GABA a AGONISTS AS TARGETS FOR DRUG DEVELOPMENT

Clinical and Experimental Pharmacology and Physiology (1992) 19,73-78

GABAAAGONISTS AS TARGETS FOR DRUG DEVELOPMENT Graham A. R. Johnston

The Adrien Albert Laboratory of Medicinal Chemistry, Department of Pharmacology, The University of Sydney, New South Wales, Australia (Received 9 August 1991)

SUMMARY 1. Agents with selective actions on bicuculline-sensitive GABAAreceptors have been developed by systematically restricting the conformational mobility of the GABA molecule. 2. THIP, a bicyclic isoxazole that represents GABA held in a relatively rigid and partially extended conformation, is an analgesic of potency comparable to that of morphine. THIP represents a lead compound for a novel series of analgesics acting independently of Naloxone-sensitive opiate systems. 3. ZAPA, a conformationally-restricted analogue of GABA containing an isothiouronium moiety, is a selective agonist for low affinity GABAA receptors and is a lead compound for the development of a novel series of anthelmintics. 4. (+)-TACP, a cyclopentane analogue of GABA, may activate a different subclass of GABAA receptors from THIP. 5 . Pharmacological, molecular modelling and molecular biological studies provide evidence for a heterogeneity of GABAAreceptors which might be exploited for drug development.

Key words: conformational restriction, drug development, GABA, ( -k )-TACP, THIP, ZAPA.

INTRODUCTION Receptors for the inhibitory neurotransmitter GABA ( y-aminobutyric acid) are divided into GABAA

receptors, which are antagonized by the convulsant alkaloid bicuculline, and GABABreceptors, which are activated by the muscle relaxant Baclofen, according to the classification of Hill and Bowery (1981). GABAA receptors are widely distributed in the central nervous system, being present on most, if not all, neurones in all brain regions. Great progress is being made in the molecular biology of the cDNA encoding the protein subunits of GABAA receptors. Some I5 genes have

been described t o date that can code for proteins in the pentameric oligomer structure proposed for GABAA receptors; different combinations of the protein subunits could produce a rich diversity in these receptors (see Duggan & Stephenson 1990). GABAA receptors have been implicated in a wide range of neurological conditions and disorders including anxiety, analgesia, convulsions, coma, dementia, epilepsy, hypertension and schizophrenia and in the action of a variety of centrally active drugs such as benzodiazepines, barbiturates and steroid

Correspondence: Professor Graham A. R. Johnston, The Adrien Albert Laboratory of Medicinal Chemistry, Department of Pharmacology, The University of Sydney, NSW 2006, Australia. This paper was presented at a Symposium in honour of Professor Alan L. A. Boura entitled ‘Academic and Industrial Pharmacology: Past, Present and Future Philosophies for Novel Drug Discovery‘ held at Monash University on 28 June 1991.

G. A. R. Johnston

74

anaesthetics. Agents acting on GABAAreceptors are targets for drug development, and the increasing evidence for heteogeneity of GABAAreceptors indicates that agents selective for subtypes of GABAA receptors can be developed. This article reviews progress in the development of some directly acting GABAAagonists.

GABA ANALOGUES OF RESTRICTED C 0 N FO RM AT1 ON The GABA molecule has considerable conformational flexibility as a result of free rotation around the single bonds as illustrated in Fig. 1. There is very little difference in free energy between the most favoured fully extended conformation and the least favoured fully folded conformation, such that these extreme conformations and a range of intermediate conformations are all available for binding to receptors. Different conformations of GABA may well interact with different receptors giving a complex function to what is a relatively simple molecule. The need to investigate GABA analogues of restricted conformation was recognized in the very early pharmacological studies of this inhibitory neurotransmitter. Purpura ef al. (1959) commented that 'Studies are underway with

