First Author

If the same person is found on the scene of multiple crimes, are they guilty by association? That's the question that Mark Pepys, a researcher at the Royal Free & University College Medical School in London, asked of C-reactive protein (CRP). This protein is produced in abundance in most cardiac diseases, including after heart attacks and strokes. Pepys has been on the CRP case since 1974.

There is no evidence that CRP causes cardiac problems, but Pepys says he has validated the protein as a therapeutic target by showing that high levels in the blood increase the damage caused by a heart attack. He is searching for a CRP inhibitor, but high-throughput testing of 500,000 known molecules yielded no leads. So he has drawn on his experience with a drug that targets a related protein and designed a new chemical entity. This has been synthesized and shown to bind tightly to CRP, blocking its inflammatory effect (see page 1217). Here, Pepys presents his case to Nature.

Why are you interested in targeting C-reactive protein?

Everyone has CRP. Whenever you're sick, CRP concentration increases. In everyone who has a heart attack, the CRP goes high. If it stays high, they don't get better. Clinical and experimental evidence suggests that the abundant CRP may be bad for you. So it's possible that you could reduce the damage from a heart attack if you could block the effects of the protein.

What does your finding say about drug design versus high-throughput screening?

It shows the potential for rational drug design — but the compound hasn't gone into patients yet, so we'll see.

Any big breaks along the way?

We had a big success with an inhibitor for another protein, which fed directly into the CRP project. This compound was two-handed with identical ends; it could therefore crosslink two target molecules and these were swiftly cleared from the patient's blood. Having previously solved the structure of CRP, we were able to design a two-handed compound that would crosslink this protein in a similar way.

Are there limits to rational drug design?

Yes. The structure of CRP suggested that shorter, twisted inhibitors would be better. But when we tried these they were much worse, whereas a longer, flexible molecule worked very well. The two CRP molecules crosslinked by the drug are slightly rotated on each other, and the length and flexibility of the inhibitor is evidently necessary to accommodate this.