Professor, Department of Pharmacology and Toxicology, and Director, Anticonvulsant Screening Project, University of Utah, Salt Lake City, Utah
Topiramate, a sulfamate-substituted derivative of D-fructose, has been demonstrated to possess anticonvulsant activity in a battery of well-defined animal models (Table 77.1). Topiramate is highly effective in blocking maximal electroshock seizures in rats and mice (39,50), but it is inactive against clonic seizures induced by subcutaneous administration of the chemoconvulsants pentylenetetrazol, picrotoxin, or bicuculline (39,50). Topiramate does elevate the pentylenetetrazol seizure threshold at low doses; however, this effect is reduced at higher doses of topiramate (57). In addition, topiramate has been shown to be effective against fully kindled seizures in the amygdala-kindled rat model of partial seizures. These findings in animal models support the results of clinical trials in which topiramate has been shown to be effective against both partial and generalized seizures.
Several pharmacologic properties have been identified that may account for the broad anticonvulsant preclinical profile of topiramate in animal models. For example, the drug has been found to modulate voltage-dependent sodium ion (Na+) channels, to potentiate γ-aminobutyric acid (GABA)-mediated inhibitory neurotransmission, to block excitatory neurotransmission mediated by non-N-methyl-D-aspartate (non-NMDA) receptors, to modulate voltage-gated calcium ion (Ca2+) channels, and to inhibit brain carbonic anhydrase (CA).
This chapter reviews the evidence supporting the multiple mechanisms of action thus far identified for topiramate and briefly considers the therapeutic potential of topiramate in other disease states in which the underlying pathophysiologic features suggest a role for the drug.
EFFECTS OF TOPIRAMATE IN EXPERIMENTAL MODELS OF EPILEPSY
Early investigations demonstrated that topiramate, like carbamazepine and phenytoin, possesses the ability to block tonic hind-limb extension seizures induced by maximal electroshock (MES). In the MES test in mice and rats, the potency of topiramate after oral administration was found to be similar to that of carbamazepine, phenytoin, phenobarbitone, and acetazolamide, but it was greater than that of sodium valproate (13,39). Peak anticonvulsant activity was reached within 1 hour in mice and between 1 and 4 hours in the rat, with a duration of activity of >4 and >8 hours in mice and rats, respectively (39). These observations suggest that topiramate acts primarily by blocking the spread of seizures and in this respect shares an anticonvulsant mechanism similar to that of carbamazepine and phenytoin, that is, state-dependent block of voltage-sensitive Na+ channels (39).
Early studies found topiramate to be virtually ineffective in preventing seizures induced by subcutaneous injection of pentylenetetrazol or other chemicals such as bicuculline, picrotoxin, or metrazol (39,50). However, topiramate was subsequently found to elevate seizure threshold in the intravenous pentylenetetrazol test in mice (57), an effect that also indicated possible clinical activity against generalized seizures (55).
Genetic (Spontaneous Epileptic Rat and DBA/2 Mouse) and Amygdala Kindling Models
Studies in the spontaneously epileptic rat demonstrated that topiramate is effective against both tonic extension and spike-wave seizures (30). Topiramate has also been found to
be effective in blocking audiogenic seizures in the DBA/2 audiogenic seizure-susceptible mouse (30).
TABLE 77.1. ANTICONVULSANT PROFILE OF TOPIRAMATE IN ANIMAL SEIZURE AND EPILEPSY MODELS
In addition, topiramate has been found to be highly effective in various kindling models including mouse (18), rat (18,53), rabbit (18), and cat (29). In the amygdala-kindled rat model, topiramate was found to be more effective than phenytoin (1,53). In this model, topiramate displayed a potent and dose-dependent inhibition of all measured seizure parameters (behavioral seizures, forelimb clonus, amygdala, and cortical afterdischarges). Reissmüller and colleagues (36) evaluated the anticonvulsant efficacy of topiramate in phenytoin-resistant amygdala-kindled rats, a relatively new and unique model of drug resistant partial epilepsy (24). In phenytoin responders, topiramate (40 mg/kg) increased the afterdischarge threshold and decreased seizure severity and duration. Consistent with the proven clinical efficacy of topiramate in patients with refractory partial epilepsy, topiramate also significantly increased the afterdischarge threshold in phenytoin-resistant kindled rats. Finally, topiramate, in addition to blocking the expression of fully kindled seizures, significantly delayed the acquisition of amygdaloid kindling at doses of 100 and 200 mg/kg (1).