substances that have fewer degrees of rotational freedom such as unsaturated w-amino acids'. Some 20 years later, after the design, synthesis and evaluation of a range of unsaturated and cyclic analogues of GABA, it was possible to review in a systematic way the structure and activity of analogues of restricted conformation (Johnston et aZ. 1979). The incorporation of unsaturation, ring structures or both, into the basic GABA molecule could produce compounds having more selective actions than that of GABA itself, not only on neuronal receptors but also on GABA enzymes and transport systems. There are 10 possible ways of joining atom pairs of the three carbon atoms, the amino nitrogen and the carboxyl group of GABA so as to restrict conformational mobility, these are illustrated in Fig. 2 which has become known as the 'Happy GABA Man'. Adjacent atoms are joined by rings, double or triple bonds in type a, b, c and d analogues, alternate atoms in type e, f and g analogues, while atoms separated by two or three carbon atoms are joined in type i, j or k analogues. In addition to restricting conformation by joining atom pairs, suitable substituents on the carbon atoms of GABA can restrict rotation, as in the GABAB agonist Baclofen which is the P-p-chlorophenyl derivative of GABA.

A NEW CLASS OF GABA AGONIST intermediate conformations H2N?&?$7%02?N~C02H

fully extended

H2N

H2u02H

Three conformationally restricted analogues of

GABA, whose structures are shown in relationship to dco2H GABA in Fig. 3, are particularly important. Muscimol, fully folded

Fig. 1. Different conformations of the GABA molecule produced by bond rotation.

a psychoactive isoxazole from the mushroom Amanita muscaria, is a very potent GABAAagonist but it is not selective as it has weak GABABagonist action and is a

GABA

w Fig. 2. Ten ways of joining key atoms pain in the GABA molecule to produce analogues of restricted conformation.

ISOGUVACINE

MUSCIMOL

THIP

Fig. 3. Structural development of isoguvacine and from muscimol and GABA.

TBIP

GABAAagonists and drug development

weak substrate for GABA uptake systems and for GABA aminotransferase. Muscimol is a type f GABA analogue of restricted conformation with a 3-hydroxyisoxazole moiety acting as a ‘masked carboxyl‘ group. As indicated, the aminomethyl substituent is free to rotate with respect to the plane of the isoxazole ring; incorporating this substituent into a second ring yielded THIP, a type h GABA analogue, the lead compound for what was described as a new class of GABA agonist (Krogsgaard-Larsen et al. 1977). Isoguvacine, a type i GABA analogue, was the most potent GABAA agonist of this series that was characterized by the presence of a secondary rather than a primary amino function. Until this series of compounds, potent GABAA agonists were characterized by a primary amino function or its equivalent. The favourable conformation of the new class of GABAA agonist was apparently able to overcome the unfavourable substitution on the amino function of GABA. THIP and isoguvacine were described as ‘relatively rigid‘ analogues of GABA and they had more selective actions than either GABA or muscimol. As bicucullinesensitive inhibitors of the firing of neurones in the cat spinal cord, isoguvacine was equipotent to muscimol and 2-4 times more potent than THIP and GABA; neither isoguvacine nor THIP influenced GABA uptake into brain slices or the activity of GABA transaminase (Krogsgaard-Larsen et al. 1977). The selective actions of isoguvacine and THIP as GABAA agonists indicated that GABA interacts with GABAA receptors in a partially extended and almost planar conformation. Isoguvacine has become the GABAA agonist of choice to define GABAAreceptors and is used to occlude GABAAreceptors in binding studies of GABABreceptors (Hill & Bowery 1981).

THIP While the blood-brain barrier limits the CNS effects of systemically administered amino acids such as GABA and isoguvacine, the isoxazoles muscimol and THIP have strong central actions on systemic administration. THIP is somewhat weaker than muscimol as an anticonvulsant but has a potent analgesic action (Hill et a/. 1981; Grognet et al. 1983). THIP is approximately equipotent to morphine as an analgesic and, in contrast to morphine, does not produce respiratory depression. THIP-induced analgesia is insensitive to Naloxone, indicating that it is not acting via Naloxone-sensitive opiate receptors. THIP is active clinically, for example in patients with chronic pain of malignant origin at doses of 5-30 mg intramuscularly (Kjaer & Nielson 1983).

