.61) | 0.84 (0.68–1.04) | 1.19 (0.73–1.94) |
Logistic regression odds ratios (OR) after adjusting for the confounders listed in Table 3:1 is heterozygote and 2 is homozygous for the rarer or risk allele. Significant associations are shown in bold and OR 95% confidence intervals are in parentheses. ERBB1 microsatellite and IL1RN VNTR are multiallelic so to simplify analysis; the allelic data were binned into two appropriate categories, defined by reviewing the literature for allelic functional relevance. For the ERBB1/EGFR microsatellite, repeat numbers were defined as either short (S) or long (L). S being 8–18 repeats of 20 or greater denoted as L§. The IL1RN tandem repeat lengths were categorized as 2 or X; 2 referring to the IL1RN*2 allele (previously described risk allele) and X includes all other alleles (IL1RN*1, 3, 4 and 5). N values range from 2194–2361.
Tests for gene–gene interactions
Significant interactions with respect to bronchitis were identified between MUC5AC rs1132440 and ERBB1 rs2227983 (P = 0.019), IL1RN VNTR (P = 0.009) and TNFA rs1800629 (P = 0.046). The ERBB1 and IL1RN interactions are illustrated graphically in Figure S2 in the supporting information available online. The association of MUC5AC with bronchitis is only significant in individuals who lack the IL1RN*2 (risk) allele (Fig. S2A in the supporting information available online) and only in individuals homozygous for the ERBB1 common rs2227983(R) allele—that is non-carriers of the rarer K allele (Fig. S2B in the supporting information available online).
Discussion
Abnormal expression of mucinsis a central feature of airway pathology. Here we report significant associations between a MUC5AC SNP rs1132440 and occurrence of asthma, wheeze, hay fever and bronchitis. Although this particular SNP did not show significant association with measures of respiratory function, there was a statistically significant association of forced expiratory volume in 1 s/forced vital capacity with heterozygosity for a SNP within MUC5B which itself is significantly associated with rs1132440. A novel association between bronchitis and an ERBB1/EGFR SNP rs2227983 was also identified and possible gene–gene interactions between MUC5AC and ERBB1 and IL1RN are described.
Variations in the TR regions of the mucin genes that can potentially influence mucus rheology and mucin glycosylation were initially thought to be strong candidates for influencing inflammatory disease susceptibility. The strength of association seen here between the respiratory outcomes and MUC5AC rs1132440 was not however reflected by the MUC5AC TR data, although with hindsight, the use of alternative methods to improve resolution of individual TR alleles may have been more informative.31
The deviations from Hardy–Weinberg equilibrium due to an increased number of heterozygotes in the affected groups suggested that risk may be attributable to more than a simple nucleotide polymorphism and that the region may be subject to copy number variation, with the risk allele being a duplicated region. Efforts to demonstrate this, using the copy number variation detection technique multiplex ligation-dependent probe amplification32 or examining allelic imbalance in heterozygotes were unsuccessful, suggesting that a common duplication was not involved, so the explanation remains elusive.
The evidence of interactions between MUC5AC and various functional variants of genes will need replication, but the newly described interaction between ERBB1 and MUC5AC is of particular interest. The MUC5AC association appears to be dependent upon the presence of two copies of the ancestral ERBB1R allele, which leads to increased signalling in response to ligand binding33 and consequently causes enhanced levels of inflammation, higher risk of bronchitis and presumably higher MUC5AC expression.
In this study, we have not corrected for multiple testing because the initial tests for association were driven by an a priori hypothesis and were between a series of non-independent MUC gene markers (associated because of their close proximity in a gene cluster) and non-independent outcome variables. The inflammatory markers were ones for which there was prior evidence of association with respiratory disease and/or there was evidence of function for the SNP themselves, but the evidence for interactions must be considered preliminary, because the three interactions for bronchitis that were significant at the 5% level were detected in a test involving seven loci and four (albeit non-independent) outcomes (28 tests).
In the main analysis, a total of six MUC gene markers were tested with seven categorical and five respiratory function outcomes, making a total of 72 initial/unadjusted tests for which six were significant at the 5% level. Thus a false positive association of MUC5AC with respiratory outcomes certainly cannot be excluded, but it is noteworthy that this gene region is now repeatedly showing association with respiratory disease. Statistically significant associations have been recently noted between several SNP in the region of MUC5AC and MUC5B and the respiratory diseases familial interstitial pneumonia and idiopathic pulmonary fibrosis.8 These authors claim that a single SNP about 3 kb upstream from the start of transcription of MUC5B is causal, because it is associated with increased expression and the rarer allele is four times more frequent in patients than controls.
