This isn't anything I came up with. It is a chemotherapeutic drug called lidamycin.
It causes double strand breaks in DNA from that central enediyene ring rearranging into a phenyl ring, which forms an extremely reactive benzenoid diradical, which cleaves both strands of DNA (a Bergman cyclization).
Mechanism below
This class of compounds is extremely potent, killing cultured cancer cells in low picomolar quantities.
This is impressive to me, because most things this toxic are catalysts; that is, they exert their damage in a cycle, that regenerates the damaging potential. Heavy metals for example can just repeatedly grab electrons and then give them away to water molecules, creating reactive oxygen species that damage biomolecules. Similar are enzymatic toxins like ricin, that slice off one RNA base on a ribosome, disabling them from translating proteins.
These enediyenes are different. They are stoichiometric toxins. The majority of toxins are like this, where their toxicity comes from the formation of a chemical bond that breaks something inside of a cell. Cyanide binds iron in blood cells and mitochondria so makes you unable to make ATP. Organophosphates bind to acetylcholinesterase and prevent acetylcholine hydrolysis. Alkylating agents like mustard gas simply react with proteins that have nucleophilic hot spots, like the sulfur in cystienes.
The enediyene ring opening can only occur once, so at best one molecule of lidamycin could cut one duplex of DNA.
To have such a high potency (think three orders of magnitude more potent than common chemotherapeutics) while not being catalytic (or doing something like activating a hormone receptor potently like dioxin) boggles my mind.