Medicinal compounds with antiepileptic/anticonvulsant activities




Epilepsy is a serious neural disease that affects around 50 million people all over the world. Although for the majority patients with epilepsy, seizures are well controlled by currently available antiepileptic drugs (AEDs), there are still >30% of patients suffered from medically refractory epilepsy and approximately 30–40% of all epileptic patients affected by numerous side effects and seizure resistance to the current AEDs. Therefore, many researchers try to develop novel approaches to treat epilepsy, for example, to discover new antiepileptic constituents from herbal medicines. Although there are already several reviews on phytotherapy in epilepsy, most of them placed emphasis on the plant crude extracts or their isolated fractions, not pure active compounds derived from herbal medicines. This article aims to review components in herbal medicines that have shown antiepileptic or anticonvulsant properties.


We searched online databases and identified articles using the preset searching syntax and inclusion criteria. The active medicinal compounds that have shown anticonvulsant or antiepileptic activity were included and classified according to structural types.


We have reviewed herein the active constituents including alkaloids, flavonoids, terpenoids, saponins, and coumarins. The screening models, the seizures-inducing factors and response, the effective dose, the potential mechanisms, as well as the structure-activity relationships in some of these active components have also been discussed.


The in vitro and in vivo experimental data reviewed in this paper would supply the basic science evidence for research and development of novel AEDs from medicinal plants.

Epilepsy, which has a high prevalence among people of all ages, is a serious and diverse set of chronic neurologic disorders characterized by spontaneously occurring seizures. It has been estimated that 50 million people worldwide live with epilepsy, and that >85% of this disease occurs in low-income and lower middle-income countries.[1, 2] Although seizures are controlled successfully with currently available antiepileptic drugs (AEDs) in most patients with epilepsy, >30% of patients still have medically refractory epilepsy.[3] Furthermore, there are still about 30–40% of all epileptic patients affected by numerous side effects and seizure resistance to the current AEDs[4]). These facts motivate researchers to develop novel approaches to treat epilepsy, for example, to discover new antiepileptic constituents from herbal medicines.[5, 6]

Since antiquity, many kinds of herbs have been used in folk medicines to treat epilepsy and convulsions. In nine of the most important European herbals of the 16th and 17th centuries, 221 plants from 53 plant families were described for their use as remedies for treating epilepsy.[7] By evaluating the prophylactic and therapeutic remedies used by folk medicine to cure epilepsy in Italy between the late 19th and mid-20th century, the potential role of 12 plants in the treatment of epilepsy was identified.[8] Their antiepileptic effects can be explained by direct or indirect pharmacologic mechanisms, that is, the ability to interact with central nervous system (CNS) receptors like γ-aminobutyric acid receptor type A (GABAA), anxiolytic or sedative properties, and the intervention of neurotrophic factors in the recovery of the physiologic excitability of neurons membrane after seizure. In China, some traditional Chinese medicines have been used for the treatment of a variety of epileptic syndromes for thousands of years. By reviewing and analyzing the related experimental studies, 15 herbal drugs, such as Acorus tatarinowii Schott., Bupleurum chinense DC., Ligusticum chuanxiong Hort., Paeonia suffruticosa Andr., and Salvia miltiorrhiza Bge. have been discovered with definite antiepileptic activity.[9] And in Africa, among the 60 plants that are commonly used to treat epilepsy by Tanzanian traditional healers, Abrus precatorius L., Clausena anisata (Willd.) Oliv., and Hoslundia opposita Vahl. have shown definitely anticonvulsant activity from the experimental literature reports.[10] Although there are already several reviews on phytotherapy in epilepsy, most of them placed emphasis on the plant crude extracts or their isolated fractions, not pure active compounds derived from herbal medicines.[5, 6, 10] Probably due to the complexity of extracts, few of these reviews paid attention to the mechanisms with respect to the antiepileptic activities, which are important for research and development of novel AEDs from medicinal plants.

Therefore, the main purpose of the present work is to review single or one kind of components in herbal medicines that have been reported to treat epilepsy. Components with similar structures are gathered to see if their antiepileptic/anticonvulsant activities and proposed mechanisms have something in common. In this context, the active constituents that have shown anticonvulsant or antiepileptic activity are classified by structural types, namely alkaloids, flavonoids, terpenoids, saponins, and coumarins.

