Does the relationship between IgE and the CD14 gene depend on ethnicity?
School of Paediatrics and Child Health
University of Western Australia
This review considers the data from studies analysing associations between the CD14C−159T single nucleotide polymorphism (SNP) and asthmatic phenotypes and discusses the variability of the conclusions. By searching PubMed and EMBASE for articles on CD14C−159T -related population or family-based association studies, 47 were identified up till September 2007. Collectively, the studies reviewed herein consistently showed population differences in frequencies of the alleles of the SNP, with African descent having the highest C allele frequencies, followed by Caucasians and Asians. The T allele of the SNP was associated with increased sCD14 in some studies but not in others. Inconsistently, the C allele, or even occasionally the T allele, was associated with atopic phenotypes in a population subgroup. There are several explanations for these inconsistencies, including lack of power, linkage disequilibrium, gene–gene interactions, population admixture and gene–environment interactions. If the SNP was associated with functional changes to the coded protein and thus modulating susceptibility to allergic disease, its effect may be modest and dependent on other co-existent, ethnicity-specific, genetic or environmental risk factors.
The gene encoding CD14 is located in the chromosome region 5q31.1 (1), which has been found to be linked to atopy and asthma in several populations (2). The CD14 protein, which is a receptor for lipopolysaccharides (LPS) and other bacterial wall-derived components, is constitutively expressed primarily on the surface of monocytes, macrophages and neutrophils as membrane CD14 (mCD14) (3). Also abundant in serum, CD14 has a soluble form (sCD14), which is derived both from secretion of CD14 and from enzymatically cleaved glycosyl-phosphatidylinositol-anchored mCD14 (3). As an important high-affinity receptor binding to LPS, CD14 has been suggested to play a critical role in favouring Th1 differentiation of naive T cells, with the likelihood of developing a T helper 1 (Th1)-type response by up-regulating the expression of IL-12, which in turn decreases IgE-mediated immunity (4).
A genetic variant [CD14C−159T (rs2569190)] was found in the promoter region of the CD14 gene and this single nucleotide polymorphism (SNP) was associated with higher levels of sCD14 expression, and inversely associated with levels of total serum IgE, in an hypothesis driven study (4). Since this initial report, many studies have been published on relationships between the CD14 polymorphism and asthma and associated traits. However, the functional significance of CD14C−159T SNP in the causation of atopy and atopic phenotypes is unclear as the same allele has been associated with opposite effects in different studies. It is possible that these variable associations with the SNP result from ethnic related differences.
This review will summarize the published population and family-based association genetic studies on the SNP CD14C−159T in relation to atopic phenotypes in humans. Publications were identified by the systematic searches of PubMed and EMBASE with the key word of ‘CD14’ in combination with either of the following terms: ‘asthma’, ‘atopy’, ‘rhinitis’, ‘IgE’, ‘specific IgE’, ‘dermatitis’, ‘atopic’, ‘eosinophils’, ‘bronchial hyperresponsiveness’, ‘airway hyper-reactivity’ and ‘SNP’. The downloaded articles were manually screened by reading the titles and their abstracts. In order to review the related-articles as thoroughly as possible we also screened the references of the identified publications. With these paper screening strategies, we identified 47 articles reporting population or family-based association studies, up till September 2007, which involved associations between the SNP and asthma-associated phenotypes. Forty-five of the publications were in English and one each in Polish and Chinese. In the present review, we do not intend to pool all the data for a meta-analysis as the study designs, phenotypic end points, ethnic backgrounds and environments were significantly different among the studies. This review will critically analyse these reports, with the aim of clarifying the discordant data with respect to the association between the SNP CD14C−159T and atopic phenotypes in humans.
C allele frequency in different ethnic groups
On account of the historical isolation and environmental positive selections major racial/ethnic groups differ in allele frequencies (5). Consequently, ethnic differences may contribute to the variations in allele frequencies of phenotype-associated genes in humans. With respect to the SNP CD14C−159T in the promoter region of CD14, there are significant variations in allele frequencies in different ethnic groups. Table 1 summarizes C allele frequencies for the reviewed studies. There is a consistent pattern in C allele frequencies in the different ethnic groups, highest in those of African descent, followed by Caucasians and Asians. This indicates that there are significant genetic variations between these ethnic groups in terms of predominant alleles and linkage disequilibrium structure of the SNP CD14C−159T. It is generally agreed that the ethnicity of study subjects is a key factor in interpreting genetic polymorphism studies. Indeed, several studies reviewed herein have attributed their inconsistent findings to ethnic differences (6–8), thus emphasizing caution in generalizing the findings of genetic association studies in a specific ethnic group, as racial and ethnicity-specific genetic modifiers may modulate relative risk.
