gabaergic mechanisms in epilepsy

GABAergic Mechanisms in Epilepsy

David M. Treiman

Department of Neurology, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School,New Brunswick, New Jersey, U.S.A.

Summary: g-Aminobutyric acid (GABA), the principal in-hibitory neurotransmitter in the cerebral cortex, maintains theinhibitory tone that counterbalances neuronal excitation. Whenthis balance is perturbed, seizures may ensue. GABA is formedwithin GABAergic axon terminals and released into the syn-apse, where it acts at one of two types of receptor: GABAA,which controls chloride entry into the cell, and GABAB, whichincreases potassium conductance, decreases calcium entry, andinhibits the presynaptic release of other transmitters. GABAA-receptor binding influences the early portion of the GABA-mediated inhibitory postsynaptic potential, whereas GABAB

binding influences the late portion. GABA is rapidly removedby uptake into both glia and presynaptic nerve terminals andthen catabolized by GABA transaminase. Experimental andclinical study evidence indicates that GABA has an importantrole in the mechanism and treatment of epilepsy: (a) Abnor-malities of GABAergic function have been observed in geneticand acquired animal models of epilepsy; (b) Reductions of

GABA-mediated inhibition, activity of glutamate decarboxyl-ase, binding to GABAA and benzodiazepine sites, GABA incerebrospinal fluid and brain tissue, and GABA detected duringmicrodialysis studies have been reported in studies of humanepileptic brain tissue; (c) GABA agonists suppress seizures,and GABA antagonists produce seizures; (d) Drugs that inhibitGABA synthesis cause seizures; and (e) Benzodiazepines andbarbiturates work by enhancing GABA-mediated inhibition.Finally, drugs that increase synaptic GABA are potent anticon-vulsants. Two recently developed antiepileptic drugs (AEDs),vigabatrin (VGB) and tiagabine (TGB), are examples of suchagents. However, their mechanisms of action are quite different(VGB is an irreversible suicide inhibitor of GABA transami-nase, whereas TGB blocks GABA reuptake into neurons andglia), which may account for observed differences in drug side-effect profile.Key Words: Antiepileptic drug—GABA—Model—Neurotransmitters—Seizures.

“Abnormal discharges are due to potentiation of excitatorymechanisms or to a failure of intrinsic cerebral inhibitorysystems.”

Gowers, 1881

Epileptic seizures can be thought of as paroxysmalhypersynchronous transient electrical discharges in thebrain that result from too much excitation or too littleinhibition in the area in which the abnormal dischargestarts. Excitation and inhibition of neurons may be me-diated by many different neurotransmitters.g-Aminobutyricacid (GABA) is now recognized as the principal inhibi-tory neurotransmitter in the cerebral cortex (1). Thestructure of GABA is shown in Fig. 1. GABA is local-ized primarily in short-axon interneurons that synapse oncell bodies and proximal axons, and serves to maintain

inhibitory tone that counterbalances neuronal excitation.When this balance is perturbed, seizures may ensue.

GABA is formed within GABAergic axon terminalsby transamination ofa-ketoglutarate to glutamic acid,which is then decarboxylated by glutamic acid decarbox-ylase (GAD) to GABA. It is released into the synapseand then acts at one of two types of GABA receptors:GABAA receptors and GABAB receptors. GABAA re-ceptors are ligand-gated ion channels that hyperpolarizethe neuron by increasing inward chloride conductanceand have a rapid inhibitory effect. The GABAA-receptorcomplex is a pentameric heterooligomer that containsbinding sites for GABA, barbiturates, benzodiazepines,picrotoxin, and neurosteroids. A number of subunits ofthe GABAA complex have been isolated (a1–6, b1–4,g1–3,d, andr), and multiple GABAA receptor subtypesappear to exist in vivo. GABAA-receptor subtypes in thebrain have been reviewed recently by McKernan andWhiting (2). GABAB receptors are G protein–linkedreceptors that hyperpolarize the neuron by increasing po-tassium conductance. GABAB receptors decrease cal-cium entry and have a slow inhibitory effect. GABAB

receptors are present on both excitatory and inhibitory

Address correspondence and reprint requests to Dr. D.M. Treiman atDepartment of Neurology, University of Medicine and Dentistry ofNew Jersey, Robert Wood Johnson Medical School, 97 Paterson Street,New Brunswick, NJ 08901, U.S.A. E-mail: [email protected]

