Functional pharmacology of cloned heterodimeric GABA_B receptors expressed in mammalian cells

Abstract

1. In this study we report a new assay of heterodimeric γ-amino-butanoic acid subtype B (GABA_B) receptors where either GABA_BR1a or GABA_BR1b are co-expressed with GABA_BR2 and the chimeric G-protein Gαq-z5 in tsA cells. In this manner we obtained a robust response to GABA_B agonists measured as increase in phosphoinositide hydrolysis.

2. We used this assay to characterize a number of commonly used GABA_B receptor ligands. Both splice variants displayed the same rank order of agonist potency; 3-aminopropyl(methyl)phosphinic acid (SKF-97541)>GABA>(R)-4-amino-3-(4-chlorophenyl)butanoic acid ((R)-baclofen)>(RS)-4-amino-3-(5-chloro-2-thienyl)butanoic acid (BCTG)>3-aminopropylphosphonic acid (3-APPA) and furthermore, the absolute agonist potency values were very close to each other.

3. 3-APPA was a partial agonist displaying maximal responses of 41 and 61% compared to GABA at GABA_BR1a and GABA_BR1b, respectively. The antagonist (RS)-3-amino-2-(4-chlorophenyl)-2-hydroxypropylsulphonic acid (2-OH-saclofen) displayed K_B values of 15 and 7.8 μM at GABA_BR1a and GABA_BR1b, respectively.

4. The rank order of agonist potency as well as the absolute ligand potencies correspond very well with those previously reported in different tissues, and this study thus provides a functional assay of cloned GABA_B receptors which should be a valuable tool for further characterization of GABA_B ligands. Finally, we can conclude that the functional pharmacological profiles of the two GABA_BR1 splice variants are very similar.

Introduction

γ-Aminobutanoic acid (GABA) is the major inhibitory amino acid in the central and peripheral nervous system. The physiological effects of GABA are mediated by ionotropic GABA_A receptors and metabotropic GABA_B receptors, the latter being coupled to the Gαi class of G-proteins (Morishita et al., 1990). The GABA_B receptor was first defined by its selective activation by the GABA analogue (R)-4-amino-3-(4-chlorophenyl)butanoic acid ((R)-baclofen) (Bowery et al., 1980; Hill & Bowery, 1981) and more recently a number of selective and very potent antagonists have been developed (Bittiger et al., 1993). Using one of these radiolabelled antagonists, two different splice variants of the GABA_B receptor (named GABA_BR1a and GABA_BR1b, respectively) were isolated by expression cloning (Kaupmann et al., 1997) which led several groups to identify a second subunit (Jones et al., 1998; White et al., 1998; Kaupmann et al., 1998a; Kuner et al., 1999; Ng et al., 1999a; Martin et al., 1999). Interestingly, although some of these groups have reported functional responses from cells transfected with either GABA_BR1 or GABA_BR2 alone (Kaupmann et al., 1997; 1998b; Kuner et al., 1999; Martin et al., 1999), several lines of evidence have shown that the two receptor subunits dimerize in order to form fully functional receptors with the expected pharmacological responses by agonists (Jones et al., 1998; White et al., 1998; Kaupmann et al., 1998a; Kuner et al., 1999; Ng et al., 1999b; Möhler & Fritschy, 1999). However, although transient expression functional assays are now available, the assays reported so far have either been limited by a low throughput or relatively weak responses. Thus, to this point only a couple of ligands have been characterized in functional assays based on heterodimeric cloned GABA_B receptors compared to a much larger number of compounds characterized in binding experiments (Kaupmann et al., 1997; 1998a; Jones et al., 1998; White et al., 1998; Kuner et al., 1999; Ng et al., 1999b). In order to overcome these limitations on functional assays by designing a ‘more sensitive and much-needed' functional assay system (Bettler et al., 1998), we used a strategy of co-transfection of the receptors with the chimeric G-protein named Gαq-z5 in which the five C-terminal aminoacids of Gαq have been replaced by the C-terminus of Gαz (a member of the Gαi subgroup). In this manner, receptors coupled to the Gαi class of G-proteins have previously been shown to mediate robust stimulative responses with the expected pharmacological profiles on e.g., monoamine receptors (Conklin et al., 1993; Bräuner-Osborne & Brann, 1996) and metabotropic glutamate receptors (Gomeza et al., 1996) and as shown in this paper a similar robust signal (generation of inositol phosphates) can be obtained from heterodimeric GABA_B receptors when these are co-transfected with Gαq-z5.