Medicinal Compounds with Antiepileptic/Anticonvulsant Activities


Aconitum alkaloids are a series of diterpene alkaloid neurotoxins that interact with the voltage-dependent Na+ channel. Because this channel has an exceptional relationship with neuronal excitability, Ameri's team has investigated the effects of Aconitum alkaloids on the CNS, especially their potent antiepileptiform activity in vitro on rat hippocampal slices.[11-18] These experiments revealed that benzoyl ester may be an important active center for this kind of anticonvulsant. Most of the tested Aconitum alkaloids displayed inhibition of the epileptiform activity, except for those that lack a benzoyl ester side chain, such as napelline,[11] heteratisine,[14] lappaconidine,[13] and talitasamine.[16] It was interesting to note that the position of benzoyl ester at the diterpene skeleton was also an important factor in determining the antiepileptiform effect and mechanism. For example, aconitine[17] and its derivative 3-acetylaconitine,[12] the main alkaloids in plants of the Aconitum genus, bear a benzoyl ester in the C-14 position. They suppressed both normal neuronal activity and epileptiform activity, whereas 1-benzoylnapelline,[11] lappaconitine,[13, 18] N-deacetyllappaconitine,[13] and 6-benzoylheteratisine[14] possess a benzoylester side chain at C1, C4, C4, and C6, respectively, and mainly attenuated epileptiform activity. Particularly, due to the inactivation of the Na+ channel and the activation of current already at resting membrane potential, the inhibitory action of aconitine or 3-acetylaconitine was preceded by a transient initial excitation. In addition, recent study demonstrated the potential proconvulsant effect of aconitine, the highly potent neurotoxin, which showed acute and prolonged excitatory effects on neocortical activity in neocortical brain slices of the juvenile Sprague-Dawley (SD) rat.[19]

Apart from diterpenoid Aconitum alkaloids, there are many other kinds of alkaloids that have been reported with antiepileptic or anticonvulsant activities, including isoquinoline alkaloids such as berberine,[20] montanine,[21] and tetrahydropalmatine[22]; indole alkaloids such as ibogaine,[23] piperidine alkaloids such as piperine,[24] amide alkaloids such as piplartine,[25] tetracyclic oxindole alkaloids such as rhynchophylline and isorhynchophylline,[26] aporphine alkaloids such as nantenine,[27] erythrine byproducts such as erysothrine,[28] (+)-erythravine and (+)-11-α-hydroxy-erythravine,[29] as well as raubasine.[30] Anticonvulsant activities of the above-mentioned alkaloids have been investigated in one or more of the models including pentylenetetrazole (PTZ), maximal electroshock (MES), kainic acid (KA), and bicuculline, pilocarpine, or N-methyl-d-aspartate (NMDA)–induced convulsions. Although both montanine and berberine exhibited anticonvulsant activity by modulating neurotransmitter systems, montanine protected against PTZ-provoked convulsion while berberine did not show activity in the PTZ test.[20, 21] Faggion et al. have evaluated the action of erythrine byproducts against seizures evoked by bicuculline, PTZ, NMDA, and KA in Wistar rats. Erysothrine reduced all of these convulsions,[28] whereas (+)-erythravine and (+)-11-α-hydroxy-erythravine were inactive in NMDA- and PTZ-induced seizures, respectively.[29] This indicated that even those compounds who share the same basic structure may have different mechanisms or activities. Piplartine in the PTZ-induced convulsion model decreased the latency to death in mice.[25] At lower doses, nantenine reduced PTZ- and MES-induced seizures, which were probably attributed to nantenine’s stimulation and the resultant decrease of Ca2+ influx into the cell.[27] Piperine, which significantly increased latencies to first convulsion and to death, as well as percentage of survival in pilocarpine-induced convulsive mice, showed effects on inhibitory amino acids and on the GABAergic system.[24] Raubasine attenuated PTZ- and bicuculline-induced convulsions in mice; this may well be due to its benzodiazepine agonist–type activity.[30] Furthermore, anticonvulsant activity of ibogaine has been studied both in vitro through voltage clamp recording, and in vivo using MES- and NMDA-induced models. The results showed that ibogaine has NMDA-receptor blocking activity.[23] Using Xenopus oocytes injected with total RNA prepared from rat cortices or cerebelli as the model, the antiepileptic effects of rhynchophylline and isorhynchophylline were investigated; the effects were attributable to their role as noncompetitive antagonists of the NMDA receptor.[26] dl-Tetrahydropalmatine (THP), the antiseizure activity of which has been tested on the development of electrically kindled amygdala and quantified by an ultrasonic system for vertical motion measurements, was suggested to be an effective antiepileptogenic and anticonvulsant agent.[22] The structures and the action mechanisms of alkaloids with antiepileptic/anticonvulsant activities are shown in Fig. 1 and Table 1, respectively.

