A pharmacological assessment of agonists and modulators at α4β2γ2 and α4β2δ GABA_A receptors: The challenge in comparing apples with oranges

Abstract

Extrasynaptically located γ-aminobutyric acid (GABA) receptors type A are often characterized by the presence of a δ subunit in the receptor complex. δ-Containing receptors respond to low ambient concentrations of GABA, or respond to spillover of GABA from the synapse, and give rise to tonic inhibitory currents. In certain brain regions, e.g. thalamocortical neurons, tonic inhibition is estimated to represent the majority of total GABA-mediated inhibition, which has raised substantial interest in extrasynaptic receptors as potential drug targets. Thalamocortical neurons typically express α4β2/3δ receptors, however, these have proven difficult to study in recombinant in vitro expression systems due to the inherently low current levels elicited in response to GABA.

In this study, we sought to characterize a range of agonists and positive allosteric modulators at α4β2δ and α4β2γ2 receptors. All tested agonists (GABA, THIP, muscimol, and taurine) displayed between 8 and 22 fold increase in potency at the α4β2δ receptor. In contrast, modulatory potencies of steroids (allopregnanolone, THDOC and alfaxalone), anesthetics (etomidate, pentobarbital) and Delta-Selective agents 1 and 2 (DS1 and DS2) were similar at α4β2δ and α4β2γ2 receptors. When evaluating modulatory efficacies, the neurosteroids and anesthetics displayed highest efficacy at α4β2γ2 receptors whereas DS1 and in particular DS2 had highest efficacy at α4β2δ receptors. Overall, several key messages emerged: (i) none of the tested compounds displayed significant selectivity and a great need for identifying new δ-selective compounds remains; (ii) α4β2δ and α4β2γ2 receptors have such divergent intrinsic activation properties that valid comparisons of modulator efficacies are at best challenging.

Introduction

γ-Aminobutyric acid (GABA), the principal inhibitory neurotransmitter in the central nervous system, mediates some of its actions via GABA_A receptors (GABA_ARs). The majority of GABA_ARs are located synaptically and predominantly contain the γ-subunit. Activation of the synaptic GABA_ARs by high concentrations of GABA mediates phasic inhibitory currents that ceases within milliseconds as GABA is cleared from the synaptic cleft. In contrast, δ-containing GABA_ARs are located peri- and extrasynaptically and are highly sensitive to GABA [1], [2]. Activation of these extrasynaptic receptors depends on exposure to ambient or low GABA concentrations, which leads to a persistent or tonic inhibitory current. The contribution of extrasynaptic GABA_AR-mediated inhibition to neurotransmission is substantial and in thalamocortical neurons [3] it is estimated to be more than 90% of the total GABA-mediated inhibition. Further, it was recently shown that δ-containing GABA_ARs in thalamus dynamically modulate neuronal inhibition by ⿿sensing⿿ presynaptic GABA activity levels, which indicate an intricate role of these receptors beyond slow-responding long-term tonic inhibition [4].

GABA_ARs are activated by a variety of agonists and likewise modulated by a variety of positive and negative allosteric modulators (PAMs and NAMs). However, there are distinct differences between the pharmacology of γ- and δ-containing GABA_ARs. For example, γ-containing receptors are modulated by benzodiazepines, δ-containing receptors are not. δ-containing receptors are efficiently modulated by the Delta Selective agents 1 and 2 (DS1 and DS2) with only moderate effects reported at γ-containing receptors [5], [6], [7]. THIP (gaboxadol) is a super-agonist at δ-containing receptors, eliciting a greater maximum response in comparison to GABA. Further, PAMs such as neurosteroids, considered to be low-efficacy modulators of γ-containing receptors [8], [9], have significant efficacies at δ-containing receptors. As δ-containing receptors exhibit relatively low level of receptor desensitization at high GABA concentrations [1], [10] and low maximum open-state probability in response to GABA, they may result in greater relative activation by agonists and PAMs compared to γ2-containing receptors [11], [12].

Within the field of pharmacology, the term selectivity is used to describe the preference of a compound for a specific target, but it is rarely defined exactly what the term denotes or in what context it is used. Obviously, a compound would be selective for e.g. δ- vs. γ2-containing GABA_AR if binding only occurs at δ-containing receptors. However, due to the high similarity between receptors of the Cys-loop receptor class, this is rarely the case and instead selectivity often become a question of a degree of differentiation. The magnitude of this differentiation to claim selectivity further depends on the specific needs. A compound with a 10-fold difference in potency might be adequately selective for research use but not for clinical development where a > 100-fold difference would be desired. Besides binding to receptors with different potencies, a compound can also display another form of selectivity based on its efficacy. In other words, a compound can be functionally selective. Using this measure, modulators have previously been described as selective despite no binding affinity difference between GABA_ARs [18], [33]. Unfortunately, this only adds to the difficulty in defining the term selectivity as a compound then can have actions at two targets simultaneously but with different efficacy levels.

α4β2δ and α4β3δ receptors are known to be mediators of tonic current in a range of cells, including thalamic relay neurons, striatal medium spiny neurons, hippocampal dentate gyrus granule cells and neocortical pyramidal cells [13]. Studies using β2 knockout and etomidate insensitive β2^N265S knock-in mice have identified α4β2δ receptors as contributing significantly to the tonic current in ventrobalis complex neurons [14], [15] and dentate gyrus granule cells and is critically involved in the regulation of sleep and wakefulness [16]. To date, the majority of reported studies characterizing extrasynaptic α4β(1-3)δ GABA_ARs have focused on β3-containing receptors despite the potential physiological importance of the α4β2δ receptor in its own right. Comparisons between α4β3δ and α4β3γ2 suggested that neurosteroids and some anesthetics modulate α4β3δ receptors with greater efficacy than α4β3γ2 [17]. However, similar comparisons of modulators or even agonists have not been performed at α4β2δ vs. α4β2γ2 receptors.

In the present study, we characterized the effects of a series of agonists (GABA, THIP, muscimol, and taurine) and modulators (allopregnanolone, alphaxalone, THDOC, etomidate, pentobarbital, DS1, DS2 and Ro15-4513) at recombinant human α4β2δ and α4β2γ2 GABA_ARs expressed in Xenopus laevis oocytes. Several findings emerged from the pharmacological data that are particular noteworthy. First, incorporation of a δ vs. γ2 subunit in the receptor complex substantially affects the intrinsic receptor properties. While this causes all agonists to display increased potency at δ-containing receptors, the derived effects complicate testing of modulators. For valid inter-receptor comparison of modulators, it is necessary to ensure equivalent testing conditions. However, δ vs. γ2 subunit incorporation causes a ⿼7-fold decrease in the gating efficiency of GABA, which means that the standard methodology of measuring modulators at an equally efficacious GABA concentration is not appropriate. Hence, a methodology using equal open-probabilities of GABA was employed instead. Second, the different intrinsic receptor properties have resulted in descriptions of steroids and DS2 selectivity for δ-containing receptors that needs to be taken with caution. Based on the studies presented here, steroids and DS2 cannot reasonably be termed selective for δ-containing receptors in the strict pharmacological sense. Third, extrapolating in vitro results into predictions for currents in a neuronal context is highly complex. Here, the intrinsic difference between γ- and δ-containing receptors combined with their different physiological role as mediators of phasic or tonic current, respectively, means that even minor preferences for δ-containing receptors can lead to usefulness of compounds that are not selective per se. As a consequence of the above, comparing the effects of modulators at α4β2δ vs. α4β2γ2 receptors can be like comparing apples with oranges.