Novel receptor
Kind Code:

The present invention relates to the novel GABAB receptor subtypes GABAB-R1c and GABAB-R2 as well as to a novel, functional GABAB receptor which comprises a heterodimer of GABAB-R1 and GABAB-R2 receptor subunits. The present invention also relates to variants of the receptors, nucleotide sequences encoding the receptors and variants thereof and novel vectors, stable cell lines, antibodies, screening methods, methods of treatment and methods of receptor production.

Barnes, Ashley Antony (Herts, GB)
Wise, Alan (Bedfordshire, GB)
Marshall, Fiona Hamilton (Hertfordshire, GB)
Fraser, Neil James (Herts, GB)
White, Julia Helen Margaret (Herts, GB)
Foord, Steven Michael (Buckinghamshire, GB)
Application Number:
Publication Date:
Filing Date:
Primary Class:
Other Classes:
435/320.1, 435/369, 514/17.4, 514/17.8, 514/18.1, 514/18.3, 514/20.6, 530/350, 530/388.22, 536/23.5, 435/69.1
International Classes:
G01N33/53; A61K38/17; C07H21/04; C07K14/705; C07K16/28; C12P21/06; A61K38/00
View Patent Images:

Primary Examiner:
Attorney, Agent or Firm:
1. 1-37. (canceled)

38. A method for identification of a compound which exhibits GABAb receptor modulating activity, comprising contacting the GABAb receptor with a test compound and detecting modulating activity or inactivity, wherein said GABAb receptor comprises a heterodimer between a GABAb-R1 receptor protein and a GABAb-R2 receptor protein wherein the GABAb-R1 comprises the no acid sequence for GABAb-R1a, GABAb-R1b or GABAb-R1c as set forth in FIG. 2 and wherein the GABAb-R2 receptor comprises the amino acid sequence set forth in FIG. 1b.

39. 39-46. (canceled)



This US patent application claims priority to GB9819420.2 filed on Sep. 7, 1998 in the United Kingdom and to U.S. provisional application 60/103,670 filed on Oct. 9, 1998 in the United States Patent Office.


The present invention relates to the novel GABAB receptor subtypes GABAB-R1c and GABAB-R2 as well as to a novel, functional GABAB receptor which comprises a heterodimer of GABAB-R1 and GABAB-R2 receptor subunits. The present invention also relates to variants of the receptors, nucleotide sequences encoding the receptors and variants thereof and novel vectors, stable cell lines, antibodies, screening methods, methods of treatment and methods of receptor production.


GABA (γ-amino-butyric acid) is the main inhibitory neurotransmitter in the central nervous system (CNS) activating two distinct families of receptors; the ionotropic GABAA and GABAC receptors for fast synaptic transmissions, and the metabotropic GABAB receptors governing a slower synaptic transmission. GABAB receptors are members of the superfamily of 7-transmembrane G protein-coupled receptors. Activation results in signal transduction through a variety of pathways mediated principally via members of the Gi/Go family of pertussis toxin-sensitive G proteins. GABAB receptors have been shown to inhibit N, P/Q and T-type Ca2+ channels in a pertussis toxin-sensitive manner (Kobrinsky et al., 1993; Menon-Johansson et al., 1993; Harayama et al., 1998) and indeed there is also some evidence for direct interactions between GABA, receptors and Ca2+ channels since Ca2+ channel ligands can modify the binding of GABAB agonists (Ohmori et al., 1990). GABAB receptor-mediated Ca2+ channel inhibition is the principle mechanism for presynaptic inhibition of neurotransmitter release. Post-synaptically the major effect of GABAB receptor activation is to open potassium channels, to generate post-synaptic inhibitory potentials.

Autoradiographic studies show that GABAB receptors are abundant and heterogeneously distributed throughout the CNS, with particularly high levels in the molecular layer of the cerebellum, interpeduncular nucleus, frontal cortex, olfactory nuclei and thalamic nuclei. GABA % receptors are also widespread in the globus pallidus, temporal cortex, raphe magnus and spinal cord (Bowery et al., 1987). GABAB receptors are an important therapeutic target in the CNS for conditions such as spasticity, epilepsy, Alzheimer's disease, pain, affective disorders and feeding. GABAB receptors are also present in the peripheral nervous system, both on sensory nerves and on parasympathetic nerves. Their ability to modulate these nerves gives them potential as targets in disorders of the lung, GI tract and bladder (Kerr and Ong, 1995; 1996; Malcangio and Bowery, 1995).

Despite the widespread abundance of GABAB receptors, considerable evidence from neurochemical, electrophysiological and behavioural studies suggests that multiple subtypes of GABA, receptors exist. This heterogeneity of GABAB receptors may allow the development of selective ligands, able to target specific aspects of GABAB receptor function. This would lead to the development of drugs with improved selectivity profiles relative to current compounds (such as baclofen) which are relatively non-selective and show a variety of undesirable behavioural actions such as sedation and respiratory depression. Multiple receptor subtypes are best classified by the differing profiles of agonist and antagonist ligands.

To date screening for GABAB ligands and subsequent structure/activity determinations has relied on radioligand binding assays to rat brain membranes. Further analysis of such ligands in animal models has indicated differences in their behavioural profile. However, due to the absence of cloned GABAB receptors the molecular basis for such differences has not been defined, and therefore it has not been possible to optimise GABAB ligands for therapeutic use.

GABAB receptors were first described nearly 20 years ago (Hill and Bowery, 1981), but despite extensive efforts using conventional expression cloning strategies, for example in Xenopus oocytes, or cloning based on sequence homology, the molecular nature of the GABAB receptor remained elusive. The development of a high affinity antagonist for the receptor finally allowed Kaupmann et al., (1997) to expression clone the receptor from a rat cerebral cortex cDNA using a radioligand binding assay. Two splice variants of the receptor were identified, GABAB-R1a encoding a 960 amino acid protein and GABAB-R1b, encoding an 844 amino acid protein, differing only in the lengths of their N-termini. These two splice variants have distinct spatial distributions within the brain, but both reside within neuronal rather than glial cells. Pharmacologically, the two splice variants are similar, showing binding affinities for a range of antagonists, but about 10 fold lower than those of native receptors, as well as agonist displacement constants which are about 100-150 fold lower than those of native receptors. These observations have led to speculation that the cloned receptor was a low affinity receptor and an additional high affinity, pharmacologically distinct GABAB receptor subtype could exist in the brain. Alternatively, it was argued that G-protein coupling was inefficient or the receptor was desensitising in the recombinant systems used.

A number of groups working in the area have, however, found that the cloned receptor fails to behave as a functional GABAB receptor either in mammalian cells or in Xenopus oocytes. The present invention describes the cloning of a novel human GABAB receptor subtype, GABAB-R2, the identification of a novel splice variant GABAB-R1c, and the surprising observation that GABAB-R1 and GABAB-R2 strongly interact via their C-termini to form heterodimers. Co-expression of GABAB-R1 and GABAB-R2 allows trafficking of GABAB-R1 to the cell surface and results in a high affinity functional GABAB receptor in both mammalian cells and Xenopus oocytes.

These surpising findings provide a unique opportunity to define GABAB subtypes at the molecular level, which in turn will lead to the identification of novel subtype-specific drugs.


According to one embodiment of the present invention there is provided an isolated GABAB-R2 receptor protein or a variant thereof.

According to another embodiment of the invention there is provided an isolated GABAB-R2 receptor protein having amino acid sequence provided in FIG. 1B, or a variant thereof.

According to a further embodiment of the invention there is provided a nucleotide sequence encoding a GABAB-R2 receptor or a variant thereof, or a nucleotide sequence which is complementary thereto.

According to a further embodiment of the invention there is provided a nucleotide sequence encoding a GABAB-R2 receptor, as shown in FIG. 1A, or a variant thereof, or a nucleotide sequence which is complementary thereto.

According to a further embodiment of the invention there is provided an expression vector comprising a nucleotide sequence as referred to above which is capable of expressing a GABAB-R2 receptor protein or a variant thereof.

According to a still further embodiment of the invention there is provided a stable cell line comprising a vector as referred to above.

According to another embodiment of the invention there is provided an antibody specific for a GABAB-R2 receptor protein or a variant thereof.

According to another embodiment of the invention there is provided an isolated GABAB-R1c receptor protein or a variant thereof.

According to another embodiment of the invention there is provided an isolated GABAB-R1c receptor protein having amino acid sequence provided in FIG. 2, or a variant thereof.

According to another embodiment of the invention there is provided a nucleotide sequence encoding a GABAB-R1c receptor protein or a variant thereof, or a nucleotide sequence which is complementary thereto.

According to another embodiment of the invention there is provided an expression vector comprising a nucleotide sequence as referred to above, which is capable of expressing a GABAB-R1c receptor protein or a variant thereof.

According to another embodiment of the invention there is provided a stable cell line comprising a vector as referred to above.

According to a further embodiment of the invention there is provided an antibody specific for a GABAB-R1c receptor protein or a variant thereof.

According to a further embodiment of the invention there is provided a GABAB receptor comprising an heterodimer between a GABAB-R1 receptor protein or a variant thereof and a GABAB-R2 receptor protein or a variant thereof.

According to a further embodiment of the invention there is provided an expression vector comprising a nucleotide sequence encoding for a GABAB-R1 receptor or a variant thereof and a nucleotide sequence encoding for a GABAB-R2-receptor or variant thereof, said vector being capable of expressing both GABAB-R1 and GABAB-R2 receptor proteins or variants thereof.

According to a further embodiment of the invention there is provided a stable cell line comprising a vector as referred to above.

According to a further embodiment of the invention there is provided a stable cell line modified to express both GABAB-R1 and GABAB-R2 receptor proteins or variants thereof.

According to a further embodiment of the invention there is provided a GABAB receptor produced by a stable cell line as referred to above.

According to a further embodiment of the invention there is provided an antibody specific for a GABAB receptor as referred to above.

According to a further embodiment of the invention there is provided a method for identification of a compound which exhibits GABAB receptor modulating activity, comprising contacting a GABAB receptor as referred to above with a test compound and detecting modulating activity or inactivity.

According to a further embodiment of the invention there is provided a compound which modulates GABAB receptor activity, identifiable by a method as referred to above.

According to a further embodiment of the invention there is provided a method of treatment or prophylaxis of a disorder which is responsive to modulation of GABAB receptor activity in a mammal, which comprises administering to said mammal an effective amount of a compound identifiable by the method referred to above.


FIG. 1A (2 Pages) and 1B. Nucleotide and Protein Sequences of Human GABAB-R2

Nucleotide sequence (a) and the translated protein sequence (b) for Human GABAB-R2 are shown.

FIG. 2 (3 Pages). Protein Alignments Between GABAB-R1a, GABAB-R1b, GABAB-R1c splice variants and GABAB-R2.

