New study challenges longstanding assumption about the cause of the genome’s most common mutation microbiologystudy

A Ludwig Cancer Research study has punctured a longstanding assumption about the source of the most common type of DNA mutation seen in the genome—one that contributes to many genetic diseases, including cancer.

Led by Ludwig Oxford Leadership Fellow Marketa Tomkova, postdoc Michael McClellan, Assistant Member Benjamin Schuster-Böckler and Associate Investigator Skirmantas Kriaucionis, the study has implications not only for basic cancer biology but also for such things as assessments of carcinogenic risk associated with environmental factors and our understanding of the emergence of drug resistance during cancer therapy. Its findings are reported in the current issue of Nature Genetics.

The mutation in question—in which cytosine (C), one of the four bases of DNA that spell out our genes, is erroneously switched to thymine (T)—was thought to be primarily the result of a spontaneous chemical reaction with water. This reaction, deamination, is about twice as likely to happen when a cytosine is chemically tagged by the addition of a molecule known as a methyl group to create 5-methylcytosine, which occurs in DNA at so-called “CpG” positions, where C is followed by the base guanine (G).

Such tagging, commonly seen across the genome, plays a fundamental role in controlling the expression of genes and is therefore essential to pretty much every aspect of appropriate cellular function from embryonic development onward.

“It has long been assumed that C to T mutations are caused by a random chemical reaction,” said Tomkova. “Our study demonstrates that this is not entirely correct. Rather, the mutation is primarily produced when the cell copies its genome to divide and is mainly caused by the tendency of a key component of the cell’s DNA-copying machinery to make editing mistakes when it encounters methylated cytosines.”

The Ludwig Oxford team got their first inkling of what was going on several years ago, when they examined sequences of cancer genomes shared with them by laboratories in the UK and Canada. They noticed in these data that cancer cells with certain genetic aberrations were far and away more likely to have CpG to TpG mutations.

These were cells known to be deficient in their ability to repair mismatched DNA sequences generated by mutations and those that bore mutations to a component of their DNA replication machinery, DNA polymerase ε (Pol ε), that proof-reads new DNA strands and edits out such errors. Both these defects interfere with DNA repair during cell division, and both are known to be associated with highly mutated tumors in cancer patients.

“Our study would not have been possible without the free and open sharing of data between researchers around the world: we first spotted a peculiar pattern at methylcytosine sites in the data we received from those laboratories, and we used public data to refine our hypothesis before we even began to think of experimentally testing it,” noted Schuster-Böckler.

To test their hypothesis, the researchers developed a new and very sensitive DNA sequencing technology that could discern genuine errors made by Pol ε during DNA replication from experimental artifacts. They applied their technique, dubbed Polymerase Error Rate Sequencing (PER-seq), to sequence over 28 billion bases across more than 130 million DNA molecules, measuring the accuracy of both normal human Pol ε and the most common cancer-associated mutant of the enzyme.

Their studies revealed that the mutant Pol ε produced CpG to TpG mutations at rates similar to those seen in cancer cells that carry that mutant. Even the normal Pol ε produced mutations at methylcytosine sites at seven times the rate observed for nonmethylated cytosines.

These findings, which directly link the incidence of CpG to TpG mutations to cell division, explain why these mutations tend to accumulate with age. They also explain why the mutation varies so much in frequency across tissues and tumors: because different types of both normal and cancerous cells proliferate at very different rates.

“This also means that the accumulation of CpG to TpG mutations can be used like a clock to determine the age of cells, which could be useful to studies exploring, for example, how fast different cancers grow before acquiring resistance to different treatments,” said Kriaucionis.

Further, the methods developed for this study also have implications for cancer prevention. To accurately gauge how likely various environmental factors—such as chemical pollutants—are to induce cancer-causing DNA mutations, it helps to know what proportion of those mutations is caused by errors during normal processes, such as cell division, in relevant tissues.