PROPOSED MECHANISMS OF ACTION
Topiramate has been evaluated in numerous in vitro assays in an attempt to identify the molecular mechanisms of action underlying its broad preclinical and clinical spectrum of activity. These investigations suggest that topiramate possesses activity at various voltage- and receptor-gated ion channels that may account for its in vivo activity.
Blockade of Voltage-Sensitive Sodium Channels
The first electrophysiologic studies investigating the mechanisms responsible for the anticonvulsant activity of topiramate were conducted in cultured rat hippocampal neurons (6). At concentrations ranging from 10 to 100 µmol/L, topiramate reversibly reduced the duration and frequency of action potentials within spontaneous epileptiform bursts of neuronal firing. Topiramate, like carbamazepine and phenytoin (26), was also found to block depolarization-induced sustained repetitive firing (SRF), a finding thereby suggesting activity at use-dependent, voltage-sensitive Na+ channels (6,46). Subsequent studies found that topiramate was capable of reducing the duration and frequency of action potentials and of reducing the amplitude of inward voltage-gated Na+ currents in rat cerebellar granule cells (2). Furthermore, topiramate was demonstrated to inhibit neuronal activity in the rat hippocampal slice preparation, in a frequency-dependent manner (58).
A series of whole-cell patch-clamp recordings in a range of preparations provided further evidence of an effect of topiramate on voltage-gated Na+ channels. Zona and colleagues reported reduced amplitude of tetrodotoxin-sensitive, voltage-gated Na+ currents when topiramate was applied to cerebellar granule cells (63). Voltage-gated Na+ conductance was also reduced by topiramate in slice preparations of entorhinal cortex (3), as well as in dissociated neocortical neurons and neocortical slices (48).
Further insight into the modulatory effect of topiramate on voltage-dependent Na+ channels is provided by an investigation using an in vitro seizure model of cultured hippocampal neurons wherein spontaneous recurrent seizures
are evoked by depolarizing current injection (8). At concentrations of 10 µmol/L, topiramate produced a dose-dependent and partially reversible reduction in the frequency of action potentials induced by a sustained depolarizing current and abolished the late sustained phase of firing. Similarly, topiramate (10 to 100 µmol/L) decreased or abolished spontaneous recurrent seizures, with a marked reduction in both the number of bursts and the duration of epileptiform activity. These observations are consistent with topiramate's ability to modulate voltage-dependent Na+ and Ca2+ (see later) conductances responsible for the generation and propagation of action potentials.
A further study showed that topiramate is capable of producing a voltage-sensitive, use-dependent, and time-dependent suppression of SRF in cultured mouse spinal cord and neocortical cells (28). However, the effect of topiramate in these preparations followed a complex pattern. For example, at concentrations ≥3 µmol/L, limitation of SRF was voltage-sensitive, time-dependent, and associated with a decrease in the rapid inward current during the upstroke of the action potential. At elevated concentrations (30 to 600 µmol/L), topiramate produced a rapid block of SRF in approximately one-third of neurons and did not affect SRF in a further one-third. Of the remaining one-third, SRF was limited in an intermittent fashion or was blocked only after a delay of a few seconds. As pointed out by the authors, the activity of topiramate on Na+ channels differs from that of other antiepileptic drugs in which a rapid limitation or complete block of SRF occurs. Based on these findings, the authors concluded that although topiramate appears to exert some of its anticonvulsant effects by blocking voltage-sensitive, use-dependent Na+ channels, it does so by a mechanism that seems to differ from that of classic Na+ channel blocking antiepileptic drugs such as phenytoin, carbamazepine, and lamotrigine. As such, it has been suggested that Na+ channel blockade may not be the primary mechanism by which topiramate exerts its anticonvulsant activity (28).
