Neurophysiology, VoL 28, No. 6, November-December, 1996

Complex Effects of GABA-ergic Agents on Anxiety A. V. K a l u e f f I

Neirofiziologiya/Neurophysiology, Vol. 28, No. 6, pp. 267-272, November-December, 1996.

Received December 12, 1996. The paper reviews the role of GABA-ergic mechanisms in anxiety and summarizes the dam on complex bimodal effects PrOduced by GABA-ergic agents on receptors and behavioral measures of anxiety. Possible physiological mechanisms of such effects on anxiety have been discussed. The paper reviews some paradoxical anxiotropic effects produced by certain GABA-ergic agents in behavioral tests of anxiety. Currently existing traditional views on GABA-ergic mechanisms that underlie anxiety and anxiety-related states are critically re-considered.

10]. The important role of GABA-A receptors in modulation of different forms of anxiety, fears, phobias, and depression has been discussed in many studies [8, 9, I I, 12 ]. Briefly, inhibition of the GABA-ergic system by certain drugs is known to produce severe anxiety revealed by various tests both in humans and in animals, whereas activation of this system generally results in a reduction of anxiety [3 ] (Table 1). The physiology and pharmacology of anxiety are developing rapidly, and more sophisticated models are now being extensively used in the discovery of novel anxiety-active agents [13]. However, recent studies have often demonstrated inconsistent results, including those relating to the dose-dependent effects of the drugs [14 ]. This has resulted in the important observation of Wettstein [15] that in the behavioral pharmacology of both animals and humans complex dose-dependent reactions (when a specific behavior is activated at one dose and suppressed at others) can often be observed. However, the physiological mechanisms of such "unusual" complex behavioral alterations remain unclear. Importantly, this also may include a possibility of complex interactions within other neuronal systems that regulate behavior unrelated to anxiety [16]. Nevertheless, despite inconsistencies in the published data, an analysis of dose-dependent responses produced by different GABA-ergic drugs (or produced by different doses of one drug tested in different models) seems to be of importance for further understanding the neuronal mechanisms that underlie the physiology of anxiety and anxiety-related behaviors (Table 2).

INTRODUCTION Gamma-aminobutyric acid (GABA) is the primary mediator of inhibitory neurotransmission in the mammalian central nervous system (CNS). GABA-A receptors have been found especially in the brain and other nervous tissues. T h e y represent hetero-oligomeric protein complexes consisting of GABA and benzodiazepine (BDZ) receptors coupled to an integral chloride channel [1, 2]. GABA-A-receptor activation increases the chloride conductance and usually inhibits neuronal activity by hyperpolarization of the membrane. Recent findings, including the data on the numerous effects produced by drugs binding to different sites at the receptor, are briefly summarized in Table 1 (for details, see also [3, 4]). Interestingly, some data point to the fact that substances belonging to at least three classes of these ligands (GABA itself, beta-carboline BDZ-ligands, and neurosteroids) are naturally occurring in the organism [2, 5, 6]. Thus, being a target for various natural ligands or exogenous psychotropic agents, GABA-A receptor is believed to be involved in regulation of a number of normal and pathological central mechanisms, including those related to epileptogenesis and various behavioral phenomena or emotions [3, 6-8 ]. The involvement of the central GABA-ergic system in the regulation of physiology of behavior is well documented [9,

t Center for Physiological and Biochemical Research, Moscow, Russia.

208 0090-297'7/96/2806-0208515.00 ~

1997 Plenum PubUshin& Corporation

Complex Effects of GABA-ergic Agents on Anxiety

209

TABLE 1. Properties of GABA-A Receptors Typical mode of effect on the chloride influx

Traditional anxiolytic (AL) or anxiogenic (AG) properties

agonists: GABA* muscimol

+

AL

antagonists: bicuculline

-

AG

agonists: diazepam

+

AL

antagonists: flumazenil

0

?

