We have continued our investigations of the basic mechanisms of receptor (GPCR) activation of G-proteins using a combination of surface plasmon resonance (SPR) measurement of GPCR-G-protein binding and in vitro assay of GPCR-catalyzed guanine nucleotide turnover. In this past year we have completed investigations of the interactions of the alpha-i family with bovine rhodopsin. We produced and purified homogeneously myristoylated alpha-i1 and alpha-o from e.coli in comparison with the endogenous retinal G-protein, transducin (Gt). All of these G-alpha subunits are members of the structurally-related alpha-i family, but they display distinct endogenous GTP exchange rates, beta-gamma dimer affinities, and receptor specificities in addition to cellular expression. Our initial findings, which confirmed published reports, suggested rhodopsin could catalyze guanine nucleotide exchange on all three alpha subunits with nearly identical kinetics in vitro. However, examined by SPR with immobilized bovine rhodopsin, the alpha-i1 and alpha-o show dramatically decreased dissociation rates in comparison to alpha-t. Whereas alpha-t combined with the retinal beta1-gamma1 dissociates with a half-time less than 5 sec, the half-life of alpha-o is about 60-90 sec, while that for alpha-i1 is on the order of 1000 sec. We determined that this apparent discrepancy arises from a biochemical artifact of the myr-alpha-i1, which is strongly enhanced in spontaneous GDP-dissociation by lipid bilayer structures. The actual rhodopsin contribution to GDP-dissociation of this protein is marginal, consistent with the very prolonged lifetime of the rhodopsin-alpha-i1 complex. Detailed kinetic examination of the rhodopsin-catalyzed nucleotide exchange on alpha-t suggests that the currently held mechanism of GTP-assisted dissociation of activated alpha from receptor may not explain these data. As a test of this, we constructed point mutations of alpha-i1 at amino acid residues that undergo changes in guanine nucleotide interaction between GDP-bound and GTP-bound conformations of alpha. Our data for alpha-i1 show that the G203A mutant spontaneously exchanges GDP similar to wild-type, but is impaired in rhodopsin-catalyzed exchange. Mutation of R208A leads to an alpha that exchanges GDP similar to wild-type, but shows diminished affinity for GTP. SPR studies find that G203A alpha-i1 binds to rhodopsin with a prolonged life-time, independent of exogenous guanine nucleotide. Together, these data suggest that GTP is not required for dissociation of activated alpha from receptor; rather, the receptor catalyzes the activation and dissociation of an un-liganded alpha subunit, which subsequently binds GTP. We are continuing these studies using three additional GPCR structures: a structurally distinct cephalopod rhodopsin, recombinant M1-muscarinic acetylcholine receptors, and the 5HT2c subtype of human serotonin receptors. We have also continued examination of the unique properties of the family3 GPCR structures, examining mutant constructs of metabotropic glutamate receptors (mGluR1) and calcium-sensing receptors (CaR) as well as a gold-fish taste receptor (5.24), reportedly an arginine receptor (Arg-R). Previously we have reported that the seven-transmembrane helix bundle (7TM) of the human CaR (t903-rhoC) without the amino-terminal calcium binding domain (ECD) can be activated by three allosterically interacting sites for divalent cations, polyvalent organic cations (poly-Arg, spermine) and the synthetic ligand NPS 568. Mutation of all five acidic residues in the second extracellular loop of t903-RhoC abrogated the NPS568, but not PolyArg synergy of calcium activation. We have completed examination of the importance of the second extracellular loop and other loci of charged residues in the extracelluar sequences of the hCaR 7TM in the context of the full-length structure. The 5-alanine mutant homolog was found to be strongly activating. Of particular interest, mutation of a single residue in the second extracellular loop (E767) displays a high intrinsic activity, and abrogates the steeply cooperative activation of the hCaR by calcium. Mutation of the single residue K831 of the third extracellular loop is similarly, but not so strongly activating. Analysis of various amino acid substitutions at these two positions clearly identifies them as participating in ionic interactions. However, this pair of oppositely charged residues does not appear to form an internal salt-bridge. We expect that they may be interacting with charged residues within the N-terminal extracellular domain (ECD) of the hCaR. These studies amplify on our initial observations of the interacting allosteric sites within the 7TM core of the hCaR, suggesting that the calcium-binding ECD interacts with the 7TM core through a contact at this residue. Further, these data strongly imply that the ECD may be an inhibitory constraint on the activity of the 7TM core. We have set out to test this suggestion by independently expressing the ECD and 7TM core structures for mGluR1, hCaR, and 5.24. Currently, we have high level expression of the mGluR1 and hCaR 7TM cores and baculoviral constructs for secreted ECD proteins from mGluR1 and 5.24. When our complement of molecular reagents is completed, we will undertake the isolation to homogeneity of the three ECD structures, and we will examine the ligand binding and 7TM core regulating properties of these constructs. The 5.24 ECD is currently expressed with sufficient yield to entertain obtaining a crystal structure for this protein, which would be of immense interest to compare with the available structure of the rat mGluR1 ECD. However, our initial biochemical evaluation of both the full-length 5.24 receptor and the autonomously expressed ECD of 5.24 has refuted the published report that this receptor responds to L-arginine. We are currently trying to identify the ligand for this receptor. Our project in collaboration with Dr. Susan Sullivan, NIDCD seeking to identify the tastant compounds recognized by the entire repertoire of human genes encoding bitter taste receptors has made considerable progress this past year. Dr. Sullivan has identified 23 candidate genes from the human genetic databases with high similarity to the known mouse and rat bitter taste receptors but which are not olfactory receptors. At present we have constructed baculoviral vectors for expressing these, and have characterized 10 of them as directing the expression of a cell-surface localized receptor. In parallel, Dr. Dennis Drayna, also of NIDCD, has identified the human gene locus encoding the phenylthiocarbimide (PTC) trait. His studies revealed five allelic variants (2 tasting, 3 non-tasting) within the human population. Baculoviral vectors expressing all of these gene products have also been constructed. We are continuing the development of a novel cell-based screening strategy to identify the ligands for these receptors. We have constructed baculoviral vectors for the taste-enriched G-beta-3 and gamma-13 subunits, the myeloid alpha-15/16 proteins and jellyfish aquorin. By simultaneous infection of Sf9 cells with these and viruses encoding the taste receptors, we expect to re-direct the G-protein signaling pathways initiated by the tastants to the production of chemiluminescence. Our initial tests of this have utilized several well-characterized receptors (5-HT1A, 5-HT2C and GRP-R) to confirm that the strategy can succeed. Further, infection with the mouse bitter taste receptor T2R5 succeeds in producing chemiluminescent Sf9 cells in response to cyclohexamide. To date, we have not succeeded with this strategy to identify any of the tastants for the human bitter taste receptors. We have, however, succeeded with an alternative strategy of in vitro assay for these recept
Showing the most recent 10 out of 19 publications