Enhancement of GABA-Evoked Chloride Currents
Topiramate has also been demonstrated to potentiate GABAergic transmission. Topiramate (10 to 100 µmol/L) was originally found to enhance GABA-induced chloride (C1-) flux into cultured mouse cerebellar granule cells (5). Subsequent studies demonstrated that topiramate rapidly and reversibly increased the frequency of opening of GABA-mediated C1-channels in mouse cortical neurons at 1 to 30 µmol/L (54). At concentrations of 1 to 30 µmol/L, topiramate increased bursting frequency and the frequency of channel opening but did not affect the duration of channel open time. The effects of topiramate on GABAA channel kinetics suggest that the drug is acting much like a benzodiazepine; however, topiramate's ability to enhance GABAA-mediated C1- currents is not blocked by the benzodiazepine antagonist, flumazenil (54). In addition, topiramate has been shown to enhance GABAA -mediated C1- currents in clonazepam-insensitive neurons and to be ineffective in some clonazepam-sensitive neuronal populations (57) These findings have suggested that topiramate may exert its positive modulatory effects by binding to a different site on the GABA receptor or to a novel site on the GABAA receptor complex. Furthermore, clonazepaminduced elevation of pentylenetetrazol seizure threshold was reversed by the benzodiazepine antagonist flumazenil, whereas topiramate-induced elevation of seizure threshold was not.
To investigate the modulatory effects of topiramate on various GABAA receptor subtypes further, Gordey and colleagues (15) expressed specific subunits of the GABAA receptor inXenopus oocytes. Topiramate reversibly inhibited GABA-evoked C1- currents in oocytes expressing the GABAA receptor subunits α1β2γ2S and α2β2γ2S, it potentiated currents in oocytes expressing the α2β2γ2S subunit combination, but it had no effect on oocytes containing α4β2γ2S GABAA receptors. The observed negative modulatory effect of topiramate on C1- currents (as mediated by the α1β2γ2S and α2β2γ2S receptors) is a novel finding—all other studies reported to date found only a positive modulatory effect whereby topiramate increases GABA-induced C1- currents (5,56). Although no obvious explanation exists, the investigators hypothesized that topiramate may modulate GABA activity through desensitization of GABAA receptors, possibly by effects on secondary messenger systems. Additional findings using the Xenopus oocyte expression system have further demonstrated that the effects of topiramate on the GABAA receptor system are subunit selective. Simeone and colleagues (44) reported that GABAA receptors containing α4, β3, and γ2S subunits are particularly sensitive to direct activation with topiramate at concentrations as low as 10 µmol/L.
Additional findings that may or may not be associated with the modulatory effects of topiramate on GABAergic transmission relate to the detection of increased GABA concentrations in the brain after topiramate administration. One study, using a low-resolution Tesla magnetic resonance spectrometer, detected increased brain concentrations of GABA in humans after the administration of topiramate (32). However, further investigation in rats failed to support this observation (43). Both single and multiple doses of topiramate were without effect on GABA brain concentration in the rat. The investigators noted that although topiramate was detectable in the brain after a dose of ≥10 mg/kg, there was no accumulation of GABA in the retina or in any of the other brain regions analyzed. Further studies are required to replicate the initial findings in human brains using high-resolution Tesla magnets.
Effects on Glutamate-Mediated Excitatory Neurotransmission
The excitatory neurotransmitter glutamate exerts its excitatory actions by binding to both ionotropic (NMDA, α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid [AMPA], and kainate) and metabotropic (mGluR1-8) glutamate receptors. In addition to the actions described earlier, topiramate has been found to exert a negative modulatory effect on kainate-elicited excitatory currents. This activity was first described after a series of electrophysiologic studies using cultured rat hippocampal neurons. Coulter and colleagues (6,7,38) reported that topiramate (10 to 100 µmol/L) blocked membrane currents evoked by kainate but had no activity at the NMDA-mediated glutamate receptor subtype. Subsequent studies demonstrated that topiramate exerts a biphasic effect on kainate-evoked currents (14). Thus, topiramate was found to produce an initial (phase I), followed by a delayed (phase II), inhibition of kainate currents. The delayed phase II effect was observed after the constant application of topiramate for >10 minutes, and it was not readily reversed during a 2- to 4-hour topiramate washout period. The authors postulated that the delayed effect may be associated with an alteration of the phosphorylation state of kainate-activated channels. Consistent with this hypothesis is the finding that treatment of channels with dibutyryl cyclic adenosine monophosphate enhanced reversal of the phase II topiramate block. In addition, the phase II block, but not the phase I block, was prevented by the nonspecific protein phosphatase inhibitor okadaic acid.