"inverse" agonists: beta-carbolines*

-

AG

Alcohol site Barbiturate site

ethanol pentobarbital

+ +

AL AL

Steroid site

agonists: tetrahydroxyprogesterone* pregnenolone* progesterone*

+

AL

antagonists: pregnenolone-sulfate*

-

AG

picrotoxin

-

AG

Major binding sites at GABA-Abenzodiazepine

receptor complex

(BDZ)

GABA site

BDZ site

Convulsant site

Example

ligands

Mechanisms of action of principal agonists

activation of chloride influx (increases in the open channel lifespan and frequency of channel opening), positive modulation of BDZ site

modulation of GABA function: an increase in the affinity of receptors for GABA, increases in the GABA-activated open channel lifespan and channel opening frequency

enhancing of chloride conductance through the reseptors' channels dual do~e-dependent effect on chloride influx: low doses - - positive modulation of GABA site and an increase in the GABA-activated open channel lifespan; high doses - - direct increase in the number of open channels activation of chloride influx (an increase in the GABA-activated open channel lifespan and frequency of channel opening) ; positive modulation of binding sites for GABA, BDZ, and barbiturates; and modulation of convulsant binding to the ionophore

physical block of chloride channels, inhibition of BDZ and barbiturate sites (picrotoxin)

pentylenetetrazole penicillin bicyclophosphates Footnotes. Ligands found in the brain are indicated by asterisks. The symbols of " + " , " - " , and " 0 " indicate activation and inhibition of the chloride influx, or the absence of the effect, respectively. The symbol of " ? " indicates mixed or unidentified effects on anxiety.

DISCUSSION OF GENERAL PROBLEMS To assess the levels of anxiety in animals, a number of original and simple behavioral tests and models has been used, and a great many physiologically active agents have been tested. Although dose-effect curves in such behavioral studies may appear relatively flat, they can also be bell-shaped [15]. Such complex dose dependence curves are reported for a large number of drugs affecting neuronal systems, not necessarily only those involving GABA-ergic mechanisms. For example, bell-shaped curves have been found for buspirone, gepirone, and ipsapirone (active at serotonergic 5-HTI receptors) [17 ], or devazepide and L365-260 (antagonists of cholecystokinin receptors) [ 18 l. Two-phase effects are

produced by a serotonin precursor L-5-HTP [17]. Interestingly, a large body of findings indicate that polyphasic dose-dependent action is typical of the drugs affecting 5-HT3 receptors [15], ionotropic receptors of the same channel superfamily as GABA-A receptors, both of which have been reported to play an important role in regulation of the physiology of anxiety-related behaviors [ 17 1. Interpretation of such polyphasic behavioral effects produced by anxiety-active drugs is a difficult task. In general, we do not know enough about the physiological potentials of this particular group of drugs, and, therefore, cannot rule out the possibility that drugs with well-studied pharmacology may demonstrate new properties, binding sites, etc., which underlie multiple

210

A.V. Kalueff

TABLE 2. Complex D(me-Dependent Behavioral Effects of GABA-ergic Drugs on Anxiety

Agents and mechanisms of their effects

Facilitation of GABA-ergic transmission sodium valproate Effect on GABA-A site agonists: muscimol

Effect on BDZ site agonists: diazepam chlordiazepoxide

"inverse" agonists: ethyl-B-carboline 3-carboxylate FG7142 Effect on alcohol site ethanol

Possible complex dose-dependent effects on anxiety Demonstrated complex effects

Shape of the do~*--effect curves

Doses, mg/kg, l.p. (see Footnotes)

References

+

two-phase

100-600

elevated plus-maze (mice)

[26]

+ +

bell-shaped two-phase

0.5-3.0 100" 5**

elevated plus-maze (mice) punished responding (rats)

[26] [27]

+ +

0.5-20 2.5-20

light/dark transition (mice) punished operant responding (rats)

[20] [281

+ + +

be/l-shaped bell-shaped or two-phase two-phase bell-shaped bell-shaped

non-punished responding (rats) elevated plus-maze (mice) light/dark transition (mice)

[281 [ 19] [20]

+

bell-shaped

0.01-3.0

punished operant responding (monkeys)

[29}

0.01-3.0 5-10

non-punished responding (monkeys) elevated plus-maze (mice)

[29} II 4}

defensive attacks in natural antipredator behavior discrimination (rats)

[30]

25-20 2.5-15 0.6-60

+

two-phase

2000-4000 30(0)-4000 800-2400

ethylalcohol

Anxiety test (objects)

