A lab dish of cancer cells looks like a city after an earthquake: some buildings held, others collapsed, and the cracks show where the blow landed. Chemotherapy strikes DNA, but seeing that blow directly is hard: the damage is scattered across the genome and across many cells. Previously, scientists caught only small substitutions of DNA "letters" or cloned cells, which distorted the picture. Now researchers have taken a different approach: they read the DNA of tens of thousands of individual cells and compared them the way detectives compare shoe prints at a crime scene.
To isolate this footprint, the authors built a genealogical tree for each cell and subtracted mutations inherited from a common ancestor. What remained was a "mutational foreground": events that arose anew under the drug's action. They then decomposed these events into eight signatures, the way a mixture of paints is broken down into primary colors. It turned out that all five drugs increase the same signature, consisting of losses of large chunks of DNA at telomeres. In cells deficient in the TP53 protein, this footprint was 10-11 times more pronounced than in controls. And drugs that do not damage DNA left no such footprint.
Interestingly, the footprint persisted for up to 56 days after the drug was withdrawn. It is like a crack in a wall that stays visible long after the earthquake has ended. In tumors transplanted into mice, the signature appeared with the first treatment but faded as resistance developed. This means the tumor stopped accumulating genomic damage rather than simply getting used to mutations.
Today this approach can be used in preclinical research: to check how hard a new substance hits the genome and to select doses. In the future, it may become possible to predict which tumors will respond to treatment and which will quickly become resistant. But it is already clear: different drugs converge on a common footprint at the edges of chromosomes, just as different boots leave the same print on soft ground.
Telomere-bounded events arise within 2.5 megabases of the chromosome end, but short reads cannot resolve the telomeric repeats themselves, so the authors call them "telomere-bounded" rather than "telomeric."
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.