Exome sequencing of filaggrin and related genes in African-American children with atopic dermatitis.

TO THE EDITOR Atopic dermatitis (AD) is a common chronic relapsing disease. There is a considerable body of evidence supporting a genetic basis for AD (Bussman et al., 2011; Ellinghaus et al., 2013). Mutations in the Filaggrin (FLG) gene have been consistently found to be associated with AD in people of European and Asian ancestry (Brown and McLean, 2012). More than 40 FLG loss-of-function mutations have been described in Europeans and Asians (Brown and McLean, 2012). However, FLG loss-of-function mutations have not commonly been found in Africans or African Americans (Winge et al., 2011a; Brown and McLean, 2012; Margolis et al., 2012). Loss-of-function mutations in exon 3 of FLG result in diminished or absent filaggrin protein, most often due to a premature stop codon or a frameshift mutation resulting in a stop codon further downstream. Interestingly, the absence of profilaggrin protein (precursor of filaggrin) has also been noted in keratohyalin granules in the majority of individuals with ichthyosis vulgaris of European and Asian ancestry (Fleckman and Brumbaugh, 2002; Perusquia-Ortiz et al., 2013; Thyssen et al., 2013). FLG is located on chromosome 1q21 in a region called the epidermal differentiation complex. It is part of a family of genes that code for S100-fused-like proteins (SFTP). The SFTPs include the proteins profilaggrin (coded by FLG), hornerin (HRNR), filaggrin-2 (FLG2), repetin (RPTN), cornulin (CRNN), trichohyalin (TCHH), and trichohyalin-like 1(TCHHL1; Henry et al., 2012). These genes are very similar to one another with respect to structure and function, and lie in close proximity to each other in the epidermal differentiation complex (Marenholz et al., 2011; Henry et al., 2012; Pellerin et al., 2013). On the basis of previous experiences with FLG, it has been hypothesized that a stop-gain (null) mutation in exon 3 of any of the SFTP genes will result in decreased or absent protein production (Marenholz et al., 2011; Henry et al., 2012; Margolis et al., 2014). The goal of this study was to identify stop-gain variants in FLG and closely related genes in African Americans with AD from the Pediatric Eczema Elective Registry (PEER; Margolis et al., 2012). From this cohort we randomly selected 60 subjects for whole-exome sequencing to ensure sufficient power to detect variants with a minor allelic frequency (MAF) of greater than 3%. Sequencing was performed by Ambry Genetics (Aliso Viejo, CA) using whole exome–targeted enrichment by Agilent SureSelectXT Human All Exon 50Mb kit. Quality assessment revealed that most samples were above 50% on target and mean coverage per gene was excellent. The libraries were indexed using 100 base paired ends and processed using Illumina HiSeq2000 at × 100 coverage per exon. Data were assessed using a pipeline generated at the University of Pennsylvania based on the best practices protocol from the Broad Institute (Cambridge, MA). This report focused on stop-gain mutations of exon 3 (i.e., loss-of-function mutations) in the SFTP genes because of their likely functional relevance (Marenholz et al., 2011; Brown and McLean, 2012; Henry et al., 2012). Taqman allelic discrimination assays were created for any newly identified FLG loss-of-function mutations, which were then used to genotype an additional random sample of 100 African-American PEER children. Sequencing of the SFTP genes in 60 self-reported (ancestry previously confirmed with ancestral informative markers (Margolis et al., 2012)) African-American children with AD revealed a total of 289 variants in FLG, 107 variants in FLG2, 339 variants in HRNR, 4 variants in RPTN, 37 variants in CRNN, 88 variants in TCHH, and 14 variants in TCHHL1. However, very few variants resulted in a premature stop codon in exon 3 (Table 1). Each of the three newly identified FLG stop-gain mutations, Q570X, R3409X and S3707X, were observed only once. S2392X and S2377X in FLG2 were noted 1 and 16 times, respectively. In TCHHL1, the variant Q294X was noted twice. All subjects were heterozygous for the mutations. The MAF for variants noted once, twice and 16 times were 0.008, 0.017, and 0.133, respectively. Next, we used Taqman-based allelic discrimination assays to evaluate the three FLG mutations, Q570X, R3409X, and S3707X, in an additional 100 African-American PEER children. However, none of these variants could be detected in other members of our cohort. This report is from the largest whole-exome sequencing study of African Americans with AD performed to date. Here we have reported results specific to the SFTP genes. We identified a few new null mutations, albeit the ones in FLG had low MAFs. The MAFs noted for S2377X (FLG2) and Q294X (TCHHL1) vary from the healthy subjects in the 1000 Genomes database (The 1000 Genomes Project Consortium, 2012; Table 1). S2377X was seen about half as frequently in our cohort compared with that in the healthy 1000 Genomes African population; whereas Q294X, although more common in our cohort, still had a low MAF (i.e., MAF=0.017), suggesting that these variants may not be clinically important with respect to incident AD. Our findings are in agreement with those of Winge et al. (2011a), who also failed to detect common FLG loss-of-function mutations in people of African ancestry with AD. Our study does have limitations in that we focused only on exon 3 stop-gain mutations in genes. We did not assess copy number variations. We also did not assay protein function. Another point to be noted is that as most African Americans have their origins in West Africa; our findings may not generalize to everyone with African ancestry. However, based on the experiences of others as well as those from our study, which is the largest whole-exome study of African Americans with AD, it seems unlikely that FLG stop-gain mutations have a prominent role with respect to incident AD in African Americans children (Winge et al., 2011b; Thaswer-Esmail et al., 2014). The study was approved by the University of Pennsylvania Institutional Review Board. The study was conducted according to Declaration of Helsinki Principles. All participants provided written informed consent. The authors state no conflict of interest. This study was funded by R01-AR0056755 from the National Institute of Arthritis Musculoskeletal and Skin Diseases, a grant from Valeant Pharmaceuticals for the PEER study, and a grant from the Breast Cancer Research Foundation (Nathanson) for informatics analysis.