+

two-phase

500-2000 500-2000

chlordiazepoxide- vs-PTZ discrimination task (rats) aggressive behavior in social interaction (mice) resident's aggressive attacks vs intruder (mice)

[31 ] [32] [33] [33]

Effect on barbiturate site pentobarbital

+

boll-shaped

10-80

light/dark transition (mice)

[20] (see also a review in [34] )

Effect on steroid site agonists: pregnan steroids

+

two-phase

5-40

light/dark transition (mice)

[35] (see also a review in [361 )

Blocking of chloride channel PTZ penicillin G

+ +

two-phase two-phase

1.8-30 6-840

elevated plus-maze (mice) elevated plus-maze (rats)

[141 [251 (also our own unpublished data)

Footnotes. One and two asterisks indicate the cases when the doses of agents injected into the brain are shown in pg and ng, respectively. The symbols of "+" and ..... indicate the presence and the absence of complex effects, respectively.

m e c h a n i s m s of a c t i o n o n a n x i e t y . M o r e o v e r , a s m e n t i o n e d a b o v e , t h e r e is a l i k e l i h o o d t h a t t h e a g e n t s at higher e x t r e m e doses may increase behavioral nons p e c i f i c i t y in a n x i e t y t e s t s [19, 20 ]. T h u s , t h e t e s t s m a y

y i e l d b o t h f a l s e p o s i t i v e s a n d f a l s e n e g a t i v e s [13], t h e r e b y r e s u l t i n g in c o n f l i c t i n g d a t a t h a t a r e difficult to i n t e r p r e t . H o w e v e r , a c c o r d i n g to W e t t s t e i n [ 1 5 ] , d e s p i t e t h e fact t h a t h i g h d o s e e f f e c t s a r e e a s i l y a t t r i b u t a b l e to

Complex Effects of GABA-ergic Agents on Anxiety

secondary physiological effects (e.g., incoordination, stereotypy, toxicity, etc.), in some experiments anxiolytic effects may disappear at high do~es with no obvious secondary effects observed. Thus, it seems that inconsistencies in the published data may have some fundamental basis, and analysis of physiological aspects of the latter will be of some interest. One possible explanation for such inconsistent findings is that the differences in the physiological active properties of the drugs themselves may be a cause of conflicting results. For example, specific brain regions may be targeted in different ways by substances that have different levels of tissue absorption or blood-brain barrier penetration. This raises the question as to whether a sufficient quantity gets into the brain to produce a measurable behavioral effect [15]. Moreover, like different types of GABA-ergic-dependent cognition, it cannot be excluded that certain types of anxiety may have different specific "brain topography" [9 ]. The latter fact is particularly important in experiments measuring specific behaviors and employing local drug application within the CNS. Secondly, the data on complex psychophysiological profiles of GABA-ergic drugs that produce alterations in anxiety should be considered [21, 22]. Briefly, it has been suggested that some GABA-ergic substances (including possible endogenous regulators released in stress or anxiety), may have dissimilar effects on various forms of behavior (including anxiety and cognition) due to complex dose-dependent actions on a GABA-ergic receptor population [22]. For example, a slight suppression of the GABA level was reported to improve memory, whereas medium inhibition produced anxiogenic effects [22]. In line with this hypothesis, interesting results were obtained by Potier et al. [21 ], who showed that different levels of BDZ receptor occupancy could produce different behavioral effects. For example, 5% occupancy results in memory facilitation, whereas 30Yo produces a proconflict, and 40Yo evokes convulsant effects [21 ]. It therefore seems possible that interference between various simultaneous processes (for example, learning and anxiogenesis) may induce alterations of predicted behavioral patterns and, thus, result in misinterpretation of observed animal behavior in the standard behavioral tests. Thirdly, it is a known fact that in addition to traditionally accepted behavioral indices, different laboratories often use some novel parameters. Generally, such measures may have different sensitivity to anxiety (e.g., they could be more sensitive to certain anxiety-active drugs, or certain types of anxiety, or different anxietyproducing factors). The latter is also of interest because of the ability of current animal models to induce and detect rather different types of anxiety [14 ]. Moreover, it is also possible that such parameters may demonstrate "nonlinear" response to the gradually changing anxiety. In

