Deorphanizing novel modulator binding sites in Cys-loop receptors

Cys-loop receptors are ligand-gated ion channels that play a significant role in the central nervous system regulating numerous central nervous diseases. Members of the Cys-loop receptor family include nicotinic acetylcholine (nACh), γ-amino butyric acid (GABAA), 5-hydroxytryptamine(5-HT3) and glycine (Gly) receptors. These receptors share an overall similar structure and mode of action. As targets for drugs, these receptors modulate and relieve symptoms of diseases including epilepsy and Alzheimer’s Disease. AM-3607 and AM-1488 are two modulators of Glycine receptors (GlyRs) that were reported recently by Amgen. Structural studies and co-crystallisation with the human α3 GlyR identified these compounds as belonging to a novel class of positive allosteric modulators (PAMs) with a binding site not previously reported in any Cys-loop receptor. The compounds bind to a pocket just above the glycine binding pocket from where they enhance the binding affinity and potency of glycine. In the present project we aimed to investigate if a site in a similar position in other Cys-loop receptors also could act as a PAM site. Overall, this thesis aims to identify the existence of this binding site on other Cys-loop receptors. To achieve this, the 3D structure of AM-3607 co-crystallised with human α3 glycine receptor was analysed to identify residues lining the binding site. Subsequently, corresponding residues in all other Cys-loop receptor subunits were identified based on a sequence and protein structure alignment to identify the subunits and subtype most similar to the glycine α3 subunit. Together with considerations regarding common existence and ease of handling in pharmacological experiments, the α1β2 GABAA receptor was selected for further studies. A total of 6 residues were then mutated to introduce a binding site on α1β2 GABAA receptor similar to the AM-3607 binding site on the glycine receptor. In an alternative attempt, due to a structural difference result from the different length of a Tyr78 containing loop connecting two β strands, a chimeric α1 GABAA receptor was prepared incorporating the loop from the glycine receptor. The effect of mutant and chimeric receptors was subsequently evaluated with GABA and GABA in combination with AM-1488 using two-electrode voltage clamp electrophysiology. The concentration response curves revealed EC50 values for (α1)3(β2)2 and (α1)2(β2)3 GABAA receptors of 2.59 μM and 1.08 μM in agreement with literature values. Mutant receptors, GABAA (α1M)3(β2)2, (α1M)2(β2)3, (α1)3(β2M)2, (α1M)3(β2M)2, (α1M)2(β2M)3 and chimeric receptors (α1C)3(β2M)2 and (α1C)2(β2M)3 showed EC50 values of 0.65 mM, 0.65 mM, 11.8 μM, 3.38 mM, 5.60 mM, 0.62 mM, 0.92 mM, respectively. For all constructs involving mutant β2 subunit, potency was significantly impaired compared to wild type receptors. Further, none of the constructs were potentiated when AM-1488 was co-applied with GABA. This research identified that AM-1488 had no potentiating effect on the mutated GABAA receptors but in most cases, mutations affected the potency and efficacy of GABA. Despite the negative results, evaluation of the mutant receptors showed that residues in the area corresponding to the AM-1488 pocket had great effect on GABA potency and efficacy and it cannot be excluded that the site has modulatory properties in wild type receptors if a complementary molecule to fit the pocket can be identified.