Amino-acid sequences of the human GABAB-R1a, GABAB-R1b and GABAB-R2 receptors aligned for comparison. Signal sequences and predicted cleavage point (custom character), together with the N-terminal splice points for GABAB-R1a and GABAB-R1b are shown. GABAB-R1c sequence is exactly that of GABAB-R1a, except for the deletion of 63 amino acids (open box). Amino acids conserved between GABAB-R1a and GABAB-R1b are in bold type and potential N-glycosylation sites (*) are shown. Lines beneath the text show positions of the seven predicted TM domains and regions encoding coiled coil structure are indicated by shading. The C-terminal region of GABAB-R1 used as the bait in the yeast two hybrid analysis is marked as ‘BAIT→’, and GABAB-R2 C-terminal domains recovered from the library screen against GABAB-R1 C-terminus are shown as ‘YTH HITS→’.

FIG. 3. Hydrophobicity Profile of GABAB-R2.

Hydrophobicity profiles of GABAB-R2 sequence were determined using the Kyte-Doolittle algorithm, whereby positive values indicate hydrophobic regions. The predicted signal sequence and seven trans-membrane domains are shown.

FIGS. 4A and 4B. Tissue Distribution Studies for Human GABAB-R1 and GABAB-R2.

A Human RNA Master Blot (Clontech), containing normalised polyA+ mRNA from multiple tissues of adult and fetal origin, were probed sequentially with a pan specific probe for GABA-R1 (all splice variants) followed by a GABAB-R2 specific probe. Resulting autoradiographic analysis of the blots are shown, together with a grid identifying tissue type. Specificity controls include yeast RNA and E. coli DNA.

FIG. 5. Heterodimerisation and Homodimerisation Between the C-Terminal Domains of the GABAB-R1 and GABAB-R2 Receptors in the Yeast Two Hybrid System.

β-galactosidase activity was measured in yeast Y190 cells expressing the GABAB-R1 or the GABA, —R2 C-termini, either against empty vector or against each other in all combinations, using ONPG. Of each pair of proteins expressed in the two hybrid system, the first always refers to the GAL4BD fusion construct whilst the second refers to the GAL4AD fusion construct. O-galactosidase activity is determined relative to cell numbers and is in arbitary units.

FIG. 6. Co-Immunoprecipitation Studies of the GABAB Heterodimer in HEK239 Cells.

HEK293T cells were transfected with 1 μg each of either Myc-GABAB-R1b or HA-GABAB-R2 alone or in combination. Cells were harvested 48 h after transfection, lysed and epitope tagged receptors immunoprecipitated using 12CA5 (HA) or 9E10 (Myc) antisera as described in Methods. Immune complexes were then subjected to SDS-PAGE, transferred to nitrocellulose, and captured Myc-GABAB-R1b and HA-GABAB-R2 identified by immunoblotting with Myc and HA, respectively. Lanes 1 and 4, immunoprecipitates of cells transfected with Myc-GABAB-R1b only; lanes 2 and 5, HA-GABAB-R2 only; lanes 3 and 6, immunoprecipitates of cells transfected with Myc-GABAB-R1b together with HA-GABAB-R2. Lanes 1-3, lysates immunoprecipitated with 9E10 (Myc) and blotted to 12CA5(HA); lanes 4-6, lysates immunoprecipitated with 12CA5(HA) and blotted with 9E10 (Myc)

FIG. 7. Cell Surface Localisation of GABAB-R1 Receptor is Dependent Upon Coexpression with GABAB-R2.

Flow cytometry was performed on HEK293T cells transfected with 1 μg of either Myc-GABAB-R1b or HA-GABAB-R2 or both receptors in combination. (A) Analysis using 9E10 (c-Myc) as primary antibody to detect Myc-GABAB-R1b; intact cells. (B) Analysis using 9E10 (c-Myc) as primary antibody to detect Myc-GABAB-R1b; permeabilised cells. (C) Analysis using 12CA5 (HA) as primary antibody to detect HA-GABAB-R2; intact cells. Mock transfected cells, reflecting background fluorescence, are shaded and the marker indicates fluorescence measured over background levels. Myc-GABAB-R1b data is shown as a grey line whereas co-expression of Myc-GABAB-R1b with HA-GABAB-R2 is shown in black. 30,000 cells were analysed in each sample. Histograms shown are from a single experiment. Quoted statistics are from mean of three separate transfections and analysis.

FIG. 8. Coexpression of GABAB-R1a and 1b splice Variants with GABAB-R2 Receptors in HEK293T Cells Results in Terminal Glycosylation of Both GABAB-R1a and GABAB-R1b.

P2 membrane fractions were derived from HEK293T cells that were transfected with 1 μg of either GABAB-R1a (lanes 1-3), GABA %-R1b (lanes 4-6) or HA-GABAB-R2 (lanes 13-15), or with 1 μg each of HA-GABAB-R2 in combination with 1 μg of either GABA-R1a (lanes 7-9, 16-18) or GABAB-R1b (lanes 10-12, 19-21). Glycosylation status of transfected receptors was assessed following treatment of P2 fractions (50 μg of membrane protein) with either vehicle (lanes 1, 4, 7, 10, 13, 16 and 19), endoglycosidase F (lanes 2, 5, 8, 11, 14, 17 and 20) or endoglycosidase H (lanes 3, 6, 9, 12, 15, 18 and 21). Samples were resolved by SDS-PAGE (10% (w/v) acrylamide), transferred to nitrocellulose, and immunoblotted. Upper panel, antiserum 501 was used as primary reagent to allow identification of both GABAB-R1a and 1b. Lower panel, 12CA5 anti-HA antiserum was employed to identify HA-GABA-R2. *, denotes terminally glycosylated forms of GABAB-R1a and 1b.

FIGS. 9A and 9B. Coexpression of GABAB-R1 and GABAB-R2 Receptors in HEK293T Cells Leads to GABA-Mediated Stimulation of [35S]GTPγS Binding Activity.

[35S]GTPγS binding activity was measured on P2 particulate fractions derived from HEK293T cells transfected with 1 μg of Go1α together with 1 μg of either GABAB-R1a, GABAB-R1b or HA-GABAB-R2; or with 1 μg each of Go1α and HA-GABAB-R2 in combination with 1 μg of either GABAB-R1a or GABAB-R1b. (A) [35S]GTPγS binding was measured in the absence (open bars) or presence (hatched bars) of GABA (10 mM) as described in Methods. (B) The ability of varying concentrations of GABA to stimulate the binding of [35S]GTPγS was measured on P2 membrane fractions from HEK293T cells expressing either Go1α and HA-GABAB-R2 alone (open circles) or in combination with either GABAB-R1a (closed squares) or GABAB-R1b (closed triangles). The data shown are the means±S.D. of triplicate measurements and are representative of three independent experiments.

FIG. 10. GABA-Mediated Stimulation of [35S]GTPγS Binding Activity in HEK293T Cells Coexpressing GABAB-R1 and GABAB-R2 Receptors Requires Cotransfection with Additional GiG Protein, Go1α.

[35S]GTPγS binding activity was measured on P2 particulate fractions derived from HEK293T cells transfected with HA-GABAB-R1b (1 μg) together with HA-GABAB-R2 (1 μg) and Go1α (1 μg) (closed triangles), or in combination with either HA-GABAB-R2 (1 μg) (open circles) or Go1α (1 μg) (closed circles). The ability of varying concentrations of GABA to stimulate the binding of [35S]GTPγS was determined. Data shown are the mean±S.D. of triplicate measurements.

FIGS. 11A and 11B. Coexpression of GABAB-R1 and GABAB-R2 Receptors in HEK293T Cells Permits GABA-MEDIATED Inhibition of Forskolin-Stimulated Adenylate Cyclase Activity

cAMP levels were measured in HEK293T cells transfected with 1 μg of Gi1α together with 1 μg of either GABAB-R1a, GABAB-R1b or HA-GABAB-R2; or with 1 μg each of Gi1α and HA-GABAB-R2 in combination with 1 μg of either GABAB-R1a or GABAB-R1b, as described in Methods. (A) cAMP levels were determined in cells treated with forskolin (50 μM) in the absence (open bars) or presence (hatched bars) of GABA (1 mM). (B) ability of varying concentrations of GABA to inhibit forskolin-elevated adenylate cyclase activity in HEK293T cells expressing Gi1α and HA-GABAB-R2 in combination with GABAB-R1b. The data shown are the means±S.D. of triplicate measurements.

FIGS. 12A and 12B. Co-Expression of GABAB-R1 and GABA-R2 Receptors in Xenopus oocytes Permits Agonist-Dependant Activation of Ion Flux through CFTR and GIRK1/4.

Xenopus oocytes were injected with cRNA encoding GABAB-R1 and GABAB-R2 receptors (in equal amounts for CFTR, 1:2 ratio for GIRK) plus either CFTR (A) or the GIRK1/GIRK4 heteromer (B). A, Time course plot for an oocyte expressing GABAB-R1, GABAB-R2 and CFTR. Application of 100 mM GABA, 100 mM SKF97541 or 1 mM Baclofen (arrows) activated a large inward CFTR current. Note the increase in CFTR response seen with repeated GABA application. B, Time course plot for an oocyte expressing GABAB-R1, GABAB-R2, GIRK1 and GIRK4. Switching from ND96 (low potassium) to 90K (high potassium) solution led to an inward shift in holding current, showing that the GIRK1/GIRK4 channel is expressed in this oocyte. Subsequent application of 100 mM GABA activated a large inward current (middle panel). Negative and positive control experiments are shown from oocytes expressing the GABAB-R2 receptor alone (left panel) and those expressing the adenosine A1 receptor (right panel).

FIG. 13. Current-Voltage Curves in an Oocyte Expressing GABAB-R1, GABAB-R2 and the Potassium Channels GIRK1 and GIRK4.

Current-voltage curves are shown for a single oocyte following application of 200 ms voltage-clamp pulses from a holding potential of −60 mV to test potentials between −100 mV and +50 mV. Steady-state current is plotted against test potential in ND96 solution (low potassium), 90K solution (90 mM potassium) and 90K plus 100 mM GABA. Note the basal GIRK1/4 current recorded in 90K solution and the large agonist-evoked activation of the GIRK potassium channel.

FIG. 14. GABA-Mediated Stimulation of [35S]GTPγS Binding Activity is Dependent on the Relative Levels of Expression of GABAB-R1 and GABAB-R2 Receptors

HEK293T cells were transfected with HA-GABAB-R2 (1 μg) and Go1α (1 μg) together with various amounts (0-1 μg) of HA-GABAB-R1b. Cells were harvested 48 h after transfection and P2 membrane fractions were prepared. (A) Agonist stimulation of [35S]GTPγS binding activity measured in transfected cell membranes in the presence of GABA (10 mM). Data are shown as stimulation above basal (cpm) and are the mean±S.D. of triplicate measurements. (B) Cell membranes were immunoblotted with anti-HA antiserum to allow the relative levels of HA-GABAB-R2 and HA-GABAB-R1b receptors to be evaluated.

