Skip to main content
Premium Trial:

Request an Annual Quote

Single-Cell Atlas Reveals Epigenetic, Transcriptomic Changes in Colorectal Cancer Development

NEW YORK — A new single-cell analysis has cataloged the composition as well as epigenetic and transcriptomic changes that take place during colorectal cancer development. Some of these changes could potentially be used to determine which polyps have progressed further than others, the Stanford University-led team said.

Colorectal cancer (CRC) develops along a known route from normal cell to atypical cell to carcinoma. As part of the Human Tumor Atlas Network, researchers led by Stanford University School of Medicine's William Greenleaf generated single-cell data on normal, polyp, and cancerous tissue from individuals with and without germline APC mutations, which predispose people to developing polyps, and cataloged the changes that occur among immune, stromal, and epithelial cells in the samples from along this path.

"We find that polyps populate an epigenetic and transcriptional continuum from normal colon to CRC characterized by sequential opening and closing of chromatin and upregulation and downregulation of genes associated with the cancer state," Greenleaf and his colleagues wrote in the paper, which appeared Monday in Nature Genetics.

The researchers generated single-cell data from 81 samples obtained from eight individuals with familial adenomatous polyposis and seven without. In all, they obtained chromatin accessibility profiles using scATAC-seq for nearly 450,000 cells and generated transcriptomic profiles using scRNA-seq for more than 201,000 cells.

Based on these profiles, the researchers could identify the cell types found among their samples. They noted differences in the portion of immune cells between sample types. Polyps, for instance, had more T regulatory cells than unaffected tissues, but unaffected tissues had higher levels of naïve B cells, memory B, and germinal center cells as compared to polyps. Exhausted T cells, though, were more common among colorectal cancer samples.

Within fibroblasts in their samples, the researchers identified what they called pre-cancer-associated fibroblasts (preCAFs), or a cluster of cells that resembled cancer-associated fibroblasts (CAFs). This cluster, though, was enriched for cells from polyps and CRC samples and it is marked not only by the expression of CAF marker genes like FAP and TWIST1 but also by a high chromatin accessibility surrounding those genes. The researchers proposed that preCAFs may develop into CAFs.

Meanwhile, among epithelial cells, a number of polyps as well as colorectal cancer cells exhibited increasingly stem cell-like phenotypes. Along this continuum, for instance, the researchers noted the increasing expression of certain genes, including GPX2, a glutathione peroxidase. Its higher expression in pre-malignant tissue indicated to the researchers that changes to the oxidative environment might be needed for cells to progress along this malignant pathway. Based on that, they further suggested that GPX2 could represent a marker of malignancy among polyps.

The researchers likewise identified a number of transcription factors associated with changes in chromatin accessibility among transforming polyps. Stem-like cells isolated from polyps had increased accessibility at TCF and LEF motifs, as compared to normal tissue, and lost accessibility at KLF motifs. CRC samples, further, had increased accessibility at HNF4A motifs.

By further folding in methylation data, the researchers found that hypo- or hypermethylation could predict accessibility changes along the transformation continuum toward malignancy.

"This work presents a strategy to order premalignant polyps by their degree of malignancy, which we speculate may be useful for staging polyps and assessing clinical risk," the researchers wrote in their paper.

They suggested that staging by determining the fraction of T regulatory cells or stem-like cells, such as through immunohistochemistry, might be the most straightforward means of gauging where a polyp might fall along the path toward malignancy. However, they noted that whether these features correlated with patients' clinical outcomes will need to be investigated.