Hanns Möhler

Decreased GABAA-receptor clustering results in enhanced anxietyand a bias for threat cues

Patients with panic disorders show a deficit of GABAA receptors in the hippocampus, parahippocampus and orbitofrontal cortex. Synaptic clustering of GABAA receptors in mice heterozygous for the γ2 subunit was reduced, mainly in hippocampus and cerebral cortex. The γ2+/– mice showed enhanced behavioral inhibition toward natural aversive stimuli and heightened responsiveness in trace fear conditioning and ambiguous cue discrimination learning. Implicit and spatial memory as well as long-term potentiation in hippocampus were unchanged. Thus γ2+/– mice represent a model of anxiety characterized by harm avoidance behavior and an explicit memory bias for threat cues, resulting in heightened sensitivity to negative associations. This model implicates GABAA-receptor dysfunction in patients as a causal predisposition to anxiety disorders.

GABAB-receptor splice variants GB1a and GB1b in rat brain: developmental regulation, cellular distribution and extrasynaptic localization

GABAB (γ‐aminobutyric acid)‐receptors have been implicated in central nervous system (CNS) functions, e.g. cognition and pain perception, and dysfunctions including spasticity and absence epilepsy. To permit an analysis of the two known GABAB‐receptor splice variants GABAB‐R1a (GB1a) and GABAB‐R1b (GB1b), their distribution pattern has been differentiated in the rat brain, using Western blotting and immunohistochemistry with isoform‐specific antisera. During postnatal maturation, the expression of the two splice variants was differentially regulated with GB1a being preponderant at birth. In adult brain, GB1b‐immunoreactivity (‐IR) was predominant, and the two isoforms largely accounted for the pattern of GABAB‐receptor binding sites in the brain. Receptor heterogeneity was pronounced in the hippocampus, where both isoforms occurred in CA1, but only GB1b in CA3. Similarly, in the cerebellum, GB1b was exclusively found in Purkinje cells in a zebrin‐like pattern. The staining was most pronounced in Purkinje cell dendrites and spines. Using electron microscopy, over 80% of the spine profiles in which a synaptic contact with a parallel fibre was visible contained GB1b‐IR at extrasynaptic sites. This subcellular localization is unrelated to GABAergic inputs, indicating that the role of GABAB‐receptors inu2003vivo extends beyond synaptic GABAergic neurotransmission and may, in the cerebellum, involve taurine as a ligand.

Mechanism of action of the hypnotic zolpidem in vivo

Zolpidem is a widely used hypnotic agent acting at the GABAA receptor benzodiazepine site. On recombinant receptors, zolpidem displays a high affinity to α1‐GABAA receptors, an intermediate affinity to α2‐ and α3‐GABAA receptors and fails to bind to α5‐GABAA receptors. However, it is not known which receptor subtype is essential for mediating the sedative‐hypnotic action in vivo. Studying α1(H101R) mice, which possess zolpidem‐insensitive α1‐GABAA receptors, we show that the sedative action of zolpidem is exclusively mediated by α1‐GABAA receptors. Similarly, the activity of zolpidem against pentylenetetrazole‐induced tonic convulsions is also completely mediated by α1‐GABAA receptors. These results establish that the sedative‐hypnotic and anticonvulsant activities of zolpidem are due to its action on α1‐GABAA receptors and not on α2‐ or α3‐GABAA receptors.

GABAA-receptor subtypes: a new pharmacology

The GABA(A) receptor is a pluripotent drug target mediating anxiolytic, sedative, anticonvulsant, muscle relaxant and amnesic activity. These drug actions have now been attributed to defined receptor subtypes. Thus, precise guidelines are available for the development of novel drugs with more selective action and less side effects than those currently in clinical use.

The rise of a new GABA pharmacology

Key developments in GABA pharmacology over the last 30 years are reviewed with special reference to the advances pioneered by Erminio Costa. His passion for innovative science, and his quest for novel therapies for psychiatric disorders are particularly apparent in his fundamental contributions to the field of GABA research, with a focus on anxiety disorders and schizophrenia. He was a cofounder of the GABAergic mechanism of action of benzodiazepines. He envisaged partial agonists as novel anxiolytics. He identified DBI (diazepam binding inhibitor) as endogenous agonist of neurosteroidogenesis with multiple CNS effects and he pointed to the developmental origin of GABAergic dysfunctions in schizophrenia through his discovery of a reelin deficit, all this in collaboration with Sandro Guidotti. Today, the GABA pharmacology comprises selective hypnotics, non-sedative anxiolytics, memory enhancers and powerful analgesics. This article is part of a Special Issue entitled Trends in neuropharmacology: in memory of Erminio Costa.

Resolving differences in GABAA receptor mutant mouse studies.