Table 1. The C allele frequencies of the SNP CD14C−159T in different ethnic populations
|Caucasian in America, Canada and Australia|
|Baldini (4)||314 (51.4)||Americans||Woo (40)||142 (50.0)||Americans|
|Jackola (12)||367 (52.7)||Americans||Kedda (41)||1011 (52.1)||Australians|
|Gern (35)||285 (52.3)||Americans*||O’Donnell (42)||305 (50.3)||Australians|
|Litonjua (14)||344 (53.1)||Americans||Bernstein (28)||137 (51.8)||Canadians|
|Caucasian in Europe|
|Kabesch (6)||2936 (53.0)||Germany||Campos (37)||415 (50.6)||Spanish|
|Bottcher (10)||115 (58.3)||Swedish||Eder (31)||624 (51.0)||Germany|
|Leynaert (36)||600 (50.6)||French||Heinzmann (43)||443 (53.4)||Germany|
|Koppelman (44)||317 (51.2)||Dutch||Sackesen (45)||613 (47.5)||Turkish|
|Simpson (32)||442 (51.5)||English||Buckova (20)||882 (54.0)||Czech|
|Japanese and Koreans|
|Takeuchi (7)||159 (45.0)||Japanese||Campos (46)||189 (45.2)||Japanese|
|Inoue (15)||54 (43.5)||Japanese||Kang (47)||324 (51.2)||Koreans|
|Nishimura (18)||466 (48.5)||Japanese||Hong (27)||779 (40.1)||Koreans|
|Tan (24)||240 (40.2)||Chinese||Liang (8)||331 (45.1)||Chinese|
|Wang (21)||200 (38.9)||Chinese||Leung (9)||350 (44.0)||Chinese|
|Zhang (48)||180 (33.0)||Chinese|| || || |
|Cardaba (11)||87 (45.5)||Spanish Gypsies||Choudhry (22)||1977 (51.7)||Latin Americans|
|Zambell-Weiner (34)||747 (67.9)||Barbados||Williams (33)||517 (59.0)||America Africa and white|
|Sharma (49)||414 (42.3)||India|| || || |
CD14C−159T and soluble CD14
The T allele of the CD14C−159T SNP has been consistently suggested to be associated with higher levels of soluble CD14 (sCD14) when compared to the C allele, on account of the genetic variation having a functional effect or being in linkage with other functional SNPs. With regard to the articles discussed in this review, the first study reported that 42 TT homozygotes had significantly higher sCD14 levels than did 67 carriers of the CC genotypes in the population of the Tucson Children’s Respiratory Study (CRS) (4). Concordant with this, in a sub-sample (n = 545) of a large population study in Germany a trend for elevated sCD14 levels in carriers of TT genotypes was found in children and adult populations of three centres (Munich, Dresden and Giessen), which was significant (P < 0.03) when all three populations were analysed together (6). In addition, another study also replicated this in 349 Chinese children (9). However, in two relatively small population studies, no association was found between the SNP and sCD14 (10, 11), which may be on account of the lack of sufficient statistical power or differences in ethnicity. Another Chinese population study (301 children) (8) in Singapore failed to replicate the significant relationship, although the sCD14 levels appeared to be higher in TT homozygotes (not significant). Furthermore, in a Caucasian population (n = 367) (12) TT homozygotes seemed to have lower sCD14 levels when compared to CC homozygotes with a marginal significance (P = 0.04). As subjects in the study were from 91 nuclear families, the significance may disappear after adjusting for random family effects. In a cohort of children with acute asthma (n = 148) in Western Australia, we have also reported that TT homozygotes had the highest levels of sCD14 followed by CT heterozygotes and CC homozygotes (13). As other studies have investigated associations between CD14C−159T SNP, sCD14 and other nonasthmatic phenotypes, we searched PubMed with the terms: ‘polymorphism’ and ‘soluble CD14’, and, in addition to the above-mentioned studies, 20 such studies were ascertained. Most of these studies (14) consistently reported that the T allele of CD14C−159T SNP was associated with elevated plasma sCD14 compared to the C allele, with nine studies failing to show the association. No study reported an inverse relationship. Considering the possible modest effects of the T allele on sCD14, and many confounders which may have an influence on sCD14, it is not unexpected that in some populations the association between the CD14C−159T SNP and sCD14 could not be replicated.