Epilepsia,42(Suppl. 3):8–12, 2001Blackwell Science, Inc.© International League Against Epilepsy

8

axon terminals. Activation is associated with a decreasein neurotransmitter release, and thus GABAB agonistdrugs, such as baclofen, under some circumstances maybe antiepileptic (3) and, in other experimental paradigms,proepileptic drugs (4). After release from the presynapticaxon terminals, GABA is rapidly removed by uptake intoboth glia and presynaptic nerve terminals and then iscatabolized by GABA transaminase to succinic semial-dehyde. Succinic semialdehyde is converted to succinicacid by succinic acid semialdehyde dehydrogenase andthen enters the Krebs cycle.

Olsen and Avoli (5) recently described the role ofGABAA and GABAB receptors in the generation ofGABA-mediated inhibitory postsynaptic potentials(IPSPs) with data generated in Avoli’s laboratory (seeFig. 2). Bicuculline methiodide, a GABAA inhibitor,inhibits the early portion of the IPSP whereas CGP-35348, a GABAB inhibitor, blocks the slow inhibitoryeffect seen as the late portion of the IPSP. Both drugsgiven together block the entire IPSP. In another experi-ment, CGP-35348 prevented the occurrence of paired-pulse inhibition, thus demonstrating a role of GABAB

autoreceptors in this GABA-mediated inhibitory phe-nomenon.

ROLE OF GABA IN EPILEPSY

What is the role of GABA in epilepsy and in epilep-togenesis? We need to consider evidence from a numberof experimental and clinical sources that provide strongsupport for a role of GABA in the mechanism and treat-ment of epilepsy.

1. Abnormalities of GABAergic function have beenobserved in genetic and acquired animal models ofepilepsy.

2. Reductions of GABA-mediated inhibition, activityof glutamate decarboxylase, binding to GABAA

and benzodiazepine sites, GABA in cerebrospinalfluid (CSF) and brain tissue, and GABA detectedduring microdialysis studies have all been reportedin studies of human epileptic brain tissue.

3. GABA agonists suppress seizures, and GABA an-tagonists produce seizures.

4. Drugs that inhibit GABA synthesis cause seizures.5. Benzodiazepines (BZDs) and barbiturates, which

are effective anticonvulsants, work by enhancingGABA-mediated inhibition.

6. Drugs that increase synaptic GABA by inhibitingGABA catabolism [vigabatrin (VGB)] or reuptake[tiagabine (TGB)] are effective anticonvulsants.

Let us now consider each of these lines of evidencein detail.

EVIDENCE FROM GENETIC EPILEPSIES ANDEPILEPSY MODELS

Several lines of evidence from genetic epilepsies andgenetic models of epilepsy support a role for GABA inepilepsy. There is impaired GABA synthesis in humanpyridoxine deficiency, a disorder characterized by theonset of seizures in infancy. Decreased [3H]-flunitraze-pam binding has been demonstrated by Olsen et al. (6) inthe substantia nigra pars reticulata and periaqueductalgray matter of seizure-susceptible gerbils studied beforethe appearance of seizures, suggesting that the decreasedBZD binding reflects an etiologic abnormality in theGABA receptor, rather than an effect of the seizures.DBA/2 mice have an increased susceptibility to audio-genic seizures. Horton et al. (7) found a reduction in thenumber of high-affinity GABA receptors, and Olsen etal. (8) found reduced BZD binding in a number of areasin brains from DBA/2 mice. El (epileptic) mice havebeen shown to have increased numbers of GABA-immunoreactive neurons (9). Similar findings have beenreported in genetically epilepsy prone (GEPR) rats (10)and in seizure-susceptible gerbils (11). These observa-tions suggest that possible synchronizing effects ofGABA interneurons may result in paradoxic facilitationof some types of epileptic discharges in some of thesemodels. Conversely, photosensitivity-producing epilepsyin the baboon has been thought to be due to deficiency in

FIG. 1. Structure of g-aminobutyricacid.

