Genome-wide meta-analysis identifies eight new susceptibility loci for cutaneous squamous cell carcinoma

Abstract

Cutaneous squamous cell carcinoma (SCC) is one of the most common cancers in the United States. Previous genome-wide association studies (GWAS) have identified 14 single nucleotide polymorphisms (SNPs) associated with cutaneous SCC. Here, we report the largest cutaneous SCC meta-analysis to date, representing six international cohorts and totaling 19,149 SCC cases and 680,049 controls. We discover eight novel loci associated with SCC, confirm all previously associated loci, and perform fine mapping of causal variants. The novel SNPs occur within skin-specific regulatory elements and implicate loci involved in cancer development, immune regulation, and keratinocyte differentiation in SCC susceptibility.

Fig. 1. Manhattan plot of the combined meta-analysis of GWAS of SCC.

The P_fixed Stage one value for all SNPs present in at least two studies have been plotted using a −log₁₀(p-value). The total Stage one meta-analysis included eight SCC GWAS, totaling 19,149 cases and 680,049 controls. p < 5 × 10^–8 (genome-wide significance) threshold is indicated by a dashed line. In total, 22 loci reached genome-wide significance, including 8 novel loci 1q21.3, 2q33.1, 6q15, 8q23.3, 9p23, 11p15.4, 12q13.3, and 12q23.1 are highlighted by *.

Fig. 2. Gene expression analysis for novel SCC susceptibility loci.

RNA-seq data were obtained from Gene Expression Omnibus (GSE84194) were analyzed by DESeq. Transcript levels (FPKM) in SCC samples were compared with levels in paired matched normal skin. Boxplot demonstrates log₂[SCC/Normal skin] expression levels for 13 genes surrounding the novel SNPs. Legend for box and whisker plots. The black center line denotes the median value (50th percentile), while the gray box contains the 25th to 75th percentiles of data set. The black whiskers mark the 5th and 95th percentiles, and values beyond these upper and lower bounds are considered outliers, marked with white circles. The red threshold line indicates the point where these is no change in gene expression between SCC tumor and normal skin. ARNT, BACH2, TYRP1, and WEE1 were significantly downregulated in SCC as compared with normal skin and CASP8 and KRT6C were upregulated in SCC relative to normal skin by DESeq.

Fig. 3. Annotation of novel SNPs with epidermal enhancer information.

a Top: Circles represent the number of SNPs considered at each stage of the workflow to identify epigenetic context of all novel SNPs. We started with 22 lead SNPs identified by meta-GWAS, then found putative causal SNPs defined as any SNPs with a PPA of >0.05 from our fine-mapping analysis. We next refined that expanded list to SNPs for which the genomic location overlapped a previously identified epigenomic feature (either the H3K27ac enhancer mark or ends of an enhancer–promoter contact). Bottom: Heatmap displaying the overlap of SNPs with enhancer–promoter contacts or H3K27ac marked regions. The blue designation indicates that the SNP overlaps at least one H3K27ac region or contact. b Genome browser tracks for the genomic locus for SCC-index SNP rs793954, PPA > 0.99, demonstrating enhancer features in primary human keratinocytes (KC). ChIP-seq signal tracks are displayed for H3K4me1 and H3K27ac (which typically mark active enhancers and promoters) as well H3K27me3 (which marks inactive loci). Yellow denotes SNP location; note this SNP falls in a region marked by H3K27ac and H3K4me1 enhancer-associated histone marks, with lack of the repressive H3K27me3 mark. CTCF sites indicate that the SNP is not involved in CTCF loops and associated TADs.