The association between the CD14C−159T SNP and sCD14 may be attributed to the linkage of the SNP with other SNPs in the CD14 gene. A Japanese study reported that, rather than the T allele of CD14C−159T, CC homozygotes of CD14C-550T had a higher level of sCD14 in cord blood and peripheral blood at age of 6 years compared with CT or TT carriers (15). However, the C allele in CD14C-550T was in linkage disequilibrium with the T allele in CD14C−159T in the Japanese population (15). As further evidence for haplotype effects, Vercelli et al. (16) reported that carriers of the -1359G/-1145G/-159T haplotype had the highest levels of sCD14, and conversely −1359T/−1145A/−159C carriers had the lowest levels of sCD14 in the CRS population.
CD14C−159T and atopic phenotypes
As the initial report suggested that TT homozygotes had lower levels of IgE than did carriers of CC/CT, there have been another 18 studies in which the T allele has been associated with a lower risk for either total or specific IgE, atopy or asthma (Table 2), although the association was confined to subgroups of subjects and with some atopic phenotypes. Those studies have covered a range of ethnic populations, including Caucasians in America, Australia and Europe, Africans, Chinese, Koreans and Japanese. These findings are consistent with the evidence that the T allele may increase CD14 expression by decreasing affinity of DNA/protein interactions at a GC box that contains a binding site for Sp1, Sp2 and Sp3 transcription factors (17), in turn favouring Th1 differentiation of naive T cells, priming a Th1 dominant immune response and decreasing IgE-mediated immunity (4). However, the association is not consistent as in the more than 20 published studies, which also comprised different ethnic groups, the relationship was not replicated. For example, one larger sample size population study in Germany failed to replicate the association between the SNP and IgE in the same Caucasian population (6). The association was also not replicated in Chinese (8) and Japanese populations (18). In contrast, there are seven studies (Table 3) in which the T allele was associated with a higher risk for total or specific IgE, atopy, wheezing or asthma. The seven study populations included Caucasian, Chinese and Hutterites and also included adults and children, who were located in farm or urban areas. Clearly the inconsistencies could not be explained by ethnic and age differences between studies. Considering the SNP has been associated with atopic phenotypes in many hypothesis-driven studies, and the possible effects of CD14 on immune systems has been generally accepted (19), it is unlikely that these inconsistencies are attributable to flaws in the studies’ designs or spurious findings by chance. Therefore, it is likely the SNP has a modest modifying effect on IgE-mediated immunity, and if it is associated with a susceptibility to asthmatic phenotypes, the effect may be confined to population sub-groups with co-existent risk factors.
Table 2. Studies on CD14C−159T indicating T allele with a lower risk for IgE, atopy and asthma
|Baldini et al. (4)||1999||Non-Hispanic whites||TT homozygotes had lower levels of IgE than did carriers of CC/CT, but differences were significant only among children who were skin test-positive to local aeroallergens.|
|Gao et al. (50)||1999||British||Mean geometric IgE level was lower among those with TT/CT genotype than among those with CC genotype. This was significant among those with negative RAST.|
|Koppelman et al. (44)||2001||Dutch||CC homozygotes had a higher number of positive SPT and higher total IgE levels (in SPT+) and subsequently, more self-reported allergic symptoms for rhinitis and hay fever when compared to CT/TT carriers. |
|Leung et al. (9)||2003||Chinese||Atopic subjects with CC genotype had the higher total IgE when compared to those with CT/TT genotypes.|
|Buckova et al. (20)||2003||Caucasians exclusively Czech||The C allele was associated with an increased risk for SPT mould positive.|
|O’Donnell et al. (42)||2004||Caucasians in Australia||CC homozygotes were at higher risk for early-onset atopy and AHR when compared to those with CT/TT genotypes.|
|Gern et al. (35)||2004||Americans with 14% ethnic minorities||Infants with TT were less likely to develop atopic dermatitis if they were exposed to a dog.|
|Sharma et al. (49)||2004||Indians||The C allele was associated with atopic asthma and high total IgE.|
|Takeuchi et al. (7)||2005||Japanese||Among the allergic rhinitis patients, TT homozygotes had significantly fewer positive CAP RAST tests than those with CC/CT genotypes.|
|Choudhry et al. (22)||2005||Latin Americans||The T allele, interacting with ETS exposure, was associated with lower total IgE when compared to the C allele.|
|Zambelli-Weiner et al. (34)||2005||African descent||The T allele appeared to be protective against asthma and asthma severity.|
|Eder et al. (31)||2005||Germany in Austria and Bavaria||The C allele was associated with higher levels of total and specific IgE in children exposed to pets.|
|Sackesen (45)||2005||Turkish||The TT genotype was associated with lower total IgE levels.|
