Why DNA damage is a special category
Most toxicities are about dose: enough harms, a little is fine. Genotoxicity — damage to DNA — is held to a stricter standard because a single mutation in the wrong place can, in principle, seed a cancer years later. There is no comfortable therapeutic window argument for an everyday medicine that scrambles the genome. For that reason, a positive genotoxicity signal is one of the few findings that can stop a program outright, and testing for it is a mandatory part of preclinical development.
There is an instructive exception that proves the rule. Some cytotoxic cancer drugs — the classic alkylating agents — are *deliberately* genotoxic: they kill dividing tumour cells precisely by damaging DNA. The toxicity is the mechanism. That trade-off is only acceptable when the disease is life-threatening. For an everyday medicine taken by healthy people, the same chemistry would be unthinkable.
How molecules attack DNA
DNA gets damaged in two main chemical ways, and both map onto structures you can recognize. The first is direct alkylation: an electrophilic group reacts with the nucleophilic bases of DNA and forms a covalent adduct, miscopying when the cell divides. Many of the same structural alerts from the last guide apply here too, since a reactive metabolite that attacks protein can just as easily attack DNA. The second way is intercalation: flat, planar polycyclic aromatics slide between the DNA base pairs and distort the helix, jamming replication.
Screening, and the impurity trap
The workhorse early screen is the Ames test: bacteria carrying a mutation that stops them growing are exposed to your compound, and if it mutates them *back* to growing, the colonies that appear reveal a mutagen. It is cheap, fast, and run both with and without liver enzymes — the second condition catches the case where it is not the drug but its reactive metabolite that does the damage. A clean Ames result is one of the first safety boxes a series needs to tick.
One subtlety surprises newcomers: genotoxicity is policed not only for the drug itself but for the trace impurities that ride along from synthesis. Because a genotoxic compound is dangerous at vanishingly small amounts, regulators set extremely low limits on genotoxic impurities — sometimes only a few parts per million. A reagent or by-product that is a known mutagen can therefore force a whole synthetic route to be redesigned, even when the final drug is perfectly clean. Safety thinking extends all the way back to the flask.