Bourne lab Home

Molecular mechanism of GPCR activation

Investigations in this and other labs have allowed us to formulate plausible and testable models of the 3D relations among GPCRs, G protein subunits (G-alpha, G-betagamma) and of how structural elements of these proteins move to relay hormonal signals.

This laboratory's recent studies of GPCRs were the work of three postdocs, Soren Sheikh, Tom Baranski, and Basil Gerber. Soren created Zn(II)-dependent salt bridges between histidines substituted at specific sites in transmembrane helices III and VI of retinal rhodopsin, the beta-adrenoceptor, and the parathyroid hormone receptor. In each case, he found that Zn(II) blocks ligand activation of the downstream protein. We inferred that: (a) these GPCRs (and, by extension, all seven-helix GPCRs) share a common 3D structure and switch mechanism; (b) all ligands activate the switch by somehow causing a separation or shift in orientation between the cytoplasmic ends of helices III and VI.

Tom devised a quite different approach, using a mammalian GPCR (the C5a receptor) and a human G-alpha subunit expressed in the budding yeast, S. cerevisiae. The yeast system allowed him to select for yeast colonies carrying either of two kinds of mutant GPCR: (a) constitutively active (in the absence of ligand); (b) able to activate the G protein only in the presence of agonist ligand. By selecting mutants of the latter class from yeast populations expressing mutant C5aR libraries, Tom found a small number of side chains that remained invariant in all functioning mutant receptors. One group of 8 such residues (yellow balls in the figure), located near the extracellular ends of helices III, V, VI, and VII, is thought to interact with the C5aR ligand. The second group (red balls), located in the transmembrane core of the helix bundle, is thought to constitute the core of the GPCR switch. From the patterns of mutations in his screen, Tom suggested that receptors are activated by a "clothespin"-like switch: ligand binding to extracellular portions of the receptor induces concerted movement of alpha-helices about a conserved core of residues, exposing to the cytoplasm a regionÑ probably located in between helices III and VI Ñ that is essential for activating the G protein.

Basil Gerber extended investigation of C5aR mutantsto elucidate docking of C5a agonists and antagonists into the ligand-binding pocket of the C5aR. He assessed effects of individual alanine substitutions for key conserved residues (chosen from TomÕs yeast screen) on receptor function in the yeast screen and in COS-7 cells. Assays of ligand binding and effect indicated that six of the yellow residues in the putative ligand-binding pocket do interact with the ligand. Indeed, one residue of a hexapeptide agonist and of a similar hexapeptide antagonist interacts with two of these yellow residues, located in helices III and VII. With Volker Dštsch, a UCSF faculty member, he obtained an NMR structure of the antagonist hexapeptide. Then, in collaboration with Elaine Meng, he docked the hexapeptide structure into the putative ligand binding site. His docking model predicted specific interactions between other groups in the ligand and side chains of receptor amino acids.

How does a trimeric G protein respond to activation by a receptor? Taroh Iiri, a postdoc (presently a faculty member at the University of Tokyo), proposed a model designed to explain how a GPCR can cause exchange of GTP for GDP bound in a pocket in G-alpha that is located very far from the plasma membrane (35 angstroms) Ð that is, too far for the GPCR to cause GDP release directly. Based on biochemical effects of G protein mutations in inherited endocrine disorders, 3D crystal structures of G protein trimers, and suggestive biochemical experiments, the model proposed that the GPCR uses both G-betagamma and the carboxy terminus of G-alpha to open the nucleotide-binding pocket. In this hypothesis, G-betagamma plays an active role, acting as a lever to trigger GDP release. By tilting slightly away from G-alpha, G-betagamma pulls open a lip of the GDP-binding pocket, allowing GDP to be released. GTP then enters the pocket, where its gamma-phosphate group deforms the lip of the pocket, thereby causing dissociation from both G-betagamma and the receptor.

Another postdoc, Philippe Rondard , tested this model by constructing mutations predicted to mimic the postulated mechanism of GDP/GTP exchange. As predicted, a G-alpha mutant, designed to mimic the tilted interface with G-betagamma, was activated by G-betagamma in the complete absence of GPCR stimulation. Moreover, G-beta mutations at sites predicted to interact with the lip of the GDP-binding pocket in G-alpha prevented G-betagamma from activating the mutant G-alpha. These experiments suggest that G-betagamma can act as a lever to open the GDP-binding site.

Last updated December 26, 2001