A recent study has also shown an association between MUC5AC and cystic fibrosis respiratory disease severity,31 where a specific TR region allele shows strongest association.
An asthma susceptibility locus has also been mapped to 11p15 in a genome-wide linkage study of Caucasian families.34 While this result has not been replicated in any asthma genome-wide association studies, methods used to analyse genome-wide association studies data are extremely conservative and therefore type II errors (false negative) are likely to be extensive.
Here we have shown the strongest associations with a synonymous SNP rs1132440 in the MUC5AC C-terminal region which is unlikely to alter function of the protein, although the G allele is predicted using the bioinformatic software known as ESEfinder to create two exonic splicing enhancer sites (http://fastsnp.ibms.sinica.edu.tw/pages/input_CandidateGeneSearch.jsp). It seems more likely from the preliminary evidence of interaction with genes that play a role in MUC5AC expression that the true functional variant is within a regulatory region and affects MUC5AC expression. This effect is probably distinct from that observed for familial interstitial pneumonia and idiopathic pulmonary fibrosis, which is suggested to involve MUC5B, unless the causal SNP also affects MUC5AC expression, or that the inflammatory mediators also affect MUC5B expression. The causal locus might more likely be the same as that recently found for cystic fibrosis severity31, where a particular TR allele shows the best association. It is of interest that using the publically available European ‘CEPH’ data and software called ‘SNAP' (http://www.broadinstitute.org/mpg/snap/ldsearch.php), it can be seen that rs28514396, which is the best SNP for defining the risk haplotype for cystic fibrosis severity31, is in strong linkage disequilibrium with rs1132440 (r2 0.93; D' of 0.963). The fact that the MUC5AC sequence is still incomplete in the ‘complete’ human genome sequence, and the complexity of the TR region means that it may be some time before the full story can be elucidated, but there seems little doubt that variation in these genes plays a role in respiratory disease susceptibility.
Acknowledgments
We thank Ranji Araseretnam and Mari Wyn Burley for technical assistance and Nik Maniatis and Xue Liang Guo for helpful discussion. This study was funded by the Medical Research Council (UK); L.J. by an MRC DTA studentship and the Annals of Human Genetics, AXL by a UCL ORS scholarship.
Glossary
- TR
tandem repeat
Footnotes
The associations between both IL13 SNP and the asthma outcome on this dataset have been reported previously.28
Supporting Information
Additional Supporting Information may be found in the online version of this article at the publisher's website:
Figure S1 Bar charts showing as percentages the MUC5AC rs1132440 genotype frequency distribution in the affected and unaffected groups for outcomes bronchitis 1989, wheeze most days and nights 1989, ever asthma (1999) and ever hay fever (1999). See Table S1 for n values. Note that in each case the heterozygotes show elevated frequencies in the yes group while the rarer homozygotes show lower frequency.
Figure S2 Bar charts showing the distributions as percentages of the MUC5AC rs 1132440 genotypes in carriers and non-carriers of minor/risk alleles of ERBB1 and IL1RN with respect to Ever Bronchitis. A IL1RN *2 a) carriers of the IL1RN risk allele (22, 23, 24, 25, 26) (NS).b) non-carriers of the IL1RN risk allele (34, 44, 45, 46, 55) P = 0.004. B ERBB1 rs2227983. (a) carriers of the rare ERBB1 K allele (AA and AG) (NS). (b) non-carriers (GG, and homozygous for the ancestral allele R) P < 0.001. P-value from Pearson chi-square; NS, not significant.
Table S1 Details of variants within the genes encoding the inflammatory mediators and functional evidence for these. Note that the marker names refer to physical positions within the gene or its regulatory sequence. ‡ rs2227983 was previously listed as rs11543848 (NBCI dbSNP). Minor allele frequencies (MAF) range from 0.17 to 0.49 and genotype distributions were in accordance with Hardy-Weinberg equilibrium, with the exception of IL1B rs16944 (P = 0.01).
Table S2 MUC5AC and MUC5B polymorphisms typed on the 1946 cohort. Note that for simplicity the MUC5AC TR alleles have been categorized into a tri-allelic model where r refers to all rare alleles.
Table S3 Pairwise linkage disequilibrium (LD) measures for MUC5AC and MUC5B markers. (a) Significance of association is shown as chi-square P-values. (b) D′ measure of LD. Loci are shown in chromosomal order from MUC5AC TR through to rs2075859 in exon 9 of MUC5B (see Table S1) Note that all adjacent SNP are highly associated with each other. Although there is breakdown of LD in between exons 2 and 9 of MUC5B, with no significant association between rs2672785 and rs2075859, LD is still detectable across the two MUC5 genes since MUC5AC rs1132440 and MUC5ACTR are each significantly associated with one of the MUC5B SNP even though the D' values are small. Significant values are shown in bold.
