AD and asthma are complex genetic diseases, which would result from integrated effect of multiple genetic and environmental factors on pathophysiology process of the entities. Some common genetic factors being shared in the patients with AD and asthma have been identified 
. Here we investigated the association of 10 SNPs within FLG
, 5q22.1, 11q13.5, 14q11.2 and 20q13.33 for asthma in the Chinese Han population, to evaluate the relationship between AD and asthma in these loci/genes.
At the locus 14q11.2, the five genes: KIAA0323
were located within a single 220 kb LD block surrounding the tag biomarker rs4982958 which was associated with asthma in this study (). Previous studies selected mast cell chymase gene (CMA1
) at 14q11.2 as a candidate gene for analysis of AD, asthma and other allergic phenotypes 
. Mast cell tryptase and chymase are key players in mediating chronic inflammation, as observed in asthmatics of the lungs. Two polymorphisms in CMA1
, including promoter polymorphism rs1800875 (−1903G/A) 
and (TG)n(GA)m repeat polymorphism 
, were identified being associated with the allergic phenotypes in different populations. However, other studies failed to replicate role of −1903G/A (rs1800875) in AD or asthma 
. Inconsistence in these studies probably was attributable to different populations, variation on phenotype definitions, ethnicity and environmental exposure.
Regional LD plots for associated SNPs.
We observed that the association of rs4982958 with asthma under additive model (P
) was of more significance than that under dominant model (P
) and recessive model (P
) by using a logistic regression analysis (). Clearly, risk allele C of rs4982958 was the major one and allele T was of protection effect. Compared with C/C, the genotype of homozygote T/T (ORHom
0.37) showed smaller OR than the heterozygote T/C (ORHet
0.89) (). This result may provide more information for futher study on this locus. Rs4982958 located at 10.4 kb upstream of CMA1
. Linkage disequilibrium (LD) test showed that the two SNPs (rs4982958 and rs1800875) were moderately correlated with each other (D'
0.5) in CHB and JPT from 1000 Genomes Project data. Imputation of promoter polymorphism rs1800875 of CMA1
was performed in our AD GWAS samples and then tested for association (P
0.85). In AD GWAS, rs4982958 at 14q11.2 was replicated in expanded AD samples, we found a weak evidence of association in the Chinese Han population (P
0.94), but did not reach genome-wide significance (P<5.0×10−8
. Statistical tests showed that the odds ratio of rs4982958 in asthma samples (OR
0.73) was smaller than in AD (OR
0.94) and the existence of significant genetic heterogeneity between AD and asthma (Phet
88.60) in Chinese Han population.
Our AD GWAS had identified 5q22.1 as a susceptibility locus for AD 
. In this study, further haplotype analysis of the 4 SNPs (rs10067777, rs7701890, rs13360927 and rs13361382) within 5q22.1 was performed in AD and asthma groups, respectively. Only two haplotypes (AAAG
) had allele frequencies >5% (). The risk haplotype GGGA
had a consistent association evidence for AD (Pcorrection
1.26) as single SNP did (). It further supported the presence of underlying causal variants within this locus for AD. None of the 4 SNPs within 5q22.1 was associated with asthma in such a small samples size and low statistical power (1%~2%) in this study. The risk allele of these SNPs at this locus was the same for AD and asthma. In asthma group, haplotype GGGA
was a suggestive risk factor in this study (P
0.014), although it was barely below significance (P
0.084) after multiple testing correction. We observed that the odds ratio of haplotype GGGA
1.38) was much higher than single SNP in asthma cases-controls samples (1.04≤OR≤1.10). Further more, heterogeneity test at 5q22.1 showed that no genetic heterogeneity was found for each SNPs (0.2313≤Phet
≤30.2) () and risk haplotype GGGA
0) () between AD and asthma groups, which means 5q22.1 might have similar effect on AD and asthma. The four SNPs (rs10067777, rs7701890, rs13360927 and rs13361382) located within a 600 kb LD block harboring two genes TMEM232
) and SLC25A46
. Little is known about the function of the two genes. At 5q22.1, TSLP
were proposed to be candidate genes for asthma in Caucasian population 
. Nevertheless, we found haplotype GGGA
fall into different LD blocks with TSLP
, which were separated by a recombination hot spot (). Fine-mapping or direct sequencing would be powerful to further search for any potential and associated susceptibility genes at 5q22.1 dedicated to both AD and asthma.
The 5 SNPs in FLG, 11q13.5 and 20q13.33 did not reached the statistical significance in this study, which might either result from low statistical power (1%~15%) () or be attributed to 10.6% of these asthma cases with previous history of AD (). Comprehensive sequencing of entire FLG in asthma patients and larger scale of asthma samples or increased related tagSNPs for 11q13.5 and 20q13.33 would reliably enable to explore the relationship between AD and asthma in these loci/gene.
In conclusion, we compared the identical SNPs from the asthma cohort and our previous AD GWAS. We comfirmed 14q11.2 was an important candidate locus for asthma, and also demonstrated that 5q22.1 might be shared by AD and asthma in Chinese Han population. However, the evidences for the associations will be required to validate in diverse populations and then gain a better understanding on AD and asthma pathogenesis.