211

line with this, it is intriguing that recent findings of Cole and Rodgers [19, 23 ] have confirmed that some of the "novel" measures in, the elevated plus-maze are bidirectionally sensitive to alterations in the anxiety level. In conclusion, it could appear that such more complex behaviors may interfere with less specific (presumably, less sensitive) traditional measures, thus resulting in inconsistent findings. Finally, with respect to the different behavioral procedures used, it is also important to know whether varying degrees of stress experienced during such experiments can produce distinguished dose-dependent alterations in the effects of GABA-ergic agents on physiology of animal behavior (Table 2, cf. curves in punished- and non-punished tasks). Thus, based on the above observations, it cannot be ruled out that at this stage some other important physiological mechanisms (e.g., hormonal or other neuromediatory processes) may be activated by stress, resulting in a complex interference or interaction between GABA-ergic and other neuronal systems in regulation of the physiology of anxiety-related behaviors.

SOME RECENT F I N D I N G S Current views on the role of the GABA-ergic system in physiology of anxiety are based on the following principles (for details, see [24 ]): (i) the central GABAergic system is involved in the regulation of anxiety and (ii) modulation of the GABA-ergic system at all stages (including GABA binding site, various sites for binding of endogenous and exogenous regulators, and drugs affecting the receptor channel) may result in significant alteration of behavior in anxiety tests. However, the latter concept may be dramatically extended by the important observation that GABA-ergic drugs, acting t h r o u g h d i f f e r e n t s i t e s at G A B A - A r e c e p t o r s , demonstrate polyphasic dose-dependent effects on anxiety (Table 2), when low doses reduce anxiety in animals while high doses have an anxiogenic influence. Importantly, these results have been replicated in many laboratories and, therefore, cannot be considered an artifact. As such, it is perhaps not surprising that complex effects on anxiety have been shown for steroids, barbiturates, BDZ, and alcohol (Table 2). This has concerned many investigators for a considerable time, supporting speculations about additional unknown properties (sites?) of GABA-ergic positive modulators. Moreover, some recent studies included the data on behavioral/physiological properties of GABA-ergic negative modulators, such as the convulsants picrotoxin and pentylenetetrazole, or other GABA-A receptor channel blockers. Despite the fact that these drugs have long been known to provoke anxiety in a variety of tests both in animals and in humans [3 ], some of them (see

212

below) have recently been reported to exert polyphasic complex effects on anxiety over a wider dose range, producing unexpected anxiolysis at lower doses. For example, exploratory behavior (e.g., open entries and percentage of open arm entries) displayed a two-phase response to pentylenetetrazole treatment in the elevated plus-maze: with an increase at low (1.875-3.75 mg/kg) and a decrease at high (20-30 mg/kg) doses [14]. Interestingly, in line with these results, another GABAinhibiting agent, penicillin G, demonstrated rather similar two-phase effects on anxiety behavio-. For example, acute psychoses, anxiety, and fears are often observed in humans as side effects of penicillin administration at high doses. In our recent studies in the elevated plus-maze, it was shown that penicillin G systemically provoked anxiety at high doses (600-840 mg/kg; our own unpublished data), although at lower doses (i.e., 6-360 mg/kg, i.p.) it demonstrated bell-shaped antianxiety effects in rats [25]. Comparison of the effects on anxiety, produced by the latter two drugs in the same test, allows one to propose some interesting generalizations. Like pentylenetetrazole, penicillin G affected almost all anxietysensitive plus-maze measures [25, 14]. Importantly, similar effects have been produced by these drugs on highly specific behavioral indicators of anxiety level (such as head dipping and stretched attend postures). These preliminary results so far provide evidence for a similar two-phase action of some GABA-inhibiting agents in the murine elevated plus-maze, a profile of physiological activity that has not previously been reported in anxiety tests [14 ].