Table S4 Chi-square P-values from contingency tables of MUC5AC and MUC5B genotypes and the categorical respiratory outcomes; Mann-Whitney P- value for MUCTR. N values given are for MUC5AC rs1132440; those for other loci, in particular MUC2 TR and MUC5AC TR are a little lower (see manuscript text). y/n signifies yes or no as indicated on Table 1. Similar empirical P-values were obtained by permutation analysis.
References
- 1.Kuyper LM, Pare PD, Hogg JC, et al. Characterization of airway plugging in fatal asthma. Am. J. Med. 2003;115:6–11. doi: 10.1016/s0002-9343(03)00241-9. [DOI] [PubMed] [Google Scholar]
- 2.Kirkham S, Sheehan JK, Knight D, et al. Heterogeneity of airways mucus: variations in the amounts and glycoforms of the major oligomeric mucins MUC5AC and MUC5B. Biochem. J. 2002;361:537–546. doi: 10.1042/0264-6021:3610537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Burgel PR, Montani D, Danel C, et al. A morphometric study of mucins and small airway plugging in cystic fibrosis. Thorax. 2007;62:153–161. doi: 10.1136/thx.2006.062190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Caramori G, Di GC, Carlstedt I, et al. Mucin expression in peripheral airways of patients with chronic obstructive pulmonary disease. Histopathology. 2004;45:477–484. doi: 10.1111/j.1365-2559.2004.01952.x. [DOI] [PubMed] [Google Scholar]
- 5.Groneberg DA, Eynott PR, Oates T, et al. Expression of MUC5AC and MUC5B mucins in normal and cystic fibrosis lung. Respir. Med. 2002;96:81–86. doi: 10.1053/rmed.2001.1221. [DOI] [PubMed] [Google Scholar]
- 6.Kamio K, Matsushita I, Hijikata M, et al. Promoter analysis and aberrant expression of the MUC5B gene in diffuse panbronchiolitis. Am. J. Respir. Crit. Care Med. 2005;171:949–957. doi: 10.1164/rccm.200409-1168OC. [DOI] [PubMed] [Google Scholar]
- 7.Young HW, Williams OW, Chandra D, et al. Central role of Muc5ac expression in mucous metaplasia and its regulation by conserved 5' elements. Am. J. Respir. Cell Mol. Biol. 2007;37:273–290. doi: 10.1165/rcmb.2005-0460OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Seibold MA, Wise AL, Speer MC, et al. A common MUC5B promoter polymorphism and pulmonary fibrosis. N. Engl. J. Med. 2011;364:1503–1512. doi: 10.1056/NEJMoa1013660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bradding P, Roberts JA, Britten KM, et al. Interleukin-4, -5, and -6 and tumor necrosis factor-alpha in normal and asthmatic airways: evidence for the human mast cell as a source of these cytokines. Am. J. Respir. Cell Mol. Biol. 1994;10:471–480. doi: 10.1165/ajrcmb.10.5.8179909. [DOI] [PubMed] [Google Scholar]
- 10.Broide DH, Lotz M, Cuomo AJ, et al. Cytokines in symptomatic asthma airways. J. Allergy Clin. Immunol. 1992;89:958–967. doi: 10.1016/0091-6749(92)90218-q. [DOI] [PubMed] [Google Scholar]
- 11.Humbert M, Durham SR, Kimmitt P, et al. Elevated expression of messenger ribonucleic acid encoding IL-13 in the bronchial mucosa of atopic and nonatopic subjects with asthma. J. Allergy Clin. Immunol. 1997;99:657–665. doi: 10.1016/s0091-6749(97)70028-9. [DOI] [PubMed] [Google Scholar]
- 12.Naseer T, Minshall EM, Leung DY, et al. Expression of IL-12 and IL-13 mRNA in asthma and their modulation in response to steroid therapy. Am. J. Respir. Crit. Care Med. 1997;155:845–851. doi: 10.1164/ajrccm.155.3.9117015. [DOI] [PubMed] [Google Scholar]
- 13.Ying S, Meng Q, Barata LT, et al. Associations between IL-13 and IL-4 (mRNA and protein), vascular cell adhesion molecule-1 expression, and the infiltration of eosinophils, macrophages, and T cells in allergen-induced late-phase cutaneous reactions in atopic subjects. J. Immunol. 1997;158:5050–5057. [PubMed] [Google Scholar]
- 14.Gray T, Nettesheim P, Loftin C, et al. Interleukin-1beta-induced mucin production in human airway epithelium is mediated by cyclooxygenase-2, prostaglandin E2 receptors, and cyclic AMP-protein kinase A signaling. Mol. Pharmacol. 2004;66:337–346. doi: 10.1124/mol.66.2.337. [DOI] [PubMed] [Google Scholar]