FIG. 15. Co-Expression of GABAB-R1 and GABAB-R2 Receptors in HEK293T Cells Generates a High Affinity GABAB Binding Site Similar to Brain GABAB Receptors.

P2 membrane fractions were prepared from HEK 293T cells transfected using the same conditions described for GTPγS binding studies. % specific binding was determined for the displacement of [3H]-CGP54626 by GABA. Data shown are the mean of minimum of triplicate studies±sem.


Throughout the present specification and the accompanying claims the words “comprise” and “include” and variations such as “comprises”, “comprising”, “includes” and “including” are to be interpreted inclusively. That is, these words are intended to convey the possible inclusion of other elements or integers not specifically recited, where the context allows.

As previously explained, the present invention includes a number of important aspects. In particular the present invention relates to isolated GABAB-R2 receptor proteins and variants thereof, isolated GABAB-R1c receptor proteins and variants thereof, GABAB receptors comprising an heterodimer between a GABAB-R1 receptor protein or a variant thereof and a GABAB-R2 receptor protein or a variant thereof, as well as other related aspects. In the context of the present invention the wording “isolated” is intended to convey that the receptor protein is not in its native state, insofar as it has been purified at least to some extent or has been synthetically produced, for example by recombinant methods. The term “isolated” therefore includes the possibility of the receptor protein being in combination with other biological or non-biological material, such as cells, suspensions of cells or cell fragments, proteins, peptides, organic or inorganic solvents, or other materials where appropriate, but excludes the situation where the receptor protein is in a state as found in nature.

Routine methods, as further explained in the subsequent experimental section, can be employed to purify and/or synthesise the receptor proteins according to the invention. Such methods are well understood by persons skilled in the art, and include techniques such as those disclosed in Sambrook, J. et al, 1989, the disclosure of which is included herein in its entirety by way of reference.

The present invention not only includes the GABAB receptor proteins specifically recited, but also variants thereof. By the term “variant” what is meant throughout the specification and claims is that other peptides or proteins which retain the same essential character of the receptor proteins for which sequence information is provided, are also intended to be included within the scope of the invention. For example, other peptides or proteins with greater than about 80%, preferably at least 90% and particularly preferably at least 95% homology with the sequences provided are considered as variants of the receptor proteins. Such variants may include the deletion, modification or addition of single amino acids or groups of amino acids within the protein sequence, as long as the biological functionality of the peptide is not adversely affected.

The invention also includes nucleotide sequences which encode for GABAB-R2 or GABAB-R1c receptors or variants thereof as well as nucleotide sequences which are complementary thereto. Preferably the nucleotide sequence is a DNA sequence and most preferably, a cDNA sequence.

The present invention also includes expression vectors which comprise nucleotide sequences encoding for the GABAB-R2 or GABAB-R1 c receptor subtypes or variants thereof. A further aspect of the invention relates to an expression vector comprising nucleotide sequences encoding for a GABAB-R1 receptor protein and a GABAB-R2 receptor protein or variants thereof. Such expression vectors are routinely constructed in the art of molecular biology and may involve the use of plasmid DNA and appropriate initiators, promoters, enhancers and other elements, which may be necessary, and which are positioned in the correct orientation, in order to allow for protein expression.

The invention also includes cell lines which have been modified to express the novel receptor. Such cell lines include transient, or preferably stable higher eukaryotic cell lines, such as mammalian cells or insect cells, lower eukaryotic cells, such as yeast or prokaryotic cells such as bacterial cells. Particular examples of cells which have been modified by insertion of vectors encoding for the receptor proteins according to the invention include HEK293T cells and oocytes. Preferably the cell line selected will be one which is not only stable, but also allows for mature glycosylation and cell surface expression of the inventive receptors. In the case of the functional GABA, receptor which comprises a heterodimer of GABAB-R1 and GABAB-R2 subunits, the cell line may include a single vector which allows for expression of both of the receptor subtypes, or alternatively separate vectors for each subunit. It is preferred however, that the receptor subtypes should be co-expressed in order to optimise the dimerisation process, which will result in full glycosylation and transport of the glycosylated dimer to the cell surface.

It is also possible for the receptors of the invention to be transiently expressed in a cell line or on a membrane, such as for example in a baculovirus expression system. Such systems, which are adapted to express the receptors according to the invention, are also included within the scope of the present invention.

A particularly preferred aspect of the invention is the heterodimer formed between the GABAB-R1 and GABAB-R2 receptor proteins which results in the formation of a functional GABAB receptor. Without wishing to be bound by theory, it appears that the formation of the heterodimer takes place via the coiled-coil domains within the receptor C-terminal tails, and that this in turn is a pre-requisite for transport and full glycosylation of a GABAB-R1, and also for generation of an high affinity GABAB receptor at the cell surface.

The heterodimer which forms a functional GABAB receptor can comprise any GABAB-R1 receptor subtype or splice variant, or variants thereof. Although we are presently only aware of only one GABAB-R2 subtype, it is envisaged that the heterodimers according to the present invention can include other GABAB-R2 subtypes or splice variants which have not yet been identified, as well as variants of the already identified GABAB-R2 receptor proteins.

In particular, the functional GABA, receptor may include GABAB-R1 receptor proteins selected from GABAB-R1a, GABAB-R1b, GABAB-R1c splice variants, variants thereof or even other GABAB-R1 receptor subtypes or splice variants which have not yet been identified.

According to another aspect, the present invention also relates to antibodies which have been raised by standard techniques and are specific for the receptor proteins or variants thereof according to the invention. Such antibodies could for example, be useful in purification, isolation or screening involving immuno precipitation techniques and may be used as tools to further ellucidate GABAB receptor function, or indeed as therapeutic agents in their own right. Antibodies may also be raised against specific epitopes of the receptors according to the invention, as opposed to the monomer subunits.

An important aspect of the present invention is the use of receptor proteins according to the invention, particularly the heterodimer GABAB receptor, in screening methods designed to identify compounds which act as receptor ligands and which may be useful to modulate receptor activity. In general terms, such screening methods will involve contacting the receptor protein concerned, preferably the heterodimeric GABAB receptor, with a test compound and then detecting modulation in the receptor activity, or indeed detecting receptor inactivity, which results. The present invention also includes within its scope those compounds which are identified as possessing useful GABAB receptor modulation activity, by the screening methods referred to above. The screening methods comprehended by the invention are generally well known to persons skilled in the art, and are further discussed in the experimental section which follows.

Another aspect of the present invention is the use of compounds which have been identified by screening techniques referred to above in the treatment or prophylaxis of disorders which are responsive to modulation of a GABAB receptor activity, in a mammal. By the term “modulation” what is meant is that there will be either agonism or antagonism at the receptor site which results from ligand binding of the compound at the receptor. GABAB receptors have been implicated in disorders of the central nervous system (CNS), gastrointestinal (GI) tract, lungs and bladder and therefore modulation of GABAB receptor activity in these tissues will result in a positive therapeutic outcome in relation to such disorders. In particular, the compounds which will be identified using the screening techniques according to the invention will have utility for treatment and/or prophylaxis of disorders such as spasticity, epilepsy, Alzheimer's disease, pain as well as affective disorders and feeding disorders. It is to be understood however, that the mention of such disorders is by way of example only, and is not intended to be limiting on the scope of the invention.

The compounds which are identified according to the screening methods outlined above may be formulated with standard pharmaceutically acceptable carriers and/or excipients as is routine in the pharmaceutical art, and as fully described in Remmington's Pharmaceutical Sciences, Mack Publishing Company, Eastern Pennsylvania, 17th Ed, 1985, the disclosure of which is included herein in its entirety by way of reference.

The compounds may be administered via enteral or parenteral routes such as via oral, buccal, anal, pulmonary, intravenous, intraarterial, intramuscular, intraperitoneal, topical or other appropriate administration routes.

Other aspects of the present invention will be further explained, by way of example, in the appended experimental section.



1. Cloning of Human GABAB-R1 and a Novel Receptor Subtype, GABAB-R2

Human homologues to the rat GABAB-R1a and 1b splice variants were identified from ESTs and subcloned from Human cerebellum cDNA, using a combination of PCR and Rapid amplification of cDNA ends (RACE) PCR. Human GABAB-R1a and 1b sequences reveal over 99% identity to the rat GABAB-R1a and GABAB-R1b (data not shown). These receptors, like their rat counterparts, both have signal sequences, followed by extended N-termini, a typical seven-transmembrane topology and short intracellular C-terminal tail. The N-terminus encodes the GABA binding domain, which is predicted by limited homology to bacterial periplasmic proteins to exist as two globular domains that capture GABA (Bettler et al, 1998), as well as three potential N-glycosylation sites. Interestingly the GABAB-R1a splice variant N-terminus encodes 129 amino acids over that of GABAB-R1b, which encode two tandem copies of the ‘short consensus repeat’ or sushi domain. Sushi domains are approximately 60 amino acids in length and exist in a wide range of proteins involved in complement and cell-cell adhesion (Chou and Heinrikson, 1997). Therefore the sushi domains within GABAB-R1a may direct protein-protein interactions, possibly through cell-cell contact and may reflect a further role for GABAB-R1a, over and above that of GABAB-R1b. Interestingly during the isolation of these clones, a novel N-terminal splice variant, GABAB-R1c was identified. GABAB-R1c differs from GABAB-R1a by a 185 bp deletion from bases 290 to 475 (see FIG. 2). This region encodes one of the two Sushi domains unique to GABAB-R1a and therefore the GABAB-R1a and GABAB-R1c splice variants, together with their cellular localisation, may be significant in the biology of GABAB receptors. Indeed, in situ hybridisations suggest that GABAB-R1a and GABAB-R1b have different sub-cellular localisations, with GABAB-R1a expressed at pre-synaptic rather than at post-synaptic sites (Bettler et al., 1998).

Database searches also identified a number of ESTs showing weaker homology to GABAB-R1, suggesting the existence of a novel GABAB receptor subtype. Using PCR on Human Brain cerebellum cDNA, we confirmed the existence of such a novel GABAB receptor which we cloned and sequenced (FIG. 1). This novel receptor, which we have called GABAB-R2, shows an overall 54% similarity and 35% identity to GABAB-R1 over the full length of the protein (FIG. 2). As expected, hydrophobicity profiles for GABAB-R2 (FIG. 3) suggested that the protein has a 42 amino acid signal peptide followed by an extracellular N-terminal domain comparable in size to that of GABAB-R1b and seven membrane spanning regions. In total five N-glycosylation sites were predicted over the N-terminal domain, three of which are conserved within GABAB-R1. Finally, the receptor encodes an intracellular C-terminal domain, which is considerably larger than that of GABAB-R1. No sushi domains were identified within GABAB-R2 sequence and we have no evidence for any splice variants to date.

