NEW YORK – Seemingly normal colon or liver tissue can contain mutations resembling those previously described in related cancer types, according to new research from two Wellcome Sanger Institute-led teams that reported findings today in Nature.
For one of these studies, investigators developed a laser capture microscopy and sequencing strategy to do whole-genome sequencing on 2,035 so-called "crypts," or very small pockets, of colon tissue in samples donated by 42 individuals between the ages of 11 and 78 years old. The participants included 15 individuals with a history of colorectal cancer and 27 unaffected individuals.
An automated version of the capture and sequencing pipeline has reportedly been applied to tens of thousands more samples since.
"The ability to find mutations in normal cells means we can now describe how many and what types of mutations there are in different tissues across the human body, potentially providing understanding of what makes some tissues more prone to cancer than others," senior author Michael Stratton, director of the Sanger Institute, said in a statement.
The team uncovered suspected cancer driver mutations in 1 percent of samples from individuals in their 60s. The available sequence data also led to mutational signatures in the normal colon tissues, including previously unappreciated signatures.
"There is a whole landscape of genetic mutations in our cells that we haven’t previously appreciated," first author Henry Lee-Six, at the Sanger Institute, said in a statement. "We've shown that these mutant cells are abundant, though the vast majority don't go on to cause cancer."
While driver alterations that can ultimately spark cancer have been well profiled, Stratton, Lee-Six, and their colleagues noted that cancer progression clues so far have largely come from fully formed tumors or tissue samples that have already taken on unusual features "that represent intermediate stages between normal cells and cancer cells."
By digging into variant allele patterns in multiple normal tissue samples from each individual, the researchers found 11 ubiquitous mutational signatures, spanning more than 85 percent of the crypt samples tested, along with nine sporadic signatures that turned up in only a few people or a subset of samples. They also compared mutations in tumor and normal tissues, identifying a dramatic rise in mutations in the colorectal cancers.
Stratton was part of a team led by Sanger Institute researchers Peter Campbell and Matthew Hoarse that looked at almost 500 small samples microdissected from hepatocyte samples from nine individuals with liver cirrhosis and five normal liver samples.
"[W]e show that cirrhotic liver has a higher mutational burden than normal liver," the authors reported, noting that samples from those with cirrhosis were far more prone to contain structural variants, point mutations, and even chromothripsis.
When members of that team turned their attention to mutational signatures across the liver samples, they again saw some signatures that were shared by both normal and cirrhotic hepatocyte samples, though a subset of mutational signatures seemed to have enhanced activity in the samples with cirrhosis. Likewise, they saw examples of clonal expansion involving driver mutation-containing clones in the immediate neighborhood of cirrhotic samples.
Along with these and other analyses, the team went on to look at the mutational signatures and types of mutations found in tumor, cirrhotic, and seemingly normal samples from seven individuals with hepatocellular carcinoma (HCC). Although driver mutations did not necessarily line up between the cirrhotic sites and HCC tumors, the authors reported, tumors tended to show a boost in representation by some of the mutational signatures detected in cirrhotic samples.
Together, these and other findings "reveal the genomic consequences of chronic liver disease — increased rates of mutation, complex structural variation … and a low burden of mutations that target known HCC genes," the authors wrote.