Cisplatin Was Found by Trying to Electrocute Bacteria

In 1965, Barnett Rosenberg at Michigan State was looking down a microscope at dividing cells and noticed the spindle fibers fanned out the way iron filings do around a magnet. He wondered whether an electric field could affect cell division the same way a magnetic field shapes filings. So he rigged up a beaker of E. coli, dropped in two platinum electrodes — chosen because platinum was supposed to be biologically inert — and ran a current.

The bacteria stopped dividing. They kept growing, stretching to as much as 300 times their normal length, but they would not split. Rosenberg's lab spent two years chasing the electric-field hypothesis before they figured out the current had nothing to do with it. The platinum electrodes, supposedly inert, were quietly releasing a compound into the broth. That compound was doing the work.

It took another two years to identify it: cis-diamminedichloroplatinum(II), or cisplatin. The molecule binds to DNA and stops cells from copying it. That is poison to a bacterium and, it turned out, to a tumor.

Cisplatin was approved by the FDA in 1978. Before it, a young man with metastatic testicular cancer had a survival rate of around 10%. With cisplatin in the regimen, cure rates climbed past 90%. The U.S. death rate from testicular cancer has dropped by roughly two-thirds since 1975.

Rosenberg never did get the experiment he set out to do. He was trying to prove something about electric fields and biology. Instead, he stumbled onto one of the most consequential cancer drugs of the twentieth century — by accidentally dissolving his own electrodes.