2. Tissue Distribution

Expression levels of both GABAB-R1 and GABAB-R2 were determined and compared in different tissues and developmental stages by probing Human RNA Master Blots (Clontech). These blots contain polyA+ RNA samples from 50 human tissues that have been normalized to the mRNA expression levels of eight different “housekeeping” genes. GABAB-R1 levels were examined using a pan-specific probe covering all splice variants (FIG. 4a) and the blots indicate that in accordance with the observations of Kaupmann et al., (1997), GABAB-R1 is highly expressed in the CNS, in all areas of the brain and spinal cord. However, in contrast to Kaupmann et al., (1997), we find that GABAB-R1 is also expressed at comparable levels in peripheral tissues, with particularly high levels of expression in the pituitary, lung, ovary, kidney, small intestine, and spleen. In marked contrast, GABAB-R2 is specifically expressed at high levels only in the CNS, with the possible exception of spinal cord where expression appears somewhat lower. No signal is seen for peripheral tissues, in either adult or fetal tissues (FIG. 4b). This markedly different distribution of mRNA levels between GABAB-R1 and GABAB-R2 suggests that the two subtypes may have distinct roles in the CNS and periphery.

3. Initial Expression Studies

We reasoned that GABAB-R2 could be a high affinity GABAB receptor and therefore, expressed the receptor in both Xenopus oocytes and HEK293T cells and looked for functional responses. However, despite repeated attempts, we were unable to detect any functional activation of GABAB-R2 or indeed, GABAB-R1a, GABAB-R1b or GABAB-R1c receptors by either GABA itself or GABAB selective agonists (See FIGS. 9, 11 and 12). Several lines of evidence clearly indicated that GABAB-R1 was not expressed as predicted in vivo. Firstly, flow cytometry of HEK293T cells, expressing GABAB-R1b, revealed that receptors were retained on internal membranes rather than expressed at the cell surface (FIG. 7). Secondly, GABAB-R1a and GABAB-R1b were expressed as immature glycoproteins, by virtue of their sensitivity to endoglycosidases F and H (FIG. 8, lanes 1-6) and finally, GABAB-R1 co-expression in oocytes with either GIRK or CFTR, gave no indication of a functional response (data not shown). We concluded that some additional co-factor must be required to promote a functional response.

4. Yeast Two Hybrid Library Screening

The calcitonin-receptor like receptor is retained as an immature glycoprotein within the endoplasmic reticulum and requires an accessory protein from the recently identified RAMP protein family to transport the receptor to the surface to generate a functional CGRP (Calcitonin gene-related peptide) or adrenomedullin receptor (McLatchie et al., 1998). We anticipated that GABAB-R1 receptors should require an analogous trafficking factor or some other protein co-factor for its transport to the cell surface to generate a high affinity receptor To identify such potential interacting proteins, a yeast two hybrid library screen was run using the C-terminal 108 amino acids of GABAB-R1 against a Human Brain cDNA library. Interestingly, motif searches revealed a strong coiled-coil domain within these 108 residues, a structure known to mediate protein-protein interactions (Lupas, 1996). From a total of 4.3×106 cDNAs, 122 positives hits were recovered, 33 of which encoded the whole C-terminal domain of GABAB-R2. This domain of the GABAB-R2 is likewise predicted to contain a coiled-coil motif, which aligns exactly with that of GABAB-R1 (see FIG. 2). This observation strongly suggests that the two receptors interact via their C-termini to form a heterodimer. Significantly, the screen did not retrieve the C-terminal domain of the GABAB-R1 itself, implying that GABAB-R1 is unable to homodimerise. This interaction was tested directly in the yeast two hybrid system using the C-termini of the two receptors (FIG. 5). GABAB-R1 and GABAB-R2 were able to strongly interact via their C-termini, whilst neither receptor was able to homodimerise. This observation suggested that GABAB-R1 and GABAB-R2 form heterodimers via their C-terminal coiled-coil domains and led to speculation that homodimerisation may bring about a functional binding site in vivo. Therefore, we next confirmed the interaction between the two receptor subtypes by immunoprecipitation studies upon whole epitope-tagged receptor in transfected HEK293T cells.

5. Co-Immunoprecipitation Studies.

Epitope tagged receptors, Myc-GABAB-R1b and HA-GABAB-R2 were transiently expressed in HEK293T cells either alone or in combination. Immunoprecipitation of Myc-GABAB-R1b from detergent-solubilised cell fractions with Myc antisera led to immunodetection of HA-GABAB-R2 within immune complexes using HA as the primary antibody, but only upon receptor co-expression (FIG. 6, lanes 1-3). GABAB-R1 and GABAB-R2 association was confirmed by co-immunodetection of Myc-GABAB-R1b from immune complexes captured using the anti-HA antibody. Once again, co-immunoprecipitation could only be seen when the two receptor forms were co-expressed (FIG. 6, lanes 4-6). Hence in agreement with the yeast two hybrid observations, these data provide compelling evidence for heterodimerisation between full-length expressed GABAB-R1 and GABAB-R2 in mammalian cells. Therefore, we next examined GABAB receptor responses following co-expression of both receptor subtypes in HEK293T cells or in Xenopus oocytes.

6. Surface Expression of the Heterodimer

HEK293T cells were transiently transfected with Myc-GABAB-R1b alone or in combination with HA-GABAB-R2 and transfectants analysed by flow cytometry (FIG. 7). Myc-immunoreactivity could not be detected on the surface of cells transfected with Myc-GABAB-R1b alone (FIG. 7a), although cell permeabilisation revealed immunoreactivity in 35% (n=3) of the cell population (FIG. 7b). This latter observation indicated that cells were efficiently transfected and suggested that expressed Myc-GABAB-R1 receptors were localised exclusively on internal membranes. In contrast, 14% (n=3) of HEK293T cells transfected with HA-GABAB-R2 showed surface immunoreactivity (FIG. 7c). However, co-transfection of both Myc-GABAB-R1b and HA-GABAB-R2 led to the appearance of Myc-GABAB-R1b on the surface of 20% (n=3) of cells analysed (FIG. 7a), strongly suggesting that co-expression of GABAB-R1b with GABAB-R2 is necessary for surface expression of GABAB-R1b.

7. Receptor Glycosylation Studies

Endoglycosidases F and H can be used to differentiate between core and terminally glycosylated N-linked glycoproteins. Therefore, these enzymes were used to examine the glycosylation status of both GABAB-R1 and GABAB-R2 following expression in HEK293T cells. Membranes from transfected cells were treated with either endoglycosidase F or endoglycosidase H and expressed GABAB receptors were characterised by immunoblotting to compare relative electrophoretic mobilities of the receptors (FIG. 8) Cell membranes expressing either GABAB-R1a or 1b produced distinct bands of Mr 130 and 100K respectively (FIG. 8, lanes 1 and 4) which following endoglycosidase F treatment, decreased in size to single immunoreactive species of M, 110 and 80K; representing GABAB-R1a and GABAB-R1b respectively (FIG. 8, lanes 2 and 5) This shows that recombinant GABAB-R1a and 1b are glycoproteins, in agreement with the observations of Kaupmann et al., (1997). However, both GABAB-R1 a and 1b splice variant forms were also sensitive to endoglycosidase H treatment, indicating that the expressed proteins are only core glycosylated (lanes 3 and 6) and lack terminal glycosylation. This observation, together with the FACS analysis, suggests that the proteins are immaturely glycosylated and retained on internal membranes. Significantly, when either GABAB-R1a (lanes 7-9) or GABAB-R1b (lanes 10-12) was co-expressed with HA-GABAB-R2, a component of GABAB-R1a or 1b was resistant to endoglycosidase H digestion suggesting that when co-expressed with GABAB-R2, a significant fraction of GABAB-R1 is now a mature glycoprotein (lanes 9 and 12).

Similar studies with HA-GABAB-R2 gave an immunoreactive species with an M, of 120 K (FIG. 8, lanes 13, 16, 19) which was sensitive to endoglycosidase F (lanes 14, 17 and 20) but resistant to endoglycosidase H (lanes 15, 18 and 21) treatment, whether expressed alone or in combination with GABAB-R1. Thus, these data indicate that expressed HA-GABAB-R2 is a mature glycoprotein whose glycosylation status is not affected by co-expression with GABAB-R1. Thus, heterodimerisation between GABAB-R1 and GABAB-R2, possibly in the Golgi complex, could be a prerequisite for maturation and transport of GABAB-R1 to the plasma membrane.

8. Functional Studies

To determine whether co-expression of GABAB-R1 and GABAB-R2 and its subsequent mature glycosylation and cell surface expression, generated a receptor complex able to functionally respond to GABA, we measured three types of signalling. We used transiently transfected HEK239T cells to examine firstly, activation of [35S]GTPγS binding in membranes and secondly, inhibition of forskolin stimulated cAMP activation in whole cells. Thirdly we expressed GABAB-R1 and GABAB-R2 in Xenopus oocytes, expressing either the cystic fibrosis transmembrane regulator (CFTR) or inwardly rectifying K+ channels (GIRK and KATP) and examined activation of ion flux in response to agonist.

i. [35S]GTPγS Binding

No GABA stimulated [35S]GTPγS binding was observed in membranes prepared from cells transfected with either GABAB-R1 or HA-GABAB-R2 in combination with Go1α. However, co-expression of GABAB-R1 and HA-GABAB-R2 together with Go1α resulted in a robust stimulation of [35S]GTPγS binding activity (FIG. 9a). This was found to be concentration-dependent with similar EC50 (mean, ±S.E.M., n=3) values determined for membranes from cells transfected with HA-GABAB-R2 and GABAB-R1 together with either GABAB-R1a (9.5±1.1×10−5M) or GABAB-R1b (7.8±0.4×10−5M) (FIG. 9b). These values are equivalent to those of GABAB-mediated stimulation of [35S]GTPγS binding to rat brain membranes (5.9±0.4×10−5M) (data not shown). We were concerned that an N-terminal HA epitope tag on GABAB-R2 could alter receptor function and so we performed parallel studies in HEK293T cells, expressing untagged versions of GABAB-R2 and GABAB-R1 together with Go1α. Similar efficacies and potencies of GABA action were observed in membranes from these cells, as reported for the epitope tagged receptors (data not shown), clearly suggesting that the addition of these peptide sequences to the N-termini of GABAB-R2 and GABAB-R1 did not significantly alter receptor function. It is noteworthy that a measurable GABA-mediated elevation of [35S]GTPγS binding activity was only observed upon co-expression of GABAB-R1 and HA-GABAB-R2 together with additional Go1α (FIG. 10). The requirement for additional G protein is most likely due to relatively low levels of endogenously expressed Gi/o family G proteins, thus precluding a discernible GABA-mediated response upon GABAB-R1 and GABAB-R2 co-expression.

ii cAMP Inhibition

Similar results were obtained from HEK293T cells transiently transfected with GBAB-R1 and GABAB-R2, using inhibition of forskolin evoked cAMP as a readout. Once again, functional responses were only observed when both GABAB-R1 and GABAB-R2 were co-expressed (FIG. 11).

iii Xenopus oocytes

Xenopus Oocytes can Assay for Three Classes of G-Protein:

1) Endogenous oocyte Ca2+-activated chloride conductance can assay for activation of Gq and a subsequent rise in intracellular calcium (Uezono et al., 1993).

