The diversity of animal species intrigued John Kuriyan as a child. By the time he reached college, Kuriyan realized that the proteins inside living cells might be just as diverse—and even more fascinating.

"To me, the modern equivalent of studying the relationship between animal form and function is to study the molecular structure and function of proteins," he says. That led him to switch his major from zoology to chemistry, and complete a Ph.D. and postdoctoral fellowship in biophysical chemistry.

Kuriyan wants to understand the workings of the diverse proteins that regulate cell interactions and cell division, including DNA replication. Using x-ray crystallography, he studies the structures of these proteins. One class he studies most intensely are the protein kinases. These molecules regulate most cell functions, and they are especially important in sending messages across the cell membrane. There are about 500 human protein kinases, and they all work by adding a phosphate to specific sites on target proteins. This chemical reaction changes the shape of the protein, affecting the way it interacts with other molecules. Ultimately, kinases can prompt cell division, movement, and death.

In the early 1990s, Kuriyan discovered how changes in protein structure affect the regulation of Src kinases, a family of proteins important in several cancers. Concurrently with a Harvard laboratory, Kuriyan worked out the three-dimensional structure of one Src kinase. His work showed how certain parts of the protein, called regulatory domains, are important in "turning on" the kinase to send messages, and then turning it back off.

"It was the first structure that showed a protein kinase with its regulatory domains in place," Kuriyan says. The way the regulatory parts of the protein worked was not intuitive; it reminded Kuriyan of a Rube Goldberg contraption. "If you were going to design a way to do it, you wouldn't do it that way. It highlights the happenstance in the evolution of function in biology."

More than a decade later, Kuriyan found a similar "molecular handshake" mechanism at work in the BCR-Abl kinase, an abnormal protein that's behind most cases of chronic myelogenous leukemia.

The idea that activity in one part of a protein can affect the behavior of other parts of it is called allostery. Allosteric interactions are common in biology; one of the best-known examples may be oxygen's interactions with hemoglobin. When one oxygen molecule binds to hemoglobin, it changes the shape of the hemoglobin molecule, making it easier for other oxygen molecules to bind.

Because all kinases carry out the same phosphate-addition reaction, and because many active kinases look similar to one another, you might expect they would all work the same way. But experts such as Kuriyan know that's not the case. In fact, he says, "Kinase mechanisms are surprisingly diverse."

Although kinases look similar when active, they look different when they're switched off. After his work on the BCR-Abl kinase, Kuriyan and colleagues discovered that the anticancer drug Gleevec works by recognizing the unique "switched-off" form of BCR-Abl. When Gleevec binds to that kinase, it stops it from triggering the growth of leukemia cells. The specificity of Gleevec—it binds to only two or three of the 500 human protein kinases—makes it a powerful drug (with few side effects) against chronic myelogenous leukemia, but not against similar cancers.

Two years after discovering why Gleevec works so well, Kuriyan and fellow HHMI investigator Charles Sawyers identified 15 different gene mutations that cause resistance to the drug. Some mutations change the BCR-Abl kinase at the binding site for Gleevec, so the drug cannot bind. But surprisingly, many mutations cause changes far from the binding site. Kuriyan says that these mutations may change the shape of the kinase just enough to keep Gleevec from recognizing it.

Kuriyan's lab also has worked out exactly how a protein called epidermal growth factor (EGF) receptor is activated. This receptor, which sits inside the cell membrane, is important for cell division and is overactive in many cancers. Kuriyan found that in order to activate, EGF receptors must pair up, or dimerize. Then, one receptor can receive a message from a protein outside the cell and relay it inside the cell. Now that the sequence of events has been delineated, intervention at any step may hold a key to new cancer treatments.

"The diversity we're used to seeing in the forms of life is also reflected in these molecular mechanisms," Kuriyan says. "A cell might have hundreds of proteins that all work in different ways. That makes for endless study."