Table 1. Alkaloids with antiepileptic/anticonvulsant activities in medicinal plants
No.CompoundsMedicinal herbsModelsSeizure-inducing factors and responseEffective doseProposed mechanismsReferences
  1. ACSF, artificial cerebrospinal fluid; GABA, γ-aminobutyric acid; KA, kainic acid; MES, maximal electroshock; NMDA, N-methyl-d-aspartate; PTX, picrotoxin; PTZ, pentylenetetrazole; +, having positive or significant response; −, having negative or no significant response.

1AconitineAconitum speciesMale Wistar rats (6–7 weeks old)Low Mg2+-ACSF, PTZ, bicuculline, PTX, penicillin (+)0.1 and 1 μmPossibly blocking GABAA-mediated inhibition and NMDA receptors [17]
2MesaconitineAconitum speciesMale Wistar rats (150–180 g)Low Mg2+-ACSF (+)0.03–1 μmActivator of α-adrenoceptors [15]
33-Acetylaconitine Aconitum flavum Male Wistar rats (150–180 g)Low Mg2+-ACSF, bicuculline (+)0.01–1 μmPossibly blocking GABAA-mediated inhibition and NMDA receptors [12]
414-BenzoyltalitasamineAconitum speciesMale Wistar rats (150–180 g)High K+/low Mg2+-ACSF (+)0.3–10 μmPossibly blocking NMDA receptors [16]
5LappaconitineAconitum speciesMale Wistar rats (150–180 g)Low Mg2+-ACSF, bicuculline (+)10 μmBlocking sodium channels [13, 18]
6N-DesacetyllappaconitineAconitum speciesMale Wistar rats (150–180 g)Low Mg2+-ACSF, bicuculline (+)3–100 μmUnknown [13]
71-BenzoylnapellineAconitum speciesMale Wistar rats (150–180 g)Low Mg2+-ACSF, bicuculline (+)1–100 μmUnknown [11]
86-BenzoylheteratisineAconitum speciesMale Wistar rats (150–180 g)Low Mg2+-ACSF, bicuculline (+)10 μmBlocking sodium channels [14]
9Berberine Berberis vulgaris Adult male Swiss albino mice

PTZ (−)

MES, KA (+)

10 and 20 mg/kgModulating neurotransmitter systems [20]
10Montanine Hippeastrum vittatum Swiss albino mice (25–30 g) and Wistar rats (250–300 g)PTZ (+)30 and 60 mg/kgModulating neurotransmitter receptor systems, including GABAA receptors [21]
11dl-Tetrahydropalmatine Corydalis Male SD rats (250–350 g)Electrical amygdaloid kindling (+)20 and 30 mg/kgReducing dopamine output [22]
12Ibogaine Tabernanthe iboga

In vivo: Male NIH Swiss mice (25–30 g)

In vitro: Cultured hippocampal neurons of SD rats (19 days old)


KA, GABA (−)

ED50 for MES 31 mg/kg

IC50 for NMDA 3.1 μm

Blocking NMDA receptors [23]
13PiperinePiper speciesMale Swiss mice (25–30 g)Pilocarpine (+)2.5, 5, 10 and 20 mg/kgAntiinflammatory, antioxidant and TNF-α reduction [24]
14Piplartine Piper tuberculatum Male Swiss mice (25 g)PTZ (+)50 and 100 mg/kgUnknown [25]
15RhynchophyllineUncaria speciesXenopus oocytesNMDA (+)30 μmNoncompetitive antagonists of the NMDA receptor [26]
17NantenineVegetal speciesMale albino mice (30–35 g)PTZ, MES (+)20 and 50 mg/kgDecreasing Ca2+ influx into the cell [27]
18Raubasine Rauwolfia serpentina Male Swiss mice (20–24 g)PTZ, bicuculline (+)ED50: 57.5 and 40.6 mg/kgInteracting at benzodiazepine sites with a benzodiazepine agonist-type activity [30]
19Erysothrine Erythrina mulungu Male Wistar rats (200–250 g)Bicuculline, PTZ, NMDA, KA (+)0.5–3 μg/μlUnknown [28]
20(+)-Erythravine Erythrina mulungu Male Wistar rats (200–250 g)Bicuculline, PTZ, KA (+); NMDA (−)2 and 3 μg/μlUnknown [29]
21(+)-11-α-Hydroxy-erythravineBicuculline, NMDA, KA (+); PTZ (−)1 and 2 μg/μl
Figure 1.