2) Cystic fibrosis transmembrane regulator (CFTR), which contains a cAMP-activated chloride channel, can assay for receptor activation via Gs or Gi/o (Uezono et al., 1993; Wotta et al., 1997).

3) G-protein regulated potassium channels GIRK1 (Kir 3.1; Kubo et al., 1993) and GIRK4 (or CIR, Kir 3.4, Kaprivinsky et al., 1995), injected in equal amounts to generate a heteromeric channel, can assay for activation of pertussis toxin sensitive G-proteins (Kovoor et al., 1997).

No functional responses to GABA or baclofen were seen when cloned GABAB-R1a, GABAB-R1b or GABAB-R2 receptors were expressed in oocytes in combination with CFTR or GIRK114 (data not shown; see FIG. 12b). When GABAB-R1 and GABAB-R2 were co-expressed with CFTR, several significant, robust responses were recorded following application of 100 μM GABA (FIG. 12a). Moreover, repeated application of GABA led to a progressive increase in the size of the CFTR response, suggesting that the functional response of the heterodimer is now sensitised to further challenge by agonist. This phenomenon has not been observed for other cloned receptors expressed in oocytes and may be related to the heterodimerisation or even oligomerisation of the GABAB receptors. Finally, two other GABAB-selective agonists, Baclofen and SKF97541 elicted similar functional responses through CFTR to that of GABA (FIG. 12a). In contrast, antagonists gave no response (data not shown).

Next, we examined the GABAB-R1/GABAB-R2 heterodimer with the G-protein regulated potassium channels GIRK1 and GIRK4 and once again found agonist dependant responses. Time course plots were examined for three individual oocytes expressing GABAB-R2 alone (left panel), GABAB-R1 plus GABAB-R2 (middle panel) and the adenosine A1 receptor (as a positive control, right panel) (FIG. 12b). In each case, switching from a low potassium physiological solution (ND96) to a high potassium extracellular solution (90 mM K+) led to an inward shift in holding current, resulting from agonist-independent influx of potassium ions through the GIRK1/4 channel. No GABA response was seen in oocytes expressing GABA6-R2 in isolation (FIG. 12b, left panel) and similarly, GABAB-R1a and GABAB-R1b expressed alone also gave no response to GABA (data not shown). Significantly, a large GABA response was recorded in oocytes co-expressing GABAB-R1 and GABAB-R2 (FIG. 12b, middle panel) of a similar magnitude to that of the adenosine A1 receptor in response to the agonist NECA (FIG. 12b, right panel). Thus, once again co-expression of the two receptor subtypes elicits a functional agonist-dependant response, whereas expression of either subtype receptor alone does not. We also examined whether co-expression of the two receptors in oocytes could activate endogenous Ca2+-activated chloride conductance. No evidence for activation was seen (data not shown) suggesting that at least in oocytes, the GABAB-R1/GABAB-R2 receptor complex does not signal through Gq. Finally, a current-voltage curve were constructed for an oocyte co-expressing GABA-R1 and GABAB-R2 (FIG. 13). This clearly demonstrates that GABA, bound to the GABAB receptor, activates a large inwardly rectifying current consistent with activation of the GIRK potassium channel in a fully dose dependant manner.

9. Stoichiometric Studies on the Heterodimer

Since co-expression of GABAB-R1 and GABAB-R2 is necessary for a functional GABAB receptor, we decided to investigate stoichiometric ratio between the two receptor subtypes in vivo. Relative levels of expression for both GABAB-R1 and GABAB-R2 were measured following transfection into HEK293T cells and compared to receptor function, as determined by GTPγS binding (FIG. 14). Increasing amounts of HA-GABAB-R1 (up to 1 μg) plasmid were transfected into HEK293T cells along with a constant (1 μg) amount of HA-GABAB-R2. GABA caused stimulation of [35S]GTPγS binding above basal levels in membranes extracted from these cells, which increased with increasing amount of transfected HA-GABAB-R1 until binding reached a plateau when levels of HA-GABAB-R1 were greater than 0.25 kg (FIG. 14a). Immunoblotting of the same membrane samples revealed equivalent levels of expression of HA-GABAB-R1 and HA-GABAB-R2 in membranes transfected with 0.25-0.5 μg of HA-GABAB-R1 (FIG. 14b). This corresponded to the plateau of GABA-mediated elevation of [35S]GTPγS binding activity and therefore strongly suggests that GABAB-R1 and GABAB-R2 functionally interact in a 1:1 stoichiometric ratio.

10. Competition Binding Studies

Finally, we determined whether the observed functional responses were due to a high affinity GABAB receptor, composed of a heterodimer of the two receptors. HEK293T cells were transfected with either 1 μg HA-GABAB-R1b and HA-GABAB-R2 individually or with increasing amounts (up to 1 μg) of HA-GABAB-R1b and a fixed amount (1 μg) of HA-GABAB-R2 together with Go1α. Competition binding assays were then performed upon purified membranes. Expression of HA-GABAB-R1b alone produced high levels of specific binding of [3H]-CGP54626 (Bittiger et al., 1992), a structural analogue of [125I]-CGP64213 and the antagonist originally used to expression clone GABAB-R1 (Kaupmann et al., 1997). However, as previously reported for [125I]-CGP64213, GABA inhibition curves were significantly shifted to the right compared with binding to rat brain membranes (FIG. 15), giving approximately 22-fold lower IC50 than rat brain binding. Significantly, co-expression of equivalent amounts of HA-GABAB-R1b and HA-GABA/B-R2 protein revealed high levels of specific binding. In a control experiment using untagged receptors similar values were obtained (data not shown). Achievement of a 1:1 stoichiometric ratio of expression of HA-GABAB-R1b and HA-GABAB-R2 led to agonist inhibition curves similar to those obtained in rat brain membranes (IC50±95% confidence intervals for 1 μg HA-GABAB-R2/0.25 μg HA-GABAB-R1b=2.29 μM (1.48-3.55 μM) and for rat brain=1.04 μM (0.69-1.58 μM). Such comparable levels of receptor expression were also shown to permit optimal agonist activation in the GTPγS assay (see FIG. 14). Alteration of receptor ratio from 1:1, such that GABAB-R1b was the most prevalent receptor, led to reduced agonist affinity, presumably due to binding at non-dimerised and immaturely glycosylated GABAB-R1b receptors (FIG. 15).

In addition, despite its apparent cell surface expression, we were unable to detect any [3H]-CGP54626 specific binding to HEK293T cells transiently transfected with HA-GABAB-R2 alone (data not shown). We conclude that heterodimerisation of the GABAB-R1 and GABAB-R2 subtypes are necessary to generate a high affinity GABAB receptor. There are a number of possible explanations for the change in GABA affinity following co-expression of the two receptor subtypes. Appearance of the GABA8 receptor complex at the cell surface would be expected to allow G protein coupling of the receptor which would increase agonist affinity. However, in previous studies is has been shown that the lack of G protein coupling alone cannot account for the difference in agonist affinity between rat brain receptors and GABAB-R1 (Kaupmann et al, 1997). Furthermore, we have noted that [3H]-CGP54626 appears to primarily bind the low affinity state of the receptor, even in rat brain membranes, as demonstrated by the fact that GTPγS is unable to shift agonist inhibition curves and actually increases the level of 3H-CGP54626 specific binding (data not shown). Therefore, a more likely explanation for the change in GABA affinity following co-expression of the two GABAB receptors is that heterodimerisation together with the mature glycosylation state of the protein, produces a binding site conformation with an inherent higher affinity.


Functional GABAB receptors within the CNS comprise a cell surface heterodimer of two distinct 7-transmembrane receptor subunits, GABAB-R1 and GABAB-R2 in a 1:1 stoichiometric ratio. In vivo, GABAB receptors may exist simply as heterodimers or form even larger multimeric complexes of many heterodimers-Formation of the heterodimer via the coiled-coil domains within the receptor C-terminal tails appears to be a pre-requisite for transport and full glycosylation of GABAB-R1, as well as for the generation of a high affinity GABAB receptor at the cell surface. Using this information, we have been able to reproduce GABAB sites in both mammalian HEK293T cells as well as in oocytes, using several functional readouts such as activation of ion flux through CFTR or GIRK in oocytes, or inhibition of adenylyl cyclase in HEK293T cells. Indeed the lack of functional responses in cells expressing GABAB-R1 alone and the need for expression of a second 7TM receptor explains why many groups have encountered extreme difficulty in expression cloning a GABAB receptor via conventional means. We believe this is the first report of receptor heterodimerisation as an obligate requirement to generate a high affinity, fully functional receptor in recombinant systems, which is fully equivalent to that of endogenous tissues.

Dimerisation has been reported for other receptor families, such as the opioid family as a part of their desensitisation process, the β2-adrenergic receptor, where homodimers may play a role in signalling, and the metabotropic glutamate receptors (mGluRs, Hebert et al, 1996; Romano et al., 1996; Cvejic et al., 1997, Hebert and Bouvier, 1998). Significantly, dimerisation in these receptor families does not appear to be an absolute requirement for functional coupling in recombinant systems. In the case of the mGluRs, which are a closely related receptor family to GABAB (Kaupmann et al., 1997), homodimerisation is mediated through disulphide bridges between the N-terminal extracellular domains rather than a C-terminal coiled-coil. Indeed, heterodimerisation between two 7-transmembrane receptors, leading to both trafficking and mature glycosylation of the proteins to yield a functional receptor is unprecedented and is unique in the GPCR field. Certainly, mGluRs have not been found to form heterodimers (Romano et al., 1996) and the fact that two such closely related receptors families have evolved such different mechanisms of dimer formation suggests that this is a fundamentally important process for receptor function.

