We have continued our investigations of the determinants of G-protein-coupled receptor (GPCR) selectivity for G-protein using recombinant expressed GPCRs and G-protein alpha subunit chimeras to map the contact site(s) on the alpha subunit responsible for receptor-selective interactions. Previously, we had constructed chimeras between G-alpha-i1 and G-alpha-q using human 5HT1A and 5HT2c receptors which are uniquely selective for the respective G-alpha subunit types. By co-expression of the chimeric G-alpha constructs with the years N-myristoyltransferase enzyme in e. coli hosts and purifying the myr-G-alpha proteins, we succeeded in producing receptor-competent G-alpha chimeras and we mapped the contact site(s) to include at least 54 residues of the carboxyl terminus of G-alpha-q. To confirm this mapping, we initiated the examination of two additional, distinct G-alpha-q coupled GPCR types, the thyrotropin releasing hormone (TRH) receptors (TRH-R1 and TRH-R2) in collaboration with the laboratory of Dr. Marvin Gershengorn, NIDDK and the neurotensin receptor (NT-R) in collaboration with the laboratory of Dr. Reinhard Grisshammer, NINDS. ? Our initial studies using in vitro G-protein reconstitution techniques refuted the apparent coupling of the human TRH-R2 with the G-alpha-i1 or G-alpha-o proteins found by examining intracellular second messenger pathways in transfected cells. Rather, the in vitro data show unique activation of G-alpha-q by this GPCR. However, this assay also revealed a quantitative difference in the agonist efficacy among synthetic analogs of TRH for the activation of G-alpha-q. This quantitative difference was confirmed for the cell-based response reporters as well. Intriguingly, the most efficacious agonists were far lower affinity than the native TRH, suggesting a possible link between ligand dissociation rates and agonist efficacy.? For the NT-R, we similarly refuted an earlier report of the coupling to the G-alphai1 protein, finding highly selective activation of G-alpha-q. Thus, both the TRH-R2 and NT-R will be used for further examination of the selective contact site(s) for G-alpha-q using the chimera set for which contacts have been mapped for 5HT2c. We have also investigated the possibility of producing chimeric G-alpha subunit chains in e. coli using the G-alpha-s as the acceptor structure and in collaboration with Dr. Brian Kobilka, Stanford Medical School we have investigated interactions with the beta-adrenergic receptor. As opposed to G-alpha-i1 which must bear a lipid modification on its amino terminus for any competent G-beta-gamma or GPCR interaction, our examination of the e. coli produced r-alpha-s which has no lipid modification reveals sufficient affinity for facile GPCR interaction. In fact, the r-alpha-s (non-modified) has a higher affinity than a chimeric structure of N-terminal alphai1-alpha-s which bears homogeneous myristoylation. This suggests that the alpha-s provides a better base structure for further studies. With alpha-s chimeras we could produce competent G-alpha in e. coli without the co-expression of the NMT enzyme and the tedious isolation of homogeneously myristoylated G-alpha structures required for the G-alphai1 chimeras we have been producing. This additional set of chimeras should also provide a strong test for the accuracy of the mapping of the selective contact surface for GPCRs.? The NT-R receptor we have used in collaboration with the laboratory of Reinhard Grisshammer is a homogeneous, purified preparation suitable for crystallization. We have further examined its suitability for the formation of stable complexes with G-protein subunits. Our initial results have identified detergent conditions that allow the productive, ligand regulated interaction of G-alpha-q and G-beta-gamma with NT-R. Under optimal conditions, NT stimulates the NT-R catalyzed rate exchange of GTP for GDP by 50-fold, with Km for G-alpha-q of about 300 nM and G-beta1-gamma2 of about 200 nM. Thus, it is feasible to prepare the complex of NT-R with G-alpha-q and/or G-beta1-gamma2 in defined detergent conditions. The biophysical characterization of these complexes under these conditions provided the surprising result that the NT-R is a dimeric complex. In the absence of G-protein, this dimer displays positive cooperative binding for NT. The apparent Kd for the monomer-dimer interaction is in the low nM range, which is 2 orders of magnitude higher affinity than any found so far for transmembrane proteins in detergent solutions. Even more surprising, while the NT-R clearly prefers a dimer organization, the monomer displays both a higher affinity and considerably higher catalytic activity towards G-alpha-q than does the monomer. Our current studies are aimed at the definition of the stoichiometry of G-protein subunit interaction with the dimeric NT-R structure and an elucidation of the molecular mechanism for G-protein activation.? We continued our investigations of the family c GPCRs with a study of the role of the linker sequences of the human calcium sensing receptor (hCaR) and the metabotropic glutamate receptor 1 (mGluR1). These studies were designed to reveal the role(s) played by this region of the family c receptors in transmitting the ligand binding signal from the autonomously folding extracellular domain (ECD) to the transmembrane helix bundle (7TM) for G-protein regulation. Our findings are that disruption of this part of the hCaR structure in most cases leads to misfolded GPCR. However, a scanning alanine mutagenesis of the highly conserved residues within this 14 amino acid sequence revealed 3 positions in which the mutations were tolerated and moderately activating of the hCaR and a fourth in which mutations produced significant misfolding and uncoupling of the ligand regulation through the ECD. These data complement our previous work suggesting essential contacts between the ECD, linker and the extracellular loops 2 and 3 of the 7TM domain.? Lastly, in collaboration with Dr. Susan Sullivan, NIDCD, we have examined the signaling properties of novel constructs produced from the human sweet and amino acid sensing T1Rs (T1R1-3). These GPCRs belong to the family C, which is characterized by an autonomously folded amino-terminal domain which is the site(s) for ligand regulation. The T1R taste receptors are heterodimeric complexes of T1R1/T1R3 or T1R2/T1R3. Thus, they present many unique challenges for expression and molecular characterization. Therefore, we have constructed truncated receptor structures from these which consist of the transmembrane helix bundle (7TM) domain, the portion which is responsible for G-protein interaction, with which to test the T1R-G-protein selectivity. We were unable to select stable mammalian cell clones expressing these constructs. Even transient expression in mammalian cell hosts lead to a lack of viability for the T1R1 and T1R2 7TM domains. Therefore, we constructed baculovirus vectors for their expression in insect cells. Using membranes derived from Sf9 cells expressing the T1R 7TM domains, we find that both T1R1 and T1R2 constructs constitutively activate G-proteins of the alphai family (alpha-i1, alpha-o, alpha-gustducin and alplha-transducin). None of the identified sweet tastants or allosteric regulators modulate the activation for these receptors. These data suggest that the extracellular domains for the T1Rs (and likely the entire family c) are inhibitory constraints on the 7TM domain. Further, these experiments define a G-protein pathway which is likely to be mediated by the common beta-gamma component of the alphai family.
Showing the most recent 10 out of 19 publications