Topiramate was subsequently found to inhibit kainate-evoked inhibition of cobalt (Co2+) flux into cultured cerebellar granule neurons in a concentration- and time-dependent manner (45). In these studies, maximal inhibition of kainate-stimulated Co2+ flux was observed after 30-minute pretreatment with topiramate. In addition, the effect of topiramate appeared to depend on the developmental stage of the cerebellar granule cells. For example, topiramate inhibited kainate-stimulated Co2+ uptake in cells grown in vitro for 9 to 11 days, but it was without effect in older neurons (13 to 14 days). In contrast, the non-NMDA antagonist DNQX was effective at all ages and did not require pretreatment. These findings are of interest because they suggest that topiramate exerts a negative modulatory effect on a Ca2+ -permeable non-NMDA receptor that is developmentally regulated.
A further effect of topiramate on glutamate was observed in microdialysis studies in which the drug reduced abnormally high (two- to threefold higher) basal hippocampal concentrations of glutamate and aspartate by approximately 45% in spontaneously epileptic rats (16). In contrast, topiramate had no effect on hippocampal concentrations of other amino acid neurotransmitters, including GABA, taurine, and glycine (16,39).
Negative Modulatory Effect on Neuronal L-Type High-Voltage-Activated Calcium Channels
In addition to the mechanisms outlined earlier, a growing body of evidence suggests that topiramate modulates neuronal Ca2+ channels. Activity has been demonstrated at both N- and L-type high-voltage-activated Ca2+ channels (HVACCs) in CAl pyramidal neurons using whole-cell patch-clamp techniques in rat dentate gyrus granule cells (61,62). After separation of the different HVACCs into non-L-type and L-type channels, topiramate was tested at concentrations of 1, 10, and 50 µmol/L. Although the drug was without effect at 1 µmol/L, topiramate 10 µmol/L consistently decreased the peak current and area under the curve of L-type Ca2+ currents. This effect typically occurred within 10 minutes of perfusion and was partially or fully reversible within 5 minutes of the removal of the drug from the bathing medium. At 50 µmol/L, topiramate was less effective in modifying Ca2+ currents, with a greater effect on peak current than area. Topiramate did not affect the voltage sensitivity or the gating properties of the L-type channels. This study provides further evidence of the diverse mechanisms by which topiramate appears to modulate neuronal excitability and suggests that an inhibitory effect on L-type HVACCs is another potential anticonvulsant mechanism.
That topiramate produced a biphasic concentration-response curve, in which a greater reduction in L-type Ca2+ currents was seen at 10 µmol/L than at 50µmol/L, suggests a mode of action different from other Ca2+ channel blockers. Zhang and colleagues hypothesized that an indirect modulation of L-type HVACCs may occur through effects on intracellular second messenger systems or neurotransmitter receptors. Indeed, emerging evidence suggests that topiramate may bind to phosphorylation sites within AMPA, kainate, GABAA, and voltage-activated Na+ channels (40). Furthermore, HVACCs are partly regulated by protein phosphorylation, and the HVACC phosphorylation site has an amino acid sequence similar to those of the Na+ channel, GABAA receptor, and AMPA-kainate receptor (40). Whether this accounts for the U-shaped concentration curve observed in these studies is not known at the present time but is currently under investigation.