|Tan et al. (24)||2006||Chinese||Subjects with the CD14−159C/−1145A/−1359T haplotype had the highest IgE compared to other haplotypes.|
|Leynaert et al. (36)||2006||French||The T allele was associated with a lower risk for nasal allergies, atopy and elevated specific IgE.|
|Campos et al. (37)||2006||Spanish||Heavy drinkers with the C allele had higher levels of total and specific IgE.|
|Williams et al. (33)||2006||Africans and Caucasians in America||The T allele was associated with decreased total and specific IgE.|
|Buckova et al. (23)||2006||Caucasians exclusively Czech||The C allele was associated with positive SPT for mites and moulds with marginal significance.|
|Kang et al. (47)||2006||Koreans||Among subjects with SPT+, the CC homozygotes were associated with higher total IgE levels and greater number of SPT+, compared with subjects with CT/TT genotypes.|
Table 3. Seven studies on CD14C−159T indicating T allele with a higher risk for IgE, atopy and asthma
|Ober et al. (25)||2000||Hutterites||The T allele was overtransmitted to individuals with atopy.|
|Woo et al. (40)||2003||Africans and Caucasians in America||Among the asthmatics, the T allele was associated with an increased risk for nonatopic asthma and food allergy.|
|Wang et al. (21)||2005||Chinese||The T allele was associated with elevated IgE when the T allele was part of a haplotype containing a D5S2011 E allele.|
|Litonjua et al. (14)||2005||Caucasian in America||Infants with CC/CT genotypes had significantly increase risk for eczema and elevated total IgE.|
|Kedda et al. (41)||2005||Caucasian in Australia|| There was a weak association (P = 0.084) with the T allele that was associated with atopy.|
|LeVan et al. (51)||2005||American farmers||The T allele was associated with decreased lung function and increased wheeze when compared to the C allele.|
|Jackola et al. (12)||2006||Caucasians in America||The TT genotype was associated increased risk for atopy with a marginal significance.|
Linkage or interaction with other polymorphisms
Inconsistencies in genetic association studies have been suggested to result from linkage of one candidate SNP with other causal SNPs in the same gene or chromosome region or from interaction with other genes in the same pathogenetic pathway. Thus CD14C−159T may be in linkage disequilibrium with one or more causal SNPs in the promoter area or other regions of the CD14 gene. Therefore, it is possible that one of these polymorphisms, rather than CD14C−159T, may be responsible for the association with atopic phenotypes. Among the reviewed papers, there are 12 studies in which other SNPs in the CD14 gene, as well as the CD14C−159T, have been investigated in terms of their effects on asthmatic phenotypes. There are more than twenty SNPs reported in the promoter areas and downstream of the CD14C−159T with strong linkage disequilibrium between them (http://www.ncbi.nlm.nih.gov/sites/entrez?db=gene). There is evidence that the CD14C−159T SNP is in the same LD block with several SNPs, which have also been associated with atopic phenotypes, such as −C2984T, −A1619G, −G1359T and −C550T in the promoter area and +C1437G and +G1342T in the downstream region (15, 20–23). Two studies, one in Caucasians (16) and the other in Chinese (24), have consistently shown that carriers of the −1359T/−1145A/−159C haplotype had the highest levels of IgE and lowest sCD14. However, another study in the Chinese found that the T allele of CD14C−159T SNP was associated with elevated IgE level when −159T was part of a haplotype containing a D5S2011 E allele (STR marker), which is approximately 1.1 MB distant from the CD14 gene (21). Consistent with this, Ober et al. (25) reported that the −159T allele was overtransmitted to Hutterites with atopy, occurring only when the allele was on a haplotype with the D5S642 STR marker and not when it was associated with other haplotypes. Considering the complexity of haplotype combinations in different ethnic groups, more studies are necessary to elucidate the association between haplotypes of the CD14 gene and asthmatic phenotypes.
Variants in other genes, which share the same pathway or phenotypes with CD14 in terms of IgE immunity and asthmatic traits, may also influence the association between the CD14C−159T SNP and asthmatic phenotypes. For example, in a population of African descent in Barbados a significant interaction was found between the acyloxyacyl hydroxylase (AOAH) SNP (rs2727831) and CD14C−159T for asthma and total IgE (26). In an Asian population, asthmatic subjects with the combination (GA+AA/CT+CC) of the two SNPs of TNF-aG308A and CD14C−159T had the lowest log PC20 when compared to those with other genotype combinations (27). In addition, in a study of occupational asthma, the combination of IL4RA (I50V) II and CD14 (C−159T) CT genotypes and IL4RA (I50V) II, IL-13 (R110Q) RR and CD14 (C−159T) CT triple genotype combinations were associated with Diisocyanate asthma among exposed individuals (28). As with other complex diseases, a single gene may only make a modest contribution to the susceptibility to asthma and allergy. Therefore, multiple, interacting, nonallelic genetic variants, in the same or related pathways, may be associated with the inherited susceptibility to these conditions.