- 15.Gray T, Coakley R, Hirsh A, et al. Regulation of MUC5AC mucin secretion and airway surface liquid metabolism by IL-1beta in human bronchial epithelia. Am. J. Physiol. Lung Cell. Mol. Physiol. 2004;286:L320–L330. doi: 10.1152/ajplung.00440.2002. [DOI] [PubMed] [Google Scholar]
- 16.Koo JS, Kim YD, Jetten AM, et al. Overexpression of mucin genes induced by interleukin-1 beta, tumor necrosis factor-alpha, lipopolysaccharide, and neutrophil elastase is inhibited by a retinoic acid receptor alpha antagonist. Exp. Lung Res. 2002;28:315–332. doi: 10.1080/01902140252964393. [DOI] [PubMed] [Google Scholar]
- 17.Smirnova MG, Birchall JP, Pearson JP. TNF-alpha in the regulation of MUC5AC secretion: some aspects of cytokine-induced mucin hypersecretion on the in vitro model. Cytokine. 2000;12:1732–1736. doi: 10.1006/cyto.2000.0763. [DOI] [PubMed] [Google Scholar]
- 18.Hewson CA, Haas JJ, Bartlett NW, et al. Rhinovirus induces MUC5AC in a human infection model and in vitro via NF-kappaB and EGFR pathways. Eur. Respir. J. 2010;36:1425–1435. doi: 10.1183/09031936.00026910. [DOI] [PubMed] [Google Scholar]
- 19.Kohri K, Ueki IF, Shim JJ, et al. Pseudomonas aeruginosa induces MUC5AC production via epidermal growth factor receptor. Eur. Respir. J. 2002;20:1263–1270. doi: 10.1183/09031936.02.00001402. [DOI] [PubMed] [Google Scholar]
- 20.Burgel PR, Lazarus SC, Tam DC, et al. Human eosinophils induce mucin production in airway epithelial cells via epidermal growth factor receptor activation. J. Immunol. 2001;167:5948–5954. doi: 10.4049/jimmunol.167.10.5948. [DOI] [PubMed] [Google Scholar]
- 21.Cervantes-Sandoval I, Serrano-Luna JJ, Meza-Cervantez P, et al. Naegleria fowleri induces MUC5AC and pro-inflammatory cytokines in human epithelial cells via ROS production and EGFR activation. Microbiology. 2009;155:3739–3747. doi: 10.1099/mic.0.030635-0. [DOI] [PubMed] [Google Scholar]
- 22.Kang JH, Lee EH, Park SW, et al. MUC5AC expression through bidirectional communication of Notch and epidermal growth factor receptor pathways. J. Immunol. 2011;187:222–229. doi: 10.4049/jimmunol.1003606. [DOI] [PubMed] [Google Scholar]
- 23.Vinall LE, Fowler JC, Jones AL, et al. Polymorphism of human mucin genes in chest disease: possible significance of MUC2. Am. J. Respir. Cell Mol. Biol. 2000;23:678–686. doi: 10.1165/ajrcmb.23.5.4176. [DOI] [PubMed] [Google Scholar]
- 24.Davies JR, Svitacheva N, Lannefors L, et al. Identification of MUC5B, MUC5AC and small amounts of MUC2 mucins in cystic fibrosis airway secretions. Biochem. J. 1999;344(Pt 2):321–330. [PMC free article] [PubMed] [Google Scholar]
- 25.Hovenberg HW, Davies JR, Herrmann A, et al. MUC5AC, but not MUC2, is a prominent mucin in respiratory secretions. Glycoconj. J. 1996;13:839–847. doi: 10.1007/BF00702348. [DOI] [PubMed] [Google Scholar]
- 26.Rousseau K, Byrne C, Griesinger G, et al. Allelic association and recombination hotspots in the mucin gene (MUC) complex on chromosome 11p15.5. Ann. Hum. Genet. 2007;71:561–569. doi: 10.1111/j.1469-1809.2007.00374.x. [DOI] [PubMed] [Google Scholar]
- 27.Wadsworth ME, Butterworth SL, Hardy RJ, et al. The life course prospective design: an example of benefits and problems associated with study longevity. Soc. Sci. Med. 2003;57:2193–2205. doi: 10.1016/s0277-9536(03)00083-2. [DOI] [PubMed] [Google Scholar]
- 28.Black S, Teixeira AS, Loh AX, et al. Contribution of functional variation in the IL13 gene to allergy, hay fever and asthma in the NSHD longitudinal 1946 birth cohort. Allergy. 2009;64:1172–1178. doi: 10.1111/j.1398-9995.2009.01988.x. [DOI] [PubMed] [Google Scholar]
- 29.Mann SL, Wadsworth ME, Colley JR. Accumulation of factors influencing respiratory illness in members of a national birth cohort and their offspring. J. Epidemiol. Community Health. 1992;46:286–292. doi: 10.1136/jech.46.3.286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McFadden E, Luben R, Wareham N, et al. How far can we explain the social class differential in respiratory function? A cross-sectional population study of 21,991 men and women from EPIC-Norfolk. Eur. J. Epidemiol. 2009;24:193–201. doi: 10.1007/s10654-009-9326-y. [DOI] [PubMed] [Google Scholar]