Inhibition of Carbonic Anhydrase Isoenzymes
At concentrations between 1 and 10 µmol/L, topiramate weakly inhibits the CA isoenzymes CA II and CA IV. When compared with acetazolamide, topiramate is approximately 10 to 100 times less potent as an inhibitor of these isoenzymes (9,39). Topiramate, like the antiepileptic drug and CA inhibitor acetazolamide, contains a sulfamate moiety that is likely to be responsible for its CA-inhibiting properties. Although inhibition of CA is generally not considered
to represent a significant anticonvulsant mechanism of topiramate, the possibility remains that this inhibition may contribute to its anticonvulsant action by modulation of pH-dependent voltage- and receptor-gated ion channels. For example, the increase in metabolic activity after high-frequency neuronal firing results in an increase in the intracellular concentration of bicarbonate that can alter net current flux and produce membrane depolarization (23,47). Given that the excitatory effect of bicarbonate may occur in the hippocampus, where some forms of epilepsy originate, it can be hypothesized that inhibitors of CA will block the excitation. The proposed mechanisms of action of topiramate are summarized in Table 77.2.
TABLE 77.2. PROPOSED MECHANISMS OF ACTION OF TOPIRAMATE
Evidence of Additional, Novel Mechanisms
Interaction with Protein Kinase Phosphorylation Sites
To date, topiramate has demonstrated activity at four main types of protein complex: voltage-activated Na+ channels; GABAA receptors, AMPA, and kainate receptors; and HVACCs and CA. The observed activity of topiramate at a wide range of ion channels and receptors, together with variable responses in different in vitro studies and preparations, has led to a hypothesis concerning an underlying common mechanism by which topiramate exerts its effects. Although each of the foregoing effects is reproducible, investigators have also observed a variable response to topiramate. For example, some studies report reversible effects, whereas others fail to observe a readily reversible action of topiramate. Topiramate's action has also been reported to be both immediate and delayed in onset, depending on the preparation studied. One hypothesis that has been proposed to help understand the variability often observed from preparation to preparation and from investigator to investigator suggests that the action of topiramate depends in part on the phosphorylation state of the receptor or ion channel (40).
In support of this hypothesis is the observation that the AMPA receptor, the GABAA receptor, the voltage sensitive sodium channel (VSSC), and the voltage-sensitive calcium channel (VSCC) are all regulated by protein phosphorylation mediated by protein kinase A, protein kinase C, and/or Ca2+-CaM-activated kinases (21,37,42,52). Furthermore, the peptide sequence at the protein kinase A-mediated phosphorylation site displays homology among these four molecular targets. According to the hypothesis, topiramate, if bound to the dephosphorylated state of the channel and/or receptor, may shift the channel and/or receptor further toward the dephosphorylated state by preventing protein kinase A from binding to the phosphorylation site. The consequence of this would be a delayed effect on channel conductance. Should this hypothesis prove accurate, the activity of topiramate would be predicted to be inversely related to the degree of phosphorylation of the protein complex (40).
Neurotherapeutic Disease-Modifying Properties
The results of the studies described earlier demonstrate that topiramate exerts either direct or indirect effects on various receptor- and voltage-gated ion channels that control neuronal excitability. In addition to epilepsy, all these molecular targets of topiramate have been associated with the underlying pathologic features of numerous central nervous system (CNS) disorders and insult-induced neuronal damage and death. To date, topiramate has been found effective in models of status epilepticus (31), cerebral ischemia (59,60), hypoxia-induced cell damage and resultant seizures (19), and periventricular leukomalacia (12). These promising results suggest that topiramate may have future utility in conditions other than epilepsy, including cerebral infarction
and hemorrhage, CNS trauma, neurodegenerative diseases such as Parkinson's disease and Alzheimer's disease, and neurologic conditions such as multiple sclerosis and encephalopathies. Furthermore, the similarities shared by epilepsy and certain CNS disorders, such as bipolar disorder, essential tremor, and migraine, and the diverse mechanisms of action of topiramate, provide a mechanistic basis supporting ongoing clinical trials in these and other disorders.