FIG. 2. Inhibitory postsynaptic potentials (IPSPs) and presynap-tic inhibition. The control tracing shows a typical monosynapticresponse recorded with a potassium acetate–filled intracellularmicroelectrode in a human cortical neuron in the presence ofionotropic excitatory amino acid–receptor antagonists. The earlyhyperpolarization (IPSP) (arrow at A in the control tracing) isabolished by the g-aminobutyric acid (GABAA)-receptor antago-nist bicuculline methiodide (BMI), as shown in the middle tracing,thus indicating this response is GABAA receptor mediated. Thelate hyperpolarization (arrow at B in the control tracing) is abol-ished by applying the GABAB receptor antagonist P-3-aminopropyl,P-diethoxymethyl phosphoric acid (CGP 35348), asshown in the bottom tracing, thus indicating the late response isGABAB receptor mediated. (From ref. 5, with permission).

GABAergic MECHANISMS 9

Epilepsia, Vol. 42, Suppl. 3, 2001

cortical GABAergic inhibition. Lloyd et al. (12) foundreduced CSF GABA concentration to be correlated withthe degree of epileptic photosensitivity in such baboons.

Acquired epilepsyA role of GABA in models of acquired epilepsy has

also been proposed and reported in some studies. Re-duced GABA-mediated inhibition has been found in sev-eral animal models. Esclapez and Trottier (13) reportedreduced GABA-immunoreactive cell density in cobalt-induced cortical lesions. The reduction in GABA immu-noreactivity was closely correlated with developmentand regression of seizure activity in this model. Ribak etal. (14) observed reduced GAD-positive axon terminalsin alumina gel–treated monkeys with epilepsy. However,no consistent changes in GABA function in a number ofanimal models of limbic epilepsy have been demon-strated. Houser et al. (15) demonstrated a loss of GAD-positive terminals in alumina gel foci in the rat. Therewas a progressive loss of GAD-positive terminals overtime, which was correlated with the development of elec-trocorticogram (EcoG) abnormalities and eventuallywith behavioral seizures.

Human tissueA number of studies have shown changes in GABA

concentrations or GABA-receptor densities in humanepileptic tissue. GABAA receptors have been found to bereduced in parts of the human hippocampus in which cellloss has been observed (16–18). Reduced BZD bindinghas been demonstrated in the mesial temporal lobe onpositron emission tomographic scanning (19,20). A lossof GAD and GABAergic neurons in human mesial tem-poral sclerosis also has been demonstrated by some in-vestigators (21–23), but not by others (24).

During and Spencer (25) studied glutamate andGABA concentrations before and after seizures in sixhuman patients using microdialysis techniques. Gluta-mate concentrations increased before seizure onset andwere higher in the epileptic hippocampus than in thenonepileptic hippocampus, suggesting a role of gluta-mate in triggering the seizures. However, the GABAincrease that occurs during seizures was greater in thenonepileptic hippocampus than in the epileptic hippo-campus (see Fig. 3). There are two types of presynapticGABA release. Vesicular GABA release is calcium de-pendent, sensitive to tetanus toxin, and triggered by highpotassium concentrations. Nonvesicular GABA releaseis calcium independent and occurs secondary to depolar-ization of the cell membrane and sodium influx. Nonve-sicular GABA release is dependent on reversal of theGABA transporter. During et al. (26) found the numberof GABA transporters to be reduced in epileptic hippo-campus, that nonvesicular GABA release from epileptichippocampus during seizures is not so robust as that fromnormal hippocampus, and that therefore insufficient

amounts of GABA are released to suppress seizure ac-tivity.

