ABSTRACT
Lioresal1, a racemic mixture of b-p-chlorophenyl-GABA (baclofen), a lipophilic, brain-penetrating analog of GABA, was first marketed in the beginning of the 1970s as a muscle-relaxant drug used in the treatment of spasticity arising from multiple sclerosis or spinal injury. It was only much later that baclofen was found to interact stereospecifically with a novel type of GABA receptor. In fact, its inhibitory effects on the release of different neurotransmitters in distinct tissue preparations were found to be bicucullineinsensitive, whereas the bicuculline-sensitive GABA receptor agonist 3aminopropane sulphonic acid did not affect neurotransmitter release (3). The existence of this novel, distinct ‘‘GABAB’’ receptor was soon thereafter confirmed in radioligand-binding experiments (4). In the subsequent decades, a great number of useful pharmacological tools, potent and selective agonists and antagonists, for the GABAB receptor became available [for review, see Bowery et al. (2)]. Also, a multitude of biochemical and physiological studies made it clear that this receptor, like the GABAA receptor, was of inhibitory nature, but unlike the former, it coupled to G-proteinmediated intracellular pathways. However, while the existence and the properties of this GABAB receptor were well established on pharmacological and physiological grounds, its molecular entity remained elusive for a long time. It was not until 1997 that a complementary DNA, encoding a quite large (ca 100 kDa) GABAB receptor protein, was successfully cloned for the first time (5). It was found to resemble metabotropic glutamate receptors (mGluRs) and exist in two distinct variants, named GABAB(1a) (containing 960 amino acids) and GABAB(1b) (844 amino acids), which differ in the presence or the absence, respectively, of an N-terminal complement protein sequence (Sushi repeats). These isoforms are not splice variants, but arise from the initiation of transcription at different sites, under the control of distinct promoters. However, true splice variants of the GABAB(1) protein have also been shown to exist [summarized in Refs. (6-8)]. Surprisingly, although heterologous expression of either GABAB receptor protein allowed measuring the binding of high-affinity antagonist radioligands, a considerably lower binding affinity for agonists compared to native receptors was found, and it did not make possible the measurement of robust functional responses. This finding remained enigmatic until several groups reported at the same time the cloning of a second GABAB receptor protein (GABAB2) containing 941 amino acids (110 kDa) (9-12). GABAB(1) and GABAB(2) are derived from different genes and share 35% sequence identity and about 50% similarity. The coexpression of GABAB(1) and GABAB(2) to form heterodimeric assemblies, a unique feature among Gprotein-coupled receptors (GPCRs), is a prerequisite to form functional
GABAB receptor entities. This breakthrough discovery has fostered intense research efforts in recent years, which have greatly increased our knowledge of the structure and function of the GABAB receptor (8,13).