As pointed out previously, two different splice variants of GABA_BR1 have been identified differing in the distal amino terminal part of the amino-terminal of the receptor. The two splice variants show distinct expression patterns and different developmental regulation, indicating that they could have different physiological functions (Kaupmann et al., 1998b; Malitschek et al., 1998). Since it has recently been shown that GABA binds to the amino terminal domain (ATD) of the receptor (Galvez et al., 1999), it is tempting to speculate that the pharmacology of the splice variants could differ. So far, binding studies of the GABA_BR1 splice variants expressed alone (Kaupmann et al., 1997; 1998b) or in the presence of GABA_BR2 (Kaupmann et al., 1998a) have not revealed any such difference in binding affinity between the two splice variants. However, this does not rule out that the splice variants could differ in their pharmacological profiles in functional assays. Thus, in order to investigate this possibility we have also compared the pharmacology of a number of commonly used GABA_B ligands (Figure 1) on the two splice variants of GABA_BR1 co-expressed with GABA_BR2 and Gαq-z5.

Figure 1.

Structure of the GABA_B ligands used in this study.

Methods

Cell culture and second messenger assay

tsA cells (a transformed HEK 293 cell line (Chahine et al., 1994)) were maintained at 37°C in a humidified 5% CO₂ incubator in Dulbecco's modified Eagle medium (DMEM) supplemented with penicillin (100 U ml⁻¹), streptomycin (100 mg  ml⁻¹) and 10% foetal calf serum. Based on previously published protocols we decided to transfect the cells with a 5 fold excess of GABA_BR2 as compared to the GABA_BR1 splice variants (White et al., 1998; Kaupmann et al., 1998a). Two million cells were split into a 15 cm tissue culture plate and transfected with 1.4 μg GABA_BR1a-pcDNA3.1 or GABA_BR1b-pcDNA3.1, 7 μg GABA_BR2-pcDNA3.1 and 1.4 μg Gαq-z5-pcDNA the following day using SuperFect as a DNA carrier according to the protocol by the manufacturer (Qiagen, Hilden, Germany). In experiments where cells were only transfected with a subset of the plasmids, total DNA was maintained at the same level with empty pcDNA3 vector. The day after transfection, cells were split into two poly-D-lysine coated 24-well tissue culture plates in inositol-free growth medium containing 1 μCi ml⁻¹ myo-[2-³H]inositol (Amersham, Buckinghamshire, U.K.). Sixteen to twenty hours later, cells were washed with phosphate buffered saline solution (PBS) and incubated for 20 min in PBS containing 0.9 mM CaCl₂ and 1.05 mM MgCl₂ (PBS⁺⁺). Cells were then incubated for another 20 min in PBS⁺⁺ containing 10 mM LiCl (PBS⁺⁺-LiCl). Finally, after the cells had been incubated in PBS⁺⁺-LiCl for 40 min with the indicated amounts of agonists, reactions were terminated by ice-cold 10 mM formic acid and total inositol phosphate (IP) generation was determined by ion-exchange chromatography as described previously (Nanevics et al., 1996). In the case of antagonist assays, cells were incubated in PBS⁺⁺ for 20 min then in PBS⁺⁺-LiCl containing antagonist for 20 min and finally in PBS⁺⁺-LiCl containing both antagonist and 5 μM GABA for 40 min.

Materials

GABA and (R)-baclofen were obtained from Sigma (St. Louis, MO, U.S.A.). 3-Aminopropyl(methyl)phosphinic acid (SKF-97541), (RS) -4- amino-3 -(5- chloro-2 - thienyl)- butanoic acid (BCTG), 3-aminopropylphosphonic acid (3-APPA) and (RS) - 3 - amino - 2 - (4 - chlorophenyl)-2-hydroxypropylsulphonic acid (2-OH-saclofen) were obtained from Tocris Cookson (Bristol, U.K). Culture media, serum and antibiotics for cell culture were obtained from Life Technologies (Paisley, U.K.). The rat GABA_B receptor plasmids and the Gαq-z5 construct were generous gifts from Dr Janet Clark (National Institute of Health, Bethesda, MD, U.S.A.) and Dr Bruce Conklin (University of California, San Francisco, CA, U.S.A.), respectively. The pcDNA3 vector was obtained from Stratagene (La Jolla, CA, U.S.A.). The tsA cells were a generous gift from Dr Penelope S.V. Jones (University of California, San Diego, CA, U.S.A.).