1059 TO THE EDITOR—The validity of genetically altered mice as models for disease or for drug target identification relies on the reproducibility of behavioral test results. In a recent paper in Nature Neuroscience1, behavioral tests were described on α1(H101R) mice carrying a histidine-toarginine point mutation in the α1 subunit of the GABAA receptor. An accompanying News and Views article2 discussed discrepancies between results of behavioral experiments for this study1 and our study of α1(H101R) mice3. The two lines of mice seemed to differ in their drug-induced behavior, although both had been constructed with the same point mutation. Now we report that these discrepencies were caused by differences in the behavioral protocols used by the two groups, not by differences in the mouse lines. Both studies compared the behavioral effects of diazepam in wild-type and pointmutated α1(H101R) mice whose α1 GABAA receptor subtype was made insensitive to diazepam by a histidine-to-arginine substitution. We assessed the sedative action of diazepam on spontaneous motor activity of mice familiar with the testing environment3. Diazepam produced sedation in wild-type mice, as measured by the dose-dependent decrease in the extent of spontaneous motor activity, whereas diazepam failed to impair the spontaneous motor activity in α1(H101R) mice up to 30 mg per kg (ref. 3). These results led us to conclude that the sedative action of diazepam as measured by this protocol is mediated by α1 GABAA receptors. In contrast, McKernan and colleagues measured locomotor activity in mice exploring an unfamiliar environment1. The authors reported that diazepam (3 mg per kg orally) did not affect wild-type mice, but induced a significant increase in locomotor activity in the α1(H101R) mice. This has been interpreted as “paradoxical hyperactivity”2 and “reduction in neophobia”1 or “reduction in GABA-mediated neuronal inhibition”1. Following transfer to a new testing room 30 minutes before drug treatment, our α1(H101R) mice also exhibit hyperactivity (Fig. 1). These findings, which correspond to McKernan and colleagues’ results1, suggest that the failure of diazepam between the α1(H101R) mice generated by the two laboratories1,3. This shows that taking environmental and technical details of test procedures into account may help to resolve interlaboratory differences in results.

Role of GABAA receptors in cognition

Complex brains have developed specialized mechanisms for the grouping of principal cells into temporal coalitions of local or distant networks: the inhibitory interneuron clocking networks. They consist of GABAergic (where GABA is gamma-aminobutyric acid) interneurons of a rich diversity. In cortical circuits, these neurons control spike timing of the principal cells, sculpt neuronal rhythms, select cell assemblies and implement brain states. On the basis of these considerations, the deficits in cognition, emotion and perception in psychiatric disorders such as anxiety, depression or schizophrenia are considered to manifest themselves through a dysregulation of the inhibitory interneuron clocking network as a final common denominator, irrespective of the diverse underlying disease pathologies. The diversity of GABAergic interneurons is paralleled by a corresponding diversity of GABA(A) receptors in network regulation. The region-, cell- and domain-specific location of these receptor subtypes offers the possibility to gain functional insights into the role of behaviourally relevant neuronal circuits. Using genetic manipulation, the regulation of anxiety behaviour was attributed to neuronal circuits characterized by the expression of alpha(2)-GABA(A) receptors. Neurons expressing alpha(3)-GABA(A) receptors, located mainly in aminergic and basal forebrain cholinergic neurons, were related to a hyperdopaminergic phenotype, typical of schizophrenic symptoms. Temporal and spatial memory were selectively modulated by extrasynaptic alpha(5)-GABA(A) receptors. Chronic pathological pain was under the regulation of spinal and cortical alpha(2)- (and alpha(3)-) GABA(A) receptors. Thus the relevance of the diversity of inhibitory GABA(A) receptor subtypes for the regulation of cognition, emotion and memory is increasingly being recognized. The clinical proof-of-concept of a subtype-specific pharmacology is most advanced for the alleviation of cognitive dysfunctions in schizophrenia, based on the treatment of patients with an alpha(2)/alpha(3)-GABA(A) receptor ligand.

GABAB-receptor isoforms: Molecular architecture and distribution

The slow component of GABAergic inhibition in the brain is mediated by the metabotropic GABA(B)-receptors. Most if not all GABA(B)-receptors are heterodimers of GABA(B)R1 (GBR1) and GABA(B)R2 (GBR2) proteins. Distinctive receptor isoforms are based on the presence of two GBR1 splice variants termed GBR1a and GBR1b. Both were found to be associated with GBR2 suggesting that the isoforms GBR1a/GBR2 and GBR1b/GBR2 represent the vast majority of GABA(B)-receptors in the brain. The two isoforms differed strikingly in their pattern of expression on the regional, cellular and subcellular level. These results point to distinct funcional roles of the two receptor isoforms.

Cognitive enhancement by pharmacological and behavioral interventions: the murine Down syndrome model

The cognitive deficits in Down syndrome (DS) are attributed to an excessive hippocampal inhibition, which obstructs neuronal plasticity and normal learning and memory, a view which is largely based on studies of Ts65Dn mice, the best characterized mouse model of DS. The cognitive behavioral deficits of Ts65Dn mice can be rescued by reducing GABAergic inhibition, most selectively by partial inverse agonists acting on α(5) GABA-A receptors, of which one compound has recently entered clinical trials in DS. Most remarkably, the improved cognitive performance of Ts65Dn can persist for weeks and months after cessation of drug treatment, as demonstrated for the non-specific GABA antagonist pentylenetetrazole. The Alzheimer drugs, memantine and donepezil largely fail to show any benefit. Finally, repeated non-invasive sensory stimulation such as over-training or enriching the environment, are able to enhance the learning performance which underlines the reversibility of an obstructed neuronal plasticity in Ts65Dn mice.