Chemical structures of alkaloids with antiepileptic/anticonvulsant activities in medicinal plants.


Flavonoids are a class of natural constituents that are widely distributed in plants with many pharmacologic properties including antiseizure properties. Medina et al.[31] have found that chrysin was able to prevent the expression of tonic–clonic seizures induced by PTZ through central benzodiazepine (BZD) receptors. Apigenin, which bears one more hydroxyl group on ring B than chrysin, could reduce the latency in the onset of PTX-induced convulsions.[32] The addition of a methoxy group to the C-8 position of chrysin produces wogonin, which could significantly block convulsion induced by PTZ and electroshock.[33] More importantly, wogonin did not produce sedation and myorelaxation in locomotor activity and the rota-rod test, whereas chrysin and apigenin exerted myorelaxant and sedative properties, respectively. From this point of view, wogonin might be a safer antiepileptic reagent because sedation and myorelaxation have been often regarded as the adverse effects of AEDs. Fisetin is a tetrahydroxy flavonoid, which contains one or more hydroxyl groups on the basic C6-C3-C6 skeleton than chrysin, apigenin, and wogonin. Recent studies found that fisetin could reduce both electrical kindling seizures in rats and acute seizures in PTZ, strychnine, isoniazid, or MES tests in mice. And the anticonvulsant effect may be ascribed to protection of endogenous enzyme level, increase of the GABA level in brain, and inhibition of oxidative injury.[34]

In addition to these aglycones, some flavonoid glycosides have showed anticonvulsant or antiepileptic effects in different manners. For example, baicalin, one of the major flavonoids in the famous traditional Chinese medicine Scutellaria baicalensis Georgi, has been discovered to possess remarkable anticonvulsant and neuroprotective effects on pilocarpine-evoked status epilepticus (SE) in adult SD rats.[35] Moreover, Abbasi et al.[36] have carried out studies on the anticonvulsant effects of vitexin in rats treated with PTZ and clarified the underlying mechanism. By increasing the seizure-onset time, vitexin affected PTZ-induced minimal clonic seizures (MCS) and generalized tonic–clonic seizures (GTCS). The results showed that vitexin exerted its anticonvulsant effects probably through binding at the benzodiazepine site of the GABAA-receptor complex. Rutin, one of the most common flavonoid glycosides that existed in many plants, fruits, and vegetables, also showed anticonvulsant effects. Intracerebroventricular (i.c.v.) injection of rutin dose-dependently diminished PTZ-induced MCS and GTCS. However, this effect could be abolished by pretreatment of flumazenil, a central benzodiazepine-receptor antagonist, which revealed that the GABAA-benzodiazepine receptor complex may be the target of rutin.[37] Because memory impairment is a common complication of patients with localized epilepsies, Nassiri-Asl further explored the effect of rutin on learning and memory.[38] Using a step-through passive avoidance task, the potential role of rutin in enhancing memory retrieval in PTZ-kindled rats was discovered. Structures and details of above-mentioned flavonoids are listed in Figure 2 and Table 2, respectively.

Table 2. Flavonoids with antiepileptic/anticonvulsant activities in medicinal plants
No.CompoundsMedicinal herbsModelsSeizure- inducing factors and responseEffective doseProposed mechanismsReferences
  1. ACSF, artificial cerebrospinal fluid; GABA, γ-aminobutyric acid; ICR, Institute of Cancer Research; KA, kainic acid; MES, maximal electroshock; NMDA, N-methyl-d-aspartate; PTX, picrotoxin; PTZ, pentylenetetrazole; +, having positive or significant response; −, having negative or no significant response.

1ChrysinPassiflora caerulea L.Swiss mice (22–28 g)PTZ (+)40 μgActing on central BZD receptors [31]
2ApigeninMatricaria chamomilla L.Male SD rats (180–200 g)PTX (+)25 and 50 mg/kgInteracting on neurotransmission systems other than GABAA [32]
3Wogonin Scutellaria baicalensis

Male SD rats (8–10 weeks old)

Male Institute of Cancer Research (ICR) mice (20–25 g)

Electroshock, PTZ (+)

strychnine (−)