Interaction with environment
Differences in exposures to environmental risk factors may modulate the effect of the CD14C−159T SNP on asthmatic phenotypes. Vercelli (29) proposed an endotoxin-switch model, in which the CD14 promoter polymorphism changes the threshold at which environmental endotoxin stimulation leads to a Th2 immune response. It is tempting to use the endotoxin-switch model to explain the inconsistencies with the reported associations between the CD14C−159T SNP and asthmatic phenotypes, as the polymorphism may have a different influence on the development of atopy at low or high levels of environmental exposure to endotoxins. In recent reviews, Martinez (30) discussed how a genetic variant may be associated with asthmatic phenotypes in one environment and be protective against asthmatic phenotypes in another environment as a specific protein may have opposite effects when involved in a coordinated response to different environmental exposures (3). There were four model studies, three population- (31–33) and one family-based (34), that were consistent with Martinez’s hypothesis. The four studies showed that TT homozygotes of the CD14C−159T SNP, with possible higher expression of CD14, exhibited protective effects against atopy with low exposure to endotoxin, while, in contrast, being at increased risk for atopy with high exposure to endotoxin. However, among the four studies, the significant interaction (denoted as the interaction term in the model) between endotoxin exposure and the CD14C−159T SNP was not reported in two of the studies and only approached borderline significance in the other two studies. In addition, Gern et al. (35) reported that infants with the genotype −159TT were less likely to develop atopic dermatitis if they were exposed to a dog. Leynaert et al. (36) found that the risk of nasal allergies and atopy was most reduced in the subjects who had the combination of an early-life exposure to a farming environment and the −159TT genotype. These two studies differed from the four others in term of the interaction between endotoxin exposure and the CD14C−159T polymorphism.
Besides endotoxin exposure, other environmental risk factors might also modify the association. Campos et al. (37) found that alcohol consumption modified the effect of the CD14C−159T polymorphism on serum IgE levels. Choudhry et al. (22) reported that there was a significant interaction between plasma IgE levels, SNP CD14C−159T genotypes, and environmental tobacco-smoke (ETS) exposure. The variations in environmental factors across studies, which may be important in modifying genetic associations, may be partly responsible for the inconsistencies reported in the associations between the CD14C−159T polymorphism and asthmatic phenotypes.
Population stratification is another significant confounder for genetic association studies (38). As mentioned above, there are significant variations in the genotype frequencies of the CD14C−159T polymorphism across different ethnic populations. The genetic variations of the SNP and ethnic disparities observed for susceptibility to asthmatic phenotypes may work together to distort the association in a mixed population with the effects of population stratification. For example, individuals of African descent have higher frequencies of the C allele of CD14C−159T SNP and also have higher prevalences of asthma when compared to Caucasian populations (39). In an African and Caucasian mixed population, the C allele will more likely be found to associate with asthmatic phenotypes as a risk allele on account of population admixture. The mixture of populations of African and Caucasian descent may be present for several studies reviewed herein. Accordingly, strategies have been utilized to account for the effects of population admixture such as family based analysis and adjustment for the ethnic variable. For the latter, it is not possible to adjust for the hidden population admixture without sufficient numbers of ancestry-informative markers genotyped. Therefore, spurious associations caused by population stratification may also be a concern for some studies in term of the relationship between the CD14C−159T SNP and asthmatic phenotypes.
The CD14C−159T SNP shows significant allele frequency variations in different ethnic populations. More than 40 studies have been reported, which investigated the association of the SNP with asthmatic phenotypes in different human populations and family-based cohorts. The −159T allele seemed to be associated with elevated sCD14, as individuals with the T allele may have higher constitutive expression of CD14 than those with the −159C allele. Although not consistently, the T allele of the CD14C−159T polymorphism appeared to be protective against atopy in population subgroups. However, a number of studies also reported that the T allele was a risk for asthmatic phenotypes. These inconsistencies may be attributed to the lack of power, linkage disequilibrium, gene–gene interactions, population admixture, ethnicity-specific variations and gene–environment interactions. In conclusion, if the SNP results in functional alteration to the coded protein, it may modulate susceptibility to allergic disease through its involvement in IgE regulation, and its effect may be modest and dependent on other co-existent environmental risk factors.