A Genome-Wide Association Study of Basal Transepidermal Water Loss Finds that Variants at 9q34.3 Are Associated with Skin Barrier Function
A Genome-Wide Association Study of Basal Transepidermal Water Loss Finds that Variants at 9q34.3 Are Associated with Skin Barrier Function
复制标题
基础经表皮失水的全基因组关联研究发现 9q34.3 的变异与皮肤屏障功能相关
DOI:
10.1016/j.jid.2016.11.030
复制
发表时间:
2017
影响因子:
6.5
通讯作者:
Wang Sijia
中科院分区:
文献类型:
--
作者:
Zhang Manfei;Li Bingjie;Wu Sijie;Tan Jingze;Yang Yajun;Marini Aless;ra;Vierkotter Andrea;Zhang Juan;Li Hui;Schikowski Tamara;Jin Li;Krutmann Jean;Wang Sijia
Epidermal homeostasis and barrier permeability are very important properties of human skin. Transepidermal water loss (TEWL), the passive diffusion of water from the hydrated layers of the dermis and epidermis toward those layers with a lower water content (Nilsson, 1977), has been widely used to determine epidermal permeability barrier status (Fluhr et al., 2006). For example, TEWL measurement helped to establish that skin barrier function is compromised in skin diseases such as atopic dermatitis (AD)(Elias, 2008). Likewise, it has successfully been used to monitor the effects of different treatments on skin barrier function recovery (Sextius et al., 2010). Although TEWL has been reported to be affected by environmental factors such as temperature, seasonal variation, sun exposure, and smoking (Li et al., 2014, Liu et al., 2010, Xin et al., 2016), the presence of significant ethnic differences in stratum corneum permeability suggests that genetics also plays a role in epidermal barrier function (Kompaore and Tsuruta, 1993). However, to our knowledge, no genomic study has been conducted to explore the genetics of barrier function of healthy human skin. To address this, we performed a genome-wide association study (GWAS) of basal TEWL as a measure of the skin barrier function, with the aim of detecting the potential genetic variants associated with this important skin trait.We collected 611 samples from healthy Han Chinese in Taizhou, Jiangsu Province, aged between 31 and 87 years. This research was conducted with official approval from the ethics committee of Fudan University, Shanghai, China. All participants provided written informed consent. TEWL measurement was carried out with a DermaMeter Professional 100 (VASEMA GmbH, Vienna, Austria) on the right cheek (see Supplementary Materials and Supplementary Table S1 online for details). Because the obtained TEWL values did not follow the normal distribution (Shapiro-Wilk test, P< 2.2× 10–16), a logarithmic transformation was performed (see Supplementary Figure S1 online). Principal component analysis found no significant population stratification in our sample (see Supplementary Figure S2 online). Mostly consistent with previous reports, we found TEWL to be significantly correlated with temperature (r= 0.284, P= 9.54× 10–13), sex (two-tailed Student t test, P= 6.06× 10–3), and skincare habits (P= 4.74× 10–3). It was not correlated with humidity of the environment (P= 0.967), sun exposure (P= 0.247), and smoking (P= 0.089)(see Supplementary Materials). We then performed a GWAS, adjusting for age, sex, temperature, and skincare habits. Individuals were genotyped on an Illumina (San Diego, CA) Human Omni Zhonghua 8V1. 1 chip, and imputation was performed using 1000 Genomes Project data (phase 3)(1000 Genomes Project Consortium et al., 2012, Pickrell et al., 2009, Kent et al., 2002). After quality-control filters, the GWAS was carried out on 795,279 genotyped single-nucleotide polymorphisms (SNPs) and 7,203,134 imputed SNPs (see Supplementary Materials for the details). We found a variant on chromosome band 9q34. 3 to be significantly associated with TEWL (rs10858314, β=–0.211±0.038, P= 3.11× 10–8; see Supplementary Figure S3 online). To validate our finding, we performed a second GWAS using the same phenotyping protocol on a replication set including 366 healthy Han Chinese samples from Taixing, Jiangsu Province. There was no genome-wide significant signals in this second GWAS, but the SNP (rs10858314) was replicated with nominal significance (β=–0.167±0.065, P= 9.96× 10–3; see Supplementary Table S2 online).