In vivo, pharmacological evidence suggests that there are many different GABAB receptor subtypes, both within the CNS as well as in peripheral tissues. How are such pharmacological subtypes of GABAB receptors formed? Only GABAB-R1 and GABAB-R2 have been identified as separate genes to date and database trawling has not identified any further receptors homologous to known GABAB receptors. This does not exclude the possibility that more, as yet unrecognised GABAB receptors do exist. Differences in distribution exist for the two GABAB receptors, for example GABAB-R2 is specifically expressed in the CNS whereas GABAB-R1 is expressed in both central and peripheral sites. These differences in distribution clearly add further complexity leading to the pharmacologically distinct receptor subtypes. Moreover, the genes encoding the GABAB receptors may be differentially spliced. GABAB-R1 encodes three N-terminal splice variants and yet more may remain to be detected. Interestingly, these splice variants have alterations in their N-terminal extracellular domain, the region involved in GABA binding (Takahashi et al., 1993, O'Hara et al., 1993) and encode either two (GABAB-R1a), one (GABAB-R1c) or no (GABAB-R1b) sushi domains. Given that the sushi domains mediate cell-cell protein-protein contact, the differences in these three splice variants may account for yet more of the pharmacologically defined GABAB receptor subtypes. To date, we have not detected any splice variants to GABAB-R2. Furthermore there are significant differences in the distribution of the individual splice variants suggesting that they may serve different functions within the CNS. For instance, GABAB-R1a splice variant is reported as presynaptic within the brain (Bettler et al., 1998) and therefore may define presynaptic GABAB autoreceptors. It seems likely that these splice variants of GABAB-R1 may account for at least some of the pharmacologically defined subtypes. Finally, with this novel observation of obligate receptor heterodimerisation, a further level of complexity has been added since functional GABAB binding sites require a heterodimerisation partner.

Now the molecular nature of the GABAB receptor is more fully understood, recombinant systems can be established for high throughput screening for compounds against individual pharmacologically defined GABAB sites. By these means, compounds with greater specificity and with fewer unwanted side effects can be discovered. For this, GABAB-R1 and GABAB-R2 (including all spice variants, and any fragments of the receptor) should be co-expressed either stably or transiently in suitable host cells. Suitable host cells include higher eukaryotic cell lines, such as mammalian cells, insect cells, lower eukaryotic cells, such as yeast or prokaryotic cells such as a bacterial cells. Screening assays with these recombinant cell lines could involve the use of radioligand binding to the dimer or individual subunits within the dimer. The activity profile in a binding assay to the dimer is likely to be different from the activity of compounds assayed using binding assays to GABAB-R1 alone due to alterations in the glycosylation status and the conformation of the receptor as a result of co-expressing GABAB-R1 or GABAB-R2. Functional assays, which measure events downstream of receptor activation, can also be used for screening compounds. Such assays include [S]-GTPγS binding to membranes isolated from cells expressing the dimer, activation or inhibition of ion channels using electrophysiological recording or ion flux assays; mobilisation of intracellular calcium; modulation of cAMP levels; activation or inhibition of MAP kinase pathways or alterations in the activity of transcription factors with the use of reporter genes. Further to this, secondary screens can be established in a similar manner, using different heterodimer combinations to exclude unwanted activity and thereby establish subtype selective GABA8 compounds.

In addition, any approach targetting the disruption or enhancemant of dimer formation of the GABAB heterodimer could represent a novel therapeutic approach with which to target GABAB receptors. Such strategies could include peptides or proteins physically associated with the coiled-coil domain or indeed, any other interacting regions of the dimer. Small molecules could also be identified which act at the points of contact formed by interaction of the components of the dimer. These may either promote or enhance the receptor function. Finally, antibodies could be made which specifically recognise epitopes on the dimer, as opposed to the monomer subunits. These could be used as tools to further elucidate the function of GABA8 receptors in disease or as therapeutic agents in their own right.


DNA Manipulation

Standard molecular biology protocols were used throughout (Sambrook et al., 1989) and all bacterial manipulations used Escherichia coli XL-1Blue (Stratagene) according to the manufacturers instructions. Standard PCR conditions were used throughout, unless otherwise stated. PCR reaction mixture contained 10-50 ng of target DNA, 1 pmol of each primer, 200 μM dNTPs and 2.5 U of either Taq polymerase (Perkin-Elmer) or Pful polymerase (Stratagene) with the appropriate buffer as supplied by the manufacturer. Cycling parameters were 1 cycle 95° C. 2 mins; 25 cycles 95° C. 45 secs 55° C. 45 secs 72° C. 1 min; 1 cycle 72° C. 10 mins. All PCR were carried out using either a Perkin Elmer 9600 PCR machine or a Robocycler Gradient 96 (Stratagene) PCR machine.

GABAB-R1—Cloning of Human Homologues and Splice Variants

Several human EST's (X90542; X90543; D80024; AA348199; T06711; T07518 and AA38224) were identified as homologous to the rat GABAB-R1a and GABAB-R1b sequences (Y10369; Y10370). The ESTs were aligned and the predicted open reading frame was amplified by RT-PCR from human brain cerebellum polyA+ RNA (Clontech) using the Superscript Preamplification System (Life Technologies). The 3′ end of the receptor (1545-2538 bp; GABAB-R1b) was amplified using primers 5′-GCGACTGCTGTGGGCTGCTTACT GGC-3 and 5′-GCGAATTCCCTGTCCTCCCTCACCCTACCC-3′. The central section (277-1737 bp of GABAB-R1b) was amplified using 5′-CCGAGCTCAAGCTCATCCACCACG-3′ and 5′-TCTTCCTCCACTCCTCTTTTCTT-3′. PCR products were subcloned into pCR-Script SK(+) (PCR-script Amp cloning kit; Stratagene). Error free PCR product were assembled in a three-way BstEII, SacI and EcoRI ligation and subcloned into pBluescript SK (−) (Stratagene).

The N-termini of the splice variants were generated using RACE (rapid amplification of cDNA ends) PCR with the Marathon cDNA amplification kit against Marathon-Ready human cerebellum cDNA (Clontech). RACE PCR was primed from a conserved sequence within GABAB-R1 using primer 5′-TGAGCTGGAGCCATAGGAAAGCACAAT-3′ to generate a 700 bp product. This further PCR amplified using the AP2 primer (Marathon) and a second internal GABAB-R1 primer 5′-GATCTTGATAGGGTCGTTGTAGAGCA-3′. The resulting 600 bp product was subcloned using the Zero blunt PCR cloning kit (Invitrogen). Sequence information achieved from this RACE PCR was used to clone the N-terminus of the GABAB-R1b splice variant, using primers 5′-GCTCCTAACGCTCCCCAACA-3′ and 5′-GGCCTGGATCACACTTGCTG-3′ into pCR-Script SK (+)(Stratagene). Human GABAB-R1a 5′ sequences were retrieved from Incyte database EST's (1005101;3289832) and used to design primers 5′-CCCAACGCCACCTCAGAAG-3′ and 5′-CCGCTCATGGGAAACAGTG C-3′. PCR on cerebellum cDNA and KELLY neuroblastoma cell line cDNA produced two discreet bands at 300 bp and 400 bp, which were cloned into pCR-Script SK (+) (Stratagene). Sequencing revealed that the 400 bp product encoded some of the Human GABAB-R1a 5′ sequences and the 300 bp product encoded the novel splice variant, GABAB-R1c. Next, primer, 5′-CCCCGGCACACATACTCAATCTCATAG-3′ was designed to RACE PCR the missing ˜225 bp of GABAB-R1a. A 250 bp product was obtained and reamplified using primer 5′-CCGGTACCTGATGCCCCCTTCC-3′ with primer AP2 (Marathon). A ˜250 bp band was once again generated, subcloned into pCR-Script SK (+) and when sequenced, encoded the 5′ end of GABAB-R1a. Next, clones spanning both the conserved receptor sequence and the %′ ends of the splice variants GABAB-R1a and GABAB-R1c were generated. Primer 5′-CGAGATGTTGCTGCTGCTGCTA-3′, priming from the start codon and the reverse RACE primer generated a predicted ˜800 bp band and this was subcloned into pCR-Script SK(+). Now, full-length GABAB-R1a, GABAB-R1b and GABAB-R1c clones can be assembled in pcDNA3.1(−) (Invitrogen). For GABAB-R1b, 5′ sequences, restricted NotI/SacI, and the conserved region of the receptor, cut EcoRI/SacI were both co-ligated into pcDNA3.1(−), restricted NotI/EcoRI. Likewise, the GABAB-R1a and GABAB-R1c 5′ fragments were subcloned XhoI/SacI with the EcoRI/SacI conserved fragment and co-ligated into pcDNA3.1(−), cut XhoI/EcoRI to reconstitute full length clones.

Tagging of GABAB-R1b

GABAB-R1b was tagged with either myc or HA epitopes. PCR primers 5′-TAGGATCCCACTCCCCCCATCCC-3′ and 5′-CCAGCGTGGAGACAGAGCTG-3′ were used to amplify a region immediately following the proposed signal sequence (position 88) to approx. 20 bp downstream of a unique PstI site at position 389 of the coding sequence, creating a unique 5′ in-frame BamHI site. This fragment was cloned BamHI/PstI, into a vector containing the CD97 signal sequence, the myc epitope and an in-frame BamHI site. This construct also contains a NotI site 5′ to the CD97 signal sequence and an EcoRI site downstream of the PstI site. GABA %-R1b sequences downstream to the PstI site and upto an external EcoRI site were subcloned from full length receptor into the vector described above likewise cut with PstI/EcoRI, to assemble full length tagged GABAB-R1b. CD97 signal sequence, myc epitope and GABAB-R1b coding sequence were subcloned, NotI/EcoRI, into pcDNA3.1(−) (Invitrogen). HA epitope was added to GABAB-R1b by co-ligation of the 5′ BamHI/PstI and 3′ PstI/EcoRI fragments into pCIN6 cut with BamHI/EcoRI. This vector contains a T8 signal sequence and 12CA5 HA epitope immediately preceding an in-frame BamHI site.

Cloning of GABAB-R2, the novel GABAB Receptor Subtype

EST clones (H14151, R76089, R80651, AA324303, T07621, Z43654) were identified with approximately 50% nucleotide identity to GABAB-R1. PCR revealed that H14151 contained a 1.5 Kb insert and encoded sufficient sequence for a substantial portion the novel GABAB receptor. PCR between the 3′ end of H14151 and the 5′ end of AA324303, using a cerebellum cDNA library as template, produced a ˜700 bp product, which when cloned into the T-vector (TA cloning kit, Invitrogen) and sequenced, revealed that T07621 overlaps within AA324303. Also, Z43654 as well as genomic DNA fragments R76089 and R80651 were found to overlap AA324303 and together provided sequence data for the 3′ end of the GABAB subtype receptor. Further sequencing of H14151 provided the full sequence for the novel receptor subtype. However, because of ambiguities in the position of the stop codon in Z43654/R80448/R80651, Incyte clones 662098 and 090041, which overlap this region, were sequenced. The stop codon was identified and sequence for GABAB-R2 was confirmed as within H14151 (5′ end) and 662098 (3′ end). 5′ sequences of GABAB-R2 were PGR generated using primers 5′-ATGGCTTCCCCGCGGAG-3′ to provide the start codon of the receptor and primer 5′-GAACAGGCGTGGTTGCAG-3′, priming beyond a unique EagI site. The expected ˜250 bp product was cloned into pCRSCRIPT and sequenced. Full length receptor was then assembled with a three way ligation between H14151, cut with ApaLI/EagI; 662098, cut with ApaLI/NotI and pCRSCRIPT-GABAB-R2-5′ PCR product, restricted by EagI.Full length GABAB-R2 was removed from the pCRSCRIPT vector using EcoRI/NotI and ligated into pcDNA3 (Invitrogen) for expression studies.