GABA AGONIST AND ANTAGONIST DRUGS

A number of GABA agonist drugs including musci-mol, tetrahydroisooxazolopyridinol (THIP), cetylGABA,and progabide (PGB) are anticonvulsant in experimentalanimals, whereas GABA antagonists such as bicucullineand picrotoxin are proconvulsant. Inhibition of GABAsynthesis is frequently epileptogenic. A number ofGABA-synthesis inhibitors can cause seizures, including4-deoxypyridoxine, isoniazid, thiosemicarbazide, andL-allyglycine.

Drugs that enhance GABA-mediated inhibition alsoare anticonvulsant. These include the BZDs, which en-hance binding of GABA to the GABA receptor, thusincreasing the frequency of chloride channel openings(27,28). Barbiturates, which prolong the open time of thechloride channel (29), also are anticonvulsant.

Drugs that increase synaptic GABA also have beenshown to be potent antiepileptic drugs (AEDs). Two re-cently developed and marketed AEDs, VGB and TGB,are examples of this class of drugs. Both of these drugscan be considered “designer drugs” because they werespecifically developed with the intent of increasing syn-aptic GABA concentrations and thus inhibiting seizureactivity.

Figure 4 shows schematically how each of these drugsworks. VGB is an irreversible suicide inhibitor of GABAtransaminase and thus inhibits degradation of GABA.This in turn results in an increase in synaptic GABAconcentrations. TGB, on the other hand, blocks GABAreuptake into neurons and into glia. This also has theeffect of increasing synaptic GABA concentrations. Thenet effect of the increase in synaptic GABA concentra-

FIG. 3. Hippocampal g-aminobutyric acid (GABA) concentra-tions during complex partial seizures with secondary generaliza-tion (25). Means of percentages of basal concentrations collectedduring six seizures in six patients are presented. Samples werecollected using a microdialysis probe implanted as part of anevaluation procedure for possible surgical treatment of epilepsyrefractory to medical treatment. Sample times in relation to theseizures are indicated on the figure. (From ref. 25 with permis-sion.)

D. M. TREIMAN10


tions is reduction of seizure frequency in patients withpartial-onset seizures. Both VGB (30–33) and TGB (34–37) have been shown to be potent AEDs in the treatmentof partial-onset seizures. Both work by increasing syn-aptic GABA concentrations. However, the mechanismwhereby they produce an increase in GABA concentra-tions is quite different, and this may account for differ-ences in their side-effect profiles, discussed in otherarticles in this symposium.

REFERENCES

1. Schwartz RD. The GABAA receptor-gated ion channel: biochemi-cal and pharmacological studies of structure and function.BiochemPharmacol1988;37:3369–75.

2. McKernan RM, Whiting PJ. Which GABAA-receptor subtypesreally occur in the brain?Trends Neurosci1996;19:139–43.

3. Swartzwelder HS, Bragdon AC, Sutch CP, et al. Baclofen sup-presses hippocampal epileptiform activity at low concentrationswithout suppressing synaptic transmission.J Pharmacol Exp Ther1986;237:881–7.

4. Mott DD, Bragdon AC, Lewis DV, et al. Baclofen has a proepi-leptic effect in the rat dentate gyrus.J Pharmacol Exp Ther1989;249:721–5.

5. Olsen RW, Avoli M. GABA and epileptogenesis.Epilepsia1997;38:399–407.

6. Olsen RW, Wamsley JK, McCabe RT, et al. Benzodiazepine/gamma-aminobutyric acid receptor deficit in the midbrain of theseizure-susceptible gerbil.Proc Natl Acad Sci U S A1985;82:6701–5.

