Seeing Cisplatin

CISPLATIN HAS BEEN USED to treat cancer for decades. But the drug's detailed molecular interactions with the enzymes it aims to interrupts have remained obscure. In two separate papers, researchers are now reporting the first crystal structures of cisplatin complexed with two key enzymes.

"For the first time, we are getting deep insight on what goes on at the atomic level with cisplatin and cellular machinery," says Thomas Carell, a protein chemist at the University of Munich.

Cisplatin creates bulky platinum adducts with pairs of guanidine bases that are side-by-side or one base apart on genomic DNA. The extra bulk blocks most DNA polymerases that slide along DNA when they copy the genetic code during cell division. The adducts also obstruct RNA polymerases, which transcribe DNA to make proteins. But no structures of cisplatin in complex with either DNA or RNA polymerases have come to light until now.

In one report, Carell and Karl-Peter Hopfner announce the structure of a renegade DNA polymerase called Pol η. Normal DNA polymerases get stalled by cisplatin because their grip on DNA is so tight that there's no room for adduct riffraff. But Pol η manages to replicate over unwelcome DNA adducts.

Pol η evolved to give cells sufficient flexibility to work in the presence of sunlight-induced thymine dimers that would normally jam replication. The flexibility prevents cells from dying while waiting too long for DNA repair machinery to untangle dimers. While Pol η sustains mostly accurate replication during sunlight exposure, it unfortunately also helps tumors replicate during cisplatin chemotherapy.

The new structure reveals that Pol η replicates over cisplatin adducts by having a much looser grip on DNA than other polymerases, allowing the adducts to fit into the catalytic site. The enzyme's catalytic site also has amino acids that help recruit bases for the cisplatin adduct (Science 2007, 318, 967).

By understanding how a tumor's DNA replication machinery overcomes cisplatin adducts, the Munich researchers hope new cisplatin derivatives may be designed that avoid Pol η's relaxed grip and thereby overcome tumor resistance to cisplatin.

The work reveals "very interesting structures that will help the community understand how DNA polymerases bypass the major cisplatin lesion on DNA," comments Stephen J. Lippard, a professor of chemistry at MIT. But he suggests that understanding tumor resistance may require work beyond Pol η.

While Pol η keeps a tumor cell replicating during cisplatin therapy, the drug can still hinder the tumor's RNA polymerases, Lippard notes. Finding out how tumors also survive the transcription traffic jam would provide a complete picture of tumor resistance, he says.

A second reported crystal structure takes a step in that direction. Carell and Patrick Cramer, also at the University of Munich, show that transcription is blocked because cisplatin adducts cannot wedge themselves into the RNA polymerase catalytic site (Nat. Struct. Mol. Biol., DOI: 10.1038/nsmb1314).