Identification Of Candidate Colorectal Cancer Predisposing Gene Variants In Early-onset And Familial Cases Using Whole-exome Sequencing

Background and ObjectiveGenetic factors are estimated to account for the development of~30%of all colorectalcancers(CRC).Mendelian CRC predisposition syndromes underlie about5%-10%of allCRCs, such as HNPCC,FAP,MAP and JPS, and are caused by germline mutations in alimited set of genes. The selection of causative genes to be screened in high-risk families isbased on several phenotypic characteristics. However, many families at risk remainunexplained. CRC patients with a family history of CRC or an early age at diagnosis areespecially suggestive of a hereditary contribution and may be used in genetic associationstudies to increase the likelihood of identifying susceptibility variants. We have testedwhether whole-exome sequencing may serve as an efficient method to identify known ornovel CRC predisposing genes in early-onset or familial CRC cases.Methods1. We performed whole-exome sequencing in23Chinese patients from21CRCfamilies diagnosed at≤40years of age, or at≤55years of age with at least1first-degreerelative diagnosed with CRC. Genomic DNA was extracted from peripheral blood cells andwhole-exome sequencing was performed. Genomic DNA was fragmented and enriched forexome sequences using the SureSelect Human All Exon Kit version2and sequencing wasperformed on an Illumina HiSeq2000platform.2. Single-nucleotide variants were called by the DiBayes algorithm, and smallinsertions and deletions were detected using BWA software. All variants were annotated.For prioritization, we selected variants that were not present in dbSNPv138, The NationalHeart, Lung, and Blood Institute (NHLBI) Exome Sequencing Project (ESP), and gave riseto protein-truncation, splice site defects, or missense mutations at highly conservednucleotide positions. Identified variants were analyzed using Alamut2.0software and integrated mutation prediction software align GVDV, SIFT and Polyphen-2. Prediction ofsplicing effects were done using NNSplice and ESEfinder software. Coding variants wereanalyzed using the online tool ‘project HOPE’ that provides insight in the structural effectsof a mutation.3. Data were processed through an analytical pipeline to search for rare and novelgermline variants in known and novel candidate CRC predisposing genes. Identifiedgermline variants were validated by Sanger sequencing, screening on control cohort andco-segregation anlysis were performed on related gene variants.ResultsIn6of the21families, a total of seven CRC patients were identified with eightgermline variants in known CRC predisposing genes(MLH1,MSH2,MSH6,MUTYH). Ofthese, five variants were reported as being pathogenic in public databases.Three mismatchrepair gene mutations,observed in three unrelated patients, were not previously reported inpublic databases.We identified24rare and novel potentially deleterious variants in19genes,including seven truncating mutations,16highly conserved non-synonymous missensevariants and one in-frame insertion.These candidate predisposing genes are cancergene,DNA repair gene,WNT signaling pathway,PI3K/AKT signaling pathway, transposonstudies related gene,somatic mutated gene,GWAS related gene. Four genes (BUB1,LRP5,RPS6KB2,RYR2)were found to be recurrently affected by different rare variants. Onepreviously unreported variant identified in EIF2AK4was found to represent a local Chinesevariant, which was enriched in our early-onset CRC patient cohort compared to a controlcohort of100healthy Chinese individuals scored negative by colonoscopy. We, therefore,conclude that EIF2AK4gene may be considered as a candidate CRC predisposing gene.ConclusionWe conclude that whole-exome sequencing of early-onset or familial CRC cases servesas an efficient method to identify known and potential pathogenic variants in establishedand novel candidate CRC predisposing genes. We identified in total19novel candidateCRC susceptibility genes, our findings provide insight into the role of these variants inCRC development.