CONCLUSION Taken together, the above findings indicate that both traditional GABA-modulating and GABA-inhibiting agents may have positive anxiolytic-like effects. Importantly, the issues raised in our paper suggest that there is a need to reconsider the current "one agent - one mode of action" view on the GABA-ergic role in the mechanisms of anxiety-active agents. In view of this, it is intriguing that the agents affecting different stages of GABA-ergic receptor functioning have similar-looking (i.e., polyphasic) dose-effect curves in various tests of anxiety (for details, see Table 2). Although this would support a similar (i.e., GABA-ergic) mechanism of action, it might also suggest that, in addition to traditionally accepted modes of physiological action (Table 1), these agents (at some doses) can produce differential effects on the GABA-A receptor, which may result in bidirectional modulation of anxiety. This could include effects through some physiological mechanisms not previously known. The latter hypothesis may contribute to our further understanding of fundamental

A . V . Kalueff

physiological processes underlying anxiety and anxietyrelated behavior. Acknowledgement. The author Is grateful to Dr. R. J. Rodgers of the University of Leeds (Great Britain) for his very helpful COmments and criticism.

REFERENCES 1. S. J. Mihic, E. Sanna, P. J. Whiting, and R. A. Harris, "Pharmacology of recombinant GABA-A receptors. GABA-A receptors and anxiety," Adv. Biochent PsychopharmacoL, Set. 48, pp. 17-40 (1995). 2. D. J. Nutt, "GABA-A receptors," in: Receptor Data for Biological Experiments: a Guide to Drug Selectivity, Ellis Horwood, London (1991), pp. 225-229. 3. W. Sieghart, ~GABA-A receptors: ligand-gated CI- channels modulated by multiple drug-binding sites," Trends Pharmacog Sci., 13, 446-450 (1992). 4. W. Sieghart, "Structure and pharmacology of gamma-aminobutyric acid A receptor subtypes," PharmacoL Rev., 47, No. 2, 181-234 (1995). 5. J. J. Lamberl, D. Bilelh, C. Hill-Venning, and J. A. Peters, "Neurosteroids and GABA-A receptor function," Trends PharmacoL Sci., 16, No. 9, 295-303 (1995). 6. M. D. Majewska, ~Neurosteroids: endogenous bimodal modulators of the GABA-A receptor. Mechanism of action and physiological significance," Prog. NeurobioL, 38, No. 4, 379-395 (1992). 7. M. Davis, "A neural system approach to the study of the amygdala, fear and anxiety," in: Experimental Approaches to Anxiety and Depression, Wiley and Sons, Chichester, New York (t992), pp. 43-71. 8. W. Haefely, "The role of GABA in anxiolytic/antidepressant drug action," in: Experimental Approaches to Anxiety and Depression, Wiley and Sons, Chichesler, New York (1992), pp. 151-168. 9. M. Davis, D. Rainie, and M. Cassel, "Neurotransmission in the rat amygdala related to fear and anxiety," Trends Neurosci., 17, No. 5, 209-214 (1994). 10. J. A. Gray, "The neuropsychology of emotion and personality," in: Cognitive Neurochemistry, Oxford Univ. Press, Oxford (1987), pp. 171-190. 11. N. J. Coupland and D. J. Nutt, "Neurobiology of anxiety and panic," in: Cholecystokinin and Anxiety: from Neuron to Behavior, Springer-Verlag, New York (1995), pp. 1-31. 12. J. A. Gray, S. Quintero, J. Mellanby, et al., "Some biochemical, behavioral and electrophysiological tests of the GABA hypothesis of anti-anxiety drug action," in: Actions and Interactions of GABA and Benzodiazepines, Raven Press, New York (1984), pp. 237-262. 13. G. R. Dawson. and M. D. Tricklebank, "Use of the elevated plus-maze in the search for novel anxiolytic agents," Trends PharnLacol. Sci., 16, 33-36 (1995). 14. R. J. Rodgers, J. C. Cole, K. Aboulfa, and L. H. Stephenson, "Ethopharmacological analysis of the effects of putative 'anxiogenic' agents in the mouse elevated plus-maze," PharmacoL BiochenL Behav., 52, No. 3, 1-9 (1995). 15. J. G. Wettstein, ~5-HT3 receptor ligands as potential anxiolyrics," in: Central and Peripheral 5-HT3 Receptors, Acad. Press, New York (1992), pp. 259-273. 16. J. L. McGaugh, "Dissociating learning and performance: drug and hormone enhancement of memory storage," Brain Res. BulL, 23, No. 4/5, 339-345 (1989). 17. B. Soderpalm and J. A. Engel, "Serotonin involvement in conflict behaviour," Eur. Neuropsychopharmacol., I, No. 1, 7-13 (1990). 18. G. A. Higgins, P. Nguen, and E. M. Sellers, "Behavioural and

Complex

19.