7. Horton RW, Prestwich SA, Meldrum BS. Gamma-aminobutyricacid and benzodiazepine binding sites in audiogenic seizure-susceptible mice.J Neurochem1982;39:864–70.

8. Olsen RW, Wamsley JK, McCabe RT, et al. Midbrain GABAreceptor deficit in genetic animal models of epilepsy. In: Nistico G,Morselli PL, Lloyd KG, et al., eds.Neurotransmitters, seizures andepilepsy III. New York: Raven Press, 1986:279–91.

9. King JT Jr, LaMotte CC. VIP-, SS-, and GABA-like immunore-activity in the mid-hippocampal region of El (epileptic) andC57BL/6 mice.Brain Res1988;475:192–7.

10. Roberts RC, Ribak CE, Oertel WH. Increased numbers ofGABAergic neurons occur in the inferior colliculus of an audio-genic model of genetic epilepsy.Brain Res1985;361:324–38.

11. Peterson GM, Ribak CE, Oertel WH. A regional increase in thenumber of hippocampal GABAergic neurons and terminals in theseizure-sensitive gerbil.Brain Res1985;340:384–9.

12. Lloyd KG, Scatton B, Voltz C, et al. Amino acid and monoamine

metabolite levels in CSF ofPapio papio: correlation with photo-sensitive response. In: Nistico G, Morselli PL, Lloyd KG, et al.,eds. Neurotransmitters, seizures, and epilepsy III. New York:Raven Press, 1986:151–62.

13. Esclapez M, Trottier S. Changes in GABA-immunoreactive celldensity during motor focal epilepsy induced by cobalt in the rat.Exp Brain Res1989;76:369–85.

14. Ribak CE, Harris AB, Vaughn JE, et al. Inhibitory, GABAergicnerve terminals decrease at sites of focal epilepsy.Science1979;205:211–4.

15. Houser CR, Harris AB, Vaughn JE. Time course of the reductionof GABA terminals in a model of focal epilepsy: a glutamic aciddecarboxylase immunocytochemical study.Brain Res1986;383:129–45.

16. McDonald JW, Garofalo EA, Hood T, et al. Altered excitatory andinhibitory amino acid receptor binding in hippocampus of patientswith temporal lobe epilepsy.Ann Neurol1991;29:529–41.

17. Johnson EW, de Lanerolle NC, Kim JH, et al. “Central” and “pe-ripheral” benzodiazepine receptors: opposite changes in humanepileptogenic tissue.Neurology1992;42:811–5.

18. Olsen RW, Bureau M, Houser CR, et al. GABA/benzodiazepinereceptors in human focal epilepsy.Epilepsy Res Suppl1992;8:383–91.

19. Savic I, Persson A, Roland P, et al. In-vivo demonstration ofreduced benzodiazepine receptor binding in human epileptic foci.Lancet1988;2:863–6.

20. Henry TR, Frey KA, Sackellares JC, et al. In vivo cerebral me-tabolism and central benzodiazepine-receptor binding in temporallobe epilepsy.Neurology1993;43:1998–2006.

21. Lloyd KG, Bossi L, Morselli PL, et al. Biochemical evidence fordysfunction of GABA neurons in human epilepsy. In: Bartholini G,Bossi L, Lloyd KG, et al., eds.Epilepsy and GABA receptor ago-nists: basic and therapeutic research. New York: Raven Press,1985:43–51.

22. Houser CR, Miyashiro JE, Swartz BE, et al. Altered patterns ofdynorphin immunoreactivity suggest mossy fiber reorganization inhuman hippocampal epilepsy.J Neurosci1990;10:267–82.

23. Houser CR, Swartz BE, Walsh GO, et al. Granule cell disorgani-zation in the dentate gyrus: possible alterations of neuronal migra-tion in human temporal lobe epilepsy. In: Engel J Jr, Wasterlain C,Cavalheiro EA, et al., eds.Molecular neurobiology of epilepsy(Epilepsy Res Suppl 9). New York: Elsevier Science Publishers,1992:41–9.