HA-epitope tagged GABAB-R2 was constructed in pCIN6. A linker was constructed encoding amino acids between the GABAB-R2 signal sequence and the unique EagI site.

HindIII XhoI EagI EcoRI
Ala Trp Gly Trp Ala Arg Gly Ala Pro Arg

The linker was cloned into pUC18 (EcoRI/HindIII) followed by full length GABAB-R2, from pCRSCRIPT as an EagI/NotI fragment. Finally, the modified GABAB-R2 was cloned into pCIN6 as a XhoI fragment.
Distribution Studies

Blots were hybridized overnight at 65° C. according to the manufacturers' instructions with radioactively randomly primed cDNA probes using ExpressHyb Hybridization solution. Probe for GABAB-R1, corresponding to residues 1129-1618 of the GABAB-R1b coding sequence was PCR amplified using primers 5′-CGCCTGGAGGACTTCAACTACAA-3′ and 5′-TCCTCCCAATGTGGTAACCATCG-3′ against GABAB-R1b DNA as template. GABAB-R2 cDNA probe, corresponding to residues 1397-1800, was amplified by PCR using primers 5′-ACAAGACCATCATCCTGGA-3′ and 5′-GATCACAAGCAGTTTCTGGTC-3′ with GABAB-R2 DNA as template. DNA fragments were labelled with 32P-α-dCTP using a Rediprime DNA labelling system (Amersham). Probes were labelled to a specific activity of >109 cpm/μg and were used at a concentration of approximately 5 ng/ml hybridization solution. Following hybridization, blots were washed with 2×SSC/1% SDS at 65° C., and 0.1×SSC/0.5% SDS at 55° C. (20×SSC is 3M NaCl/0.3M Na3Citrate.2H2O pH7.0) and were exposed to X-ray film.

Yeast Two Hybrid Studies

Saccharomyces cerevisiae Y190 [MATa, gal4 gal80, ade2-101, his3, trp1-901, ura3-52, leu2-3,112, URA3::GAL1-lacZ, LYS2::GAL1-HIS3, cyhR] was used for all described yeast two hybrid work (Harper et al., 1993, Clontech Laboratories, 1996). GAL4 binding-domain (GAL4BD) fusion vectors were constructed in either pYTH9 (Fuller et al., 1998) or pYTH16, an episomal version of pYTH9. All GAL4 activation-domain fusions were made in pACT2 (Clontech Laboratories, 1998) All yeast manipulations were carried out using standard yeast media (Sherman, 1991). Human Brain MATCHMAKER library (HL4004AH) in pACT2 was purchased from Clontech Laboratories and amplified according to the manufacturers' instructions. The GABAB-R1 C-terminal domain was amplified from a full length clone, using primers 5 GTTGTCCCCATGGTGCCCAAGATGCGCA GGCTGATCACC-3′ and 5′-GTCCTGCGGCCGCGGATCCTCACTTATAAAGCAAATGCACT CG-3′. PCR product was size-fractionated on 0.8% agarose gel, purified and force-cloned NcoI/NotI into pYTH9 and subsequently into pACT2. The GABAB-R2 C-terminal domain was similarly generated with primers 5′-CTCTGCCCCATGGCCGTGCCGAAGCTCATCACCCTGA GAACAAACCC-3′ and 5′-GGCCCAGGGCGGCCGCACTTACAGGCCCGAGACCATGACTC GGAAGGAGGG-3′ and subcloned into pYTH9, pYTH16 and pACT2. All cloned PCR products were sequenced and confirmed as error free.

The GAL4BD-GABAB-R1 C-terminus fusion in pYTH9 was stably integrated into the trp1 locus of Y190 by targetted homologous recombination. Yeast expressing GAL4BD-GABAB-R1 C-terminus were selected and transformed with Human brain cDNA library under leucine selection, using a high efficiency Lithium acetate transformation protocol (Clontech Laboratories, 1998). Sufficient independent cDNAs were transformed to give a three fold representation of the library. Interacting clones were selected by growth under 20 mM 3-amino-1,2,4-triazole (Sigma) selection, followed by production of β-galactosidase, as determined by a freeze-fracture assay (Clontech Laboratories, 1998). Plasmid DNA was recovered from yeast cells following digestion of the cell wall by 400 μg/ml Zymolase 100T (ICN Biochemicals) in 250 μl 1.2M Sorbitol; 0.1M potassium phosphate buffer (pH 7.4) at 37° for 2 h. Plasmid DNA was extracted by standard Qiagen alkaline lysis miniprep as per manufacturers' instructions and transformed into Ultracompetent XL-2Blue cells (Stratagene). Plasmid DNA was sequenced using primer 5′-CAGGGATGTTTAATACCACTACAATGG-3′ using automated ABI sequencing and resulting sequences were blasted against the databases.

Yeast Y190 was transformed with pYTH16 and pACT2 expressing GABAB-R1 C-terminal domain and the GABAB-R2 C-terminal domain in all combinations, as well as against empty vectors. Transformants were grown in liquid media to mid-logarithmic phase and approximately 1.5 ml harvested. β-galactosidase activity was quantified using substrate o-nitrophenyl β-D-galactopyranoside (ONPG; Sigma) using a liquid nitrogen freeze fracture regime essentially as described by Harshman et al., (1988).

Two-Microelectrode Voltage-Clamp in Xenopus oocytes

Adult female Xenopus laevis (Blades Biologicals) were anaesthetised using 0.2% tricaine (3-aminobenzoic acid ethyl ester), killed and the ovaries rapidly removed. Oocytes were de-folliculated by collagenase digestion (Sigma type I, 1.5 mg ml−1) in divalent cation-free OR2 solution (82.5 mM NaCl, 2.5 mM KCl, 1.2 mM NaH2PO4, 5 mM HEPES; pH 7.5 at 25° C.). Single stage V and VI oocytes were transferred to ND96 solution (96 mM NaCl, 2 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 5 mM HEPES; pH 7.5 at 25° C.) which contained 50 μg ml−1 gentamycin and stored at 18° C.

GABAB-R1a, GABAB-R1b (both in pcDNA3.1rev, Invitrogen), GABAB-R2, GIRK1, GIRK4 (in pcDNA3) and cystic fibrosis transmembrane regulator (CFTR; in pBluescript, Stratagene) were linearised and transcribed to RNA using T7 or T3 polymerase (Promega Wizard kit). m′G(5′)pp(5′)GTP capped cRNA was injected into oocytes (20-50 nl of 1 μgμl−1 RNA per oocyte) and whole-cell currents were recorded using two-microelectrode voltage-clamp (Geneclamp amplifier, Axon instruments Inc.) 3 to 7 days post-RNA injection. Microelectrodes had a resistance of 0.5 to 2MΩ when filled with 3M KCl. In all experiments oocytes were voltage-clamped at a holding potential of −60 mV in ND96 solution (superfused at 2 ml per min.) and agonists were applied by addition to this extracellular solution. In GIRK experiments the extracellular solution was changed to a high potassium solution prior to agonist application, to facilitate the recording of inward potassium currents. Current-voltage curves were constructed by applying 200 ms voltage-clamp pulses from the holding potential of −60 mV to test potentials between −100 V and +50 mV.

Mammalian Cell Culture and Transfections

HEK293T cells (HEK293 cells stably expressing the SV40 large T-antigen) were maintained in DMEM containing 10% (v/v) foetal calf serum and 2 mM glutamine. Cells were seeded in 60 mm culture dishes and grown to 60-80% confluency (18-24 h) prior to transfection with pcDNA3 containing the relevant DNA species using Lipofectamine reagent. For transfection, 3 μg of DNA was mixed with 10 μl of Lipofectamine in 0.2 ml of Opti-MEM (Life Technologies Inc.) and was incubated at room temperature for 30 min prior to the addition of 1.6 ml of Opti-MEM. Cells were exposed to the Lipofectamine/DNA mixture for 5 h and 2 ml of 20% (v/v) newborn calf serum in DMEM was then added. Cells were harvested 48-72 h after transfection.

Preparation of Membranes

Plasma membrane-containing P2 particulate fractions were prepared from cell pastes frozen at −80° C. after harvest. All procedures were carried out at 4° C. Cell pellets were resuspended in 1 ml of 10 mM Tris-HCl and 0.1 mM EDTA, pH 7.5 (buffer A) and by homogenisation for 20 s with a polytron homogeniser followed by passage (5 times) through a 25-guage needle. Cell lysates were centrifuged at 1,000 g for 10 min in a microcentrifuge to pellet the nuclei and unbroken cells and P2 particulate fractions were recovered by microcentrifugation at 16,000 g for 30 min. P2 particulate fractions were resuspended in buffer A and stored at −80° C. until required. Protein concentrations were determined using the bicinchoninic acid (BCA) procedure (Smith et al., 1985) using BSA as a standard.

High Affinity [35S]GTPγS Binding

Assays were performed in 96-well format using a method modified from Wieland and Jakobs, 1994. Membranes (10 mg per point) were diluted to 0.083 mg/ml in assay buffer (20 mM HEPES, 100 mM NaCl, 10 mM MgCl2, pH7.4) supplemented with saponin (10 mg/l) and pre-incubated with 40 mM GDP. Various concentrations of GABA were added, followed by [35S]GTPgS (1170 Ci/mmol, Amersham) at 0.3 nM (total vol. of 100 ml) and binding was allowed to proceed at room temperature for 30 min. Non-specific binding was determined by the inclusion of 0.6 mM GTP. Wheatgerm agglutinin SPA beads (Amersham) (0.5 mg) in 25 ml assay buffer were added and the whole was incubated at room temperature for 30 min with agitation. Plates were centrifuged at 1500 g for 5 min and bound [35S]GTPgS was determined by scintillation counting on a Wallac 1450 microbeta Trilux scintillation counter.

Measurement of cAMP Levels

24 hours following transfection, each 60 mm dish of HEK293T cells was split into 36 wells of a 96-well plate and the cells were allowed to reattach overnight. Cells were washed with PBS and pre-incubated in DMEM medium containing 300 μM IBMX for 30 minutes at 37° C. Forskolin (50 μM) and varying concentrations of GABA were added and cells incubated for a further 30 min prior to CAMP extraction with 0.1M HCl for 1 h at 4° C. Assays were neutralised with 0.1 M KHCO3 and CAMP levels determined using scintillation proximity assays (Biotrak Kit, Amersham).

