Investigations on neuropeptide Y and histamine receptors by fluorescence-based and radiochemical methods

In the last three decades, a large amount of experimental data (for example Förster resonance energy transfer (FRET) at fluorescently labelled receptors) supported the hypothesis of the dimerization of G-protein-coupled receptors (GPCR), including neuropeptide Y (NPY YxR) and histamine (HxR) receptors. Targeting GPCR dimers could be helpful to understand (patho)physiological processes and may be applied therapeutically. This work aimed at the exploration of the applicability of FRET-based and ligand-based approaches to the investigation of NPY Y1R and histamine H2R dimers. Special attention was paid to bivalent NG-acylated hetarylpropylguanidines, the most potent H2R agonists, known so far, which were initially designed to bind to the orthosteric recognition sites of dimerized H2R protomers. Surprisingly, agonists having octamethylene linkers, insufficient in length to bridge two adjacent receptor molecules, proved to be most active. Thus, the intent of this thesis was to investigate the assumed mode of binding of such “twin compounds”, e.g. with respect to the ternary complex model, allosterism and functional selectivity. Fluorescently labelled NPY Y1 receptors: Human Y1Rs, tagged either with enhanced cyan (ECFP) or with yellow fluorescent proteins (EYFP), were stably expressed in human U-373 MG glioblastoma cells (plasmids from Prof. Dr. A. Beck-Sickinger, University of Leipzig). Membranal localization of the labelled receptors was confirmed by radioligand and flow cytometric binding assays. However, imaging of the transfected U-373 MG and CHO cells by means of confocal microscopy (CLSM) revealed primarily intracellular distribution of the fusion proteins, precluding prospective FRET-based investigations of receptor dimerization. Additionally, bivalent argininamide-type Y1R antagonists, developed in our group, did not support the hypothesis of receptor dimerization. Thus, all further work was focused on bivalent H2R/H4R ligands. Hill slopes of bivalent H2R/H4R ligands: Radioligand binding assays were performed on Sf9 membranes, expressing either the human (h) or guinea-pig (gp) H2R-GsαS fusion protein or the hH4R, because the imidazolylpropylguanidine moiety turned out to be a privileged structure with respect to histamine receptors (cooperation with Prof. Dr. Roland Seifert, Medical School of Hannover). Saturation binding experiments confirmed that tiotidine is an inverse agonist of H2R, and that fusing of the H2Rs and GsαS proteins did not alter ligand-receptor interactions. GTPγS-dependent shifts of the competition binding curves, predicted by the ternary complex model, were found for monovalent H2R agonists, but not for the bivalent ligands, tested at the gpH2R-GsαS, suggesting antagonism rather than agonism. Additionally, Hill coefficients (nH) 1 (positive cooperativity) for the bivalent compound with a 20 mem¬bered linker, approximated unity after very long periods of incubation (gpH2R-GsαS and hH4R), disclosing the failure of having reached equilibrium. Radioligand off-rate (k2): Radioligand dissociation kinetics at the gpH2R-GsαS and the hH4R (k2 of [3H]histamine at hH4R: 0.0108 ± 0.0007 min-1, reported for the first time) revealed a concentration-dependent increase in k2 in the presence of bivalent ligands. This could result from amphiphilic properties (for the “longer” bivalent compounds) or from an existence of the allosteric binding site either at the same protomer or at a neighboring receptor. H2R internalization: Clathrin-mediated receptor endocytosis was visualized in transfected CHO cells by confocal microscopy (CLSM). Both mono- and bivalent H2R agonists, but not bivalent antagonists (with respect to G-protein activation), induced the internalization of human H2Rs fused to the monomeric far-red fluorescent protein mKate2. These observations were in line with previously obtained functional data from GTPase assays, and there is no evidence for unconventional (biased) signalling of the hH2R. By contrast, agonist-dependent activities of standard antagonists were found in GTPase assays at the hH2R-GsαS in presence of a bivalent agonist with an 8-membered linker. In conclusion, the obtained results cannot be solely explained by the competitive binding model. Considering the size of the bivalent H2R/H4R ligands, additional temporary interactions with the extracellular domains of the receptors seem to be inevitable. Provided that the ortho- and the allosteric binding sites were occupied by two pharmacophoric moieties simultaneously, these bivalent compounds might act as true dualsteric ligands. Future work at the H2R and the H4R should be focused on the differentiation between G protein-dependent and β-arrestin-mediated signalling. Additionally, the preparation of a bivalent radioligand is suggested to determine the ligand-receptor stoichiometry