5 and 10 mg/kgPotentiation of the activity of GABA [33]
4RutinPlants such as buckwheat, apples, and black teaMale Wistar rats (250–300 g)PTZ (+)50 and 150 nmInteracting with GABAA-benzodiazepine receptor complex [37, 38]
5VitexinPlants such as Passiflora sp., bamboo leaves, pigeon pea leaves and mung beanMale Wistar rats (250–300 g)PTZ (+)100 and 200 μmInteracting with GABAA-benzodiazepine receptor complex [36]
6BaicalinScutellaria baicalensis GeorgiMale SD rats (180–200 g)Pilocarpine (+)100 mg/kgAntioxidative effect [35]
7FisetinStrawberries, apple, grape onion, persimmon and cucumber

Male Swiss albino mice (18–22 g)

Male Wistar rats (180–200 g)

PTZ, strychnine, isoniazid, MES, electrical kindling (+)10 and 25 mg/kgProtecting endogenous enzyme level, inhibiting oxidative damage and modulating GABAergic transmission [34]
Figure 2.

Chemical structures of flavonoids with antiepileptic/anticonvulsant activities in medicinal plants.


Terpenoids, sometimes called isoprenoids, are a large class of natural products consisting of five-carbon isoprene units. Plant terpenoids are now under investigation for multiple pharmaceutical functions including anticonvulsant activities (as indicated in Fig. 3). Most of these compounds are monoterpenes. Citronellol, an acyclic monoterpene alcohol, existed in essential oils of many aromatic plants, and showed protective effects against PTZ- and PTX-induced convulsions and MES-induced seizures in mice. Using the single sucrose-gap technique, de Sousa team deduced that reduction of neuronal excitability mainly via the voltage-dependent Na+ channels may be one of its anticonvulsant mechanisms.[39] They found that (S)-(+)-carvone at the dose of 200 mg/kg increased significantly the latency of PTZ- and PTX-induced convulsions, whereas its enantiomer (R)-(-)-carvone had no effect on the onset of convulsions.[40] α-Terpineol, a relatively common monoterpene present in many medicinal plants, showed protective effects on PTZ- and MES-induced convulsive seizures in mice.[41] Using the single sucrose gap technique and PTX test, de Sousa’s team revealed that the anticonvulsant activity of α,β-epoxy-carvone probably due to the modulation of GABAergic system and reduction of neuronal excitability by blocking the voltage dependent Na+ channels.[42] They also compared the anticonvulsant activities of linalool enantiomers in PTZ, PTX, and MES models and demonstrated that (R)-(-)-linalool was more potent than its enantiomeric form.[43] The administration of γ-decanolactone at a dose of 0.3 g/kg showed anticonvulsant properties and antiepileptogenic effects in the PTZ-kindling model, which could also decrease the DNA damage in brain tissue both at 0.1 and 0.3 g/kg.[44] Isopulegol, the product by cyclization of citronellol, seems to interact with a different target. Isopulegol exhibited anticonvulsive effects in PTZ-induced models by positive modulation of benzodiazepine-sensitive GABAA receptors and antioxidant properties.[45] Safranal, an active monoterpene aldehyde in Crocus sativus L., had beneficial effect on PTZ-induced convulsions in mice by interacting with GABAA-benzodiazepine receptor complex.[46, 47] Carvacrol and borneol, two monoterpenes present in the essential oils of numerous medicinal plants, have been demonstrated to be effective in preventing clonic seizures induced by PTZ and tonic convulsions induced by MES. And the effects of borneol were probably observed through the modulation of the GABAergic system.[48] Eugenol, a phenolic monoterpene extracted from cloves, has been found to have various uses in medicine including ameliorating epileptic seizures because of its modulating effect on neuronal excitability. Huang et al.[49] further revealed that the anticonvulsant effect of eugenol may be associated primarily with its specific effect on ionic currents, namely increasing the degree of voltage-gated Na+ currents (INa) inactivation and suppressing the noninactivating INa (INa (NI)).

Figure 3.

Chemical structures of terpenoids with antiepileptic/anticonvulsant activities in medicinal plants.