Data analysis

All experiments were performed in triplicate and the results are given as mean±s.e.mean of 3–5 experiments. Antagonist potencies were calculated from inhibition curves using the ‘functional equivalent' of the Cheng-Prusoff equation K_B=IC₅₀/(1+([A]/EC₅₀)) (Craig, 1993), where [A] is the fixed agonist concentration.

Results

Initially, we tested whether it was necessary to express GABA_BR1, GABA_BR2 and Gαq-z5 simultaneously in order to stimulate IP formation. As expected, cells transfected with GABA_BR1a and GABA_BR2 were unable to stimulate IP formation when exposed to the agonist (R)-baclofen (Figure 2). In contrary, cells further co-transfected with Gαq-z5 gave a robust response to the agonist as a 4–6 fold increase in IP formation. As seen in Figure 2, the monomeric GABA_BR1a or GABA_BR2 co-expressed with Gαq-z5 were also unable to respond to the agonist. Thus, in conclusion GABA_BR1, GABA_BR2 and Gαq-z5 all needed to be expressed in order to obtain a functional IP response. Cells transfected with Gαq-z5 alone did also not show any increase in IP formation to (R)-baclofen or any of the other ligands used in this study (data not shown).

Figure 2.

Generation of inositol phosphates by the GABA_B agonist (R)-baclofen in tsA cells transfected with varying subsets of the GABA_B receptor subunits and the chimeric G-protein Gαq-z5. After a 20 min preincubation in phosphate-buffered saline containing 0.9 mM CaCl₂ and 1.05 mM MgCl₂ (PBS⁺⁺) cells were incubated for further 20 min in the same buffer containing 10 mM LiCl (PBS⁺⁺-LiCl). Finally, after the cells had been incubated in PBS⁺⁺-LiCl for 40 min in the absence or presence of 100 μM (R)-baclofen, reactions were terminated by ice-cold 10 mM formic acid and total inositol phosphate generation was determined by ion-exchange chromatography. Data are the means±s.d.mean of representative experiments performed in triplicate.

We then went on to test a series of commonly used GABA_B receptor ligands (shown in Figure 1) in greater detail on tsA cells transiently transfected with either GABA_BR1a, GABA_BR2 and Gαq-z5 or GABA_BR1b, GABA_BR2 and Gαq-z5. As can be seen on the examples of concentration-response curves shown in Figure 3 and the derived pEC₅₀ and pK_B values shown in Table 1, the pharmacological profiles of the two GABA_BR1 splice variants were almost identical. They both showed a similar rank order of agonist potency; SKF-97541>GABA>(R)-baclofen>BCTG>3-APPA. Interestingly, whereas SKF-97541, (R)-baclofen and BCTG were full agonists, 3-APPA was a partial agonist producing responses of 41 and 61% of a full GABA response on the GABA_BR1a and GABA_BR1b splice variants, respectively (Figure 3 and Table 1).

Figure 3.

Concentration-response curves of some GABA_B receptor agonists at tsA cells transfected with GABA_BR1a, GABA_BR2 and Gαq-z5 (a) or GABA_BR1b, GABA_BR2 and Gαq-z5 (b). Cell treatment and measurement of total inositol phosphates generated was performed as described in Figure 2.

Table 1.

Agonist and antagonist potencies on heterodimeric GABA_BR1 / GABA_BR2 receptors in functional assay

Discussion

The first goal of these studies was to develop a functional assay of cloned GABA_B receptors which could circumvent some of the limitations (e.g. low throughput and low signal) of previously reported assays (Kaupmann et al., 1997; 1998a; Jones et al., 1998; White et al., 1998; Kuner et al., 1999; Ng et al., 1999b; Martin et al., 1999) by using a strategy of co-transfection with the chimeric G-protein Gαq-z5. We and others have previously used this strategy to obtain robust stimulative responses from receptors coupled to the Gαi class of G-proteins on monoamine receptors (Conklin et al., 1993; Bräuner-Osborne & Brann, 1996) and metabotropic glutamate receptors (Gomeza et al., 1996), and in agreement with these results we obtained 4–6 fold increases in IP generation by agonist treatment of cells transfected with both subunits of the GABA_B receptor and the chimeric G-protein (Figures 2 and 3). In agreement with several groups, we were only able to detect agonist responses in cells transfected with both subunits of the GABA_B receptor (Jones et al., 1998; White et al., 1998; Kaupmann et al., 1998a; Ng et al., 1999b) whereas no response could be detected in cells transfected with either GABA_BR1/Gαq-z5 or GABA_BR2/Gαq-z5. Some groups have reported that weak functional responses can be obtained from cells transfected with either GABA_BR1 or GABA_BR2 alone (Kaupmann et al., 1997; 1998b; Martin et al., 1999; Kuner et al., 1999), but most results point to the fact that both GABA_B receptor subunits have to be co-expressed in order to obtain the expected agonist potency and effector coupling (Jones et al., 1998; White et al., 1998; Kaupmann et al., 1998a; Kuner et al., 1999; Ng et al., 1999b; Möhler & Fritschy, 1999). Our results are in agreement with these observations as correct coupling to Gαq-z5 only occur when both GABA_B subunits are present.

