Investigating the germline deleterious rare variants for genetic susceptibility of esophagus squamous cell carcinomas by target sequencing

Abstract

Purpose: The genetic basis of esophageal squamous cell carcinoma (ESCC) still remains unclear. Several genome-wide association studies identified some common variants associated with small effects. Here we aimed to identify some rare deleterious variants in cancer predisposition genes, which may contribute to the genetic susceptibility of ESCC with a large effect, using family history-positive (FH+) samples from the Mainland China Henan high-risk region by targeted sequencing. Methods: A two-phase study was conducted, the discovery phase and the validation phase. In the discovery phase, we sequenced the blood DNAs of 214 independent family history ESCC cases and 123 non-ESCC hospital controls by Illumina HiSeq platform using a custom capture kit. The capture kit included more than 1000 genes, which were cancer-related genes, DNA repair genes, and other genes reported in some ESCC functional studies. For each gene, the exonic region plus 3’-UTR, 5’-UTR, upstream and downstream regions were designed for targeted capture with a total capture size of ~11 Mb. The sequencing reads were aligned to the human reference genome (GRCh37/hg19) using burrows-wheeler aligner (bwa). PCR duplication was marked by Picards. Local realignment was performed with genome analysis toolkit (GAKT). Annovar was used to do the annotation. Loss-of-function (LoF) variants, including splicing, stopgain, and frameshift InDel variants in the cases and controls were identified. The damaging effect of the missense variants was predicted by the combination of 4 different bioinformatics tools (CADD score, MutationTaster, SIFT, Polyphen2). We also used the 1000genomes or ExAC dataset as public controls. All the identified LoF variants in BRCA2 were validated by Sanger sequencing. Results: There was a total of 110 LoF genes identified. Pathway enrichment analysis of these genes found that they were enriched in the Fanconi Anemia (FA) pathway after the background correction of all captured genes. BRCA2 is in the FA pathway and was found to have the most LoF variants; this included 6 individuals from the 214 FH+ cases, and only 1 in the control cohort. We did not find many somatic mutations in ESCC tissues in BRCA2. BRCA2 expression was also significantly upregulated in ESCC tumor tissues, as compared to adjacent normal tissues from in-house and 3 other publicly available RNA-seq datasets. The variant spectra of candidate genes were compared with the public controls to filter out some genes. Conclusions: In the discovery study, we identified some LoF variants in BRCA2 and other genes in the FA pathway that may contribute to ESCC genetic susceptibility. Further validation with a larger sample size is now underway.