75

These observations with THIP have drawn attention to the likely importance of GABA mechanisms in analgesia and opened up the prospect of the development of a new class of clinically useful analgesics based on selective GABA receptor activation (Krogsgaard-Larsen et al. 1985). Naloxone-insensitive, stress-induced analgesia in rodents has been shown to be associated with a large increase in the availability of forebrain GABAA receptors (Skerritt et al. 1981; Schwartz et al. 1987),which is likely to be the result of alterations in the modulation of GABAAreceptors by corticosteroids that are released from the adrenals during stress (Johnston 1990). THIP represented a novel analgesic of significant therapeutic potential (Krogsgaard-Larsen et al. 1985). Under the name Gaboxadol it went through many years of development by Sandoz before being dropped due to its sedative action. Grognet et al. (1983) reported that, in mice, the analgesic action of THIP is not readily dissociated from its sedative or muscle relaxant properties. Although the conformational restriction of the GABA and muscimol molecules as illustrated in Fig. 3 gave rise to THIP as a relatively rigid GABA analogue which had selective actions, these actions apparently were not sufficiently selective to yield a therapeutically useful agent. Further development of the ‘sons and daughters’ of THIP may yield such agents (Falch et al. 1990). Such development is not confined to isoxazole analogues of GABA and a variety of different approaches are being made to discover sufficiently selective agonists for subtypes of GABAAreceptors.

ZAPA One example of another approach to the design and development of GABAA agonists is a series of isothiouronium derivatives based on the lead compound ZAPA, (Z)-3-[(aminoiminomethyl)thio]prop2-enoic acid (Fig. 4). These compounds incorporate conformational restriction of the GABA molecule via a type c unsaturation, together with a modified amino function, namely the isothiouronium group. ZAPA is a selective agonist for low affinity GABA receptors

H2N\

+;c-s

H&’

A ct-

COOH

Fig. 4. Structure of ZAPA hydrochloride, a selective agonist for low affinity GABAAreceptors.

G. A. R.Johnston

76

which are associated with benzodiazepine receptors (Allan et al. 1986). In addition it is a substrate for the neuronal GABA transport system and could act in GABA replacement therapy (Allan et al. 1991). ZAPA does not cross the blood-brain barrier and a suitable prodrug would have to be developed before any clinically useful CNS agent could emerge from this series of compounds. In another context the inability of ZAPA to cross the mammalian blood-brain barrier is advantageous, as ZAPA has a potent GABA agonist action in nematodes and is regarded as an important lead compound for the design of novel anthelmintics (Holden-Dye & Walker 1988).

(

+ )-TACP

Computer modelling studies indicate structural similarities between conformationally restricted GABAA agonists and the GABAA antagonists bicuculline, consistent with observations that these substances share some common binding sites on brain membranes (Andrews & Johnston 1979). On the basis of these modelling studies it was proposed that the limited range of active conformations of GABA at GABAA receptors could be defined by the ‘bicuculline conformation’, based on the structure of bicuculline, and the ‘muscimol conformation’, based on the structure of muscimol. Within this range of conformations, THIP would adopt a conformation close to the ‘muscimol conformation’, whereas ( 4- )-TACP (Fig. 5 ) would adopt the ‘bicuculline conformation’. (+)-TACP is one of the four possible stereoisomers of the cyclopentane analogue of GABA and a potent bicuculline-sensitive neuronal inhibitor (Allan et af. 1979). It is a type g conformationally restricted analogue of GABA and does not act on GABA enzymes or uptake systems. Recent ligand binding studies indicate that THIP and ( -t)-TACP may each bind to a different subtype of GABAA receptors (Dickenson et al. I990), and thus these two conformationallyrestricted analogues may represent key ligands with which to sort out a pharmacologically

Fig. 5. Structure of (+)-TACP, (+)-truns-(lS,3S)-aminocyclopentane-l-carboxylicacid, a selective GABAAagonist.

based subclassification of GABAA receptors. (+)TACP is an important lead compound for the development of selective GABAAagonists.

GABAPENTIN The anticonvulsant drug, Gabapentin (Fig. 6), is an analogue of GABA in which conformational restriction is achieved via a spiro linkage at the 8-carbon atom of the GABA moiety; this prevents the GABA moiety achieving the fully extended conformation available to GABA itself. Gabapentin appears to be an important new anticonvulsant drug (Schmidt 1989). It does not appear to act directly on any known GABA receptor, enzyme or transport system, and its structural resemblance to GABA could be merely coincidental. More likely it is an important lead to a new class of receptors.