- 31.Guo X, Pace RG, Stonebraker JR, et al. Mucin variable number tandem repeat polymorphisms and severity of cystic fibrosis lung disease: significant association with MUC5AC. PLoS ONE. 2011;6:e25452. doi: 10.1371/journal.pone.0025452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Schouten JP, McElgunn CJ, Waaijer R, et al. Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res. 2002;30:e57. doi: 10.1093/nar/gnf056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Moriai T, Kobrin MS, Hope C, et al. A variant epidermal growth factor receptor exhibits altered type alpha transforming growth factor binding and transmembrane signaling. Proc. Natl Acad. Sci. U.S.A. 1994;91:10217–10221. doi: 10.1073/pnas.91.21.10217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.CSGA. A genome-wide search for asthma susceptibility loci in ethnically diverse populations. The Collaborative Study on the Genetics of Asthma (CSGA) Nat. Genet. 1997;15:389–392. doi: 10.1038/ng0497-389. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1 Bar charts showing as percentages the MUC5AC rs1132440 genotype frequency distribution in the affected and unaffected groups for outcomes bronchitis 1989, wheeze most days and nights 1989, ever asthma (1999) and ever hay fever (1999). See Table S1 for n values. Note that in each case the heterozygotes show elevated frequencies in the yes group while the rarer homozygotes show lower frequency.
Figure S2 Bar charts showing the distributions as percentages of the MUC5AC rs 1132440 genotypes in carriers and non-carriers of minor/risk alleles of ERBB1 and IL1RN with respect to Ever Bronchitis. A IL1RN *2 a) carriers of the IL1RN risk allele (22, 23, 24, 25, 26) (NS).b) non-carriers of the IL1RN risk allele (34, 44, 45, 46, 55) P = 0.004. B ERBB1 rs2227983. (a) carriers of the rare ERBB1 K allele (AA and AG) (NS). (b) non-carriers (GG, and homozygous for the ancestral allele R) P < 0.001. P-value from Pearson chi-square; NS, not significant.
Table S1 Details of variants within the genes encoding the inflammatory mediators and functional evidence for these. Note that the marker names refer to physical positions within the gene or its regulatory sequence. ‡ rs2227983 was previously listed as rs11543848 (NBCI dbSNP). Minor allele frequencies (MAF) range from 0.17 to 0.49 and genotype distributions were in accordance with Hardy-Weinberg equilibrium, with the exception of IL1B rs16944 (P = 0.01).
Table S2 MUC5AC and MUC5B polymorphisms typed on the 1946 cohort. Note that for simplicity the MUC5AC TR alleles have been categorized into a tri-allelic model where r refers to all rare alleles.
Table S3 Pairwise linkage disequilibrium (LD) measures for MUC5AC and MUC5B markers. (a) Significance of association is shown as chi-square P-values. (b) D′ measure of LD. Loci are shown in chromosomal order from MUC5AC TR through to rs2075859 in exon 9 of MUC5B (see Table S1) Note that all adjacent SNP are highly associated with each other. Although there is breakdown of LD in between exons 2 and 9 of MUC5B, with no significant association between rs2672785 and rs2075859, LD is still detectable across the two MUC5 genes since MUC5AC rs1132440 and MUC5ACTR are each significantly associated with one of the MUC5B SNP even though the D' values are small. Significant values are shown in bold.
Table S4 Chi-square P-values from contingency tables of MUC5AC and MUC5B genotypes and the categorical respiratory outcomes; Mann-Whitney P- value for MUCTR. N values given are for MUC5AC rs1132440; those for other loci, in particular MUC2 TR and MUC5AC TR are a little lower (see manuscript text). y/n signifies yes or no as indicated on Table 1. Similar empirical P-values were obtained by permutation analysis.