Investigators have suggested that GABAergic neurons may play a role in mood disorders, including both major depression and bipolar disorder. Preliminary evidence for this proposition came from a study in which progabide, a GABA agonist, was found to have a marked antidepressant effect (4). Further studies into the disease mechanisms involved in bipolar disorder (33,34) revealed that a kindling process occurs, presumed to be analogous to that seen in epilepsy. Patients initially experience mood-related episodes in response to life events, but eventually the neurologic and biochemical pathways responsible for these episodes are sufficiently reinforced to allow the autonomous initiation of further episodes. Because bipolar disorder and epilepsy are both episodic, it has been hypothesized that the anticonvulsant mechanisms of action of topiramate may also contribute to clinical activity in bipolar disorder. Indeed, results from an open-label study support ongoing clinical evaluation of topiramate in patients with mania (27).
Distinct similarities also exist between the pathophysiologic and biochemical processes responsible for epilepsy and neuropathic pain. In neuropathic pain, a phenomenon known as windup involves an increasing responsiveness to noxious stimuli in the pain-transmitting dorsal horn neurons and results in hyperexcitability of these neurons (49). This hyperexcitability then results in central sensitization, that is, an increased responsiveness of the spinal cord to neuronal pain impulses, which, in turn, leads to chronic pain at normally subthreshold levels of initial stimulation. In part, this hyperexcitability, similarly to kindling in epilepsy, results from activation of both NMDA and non-NMDA glutamate receptors by neuronally released glutamate, which can result from massive membrane depolarization mediated by VSSCs and VSCCs. From a mechanistic perspective, topiramate's demonstrated efficacy in a double-blind controlled study in diabetic neuropathy (10) could result from the drug's ability to normalize glutamate release (17), inhibit VSSCs and VSCCs, and block non-NMDA glutamate receptors.
Two further areas in which topiramate may be useful are essential tremor and migraine prophylaxis. Essential tremor is not yet widely understood at a pathophysiologic level, but it is thought that the illness may result from a dysfunction in GABA neurotransmission (25). Additional evidence suggests that inhibition of CA is a potentially useful strategy in this disorder (20). Because topiramate has distinct GABAergic actions and is also a weak inhibitor of CA, it may prove useful in essential tremor. Patients with migraine are also thought to benefit from drugs that act at GABAergic neurons, a finding suggesting that topiramate may also be useful in this population. In fact, two double-blind trials (11,35) and three open-label studies with topiramate (41, 22,51) have shown that the drug is effective both in the prophylaxis of migraine (with or without aura) and in the treatment of cluster headaches.
Investigations into the activity of topiramate in experimental epilepsy models and complementary electrophysiologic studies have suggested a range of mechanisms by which topiramate exerts its therapeutic effects. The broad anticonvulsant profile of topiramate that has emerged in both animal studies and clinical use is consistent with the multiple mechanisms of action described. Investigators have suggested that the effects of topiramate against SRF, voltage-sensitive Na+ channels, and non-NMDA receptors possibly account for an ability to prevent seizure spread and efficacy against secondary generalized partial seizures. Likewise, enhancement of GABA-mediated inhibition by topiramate may contribute to elevation of seizure threshold and possible efficacy against spike-wave seizures and partial epilepsy.
Much of this chapter focuses on the proposed molecular activities relating to the anticonvulsant action of topiramate in children and adults. However, the same mechanisms may also contribute to the growing body of clinical data suggesting efficacy against many different CNS disorders including migraine, bipolar disorder, essential tremor, and diabetic neuropathy. Furthermore, given the similarities that have been identified between the molecular targets of topiramate and the underlying mechanisms of epileptogenesis, focal and global ischemia, and neonatal hypoxia, it seems reasonable, based on the data obtained to date, to suggest that topiramate may have potential disease-modifying properties.
As the specific molecular actions of topiramate are further elucidated and more is understood regarding the relationship that appears to exist among the individual receptor and ion channel effects, protein phosphorylation, and modulation of intracellular second messenger systems, the potential clinical applications of the compound will become increasingly apparent. Early indications are that topiramate will have therapeutic potential in many different disease states in which effective treatment is urgently needed. More important is the potential utility of topiramate as a disease-modifying neurotherapeutic drug that targets the underlying pathologic process contributing to the expression of numerous progressive CNS disorders.