[Abstract]

Molecular pharmacology of metabotropic glutamate receptors - focus on group III and subtype selectivity

Mette Hermit

Metabotropic glutamate receptors (mGluRs) modulate neurotransmission in the brain and have been suggested as drug targets for the treatment a diverse set of diseases, including Alzheimer's disease, Parkinson's disease, epilepsy and anxiety. So far eight mGluRs have been cloned and subdivided into three groups. A prerequisite for the determination of the therapeutic prospects for the individual mGluRs is the availability of highly selective ligands for each subtype useable as pharmacological tools. In recent years great progress has been made in the fields of group I and II mGluRs, whereas there still is a lack of potent subtype selective group III mGluR ligands, which do not interact with the rest of the glutamatergic system.

This PhD thesis presents results obtained in studies of mGluR subtypes using homology modeling, molecular pharmacology and in vitro pharmacology. The project aimed at facilitating the development of subtype selective pharmacological tools, in particular for group III receptors, by elucidating orthosteric agonist-receptor interactions at the molecular level, by pharmacological evaluation of novel compounds at recombinant mGluRs and by detecting the effect of group III mGluR ligands in an in vitro model of neuroinflammation. Unfortunately, the attempt to set up an assay suitable for screening mGluR ligands ability to modulate neuroinflammation was not successful.

The basic part of the PhD study has been to set up pharmacological assays that enables test of compounds on recombinant mGluR subtypes. Implementation of newly developed pharmacological assays suitable for high throughput screening, such as measurement of intracellular Ca²⁺ release, and scintillation proximity assay (SPA) based [³⁵S]-GTP_yS measurements and SPA based radioligand binding, have facilitated characterization of novel compounds at mGluRs and have also proved valuable for characterization of mutant mGluRs. The pharmacological profiles of ibotenic acid and its sulfur analogue, thioibotenic acid, at mGluRs were determined and compared. Unexpectedly, thioibotenic acid displayed a distinct pharmacological profile at group III mGluRs being a low µM potent agonist, whereas ibotenic acid was inactive. This demonstrates how subtle differences in chemical structures can result in profound differences in pharmacological activity.

In search of potential pharmacological tools for group III mGluRs a series of novel phosphono-glycines were evaluated at mGluRs. Several phosphonoglycines with a cyclic skeleton were identified as low µM potent group III selective agonists. Counter-screens on the ionotropic glutamate receptors are ongoing. Since these compounds have molecular structures different from known ligands they should be useful prototypes for further investigation of the structure activity relationship and function of group III mGluRs.

The mGluR is a seven transmembrane spanning protein with a large extracellular amino terminal domain (ATD). The recent crystallization of the ATD of mGluR1 has brought valuable information about the endogenous agonist binding site, which reside in the ATD. In the PhD study, homology models of all mGluR subtypes were created using the mGluR1 ATD as template. Ligand docking to the homology models has shed light on agonist-receptor interactions that lead to subgroup/subtype selectivity. Residues corresponding to Y74, S164, E292, G293 and R323 in mGluR1 are predicted to have the highest impact on subtype/subgroup selective agonist binding, but additional residues also play roles in subtype selectivity.

The homology models have successfully been used in and supported by mutagenesis studies. A mutagenesis strategy, where residues were transferred from one subtype to another thereby switching the ligand selectivity was used. Residues important for differentiating group I mGluR1 from group III mGluR4 with respect to the group I selective agonists quisqualic acid and ibotenic acid were identified. The mGluR4 K74Y, E287G, S313G, and K317R mutations, which mimic mGluR1, all proved capable of increasing the affinity of the ligands at mGluR4 as predicted from the modeling study. The study shows that subtype selectivity results from a synergy of several residues within the mGluR binding pocket rather than from a specific interaction between single amino acid residues and the ligands.

The homology models have also offered plausible explanations for the characteristic pharmacological profiles of several standard mGluR agonists and for the novel compounds thioibotenic acid and (R)- and (S)-homo-TDPA. Homo-TDPA is unique because both enantiomers are equipotent at mGluR2. By use of comprehensive conformational analysis of the compounds and flexible docking procedure, the homology models should be useable for future ligand design.

Together, the results presented in this thesis have shed light on the molecular interactions between orthosteric agonists and mGluR proteins conferring subgroup/subtype selectivity. In addition, novel potent group III mGluR selective agonists have been identified.