20. 21.

22.

23.

24. 25.

26.

27.

Effects of GABA-ergic

Agents on Anxiety

biochemical consequences following activation," Brain Res., 572, No. 1/2, 208-215 (1992). J. C. Cole and R. J. Rodgers, "An ethological analysis of the effects of chlordiazepoxlde and bretazenil (Ro 16-6028) in the routine elevated plus-maze," Behov. PharmacoL, 4, 573-580 (1993). J. N. Crawley, "Exploratory behavior models of anxiety in mice," Neurosci. Behav. Rev., 9, 37-44 (1985). M. C. Potier, L. Prado de Cavalho, and R. H. Dodd, ~In vivo binding of beta-carbolines in mice: regional differences and correlation of occupancy to pharmacological effects," Mol Plu~rmacoL, 34, No. 2, 124-128 (1988). P. Venault, G. Chapouthier, L. Prado de Carvalho, and J. Ro~ier, "Effects of convulsant ligands of the GABA-benzediazepine receptor complex in conflict and learning tasks in mice," Encephale, 18, No. 6, 655-660 (1992). R. J. Rodgers and J. C. Cole, "The elevated plus-maze: pharmacology, methodology and ethology," in: Ethology and Psychopharmacology, Wiley and Sons, Chichester, New York (1994), pp. 9-44. J. D. Vrioni, "Role of GABA during multiple consolidation of memory," Drug Dev. Res., 28, No. 1, 3-27 (1993). A. V. Kaluev, G. E. Samonina, and I. P. Ashmarin, "Effects of penicillin in behavioral tests on anxiety in rats," Byull. Eksp. BioL Med., No. 9, 26-29 0 9 9 5 ) . A. Dalvi and R. J. Rodgers, "Anxiolytic-like effects of sodium valproate and muscimol in mice," J. Psychopharmacol., 9, No. 3, Suppl., 117 (1995). R. C. Drugan, P. Skolnik, S. M. Paul, and J. N. Crawtey, ~Contrasting behavioural effects," Neuropharmacology, 25,

213

No. 2, 203-205 (1986). 28. D. J. Sanger, ~Behavioural effects of novel benzodiazepine (~o) receptor agonists and partial agonists: increases in punished responding and antagonism of the pentylenetetrazole cue," Behav. Pharmacol., 6, 116-126 (1995). 29. D. J. Sanger and C. Cohen, Benzodiazepine Receptor Inverse Agonists, Wiley-Liss, Chichester, New York (1995), pp. 185212, 30. R. J. Blanchard, L. Magee, R. Vaniegas, and D. C. Blanchard, ~Alcohol and anxiety: ethopharmacological approaches," Pros. Neuro-Psychopharmacol. Biol. PsychoL, 17, No. 2, 171-182 (1993). 31. D. V. Gauvin, "Attxiolytic drugs," Drugs Alcohol Depend., 24, No. 2, 103-113 (1989). 32. R. G. Lister and L. A. Hilakivi, "The effects of novelty, isolation and ethanol on social behaviour of mice," Drugs Alcohol Depend., 96, No. 2, 181-187 (1988). 33. H. Yoshimura and N. Ogawa, "Behavioural pharmacology," Nippon Yakurigaku Zasshi, 81, No. 2, 135-141 (1983). 34. E. R. Volkers, Performance Studies in the Field o1" Human Psychopharmacology, IDDR, Utrecht (1995). 35. K. W. Gee, S. Wieland, and N. C. Lan, "Neurosteroids. Relevance to the mechanisms of action of anxiolytic drugs," in: Imidazopyridines in An.riety Disorders, Raven Press, New York (1993), pp. 33-37. 36. S. Wieland, J. D. Beluzzi, L. Stein, and N. C. Lan, "Comparative behavioural characterization of the neuroactive steroids 3-alpha-OH, 5-alpha-pregnan-20-one and 3-alpha-OH, 5-betapregnan-20-one in rodents," Psychopharmacology, 118, No. 1, 65-71 (1995).