The Johns Hopkins Gazette: March 3, 2003
March 3, 2003
VOL. 32, NO. 24


Scientists Unlock 3-D Structure of Key Breast Cancer Receptor

By Joanna Downer
Johns Hopkins Medicine

Johns Hopkins Gazette Online Edition

A team of scientists from Johns Hopkins and the biotechnology company Genitope has unlocked the 3-D structure of a receptor that goes awry in 20 percent to 30 percent of breast cancers. Reporting in the Feb. 13 issue of the journal Nature, the scientists also figured out how the receptor, known as HER2, interacts with an antibody, sold as Herceptin, that is used to treat thousands of breast cancer patients each year.

"Now we know exactly which building blocks of the Herceptin antibody interact with which building blocks of HER2," says Dan Leahy, professor of biophysics and a Howard Hughes Medical Institute investigator at the Johns Hopkins School of Medicine. "When you understand the properties of receptors and antibodies in terms of their structural interaction, you can begin to explain their effects and use the information to design better drugs."

Developed by Genentech, Herceptin kills cancer cells carrying excess HER2. But even though Herceptin was approved as a breast cancer treatment by the U.S. Food and Drug Administration in September 1998, until now no one has known precisely how it interacted with the receptor.

The findings also explain why the HER2 receptor behaves so differently from its relatives HER1, HER3 and HER4, says Leahy, in whose laboratory the structure of HER3 was deciphered last summer. While all four proteins are similar in the sequence of their building blocks, only excess HER2 leads to uncontrolled cell growth in the lab and breast cancer in people.

Like its relatives, the HER2 receptor is stuck in the cell membrane, partially outside the cell, and partially inside. The extracellular part, whose structure the scientists determined, is the receptor's "on switch." Through this region, HER receptors join into pairs to become fully active and trigger events that eventually result in cell division.

Comparisons of HER family structures reveal that a few key changes in the sequence make all the difference for HER2. Specifically, HER2 doesn't need to be "opened" before it can pair with another HER, which is why extra HER2 can cause cancer, and why no one has found any small molecules, or ligands, that bind to it, Leahy says.

"We can see now that it's unlikely any natural ligands for HER2 exist--it just doesn't need one to work," Leahy says.

Led by postdoctoral fellow Hyun-Soo Cho, the research team grew crystals of the extracellular region of the HER2 protein, provided by scientists at Genitope Corp. in California. After bombarding the crystals with radiation at the National Synchnotron Light Source at Brookhaven National Lab on Long Island, N.Y., Cho and Leahy interpreted the data to create the protein's structure. The team did the same with crystals of HER2 bound to Herceptin antibody.

The extracellular regions of all the HER proteins are made up of four structurally distinct "domains," I, II, III and IV. In HER1 and 3, a fingerlike projection in domain II keeps it connected to domain IV, forming a bracelet-like loop. In HER2, however, a tight interaction between domains I and III, which Leahy likens to a spot weld, prevents the bracelet formation. Thus domain IV is available for binding to Herceptin, and domain II for pairing with other HER proteins.

Authors on the paper are Cho, Leahy, Kasra Ramyar, Ann Marie Stanley and Sandra Gabelli of Johns Hopkins; and Karen Mason and Dan Denney Jr. of Genitope. The Johns Hopkins scientists were funded by the National Institutes of Health and the Howard Hughes Medical Institute.