In order to determine whether the GABA_B receptors were able to activate Gαq-z5 in the same manner as the endogenous G-proteins, we tested a number of commonly used GABA_B ligands in greater detail. Furthermore, we also wanted to test whether the two GABA_BR1 splice variants differed in their functional pharmacological profile. First of all we found that both splice variants displayed very similar pharmacology when co-expressed with GABA_BR2 and Gαq-z5. This is in agreement with binding studies on cells expressing either GABA_BR1a or GABA_BR1b alone or in the presence of GABA_BR2 which were also unable to detect any pharmacological difference between the two splice variants (Kaupmann et al., 1997; 1998a,1998b; Malitschek et al., 1998). Thus, the function of the splice variation in the distal part of the ATD is still unknown. As pointed out recently, it is also possible that the splice variants interact with different membrane anchoring proteins causing distinct localization at the subcellular level (Möhler & Fritschy, 1999), but this remains to be proven by a different set of experiments than those presented in this study. The rank order of agonist potency was thus similar on both splice-variants; SKF-97541>GABA>(R)-baclofen>BCTG >3-APPA. Although, no previous study has looked at all these agonists simultaneously in functional experiments it can still be concluded that the rank order is in agreement with studies using tissue pharmacology or receptor binding. Thus, SKF-97541 has been reported to be more potent than baclofen (Knight & Bowery, 1996; Seabrook et al., 1990), GABA has been reported to be more potent than baclofen (Berthelot et al., 1991) which is more potent than BCTG (Berthelot et al., 1991; Lacey et al., 1993; Ong et al., 1997), and 3-APPA has been shown to be less potent than baclofen (Luzzi et al., 1986; Kerr et al., 1987). The actual pEC₅₀ values of our study are also in agreement with those reported in tissues previously. Furthermore, 3-APPA has been shown to be a partial agonist in guinea-pig ileum where it both displayed agonist and antagonist actions (Luzzi et al., 1986; Kerr et al., 1987), which is in agreement with our findings of 3-APPA being a partial agonist. It should be noted that 3-APPA is a full agonist in the cat spinal cord (Kerr et al., 1987). We have previously shown that the intrinsic activity of partial agonists is highly dependent on the expression levels of receptors and G-proteins (Bräuner-Osborne et al., 1996) and this could explain why 3-APPA is a partial and full agonist in the ileum and spinal cord, respectively. We also tested the classical GABA_B competitive antagonist 2-OH-saclofen. As shown in Table 1, this compound could also not discriminate between the GABA_BR1a (pK_B=4.9) and GABA_BR1b (p_KB=5.1) splice variants. However, the potency of 2-OH-saclofen found in our assay is very similar to those found in tissue preparations such as the guinea-pig ileum (pA2=5.0) (Kerr et al., 1988).

In conclusion, we have reported a new method for functional testing of cloned heterodimeric GABA_B receptors expressed in mammalian cells. The results using the method are in very nice agreement with previously published results from tissue preparations and since the assay has a fairly high throughput and a robust signal, the assay seems to be superior to previously reported functional assays. We can also conclude that at least in our functional assay, we were unable to detect any significant difference between the two GABA_BR1 splice variants. This is in agreement with previous reports using binding assays (Kaupmann et al., 1997; 1998a,1998b; Malitschek et al., 1998) and the physiological significance of the splice variation in the amino-terminal thus remains a puzzle which warrants further experiments to solve.

Abbreviations

2-OH-saclofen

(RS)-3-amino-2-(4-chlorophenyl)-2-hydroxypropylsulphonic acid

3-APPA

3-aminopropylphosphonic acid

ATD

amino terminal domain

(R)-baclofen

(R)-4-amino-3-(4-chlorophenyl)butanoic acid

BCTG

(RS)-4-amino-3-(5-chloro-2-thienyl)butanoic acid

DMEM

Dulbecco's modified Eagle medium

IP

inositol phosphate

GABA

γ-amino-butanoic acid

PBS

phosphate buffered saline solution

SKF-97541

3-aminopropyl(methyl)phosphinic acid