New GABA amides activating GABA_A-receptors

Introduction

γ-Aminobutyric acid (GABA) is the major inhibitory amino acid transmitter of the central nervous system (CNS) of vertebrates (Figure 1).

It plays an important role in a variety of physiological functions, including motion control, vision, relaxation, sleep and many other brain functions [1-3]. In vertebrates GABA activates specific receptors of several classes: GABA_A, GABA_B and GABA_C [4]. GABA_A-receptors are transmembrane heterooligomeric proteins, forming chloride (Cl⁻) selective channels, and are composed of five subunits: two α, two β and one γ subunit [5]. Being highly expressed in the peripheral and central nervous system, GABA_A-receptors represent a key therapeutic target for benzodiazepines, barbiturates, neurosteroids and general anesthetics [6-8]. Therefore, in spite of a large variety of existing agonists, antagonists and modulators of GABA-receptors, there is a high interest in the development of new drugs that can interact with these targets. Many compounds of different substance classes are known to modulate the activity of GABA_A-receptors [9-12]. One substance group that has been less explored is GABA-amides. Compounds 4a, 4b, 4c, 7a (as triflate salt) and 7c were previously reported to potently activate the GABA_A-channels, being "partial, full or superagonists" [13]. However, in their work the authors studied the activity of GABA-amides on GABA_A-receptors using chloride-flux assays on synaptoneurosomes. Such preparations can contain many damaged cells leading to a highly variable intracellular Cl⁻ distribution in different cells. In our study we used the patch-clamp technique, which is much more reliable and informative in comparison to the approach based on chloride-flux assays. Like Carlier et al. [13], we noticed that in the course of the GABA-amide synthesis, GABA impurities are generated in the deprotection step. However, in contrast to Carlier et al. [13], who used a modification of Sallers’s procedure [14] (detection limit <0.1 wt %) to check the purity of the compounds, we developed an improved purification procedure and used a more sensitive HPLC–MS analysis to ensure that the GABA-amide products do not contain any detectable amount of GABA (detection limit <0.002 wt %). Our observations may help to give a more complete picture of the ability of GABA-amides to activate GABA_A receptors.

Results and Discussion

Boc-Pyrrolidone 1 was nucleophilic-ring-opened by diamines 2a–c or the mono Boc-protected diamines 5a–f in THF (Scheme 1, See Supporting Information File 1 for full experimental data).

To cleave the Boc-protecting groups, the compounds were dissolved in dilute HCl (5%, EtOH/H₂O). It was observed that the use of more concentrated HCl can cause cleavage of the amide bonds and the release of free GABA. To remove any impurities of GABA, which may lead to false test results, all substances were carefully purified: In a first purification step the GABA-amide hydrochlorides were dissolved in MeOH and precipitated by slow addition of Et₂O. The precipitates were centrifuged and recrystallized several times from MeOH until no traces of GABA could be detected by HPLC–MS analysis. To determine the sensitivity of the analysis method three stock solutions of GABA were prepared (10⁻³, 10⁻⁵ and 10⁻⁷ mol/L) and 1 µL was injected into the HPLC-coupled mass spectrometer. On the basis of the obtained mass spectra, a GABA detection limit of 0.14 pmol (14 pg) was determined.

By examining the Cl⁻ uptake elicited by different GABA_A-receptor agonists, Carlier et al. demonstrated that compounds with general structure 4a–c were capable of stimulating Cl⁻ uptake with different efficacy [13]. Moreover, the authors described compound 4b as "superagonist", because it induced a maximum uptake about 50% higher than that achieved by GABA. Surprisingly, the EC₅₀ value for this compound was more than 30-fold higher than for GABA, i.e., at 733 µM and 14.3 µM, respectively. The investigations in this study were performed by using a standard ³⁶Cl⁻-flux assay in mouse brain synaptoneurosomes. This technique does not allow a comparative analysis on one cell. Moreover, ³⁶Cl⁻-flux assays were carried out by using a 15 s incubation time. As GABA-receptors usually exhibit strong desensitization at long exposures to agonists, we decided to re-examine these results using patch-clamp recordings and fast perfusion technique for the application of the tested compounds. The activity of synthesized compounds was tested by using CHO cells transiently expressing GABA_A-receptors in the configuration (α₁-GFP + β₂ + γ_2Long). To analyze the functional properties of the compounds, we performed monitoring of ionic currents using whole-cell patch-clamp techniques. First, concentration-response curves for GABA were obtained and its EC₅₀ was determined. Then we applied different concentrations of the studied compounds and estimated the minimal concentration that induced currents and, if possible, their EC₅₀. The EC₅₀ for GABA varied in different cells from 4 µM to 15 µM with a mean of 9.5 ± 0.3 µM (x [number of tested cells] = 10) (Figure 2, see Supporting Information File 1 for full experimental data).

Surprisingly, in contrast to previously described observations, our purified compounds 4a–c were not able to activate GABA_A-receptors in concentrations up to 10 mM. Figure 3 illustrates this for the compound 4c. Similar results were obtained also for compounds 4a and 4b: on application of 10 mM, the changes in the current were 0 pA, (x = 4 for each compound).

In contrast, all studied compounds from series 7a–e were capable of inducing ionic currents with different efficacy (Figure 4).

Thus, compound 7a (n = 2) activated currents at 2.6 mM with an amplitude of 10–30 pA (x = 4). The compound 7b (n = 3) activated GABA-receptors more strongly and at concentrations of 10 mM induced currents comparable to those for GABA 30–100 µM (Figure 2). The EC₅₀ for compound 7b is 1080 ± 140 µM (x = 4). The efficacy of the compounds is significantly weaker than for GABA. The number of -CH₂-units between the amide nitrogen atom and the ammonium moiety of compounds 7 affects the efficacy significantly, reaching a maximum with compound 7c (n = 4). The EC₅₀ for compound 7c is 1750 ± 330 µM (x = 7). Compound 7d (n = 6) also effectively activated GABA_A-receptors with an activation threshold of about 100 µM. At concentrations of 10 mM it caused currents similar of those induced by saturated GABA concentrations (30–100 µM) with an EC₅₀ of about 5 mM (Figure 5A–C). The EC₅₀ for compound 7d is 5250 ± 560 µM (x = 6). Kinetics of desensitization and the current–voltage dependencies (Figure 5D) of the studied compounds are similar to those for GABA. Compound 7e (n = 8) weakly activated currents at 5 mM with an amplitude of 10–20 pA (x = 3).

Conclusion

Our data suggest that compounds 4a–c are not capable of activating GABA_A-receptors. Compounds 7a–e are able to stimulate these receptors and show a distinct structure–activity correlation. The compounds may become useful as molecular tools for the functional analysis of GABA_A-receptors.