Whole-genome sequencing of esophageal adenocarcinoma in Chinese patients reveals distinct mutational signatures and genomic alterations

Abstract

While the incidence of esophageal adenocarcinoma (EAC) has risen drastically in Western countries over the last 40 years, a similar trend has not been observed for EAC in China. Here, we analyzed mutational spectrum, copy number alterations, and structural variants from whole-genome sequencing of 10 Chinese EAC tumor samples and their matched normal samples, and compared them to previously reported EAC tumor specimens from Western countries. The mutational burden in Chinese EAC was significantly lower than that found in EAC from Western countries. The hallmark A>C mutational signature observed at high frequency in EAC from Western countries, which has been linked to acid reflux, is completely absent in Chinese samples. Furthermore, none of the Chinese samples showed evidence of chromothripsis and genome doubling that are often found in EAC from Western countries. In summary, Chinese EAC tumor samples had distinct genomic profiles and signatures, suggesting that EAC in Chinese individuals may arise from a different etiological pathway.

Fig. 1

Comparisons of mutation burden and types based on whole-genome sequencing data from Chinese EAC, US and UK EAC. a The number of mutations per million base pairs. b The proportion of six types of mutations. c Mutation burden of six mutation types in the trinucleotide context

Fig. 2

Mutational signatures in Chinese EAC genomes discovered by the non-negative matrix factorization algorithm. a A S1-like signature  with C>T in a *CG context, which may be associated with aging processes. b A S3-like signature with more-or-less equal representation of all 96 tri-nucleotide mutation types, possibly linked to defects in BRCA1/2-led homologous recombination pathway. c The third signature which appears to be an unknown one, likely a mix of S1 (similarity = 0.71) and U2 (similarity of 0.80), an unvalidated mutation signature defined by Alexandrov et al.

Fig. 3

Comparisons of somatic copy number alterations (SCNA) between Chinese EAC and US EAC. a The length of SCNAs by copy gain, copy loss, and loss of heterozygosity (LOH). b Recurrent focal SCNAs in Chinese EAC detected by GISTIC 2.0. c The CIRCOS plot comparing recurrent SCNAs between US EAC and Chinese EAC. The inner circle represents Chinese EAC and the outer circle represents US EAC. d Unsupervised hierarchical clustering using recurrent SCNA among 10 Chinese EAC samples and 16 US EAC samples

Fig. 4

Recurrent structural variants detected by Manta for Chinese EAC (a) and US EAC (b). The bar on the top shows the total number of structural variants in the genome and the bar on the right shows the number of samples having the variant