Some diterpenes and their derivatives also exhibited good anticonvulsive effects in preclinical experiments. Phytol could reduce pilocarpine-induced seizures and SE probably by modulating other systems neurotransmitters rather than the GABAergic system.[50] As a potential anticonvulsant agent, abietic acid has been discovered using a special and modern method. Linear discriminant analysis based on two-dimensional (2D) descriptors has selectively revealed abietic acid from a mass of candidate compounds, and MES test confirmed its effect.[51] Moreover, three isomeric cannabinoids (a group of 21-carbon–containing terpenophenolic compounds) derived from Cannabis sativa—namely delta-8-tetrahydrocannabinol (∆8-THC), delta-9-tetrahydrocannabinol (∆9-THC), and cannabidiol (CBD)—have been reported to possess anticonvulsant activities. Both ∆8-THC and ∆9-THC dose-dependently protected against tonic extension induced by MES in rats. However, only at very high or even lethal doses could they produce a protective effect on PTZ-induced clonic convulsions. Moreover, there was no significant difference in the potency of the two isomers in these tests, which revealed that the position of the double bond in the alicyclic ring might have little impact on their anticonvulsant activities.[52] Opening of the pyran ring of ∆9-THC produces CBD, the anticonvulsant and antiepileptogenic effects of which have been examined and confirmed using in vitro and in vivo models in recent years.[53, 54] Melisa et al. have assessed the role of CB1 receptors in cannabinoid anticonvulsant effects.[55] They discovered that the protective effect of ∆9-THC against MES-induced seizures in mice was cannabinoid CB1 receptor-mediated, while the anticonvulsant effect of CBD was not mediated by CB1 receptor activation. This may be attributed to the fact that CBD has much lower affinity for CB1 and CB2 receptors than ∆9-THC.[56]

Ursolic acid, a pentacyclic triterpenoid that is widely distributed in plants, has been discovered to have anticonvulsant activity because it reduced the number and lethality of PTZ-induced seizures in mice.[57] Ursolic acid stearoyl glucoside, a terpenoid isolated from Lantana camara L., also exhibited a anticonvulsive effect.[58] This effect, possibly mediated by facilitation of GABA transmission, has been evaluated in MES- and isoniazid-induced seizures in rats. Saikosaponin a (SSa), a triterpene saponin derived from Bupleurum chinensis DC., has been evaluated for its anticonvulsant activities in the hippocampal neuronal culture (HNC) models of acquired epilepsy (AE) and SE through whole-cell current-clamp recordings. The results indicated that the inhibition of NMDA-receptor current and persistent sodium current may be the anticonvulsant mechanism of SSa.[59] Another triterpenoid compound, Baccoside A, could suppress epileptic-like seizure or convulsion, probably through interacting with the T-type calcium channel (CCA-1) protein. Unlike most of the experiments using rodents as materials, anticonvulsant activity of Baccoside A has been conducted in Caenorhabditis elegans (informally known as “the worm”), and epileptic-like seizures have been induced by gradually increasing temperatures.[60]

In addition, Sasaki et al. have previously demonstrated a sesquiterpene compound, bilobalide, which can exert anticonvulsant activities against certain types of convulsions, including isoniazid or 4-O-methylpyridoxine induced seizures. To identify the underlying mechanisms, they have conducted experiments to examine the level of GABA and glutamate, indicating that bilobalide could elevate GABA levels, possibly through potentiation of GABA synthesis.[61, 62] Terpenoids with anticonvulsant or antiepileptic activities are listed in Table 3.

Table 3. Terpenoids with antiepileptic/anticonvulsant activities in medicinal plants
No.CompoundsMedicinal herbsModelsSeizure- inducing factors and responseEffective doseProposed mechanismsReferences
  1. ACSF, artificial cerebrospinal fluid; GABA, γ-aminobutyric acid; KA, kainic acid; MES, maximal electroshock; NMDA, N-methyl-d-aspartate; PTX, picrotoxin; PTZ, pentylenetetrazole; SREDs, spontaneous recurrent epileptiform discharges; +, having positive or significant response; −, having negative or no significant response.

  2. a

    Only at very high or even lethal doses.

18-THC Cannabis sativa Male hooded rats (300–480 g)MES (+)ED50 for MES 72 mg/kgUnknown [52]
29-THCPTZ (+)aED50 for MES 58 mg/kg
3Cannabidiol Cannabis sativa In vitro: Hippocampal brain slicesMg2+-free, PTZ, 4-aminopyridine (+)

