Membrane Protein Activation

THE PORTFOLIO OF tools for probing membrane proteins, which are notoriously hard to study, has just been expanded. A research team has designed helical peptides that can selectively activate membrane proteins involved in cell signaling. The achievement, a long-standing goal in the field, could ease studies of the way membrane receptors work and could lead to new drugs.

A major challenge in biology is figuring out how signaling molecules like hormones interact with cell-surface receptors to spark a range of cellular responses. Typically, binding of a signaling molecule to a cell-surface receptor causes membrane-spanning helical domains in the receptor to pair, thus propagating the signal to the cytoplasm and often to the nucleus.

It hasn't been possible to get compounds inside cell membranes for the specific purpose of selectively influencing such helix-binding interactions. Researchers can use antibodies or antibodylike reagents to selectively target soluble proteins, but they haven't had comparable tools for targeting membrane proteins.

Professor of biochemistry and biophysics William F. DeGrado, professor of medicine Joel S. Bennett, postdoc Hang (Hubert) Yin, graduate student Joanna S. Slusky, and coworkers at the University of Pennsylvania School of Medicine have now addressed that problem by designing peptides that selectively activate signaling by integrins, which are membrane proteins involved in platelet adhesion, blood clotting, and other key processes (Science 2007, 315, 1817).

"It's the first targeting of a membrane protein by a computationally designed ligand," says biochemistry professor David Baker of the University of Washington, Seattle. "The study is a real tour de force on both the computational design and experimental fronts."

"It represents an exciting breakthrough because it demonstrates that it is already possible to design useful transmembrane peptide activators using very simple principles," says professor of chemistry and biochemistry James U. Bowie of the University of California, Los Angeles. "The dream of building a reagent to modulate the activity of any membrane protein of interest simply by running a computer program now seems like a more realistic goal. Whether the design principles used will translate to other classes of interaction surfaces remains to be seen."

To design the peptides, DeGrado and coworkers used CHAMP (computed helical antimembrane protein), a computational approach that optimizes spatial fit between the peptides and specific membrane helices. So far, it has worked amazingly well: Each of three designed peptides bound to and selectively activated an integrin on the first try.

Karyn T. O'Neil, a former DeGrado grad student who is now director of protein optimization for Centocor, Radnor, Pa., says the approach "could be a valuable tool for drug discovery," except that peptides generally aren't ideal therapeutics because they're broken down and eliminated from the bloodstream quickly. However, peptide analogs with better drug properties could be designed, an idea DeGrado and coworkers are already pursuing.