Introduction
Over the last few decades, considerable focus has been on negative allosteric modulators (NAMs) (previously referred to as inverse agonists) of the benzodiazepine site of γ -aminobutyric acid receptors (GABAARs) as a potential therapeutic target for cognitive impairment in temporal lobe epilepsy (TLE), Huntington’s disease, Down’s syndrome, schizophrenia and the most common form of dementia, Alzheimer’s disease (AD), which constitutes one of the most significant health problems confronting societies with an aging population.
The ionotropic GABAA R family are heteropentameric structures consisting of a combination of five subunits (Sieghart et al., 2002) with the α−subunit being clinically relevant, as it controls the pharmacological profile of GABAA Rs (McKernan et al., 1996). Since the understanding that distinct pharmacological properties of the GABAAR are reliant on the fact that different brain regions and cell types contain various subunit compositions, NAMs of the GABAAR at the subunit level have been widely studied. In particular, GABAARs containing the α5-subunit have been of interest, given their role in learning and memory as evidenced by various studies (Caraiscos et al., 2004; Collinson et al., 2002; Crestani et al., 2002; Dawson et al., 2006; Ghafari et al., 2017; Yee et al., 2004).
The hippocampus plays a critical role in memory formation and retrieval, and is significantly affected in AD, which is characterised by short-term memory deficits as one of the first symptoms of the disease (Price et al., 2001). The strong evidence to suggest hippocampal preferential distribution of the α5-containing GABAAR sub-type (Quirk et al., 1996), together with its diverse pathology in memory deficit-related disease, and particularly, its preservation in human brains of AD patients (Howell et al., 2000; Rissman et al., 2007), has led many researchers to test several α5 subunit-selective compounds for their potential cognition-enhancing effects (Liu et al., 1996; Quirk et al., 1996; Savic et al., 2008; Sternfeld et al., 2004).
Originally, Merck, Sharp and Dohme, (MSD) developed the first GABAAR NAM, known as α5IA, with high efficacy at the GABAA α5 receptor sub-type without being an anxiogenic agent (Atack et al., 2006). Following the development of this compound by MSD, a number of other nootropic drugs (α5 sub-type selective NAMs) have been developed (e.g. RO4938581; (Ballard et al., 2009)). Many of these studies reported an impressive pharmacological profile of this compounds and their potential as cognitive enhancers without CNS-mediated adverse effects (Ballard et al., 2009; Braudeau et al., 2011; Chambers et al., 2003; Collinson et al., 2006; Dawson et al., 2006; Duchon et al., 2019; Eimerbrink et al., 2019; Martinez-Cue et al., 2014). These studies were initially implemented in rodent models, and unfortunately, these results were not reproducible in human subjects/patients to the same extent. Several key molecules consistently failed clinical trials at different phases including Basmisanil (code, RO5186582, Roche, 2019), a5IA (Atack, 2010) and MRK-016 (Atack et al., 2009). Basmisanil was taken through Phase 1 and Phase 2 of clinical trials for Down’s syndrome and although in the Phase 2, it was shown not to be efficacious in either adults or adolescents. It appears that despite a5IA and MRK-016 demonstrating tolerance in young males, some of these molecules were poorly tolerated in elderly patients with no cognitive improvement (Atack, 2010), thus reducing the viability of α5 as a therapeutic target. Although these molecules were shown to be selective for α5 subunit- containing GABAARs, the lack of efficacy and poor tolerance in human patients could be related to poor brain penetration of the molecules or an age-related effect.
Whether this failure was due to low drug potency / bioavailability or due to a general lack of understanding of the synaptic mechanisms involving α5 receptors during the pathogenesis of the disease is currently unclear. To address these issues, we synthesised a novel water soluble α5 GABAAR selective NAM. These receptor sub-types have been shown to be located in hippocampal extrasynaptic sites, as well as synaptic sites of postsynaptic pyramidal (Ali et al., 2008; Glykys et al., 2008; Serwanski et al., 2006). Although it has been shown that dendrite-targeting interneuron populations elicit α5 GABAAR-mediated inhibition in pyramidal cells (Ali et al., 2008), it is unclear whether the α5 receptor subtype was expressed on inhibitory interneurons themselves. This was of particular interest, as we have shown previously, using theAPPNL-F/NL-F mouse, the first β-amyloid precursor protein (APP ) knock-in mouse AD model that is thought to be able to recapitulate the human condition more accurately (see (Sasaguri et al., 2017), that synaptic excitability is disrupted in various cortical regions, including the CA1 region (Petrache et al., 2019), and that this could be related to the alteration of three key modulatory interneuron populations namely; calretinin- (CR), cholecystokinin- (CCK), and somatostatin- (SST) expressing interneurons (Shi et al., 2019). We investigated whether these key modulatory interneurons located in CA1 stratum radiatum (SR), together with principal pyramidal cells, expressed the α5 subunit-containing GABAARs, in the APPNL-F/NL-Fmodel, age-matched to wild-type control mice, and then characterized the synaptic effects of our newly-developed α5 compound in these 4 sub-types of neurons.
Methods: