Evolution and heterogeneity of non-hereditary colorectal cancer revealed by single-cell exome sequencing

Abstract

Recently single-cell whole-exome sequencing (scWES) has deeply expanded and sharpened our knowledge of cancer evolution and subclonality. Herein, with scWES and matched bulk whole-exome sequencing (bulk WES) on two colorectal cancer (CRC) patients with normal or adenomatous polyps, we found that both the adenoma and cancer were of monoclonal origin, and both shared partial mutations in the same signaling pathways, but each showed a specific spectrum of heterogeneous somatic mutations. In addition, the adenoma and cancer further developed intratumor heterogeneity with the accumulation of nonrandom somatic mutations specifically in GPCR, PI3K-Akt and FGFR signaling pathways. We identified novel driver mutations that developed during adenoma and cancer evolution, particularly in OR1B1 (GPCR signaling pathway) for adenoma evolution, and LAMA1 (PI3K-Akt signaling pathway) and ADCY3 (FGFR signaling pathway) for CRC evolution. In summary, we demonstrated that both colorectal adenoma and CRC are monoclonal in origin, and the CRCs further diversified into different subclones with heterogeneous mutation profiles accumulating in GPCR, PI3K-Akt and FGFR signaling pathways. ScWES provides evidence for the importance of mutations in certain pathways that would not be as apparent from bulk sequencing of tumors, and can potentially establish whether specific mutations are mutually exclusive or occur sequentially in the same subclone of cells.

Bayesian non-parametric clustering of single-cell mutation profiles

“…To demonstrate the application of BnpC on real tumor data, we analyzed the sequencing data of five patients with childhood leukemia [28], one high-grade serous ovarian cancer (HGSOC) patient [29] and two colorectal cancer (CRC) patients [30].…”

“…Colorectal Cancer Patients CRC0827 and CRC0907 from Wu et al [30] were collected by single-cell Whole Exome Sequencing on CRC tissue samples (C1 and C2) and matched normal tissue (N). Additional samples from normal polyp (NP, CRC0907) and adenomatous polyp tissue (AP, CRC0827) were sequenced for the analysis.…”

The landscape of somatic mutation in sporadic Chinese colorectal cancer

“…In general, Identification of driver mutations from passenger mutations is a challenging task due to following reasons: a) CRC is a type of cancer with high immune and stromal cells infiltration and somatic mutation signal will be diminished [20]. b) Patients may have sub-clone with different somatic mutated genes, and a mixture of the two or more sub-clone will further decrease somatic mutation signal [21,22]. c) Disease etiology for a fraction of patients may be induced by different somatic mutations and a long tail of genes was postulated to explain the CRC initiation and progression of entire population [11].…”

SiFit: A Method for Inferring Tumor Trees from Single-Cell Sequencing Data under Finite-site Models

“…SiFit was applied to single-cell exome sequencing data from a non-hereditary colorectal cancer [46] patient. The dataset consisted of 61 single cells in total, with 35 cells were sampled from colorectal cancer tissue, 13 from an adenomatous polyp tissue and 13 from normal colorectal tissue.…”

“…The genotype matrix for the non-hereditary colon cancer patient has been reproduced from Fig. 3a of [46]. The genotype matrix for the metastatic colon cancer patient has been reproduced from Supplementary figure 4a of [49].…”