Fig. 6. Gabapentin, an anticonvulsantdrug which is a conformationally restricted analogue of GABA.

GABAA RECEPTOR HETEROGENEITY It is becoming increasingly clear that the GABAA classification of bicucullimensitive GABA receptors describes a variety of receptor subtypes. Hints of this came from pharmacologktl and molecular modelling studies such aa those outlined above and recent molecular biological findings have put this issue beyond doubt. Given the current model of a pentameric collection of proteins constituting GABAA receptors, and that 15 genes have been described thus far coding for these proteins, there could be more than 2000 possible subtypes of GABAA receptors. This rich diversity is bamd on the different combinations of protein condjWnts but it is certain that many other chemical substances make up the overall functional receptor. The lipid environment and various endogenous modulators (peptides, phospholipids, purines and steroids) greatly influence GABA receptor function. It will be a major task for pharmacologists,

77

GABAAagonists and drug development physiologists, medicinal chemists and molecular biologists to sort out the functional relevance of the apparently available diversity of GABAA receptors. This offers a rich field for the development of selective therapeutic agents.

CONCLUSION: LOW MOLECULAR WEIGHT D R U G DEVELOPMENT This review outlines the design and development of simple molecules related to the amino acid neurotransmitter GABA. These molecules represent important leads for the discovery of new therapeutic agents. At the Boden Conference on Biomolecular Design and Development in 1988, Alan Boura presented a paper on ‘The case for increased low molecular weight drug research: A personal philosophy’ (Boura 1988). He pointed out that ‘the most successful novel drugs have been produced by pharmaceutical companies carrying out innovatory fundamental research in nations with strong, continuously interacting, industrial and academic disciplines of medicinal chemistry and pharmacology’. I strongly support this view and would encourage industry to increase its interaction with the strong groups of pharmacologists and medicinal chemists in Australian universities, with a view to developing low molecular weight drugs. There are a number of such groups in Australian universities, for example the Centre for Pharmacology, Medicinal Chemistry and Toxicology at the University of Tasmania; the Centre for Drug Design and Development at the University of Queensland; and the Adrien Albert Laboratory of Medicinal Chemistry at the University of Sydney. A recent article on ‘Laying foundations for an indigenous drug industry’ (Roberts 1991) in Australia gives heavy emphasis to the development of high molecular weight drugs taking advantage of the high profile of biotechnology in Australia. Interestingly, in the UK there is a trend away from biotechnology in the pharmaceutical industry. A report from the UK National Economic Development Council on ‘New life for industry: Biotechnology, industry and the community in the 1990s and beyond’ notes that the products of chemical synthesis can be patented more reliably, administered more easily, and possibly made more powerful than the products of biotechnology (SCRIP 1991). Low molecular weight compounds produced by chemical synthesis have much to offer the pharmaceutical industry and should form a significant part of the Australian initiative in drug development. This review of GABAAagonists shows that Australian initiatives

in this field have made major contributions that may lead to novel drugs with large international markets.

ACKNOWLEDGEMENTS The author is grateful to the NH&MRC for financial support and to Robin Allan (University of Sydney), Peter Andrews (University of Queensland), David Curtis (Australian National University) and Povl Krogsgaard-Larsen (Royal Danish School of Pharmacy) for their collaboration.

