NEW YORK (GenomeWeb) – Using somatic base substitutions as a guide, researchers from the Wellcome Trust Sanger Centre and elsewhere developed a method to trace the lineages of individual cells back to the earliest divisions of the fertilized egg.
As they reported in Nature yesterday, the Wellcome Trust's Michael Stratton and colleagues performed whole-genome sequencing on clonal cell lines developed from various tissues of two male mice. By examining the patterns of somatic base changes, the researchers could observe how one early cell divided to make up certain tissues and how various lineages of cells accumulated different numbers and types of mutations.
"With this novel approach, we can peer back into an organism's development," first author Sam Behjati from the Wellcome Trust said in a statement. "If we can better understand how normal, healthy cells mutate as they divide over a person's lifetime, we will gain a fundamental insight into what can be considered normal and how this differs from what we see in cancer cells."
The researchers developed 25 clonal lines from the stomach, small bowel, and large bowel of two mice. They also developed a line from the prostate of one of the mice. They sequenced each line along with the polyclonal tail of each mouse, and aligned those reads to the mouse reference genome to identify sets of somatic mutations.
In the not-so-distant future, they said that single-cell sequencing could take the place of developing clonal lines, though the technology wasn't quite ready yet for this purpose.
To construct their early development lineage tree, Stratton and his colleagues searched for mutations present in at least two organoids and missing from at least one. These putative early embryonic mutations were resequenced in each organoid, leading to the discovery of 35 such mutations.
Some 23 cell divisions were needed to reconstruct a simple bifurcating tree for the individual cell clones for both mice, and, the researchers noted, 17 of these divisions could be reconstructed while six could not. This, they added, indicates that the intrinsic substitution rate in mice is nearly enough to reconstruct the tree.
However, they added, the earliest cell divisions they could reconstruct might not in actuality represent the first cell divisions that took place after fertilization, if none of those cells gave rise to the cells the researchers studied. But by looking at the polyclonal tail, they found that the mutations from the putative first daughter cells were present, suggesting that the reconstructed division indeed reflects what occurred.
Daughter cells from those first divisions, the researchers noted, contributed unequally to the resulting adult tissues. For instance, in one mouse, the researchers reported that one of the daughter cells gave rise to 12 organoid clones while the other gave rise to just one.
This asymmetry in division, they added, continues beyond the first division. Cell 'i' in mouse two, which had undergone three cell divisions, gave rise to prostate, large bowel, and tail tissue cells, while cell 'h' which also had undergone three divisions, gave rise to stomach and small bowel cells.
Small bowel cells, Stratton and his colleagues found, accumulated more base substitutions than any of the other tissues they examined. Prostate and stomach cells, meanwhile, had the fewest.
These differences, the researchers said, could be due to variations in mutation rates between tissue types, the number of cell divisions in the lineage, or both. However, they estimated that about 1.1 substitution mutations arise for every division in small bowel cells, and noted that this rate is similar to the overall mutation rate of about 1.5 during early embryogenesis.
The mutational burden, though, was higher for regions of repressed chromatin and late-replicating regions — a pattern that is reminiscent of mutations in human cancers, they noted.
Additionally, Stratton and his colleagues said that there were differences in mutation patterns between tissues. Using a non-negative matrix factorization approach, they uncovered two mutational signatures: one marked by C-to-T mutations, and the other by C-to-A substitutions at certain trinucleotide spots along with some C-to-T mutations.
However, small bowel cells had an even different signature. Those cells had T-to-G mutations, commonly at XpTpT trinucleotides.
The mechanism underlying these mutations remains unclear, though the researchers said that the C-to-A substitutions could be generated by reactive oxygen species.
This sort of method to catalog somatic mutations has been used to trace the lineages of cancer cells, and this, the researchers said, showed that a similar approach could be used to examine the history of normal cells.
"Much more extensive application of this approach will allow us to provide a clear picture of how adult cells have developed from the fertilized egg," Stratton said. "Furthermore, by looking at the numbers and types of mutation in each cell we will be able to obtain a diary, writ in DNA, of what each healthy cell has experienced during its lifetime, and then explore how this changes in the range of human diseases," he added.