Highly potent modulation of GABA(A) receptors by valerenic acid derivatives.

g-Aminobutyric acid (GABA) is the most abundant inhibitory neurotransmitter in the mammalian CNS, with as many as 30 % of synapses being GABA-ergic in nature. While two distinct receptor types for GABA exist (GABAA and GABAB) many of its important physiological effects are mediated through interactions with ionotropic GABAA receptors, which largely determine the timing of neuronal firing and the sculpting of neuronal oscillations. In light of the central role of the GABA-ergic system in the human brain, it is not surprising that drugs enhancing the GABA-induced chloride ion flux at GABAA receptors play a major role in the treatment of a variety of CNS-related disorders, such as generalized anxiety and panic disorders or sleep disturbances. 2] However, current GABAA modulators, most of which are benzodiazepines (BZ), are generally nonselective and act indiscriminately on all BZ sensitive GABAA receptor subtypes. 3] In contrast, based on recent advances in our understanding of GABAA receptor physiology and pharmacology, subtype selective agents would be expected to exhibit a more selective therapeutic profile with fewer side effects. 4] Thus, the identification of new types of lead structures for GABAA receptor modulation represents an important objective in drug discovery directed at different types of CNS-based disorders, in particular in the area of anxiety. 4] In this context we and others have recently reported that valerenic acid (1), which is a major constituent of common valerian (Valeriana officinalis), is a potent modulator of GABAA receptors expressed in Xenopus oocytes or HEK293 cells with EC50 values in the 5–20 mm range; concentrations >30 mm were found to lead to direct receptor activation. Valerenic acid (1), at high mm concentrations, was found to enhance GABA-induced ion currents in oocyte and HEK293 cell membranes up to fivefold and > tenfold, respectively. While the binding of 1 to GABAA receptors occurs with nm binding constants, its binding site has not been characterized in detail. It is clear, however, that no direct interaction occurs with the BZ binding site; 6] rather, 1 may exploit the site for loreclezole or a hitherto unrecognized site that is allosterically linked to binding sites for loreclezole, anesthetics and mefenamic acid. In addition to these in vitro effects, both 1 as well as its natural congener valerenol (3) have been shown to exhibit anxiolytic activity in vivo either after i.p. or p.o. administration. Collectively, these findings suggest that 1 is an attractive lead structure for the development of new GABAA receptor modulators. At the same time, no structure–activity relationship (SAR) has yet been established for the GABAA-modulatory activity of 1, except that 3 has been shown to exhibit comparable activity to 1, while valerenal (2), which is a minor constituent of V. officinalis, and the methyl ether of 3 (11, vide infra) do not bind to GABAA receptors. [5] In order to develop a broader understanding of the structural requirements for GABAAmodulatory activity by 1 and related derivatives, we have embarked on a program to synthesize and biologically evaluate analogues of 1 with the ultimate objective to identify more potent and also subtype specific allosteric modulators at the GABAA receptor. In this report we describe a first series of analogues of 1, which serve to probe the potential for modifications of the negatively charged carboxyl group of 1 and to assess the importance of the side chain double bond and the C11 methyl group for biological activity. As a result of this initial work we have identified a number of compounds that modulate GABA-induced ion currents with substantially higher potency than 1 itself. Starting from valerenic acid (1) we have prepared valerenol (3) through direct reduction with LAH (as previously described). 8] Swern oxidation of 3 then provided valerenal (2). 8] DCC/DMAP-mediated coupling of 1 with ammonia, methylamine, or dimethylamine gave the corresponding amides 6–8 in excellent yields (80–100 %; Scheme 1). Treatment of primary amide 6 with oxalyl chloride/DMF in THF provided nitrile 9 (Scheme 1), which could be further elaborated into tetrazole 10 in 64 % yield by heating with in situ prepared Bu3SnN3 [11] at 100 8C for 6 days. A number of azide sources were investigated for the cycloaddition reaction with 7, but Bu3SnN3 proved to be the most efficient with regard to yield and purity of the product, even though the reaction was very slow and took several days to go to comple[a] Dipl.-Chem. S. Kopp, Prof. Dr. H. Mçhler, Prof. Dr. K.-H. Altmann Swiss Federal Institute of Technology (ETH) Z rich Department of Chemistry and Applied Biosciences Institute of Pharmaceutical Sciences, HCI H405 Wolfgang-Pauli-Str. 10, 8093 Z rich (Switzerland) Fax: (+ 41) 44-6331369 E-mail : karl-heinz.altmann@pharma.ethz.ch