In vitro: 100 μm

In vivo: 100 mg/kg

Acting on multiple molecular targets in neuronal excitability [54]
In vivo: male Wistar Kyoto rats (70–110 g)  
Male Wistar Kyoto ratsPilocarpine, penicillin (+)1, 10, 100 mg/kgThree proposed mechanisms, details see the original paper [53]
4Ursolic acidNepeta sibthorpii BenthamAdult male Swiss mice (20–25 g)PTZ (+)2.3 mg/kgUnknown [57]
5Ursolic acid stearoyl glucosideLantana camara L.Wistar albino rats (150–200 g)MES, isoniazid (+)25 and 50 mg/kgIncreasing the GABA level in central nervous system [58]
6Saikosaponin aBupleurum chinensis DC.Cultured hippocampal neurons from 1 to 2 days postnatal SD ratsLow Mg2+-ACSF (+)IC50 for spontaneous recurrent epileptiform discharges (SREDs) and SE: 0.42 and 0.62 μmInhibiting NMDA receptor current and persistent sodium current [59]
7Baccoside A Bacopa monnieri Caenorhabditis elegans (worms)Temperature (+)0.25%, 0.1% and 0.01%Interacting with the T-type calcium channel (CCA-1) protein [60]
8PhytolMale Swiss mice (25–30 g; 2 months old)Pilocarpine (+)25, 50 and 75 mg/kgModulating other system neurotransmitters (serotonin, noradrenergic, and glutamatergic) [50]
9Abietic acidPine resinSwiss mice (18–23 g)

MES (+)

PTZ (−)

30 and 100 mg/kgUnknown [51]
10BilobalideGinkgo biloba L.Male ddY strain mice (4 weeks old)4-O-methylpyridoxine (+)30 mg/kgIncreasing GABA levels by potentiation of GABA synthesis [61, 62]
11CitronellolAromatic plant speciesAdult male albino Swiss mice (24–30 g)PTZ, PTX, MES (+)400 mg/kgInhibiting neuronal excitability through the voltage-dependent Na+ channels [39]
12(S)-(+)-Carvone Mentha spicata Male Swiss mice (28–34 g)PTZ, PTX (+)200 mg/kgUnknown [40]
13α-terpineolPlant essential oilsMale Swiss mice (28–35 g)PTZ, MES (+)100, 200 and 400 mg/kgUnknown [41]
14α,β-Epoxy-carvonePlant essential oils

Male Swiss mice (28–35 g)

Male Wistar rats (230–350 g)

PTZ, MES, PTX (+)200, 300 and 400 mg/kgInhibiting neuronal excitability through the voltage-dependent Na+ channels [42]
15(R)-(-)-LinaloolPlant essential oilsAdult male Swiss mice (24–30 g)PTZ, PTX, MES (+)200 and 300 mg/kgUnknown [43]
16γ-Decanolactone Aeollanthus suaveolens Male SR-1 mice (25–30 g)PTZ (+)300 mg/kgUnknown [44]
17Carvacrol borneolPlant essential oilsMale Swiss mice (30–35 g)PTZ, MES (+)

200 mg/kg

50, 100 and 200 mg/kg


Enhancing GABAA-BZD receptor

18EugenolPlant essential oilsAdult male SD rats (175–200 g)Pilocarpine (+)100 mg/kgIncreasing INa inactivation and suppressing the INa (NI) [49]
19IsopulegolPlant essential oilsMale Swiss mice (20–30 g)PTZ (+)200 mg/kgPositive modulation of benzodiazepine-sensitive GABAA receptors and antioxidant [45]
20SafranalCrocus sativus L.Adult male Wistar rats (250–300 g)PTZ (+)

145.5 mg/kg

0.15 and 0.35 ml/kg

Interacting with GABAA-benzodiazepine receptor complex [46, 47]


Cynanchum otophyllum Schneid (also known as Qingyangshen), is one of the most important medicinal plants used in southwest of China to treat epilepsy. A product based on Cynanchum otophyllum extract has already been developed by the Yunnan Baiyao Group together with the Kunming Institute of Botany, Chinese Academy of Sciences. Recent studies have demonstrated that otophylloside A and B, two C-21 steroidal saponins, are its main active constituents that exhibit anticonvulsant activities.[63, 64] Apart from the pure compounds, some saponin extracts are considered nowadays as potential bioactive agents in treating epilepsy. To test the anticonvulsant activity of extracts from different parts of American ginseng, seizures were induced in rats by KA, pilocarpine, or PTZ. Only a partially purified extract that concentrates the Rb1 and Rb3 ginsenosides (Rb extract) showed significant effect.[65] Jalsrai et al. explored the anticonvulsant effects of the saponin fraction obtained from Astragalus mongholicus (AM) on acute PTZ-induced seizures. The results showed that AM saponin components are useful in the treatment of convulsive disorders, although their action mechanisms remain unclear.[66] Saponin rich fraction (SFG) of Ficus platyphylla stem bark also showed anticonvulsive effects on PTZ- and strychnine-induced seizures in vivo, but failed to protect mice against MES test; neither abolished the spontaneous discharges induced by 4-AP in vitro. Owing to properties shared by most voltage-gated sodium channel blocking drugs, SFG of this herb showed promise for being developed into anticonvulsant drugs.[67]


