NEW YORK (GenomeWeb) – Using a newly developed single-cell sequencing approach, the University of Texas MD Anderson Cancer Center's Nicholas Navin and colleagues found that mutations crop up in different types of tumor cells at different rates.
As they reported in the online advanced issue of Nature today, Navin and his colleagues relied on the doubling step cells go through before they divide to increase the amount of DNA they have to work with when sequencing single cells. Analyzing a greater amount of DNA limits the sequencing errors that plague single-cell approaches.
They then applied their approach, dubbed nuc-seq, to analyze tumor and normal nuclei from ER+ breast cancer and triple-negative ductal carcinoma.
"We found that two distinct 'molecular clocks' were operating at different stages of tumor growth," Navin said in a statement. "Tumor cells from triple-negative breast cancer had an increased mutation rate, while tumor cells from estrogen receptor positive breast cancer did not."
These disparate rates, the researchers added, may have implications for the diagnosis, treatment, and evolution of treatment resistance in breast cancer.
In previous work, Navin and his colleagues used a combination of flow sorting, whole-genome amplification, and massively parallel sequencing to gauge copy number profiles of single cells, but they were unable to resolve mutations at the base-pair level.
To get to that resolution and coverage, the nuc-seq approach makes use of the genome duplication cells undergo during S phase as they prepare for cytokinesis.
After flow sorting those cells that have already doubled their genomes, the researchers lyse those cells before subjecting them to multiple-displacement amplification with ϕ29 DNA polymerase, though for a shortened timeframe to again limit errors. The amplified DNA is then incubated with a Tn5 transposase that both fragments the DNA and attaches adapters for either exome or whole-genome sequencing.
"By sorting and sequencing only the newly 'doubled' nuclei, nuc-seq takes advantage of this duplication to achieve lower rates of sequencing errors than most previous techniques," Edward Fox and Lawrence Loeb, both at the University of Washington, said in a related article appearing in Nature.
Navin and his colleagues then applied their method to an ER+ breast cancer and a triple-negative ductal carcinoma patient. They isolated nuclei from both tumor types and matched normal tissue for population and single-cell sequencing as well as copy-number profiling and other analyses.
In the ERBC, they uncovered some 4,162 somatic SNVs, including 12 non-synonymous mutations, some of which occurred in cancer genes like PIK3CA. Additionally, they found three distinct classes of mutations: clonal, subclonal, and de novo mutations.
Likewise in their TNBC sample, the researchers found 374 non-synonymous mutations, including ones in PTEN, NOTCH2, JAK1, and more, and uncovered two subpopulations of tumor cells.
Based on an exome sequencing analysis of 16 single TNBC tumor and 16 single normal nuclei, the researchers found that the 374 non-synonymous mutations found in the bulk sequencing part of the study were found in most of the single tumor cells, but they also noted a further 145 non-synonymous subclonal mutations that weren't detected in the bulk tumor. Clustering analysis indicated that many of the subclonal mutations occurred in only one subpopulation, the researchers added.
Duplex single molecular targeted deep sequencing of those two tumors validated many of the clonal, subclonal, and de novo mutations uncovered through single-cell sequencing.
This, Navin and his colleagues said, indicates that the subclonal and de novo mutations are likely to be real mutations, albeit ones that crop up at low frequencies.
That is, "no two individual tumor cells were genetically identical," Fox and Loeb said.
Additionally, based on the single-cell mutation frequencies, the researchers calculated the mutation rate for the two tumors, finding that the ERBC had a mutation rate of 0.6 mutations per cell division for the exome data and 0.9 mutations per cell division for the single-cell whole-genome data, which is in line with error rates reported for normal cells, they noted.
TNBCs, however, had a mutation rate of eight mutations per cell division — a more than 13-fold increase, as compared to normal cells. This, Navin and his colleagues said, supports the notion that an increased mutation rate may contribute to the number of mutations that accumulate in tumor cells.
"Nuc-seq and comparable single-cell sequencing methods will allow a more detailed understanding of mutational heterogeneity in individual tumors, and will influence our understanding of how cancers evolve and our approach to their treatment," Fox and Loeb said.
"In particular, mutational diversity within a tumor is likely to be predictive of whether resistance to a particular chemotherapy will emerge during treatment, because mutations in genes that render cells resistant to specific drugs may exist before initiation of therapy," they added.