As this publication recently reported in its survey of the "superstars of biomedicine, " McCormick ranks among the 25 most-cited scientists of the 1990s (see Science Watch, 9[1]:1-2, January/February 1998). So far in this decade, McCormick's research papers have garnered well over 10,000 citations. His 1991 Nature paper on the GTPase superfamily has over 1,300 citations alone (see the table on the next page).

   McCormick, now 48, did his undergraduate studies in biochemistry at the University of Birmingham and received his Ph.D. in 1975 from Cambridge University, working with Allison Newton. After postdoctoral stints at the State University of New York at Stony Brook and the Imperial Cancer Research Fund in London, he joined Cetus Corporation in 1981 as director of molecular biology and eventually vice president of research. In 1992, he founded Onyx Pharmaceuticals. In 1996, McCormick was made a Fellow of the Royal Society.

   McCormick: It was about 1984, and I was working at Cetus on their interferon project. We were trying to make human beta-interferon in mammalian cells as a potential source for clinical testing. It was the first time we actually managed to amplify a human gene in CHO cells, which are Chinese hamster ovary cells. At the time, the ras oncoproteins had been established as playing a causal role in human cancer. People were successful in producing these proteins in bacteria, so the possibility of doing biochemical analysis was suddenly on the table for the first time. I decided that it would be a good time for understanding the function of these proteins at a biochemical level eventually with the idea of designing therapeutics to inhibit ras oncogene function. So I got a ras oncogene pilot project started at Cetus, with a small number of people. Well, two people—Meg Trahey, who is a brilliant technician, and myself.

Why so small if you were so excited about it?

   McCormick: We had to prove that this whole idea was something worth doing in a company setting. We didn't know exactly where we'd end up or what kind of approaches we'd use.
   As it happened, it went very well. The first thing we did was to actually predict the 3-D structure of ras, based on some of our biochemical observations of the protein and on some structures solved on distantly related proteins, which turned out to be mostly correct. I was very proud of that but it was a bit of a sideshow. The main thing was trying to understand how the mutant protein differed from the normal. We knew from other people's brilliant work that the two proteins only differed by one amino acid, yet one causes cancer and the other is normal in cells. So Meg Trahey and I set about trying to understand at a biochemical level how this single amino-acid change could cause such a profound biological effect.

   There was already a beautiful model in the literature suggesting that the single amino acid change in ras prevented the ras protein from hydrolyzing a bound GTP molecule, which is what turns these kinds of proteins off. So the idea was that the normal protein would quickly hydrolyze GTP and turn itself off but the mutant protein wouldn't do that and would just stay on. The problem was that when you accurately measured the GTP hydrolysis rates of several different mutant and normal ras proteins, the differences were not significant enough to account for the tremendous difference in biology. So the basic idea might have been correct, but something was missing in the system that didn't fit. In looking for what could be missing, we discovered in cellular extracts that there is a factor that would massively accelerate the GTP hydrolysis on normal ras by thousands of fold but which had no effect on the mutant protein. We called that factor GAP for GTPase Activating Protein, because it had the property of accelerating GTP hydrolysis rates of the ras protein.

   As well as resolving this discrepancy in quantitation of the biological and biochemical properties of ras, that work also yielded the first cellular protein that actually regulated an oncogene. It is the first step from an oncogene protein into a pathway in a cell. At that time not many people were doing biochemical analysis of these proteins. Much of the field had been more genetic. So that was responsible for part of the move toward functional analysis of proteins and the proteins with which they interact and by which they're regulated. That was 1987. Now the crystal structure of GAP bound to ras has been solved, thanks to my long-time friend and collaborator Fred Wittinghofer, of the Max Planck Institute in Germany.
   It's quite remarkable. Ten years later the whole mechanism is now understood. We know that GAP contributes a single amino acid side chain into the active site, which, in a way, pushes the button that hydrolyzes GTP.

What prompted you to leave Cetus and head out on your own?