REFERENCES Allan, R. D., Dickenson, H. W., Duke, R. K. & Johnston, G, A. R. (199 1) ZAPA, a substrate for the neuronal high affinity GABA uptake system in rat brain slices. Neurochemistry International, 18, 63-67. Allan, R. D., Dickenson, H. W., Hiern, B. P., Johnston, G. A. R. & Kazlauskas, R. (1986) Isothiouronium compounds as y-aminobutyric acid agonists. British Journal of Pharmacology, 88,379-387. Allan, R. D., Johnston, G . A. R. & Twitchin, B. (1979) Synthesis of analogues of GABA. 111. All four stereoisomers of 3-aminocyclo-pentanecarboxylicacid and a stereochemical correlation with amidinomycin. Australian Journal of Chemistry, 32,2517-2521. Andrews, P. R. & Johnston, G. A . P . (1979) GABA agonists and antagonists. Biochemical Pharmacology, 28, 2697-2702, Boura, A. (1988) The case for increased low molecular weight drug research: A personal philosophy. Proceedings of the Boden Conference on Biornolecular Design & Development, 12- 13. Dickenson, H. W., Duke, R. K., Balcar, V. J., Allan, R. D. & Johnston, G. A. R. (1990) Binding to rat brain membranes of ( +)-trans-(lS,3S)-3-aminocyclopentane -1-carboxylic acid, ( )-TACP, a selective GABAA receptor agonist. Molecular Neuropharmacology, 1, I -6. Duggan, M. J. & Stephenson, F. A. (1990) Biochemical evidence for the existence of y-aminobutyrateA receptor iso-oligomers. Journal of Biological Chemistry, 265, 3831-3835. Falch, E., Larsson, 0. M., Schousboe, A. & KrogsgaardLarsen, P. (1990) GABA-A agonists and GABA uptake inhibitors. Drug Development Research, 21, 169-188. Grognet, A., Hertz, F. & DeFeudis, F. V. (1983) Comparison of the analgesic actions of THIP and morphine. General Pharmacology, 14,585-589. Hill, D. R. & Bowery, N. G. (1981) 3H-Baclofen and 3HGABA bind to bicuculline-insensitive GABABsites in rat brain. Nature, 290, 149-152. Hill, R. C., Maurer, R., Buscher, H. H. & Romer, D . (1981) Analgesic properties of the GABA-mimetic THIP. European Journal of Pharmacology, 69, 221 -224.

+

78

G. A. R.Johnston

Holden-Dye, L. & Walker, R. J. (1988) ZAPA, (2)-3[(aminoiminomethyl)thio]-2-propenoic acid hydrochloride, a potent agonist at GABA-receptors on the Ascaris muscle cell. British Journal of Pharmacology, 95, 3-5. Johnston, G. A. R. (1990) Anxiety, steroids and GABA receptors. In: Anxiety (Eds N. McNaughton BE G. Andrews), pp. 210-212. Otago University Press, Dunedin. Johnston, G. A. R., Allan, R. D., Kennedy, S. M. E. & Twitchin, B. (1979) Systematic study of GABA analogues of restricted conformation. In: GABA-Neurotransmitters (Eds P. Krogsgaard-Larsen, J. Scheel-Kruger and H. Kofod), pp. 149-164. Munksgaard, Copenhagen. Kjaer, M. & Nielson, H. (1983) The analgesic effects of the GABA-agonist THIP in patients with chronic pain of malignant origin: A phase 1-2 study. British Journal of Clinical Pharmacology, 16,447-485. Krogsgaard-Larsen, P., Falcb, E. & Hjeds, H. (1985) Heterocyclic analogues of GABA: Chemistry, molecular pharmacology and therapeutic effects. Progress in Medicinal Chemirtry, 22, 67-120.

Krogsgaard-Larsen, P., Johnston, G . A. R., Lodge, D. & Curtis, D. R. (1977) A new class of GABA agonist. Nature, 268,53-55. Purpura, D. P., Guado, M., Smith, T. G., Callan, D. A. & Grundfest, H. (1959) Structure activity determinants of pharmacological effects of amino acids and related compounds on central synapses. Journal of Neurochemistry, 3,238-268. Roberts, P. (1991) Laying foundations for an indigenous drug industry. Today’s Life Science, 2,6-10. Schmidt, B. (1989) Gabapentin. In: Antiepileptic Drugs (Ed. R. Levy), pp. 925-935. Raven Press, New York. Schwartz, R. D., Wess, M. J., Labarca, R., Skolnick, P. & Paul, S. M. (1987) Acute stress enhances the activity of the GABA receptor-gated chloride ion channel in brain. Brain Research, 411,151-155. Skerritt, J. H., Trisdikoon, P. & Johnston, G. A. R. (1981) Increased GABA binding in mouse brain following acute swim stress. Brain Research, 215,398-403. SCRIP (1991) More collaboration sought in UK pharma biotech. SCRIP, 1638,4-5.