24. Babb TL, Pretorius JK, Kupfer WR, et al. Glutamate decarboxylase-immunoreactive neurons are preserved in human epileptic hippo-campus.J Neurosci1989;9:2562–74.

25. During MJ, Spencer DD. Extracellular hippocampal glutamate andspontaneous seizure in the conscious human brain.Lancet1993;341:1607–10.

FIG. 4. g-Aminobutyric acid (GABA)ergic synapse.GABA is synthesized in the presynaptic terminal.After release it causes the GABAA receptor on thepostsynaptic neuron to increase inward chlorideconductance, as do many other drugs described inthe text, including the barbiturates and benzodiaz-epines. Synaptic GABA is taken back up into thepresynaptic terminal and into glial cells. Reuptakeinhibitors, such as tiagabine, and drugs that blockGABA metabolism, such as vigabatrin, thus in-crease synaptic GABA levels. SSA, succinic semi-aldehyde; GAD, glutamic acid decarboxylase;GABA-T, GABA transaminase. (From ref. 38, withpermission).

GABAergic MECHANISMS 11


26. During MJ, Ryder KM, Spencer DD. Hippocampal GABA trans-porter function in temporal-lobe epilepsy.Nature1995;376:174–7.

27. Macdonald RL. Mechanisms of anticonvulsant drug action. In:Pedley TA, Meldrum BS, eds.Recent advances in epilepsy I. Ed-inburgh: Churchill Livingstone, 1983:1–23.

28. Barker JL, Owen DG. Electrophysiological pharmacology ofGABA and diazepam in cultured CNS neurons. In: Olsen RW,Venter JC, eds.Benzodiazepine/GABA receptors and chloridechannels: structural and functional properties. New York: Alan R.Liss, 1986:135–65.

29. Study RE, Barker JL. Diazepam and (––)-pentobarbital: fluctua-tion analysis reveals different mechanisms for potentiation ofgamma-aminobutyric acid responses in cultured central neurons.Proc Natl Acad Sci U S A1981;78:7180–4.

30. Penry JK, Wilder BJ, Sachdeo RC, et al. Multicenter dose-responsestudy of vigabatrin in adults with focal (partial) epilepsy[Abstract].Epilepsia1993;34(suppl 6):67.

31. Grunewald RA, Thompson PJ, Corcoran R, et al. Effects of viga-batrin on partial seizures and cognitive function.J Neurol Neuro-surg Psychiatry1994;57:1057–63.

32. Beran RG, Berkovic SF, Buchanan N, et al. A double-blind,placebo-controlled crossover study of vigabatrin 2 g/day and3 g/day in uncontrolled partial seizures.Seizure1996;5:259–65.

33. French JA, Mosier M, Walker S, et al. A double-blind, placebo-controlled study of vigabatrin three g/day in patients with uncon-trolled complex partial seizures.Neurology1996;46:54–61.

34. Richens A, Chadwick DW, Duncan JS, et al. Adjunctive treatmentof partial seizures with tiagabine: a placebo-controlled trial.Epi-lepsy Res1995;21:37–42.

35. Ben-Menachem E. International experience with tiagabine add-ontherapy.Epilepsia1995;36(suppl 6):S14–21.

36. Sachdeo RC, Leroy RF, Krauss GL, et al. Tiagabine therapy forcomplex partial seizures: a dose-frequency study: the TiagabineStudy Group.Arch Neurol1997;54:595–601.

37. Uthman BM, Rowan AJ, Ahmann PA, et al. Tiagabine for complexpartial seizures: a randomized, add-on, dose-response trial.ArchNeurol 1998;55:56–62.

38. Suzdak PD, Jansen JA. A review of the preclinical pharmacologyof tiagabine: a potent and selective anticonvulsant GABA uptakeinhibitor. Epilepsia1995;36:612–26.

D. M. TREIMAN12


gabaergic mechanisms in epilepsy

Documents