Flow Cytometric Analysis

HEK293T cells were transiently transfected with cDNA as described. 48-72 h following transfection, cells were recovered and washed twice in PBS supplemented with 0.1% (w/v) NaN3 and 2.5% (v/v) foetal calf serum Cells were resuspended in buffer and incubated with primary antibodies 9E10 (c-Myc) or 12CA5 (HA) for 15 min at room temperature. Following three further washes with PBS, cells were incubated with secondary antibody (sheep anti-mouse Fab2 coupled with fluorescein isothiocyanate (FITC)) diluted 1:30 for 15 min at room temperature. For permeabilised cells, a Fix and Perm kit (Caltag) was used. Cell analysis was performed on a Coulter Elite flow-cytometer set up to detect FITC fluoresence. 30,000 cells were analysed for each sample.

Immunological Studies

Antiserum 501 was raised against a synthetic peptide corresponding to the C-terminal 15 amino acids of the GABAB-R1 receptor and was produced in a sheep, using a conjugate of this peptide and keyhole limpet hemocyanin (Calbiochem) as antigen. Membrane samples 30-60 μg) were resolved by SDS-PAGE using 10% (w/v) acrylamide. Following electrophoresis, proteins were subsequently transferred to nitrocellulose (Hybond ECL, Amersham), probed with antiserum 501 at 1:1000 dilution and visualised by enhanced chemiluminescence (ECL, Amersham). Epitope tags were visualised by immunoblotting with anti-Myc (9E10; 1:100 dilution) or anti-HA (12CA5; 1:500) monoclonal antibodies.


Enzymatic removal of asparagine-linked (N-linked) carbohydrate moieties with endoglycosidases F and H was performed essentially according to manufacturers' instructions (Boehringer Mannheim) using 50 μg of membrane protein per enzyme reaction. GABAB receptor glycosylation status was studied following SDS-PAGE/immunoblotting of samples.

Immunoprecipitation Procedures

Transiently transfected HEK293T cells were harvested as described above from 60 mm culture dishes. Cells from each dish were resuspended in 1 ml of 50 mM Tris-HCl, 150 mM NaCl, 1% (v/v) Nonidet® P40, 0.5% (w/v) sodium deoxycholate, pH 7.5 (lysis buffer) supplemented with Completes protease inhibitor cocktail tablets (1 tablet/25 ml) (Boehringer Mannheim). Cell lysis and membrane protein solubilisation was achieved by homogenisation for 20 seconds with a polytron homogeniser, followed by gentle mixing for 30 min at 4° C. Insoluble debris was removed by microcentrifugation at 16,000 g for 15 min at 4° C. and the supernatant was pre-cleared by incubating with 50 μl of Protein A-agarose (Boehringer Mannheim) for 3 h at 4° C. on a helical wheel to reduce non-specific background. Solubilised supernatant was divided into 2×500 μl aliquots and 20 μl of either HA or Myc antisera was added to each. Immunoprecipitation was allowed to proceed for 1 h at 4° C. on a helical wheel prior to the addition of 50 μl of Protein A-agarose suspension. Capture of immune complexes was progressed overnight at 4° C. on a helical wheel. Complexes were collected by microcentrifugation 12,000 g for 1 min at 4° C. and supernatant was discarded. Beads were washed by gentle resuspension and agitation sequentially in 1 ml of 50 mM Tris-HCl, pH 7.5, 500 mM NaCl, 0.1% (v/v) Nonidet® P40 and 0.05% (w/v) sodium deoxycholate followed by 1 ml of 50 mM Tris-HCl, pH 7.5, 0.1% (v/v) Nonidet® P40 and 0.05% (w/v) sodium deoxycholate. Immunoprecipitated proteins were released from Protein A-agarose by incubation in 30 μl of SDS-PAGE sample buffer at 70° C. for 10 min and analysed by SDS-PAGE followed by immunoblotting.

Binding Assays

Competition binding assays were performed in 50 mM Tris HCl buffer (pH7.4) containing 40 μM isoguvacine (Tocris Cookson) to block rat brain GABAA binding sites. P2 membrane preparations were made from HEK293T cells transfected using conditions described above. Increasing concentrations of GABA were added to displace the antagonist [3H]-CGP 54626 (Tocris Cookson, 40 Ci/mmol). Assay conditions were 0.4-0.6 nM [3H]-CGP54626, incubated with 50 μg/tube crude rat brain ‘mitochondrial’ fractions or 25 μg/tube HEK293T P2 membranes at room temperature for 20 minutes. The total volume per tube was 0.5 ml and non specific binding was determined using 1 mM GABA. Bound ligand was recovered using a Brandel 48 well harvester onto GF/B filters (Whatman) and measured by liquid scintillation using a Beckman LS6500 counter.


  • Bettler, B., Kaupmann, K and Bowery, N. Curr Opin in Neurobiol 8: 345-350 (1998)
  • Bittiger, H., Froestl, W., Gentsch, C., Jaekel., J., Mickel, S. J., Mondori, C., Olpe, H. R., Schmuz, M. (1996) in GABA: receptors, transporters and metabolism Ed: C. Tanaka and N. G. Bowery. Birkhauser Verlag Basel Switzerland.
  • Bittiger, H., Reymann, N., Froestl, W. & Mickel, S. J. Pharmacology Communications 2:1-2: 23 (1992)
  • Bowery, N. G., Hudson, A. L., Price, G. W. Neuroscience 20: 365-383 (1987)
  • Chou, K. C. and Heinrikson, R. L. J. Protein Chem 16: 765-773 (1997)
  • Clontech Laboratories. CLONTECH MATCHMAKER™ GAL4 Two Hybrid system User Manual, Protocol PT3061-1 Palo Alto, Calif. (1996)
  • Cvejic, S. and Devi, L. A. J. Biol Chem 272: 26959-26964 (1997)
  • Fuller, K. J., Morse, M. A., White, J. H. M., Dowell, S. J. and Sims, M. J. BioTechniques 25: 85-92 (1998)
  • Gemignani, A., Paudice, P., Bonanno, G., Raiteri, M. Mol Pharmacol 46: 558-562 (1994)
  • Harayama, N., Shibuya, I., Tanaka, K., Kabashima, N., Ueta, Y., Yamashita, H. J. Physiol (Lond) 509: 371-383 (1998)
  • Harper, J. W., Adami, G. R., Wei, N. Keyomarsi, K. and Elledge, S. J. Cell 75: 805-816 (1993)
  • Harshmann, K. D., MoyeRowley, W. S. and Parker, C. S: Cell 53: 321-330 (1988)
  • Hebert, T. E. and Bouvier, M. Biochem. Cell. Biol. 76: 1-11 (1998)
  • Hebert, T. E., Moffett, S., Morello, J. P., Loisel, T. P., Bichet, D. G., Barret, C., and Bouvier, M. J. Biol Chem 271: 16384-16392 (1997)
  • Hill, D. R. and Bowery, N. G. Nature (Lond) 290: 149-152 (1981)
  • Kaprivinsky, G., Gordon, E. A., Wickman, K., Velimirovic, B., Kaprivinsky, L. and Clapham, D. E. Nature (Lond) 374:135-141 (1995)
  • Kaupmann, K., Huggel, K., Heid, J., Flor, P. J., Bischoff, S., Mickel, S-J., McMaster, G., Angst, C., Bittiger, H., Froesti, W. and Bettler, B. Nature(Lond) 386, 239-246 (1997)
  • Kerr, D. I., Humeniuk, R. E., and Ong, J. Eur. J. Pharmacol. 262: 189-192 (1994)
  • Kerr, D. I. and Ong, J. Pharmacol. Ther. 67: 187-246 (1995)
  • Kerr, D. I. and Ong, J. DDT 1: 371-380 (1996)
  • Kobrinsky, E. M., Pearson, H. A. and Dolphin, A. C. Neuroscience 58: 539-552 (1994).
  • Kovoor, A., Nappey, V., Kieffer, B. L. and Chavkin, C. J. Biol. Chem. 272: 27605-27611. (1997)
  • Kubo, Y., Reuveny, E., Slesinger, P. A., Jan, Y. H. and Jan, L. Y. Nature (Lond) 364: 802-806. (1993)
  • Lovinger, D. M., Harrison, N. L. and Lambert, N. A. Eur. J. Pharmacol. 211: 337-341 (1992)
  • Lupas, A. Trends in Biol. Sci. 21: 375-382 (1996)
  • Malcangio, M. and Bowery, N. G. Clin Neuropharmacol 18: 285-305 (1995)
  • McLatchie, L. M., Fraser N. J., Main, M. J. Wise A., Brown, J., Thompson, N., Solari, R., Lee, M. G and Foord, S. M. Nature (Lond) 393: 333-339 (1998)
  • Menon-Johansson, A. S., Berrow, N. and Dolphin, A. C. Pflugers Arch 425: 335-343 (1993)
  • Ohmori, Y., Hirouchi, M., Taguchim J. and Kuriyama, K. J. Neurochem 54: 80-85 (1990)
  • Ong, J., Kerr, D. I., Doolette, D. J., Duke, R. K., Meweft, K. N., Allen, R. D. and Johnston, G. A. Eur J Pharmacol 233:169-172 (1993).
  • O'Hara, P. J., Sheppard, P. O., Thogersen, H., Venezia, D., Haldeman, B. A., McGrane, V., Houamed, K. M., Thomsen, C., Gilbert, T. L. and Mulvihill, E. R. Neuron 11: 41-52 (1993)
  • Raiteri, M., Bonanno, G., Gemignani, A., Pende, M., Vellebuona, F. and Lanza, M. Adv. Biochem. Psychopharmacol. 47: 205-216 (1992)
  • Romano, C., Yang, W-L. and O'Malley, K. L. J. Biol. Chem. 271: 28612-28616 (1996).
  • Sambrook, J., Fritsch, E. F. and Maniatis, T. Molecular Cloning: a Laboratory Manual. 2nd Edition. CSH Laboratory Press. (1989)
  • Sherman, F. Methods Enzymol. 194:3-21 (1991)
  • Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J. and Klenk, D. C. Anal. Biochem. 150: 76-85 (1985)
  • Takahashi, K, Tsuchida, K., Tanabe, Y., Masu, M. and Nakanishi, S. J. Biol. Chem. 268: 19341-19345 (1993)
  • Uezono, Y., Bradley, J., Min, C., McCarty, N. A., Quick, M., Riordan, J. R., Chavkin, C., Zinn, K., Lester, A. and Davidson, N. Receptors and Channels 1: 233-241 (1993)
  • Wotta, D. R., Birnbaum, A. K., Wilcox, G. L., Elde, R. and Law, P. Y. Brain Res. Mol. Brain Res. 44: 5565 (1997)