Coumarins are a group of plant-derived polyphenolic compounds and are made of fused benzene and α-pyrone rings, many of which possess a wide range of pharmacologic and biochemical applications. Esculetin (6,7-dihydroxy-coumarin) could significantly decrease seizure response induced by electroshock, whereas it did not cause a sedative and myorelaxation effect, and it exerted anticonvulsant effect probably through the GABAergic neuron.[68] Luszczki et al. investigated the protective effects of imperatorin and osthole (two natural coumarin derivatives) against MES-induced convulsions in mice. First, they discovered that the antiseizure effects produced by imperatorin was dose dependent, and produced its maximum anti-electroshock action at 30 min after its i.p. administration.[69] A similar time-course of anti-electroshock action was also observed in the osthole test. The maximum of its anticonvulsant effect was observed at times ranging between 15 and 30 min after its i.p. administration.[70] Further investigation found that the protective index value (median toxic dose/median effective dose; TD50/ED50) of imperatorin or osthole was comparable to that of valproate, which revealed that these two active coumarins warrant further evaluation as promising therapeutic agents against seizures.[71] Of interest, a comparative study on the anticonvulsant effects of four linear furanocoumarins revealed that the compounds contained substitutions at the C-8 position of the psoralen ring (imperatorin and xanthotoxin) and exerted strong anticonvulsant activity, whereas bergapten and oxypeucedanin (C-5 substituted derivatives of psoralen) did not produce such action.[72] A similar result was also observed in the research on the anticonvulsant activities of some furanocoumarins from the fruits of Heracleum crenatifolium. Compared with bergapten that contained a C-5 substituent, isopimpinellin and byak-angelicol both possess a substituent at the C-8 position; therefore, they showed much stronger anticonvulsant activities than bergapten.[73] Recently, Singhuber et al. [74] systematically analyzed 18 structurally diverse coumarin derivatives for GABA-induced chloride current (IGABA) enhancement. Of interest, their experiments indicated that the C-5 side chain rather than the C-8 side chain represents a structural requirement for IGABA modulation. For example, heraclenin and oxypeucedanin bear an epoxylated oxyprenyl residue at the C-8 position and C-5 position, respectively; and heraclenin showed a >10-fold loss of activity (31% vs. 547%). However, considering the effect of the tested coumarins occured at very high concentrations (100 μM), the authors especially pointed out that coumarins exert their anticonvulsant effects may additionally interact with other receptors while not exclusively via the GABAA receptor, as it was difficult to ensure if such a high concentration can enter the brain.


At present, although many kinds of herbs have been used to treat epilepsy and many researchers have already focused their attention on them, most of these studies placed emphasis on the plant crude extracts or their isolated fractions. The quality and quantity of active ingredients, as well as the mechanisms with respect to the antiepileptic activities were often unknown, which hindered the discovery of novel AEDs from medicinal plants. Based on the limited literature on single or pure components in herbal medicines that have been reported to treat epilepsy, the present work summarizes several kinds of natural components with anticonvulsant or antiepileptic activities, including alkaloids, flavonoids, terpenoids, saponins, and coumarins. The screening models, the seizure-inducing factors and response, the effective dose, as well as the potential mechanisms underlying the anticonvulsant or antiepileptic effects have been reviewed. Moreover, the structure-activity relationships in some of these active components have also been discussed herein.

However, we should also note that the limitations in some of these studies. First, although various chemically or electrically induced models have been used to evaluate the anticonvulsant activities of medicinal compounds in most studies, some still predicted the antiepileptic potential of a drug from a single model. Because different models could simulate dissimilar kinds of seizures or epilepsy, selecting the appropriate model is a key factor in AED screening in the case of false-positive or false-negative results. Furthermore, prediction of a drug's efficacy on epilepsy cannot rely on one chronic model, but rather a battery of models should be used.[75] Second, some of these studies lacked clear or definable seizure end points; they merely detected weak anticonvulsant effects, for example, effects on seizure latency.[20, 40, 45] In our opinion, more reliable end points should be chosen for anticonvulsant drug evaluation. For example, in MES-induced seizures, the criterion for the occurrence of seizure activity was the tonic hindlimb extension, with ED50 (the dose protecting 50% of animals tested against tonic hindlimb extension) as the end point of the test.[76] With respect to subcutaneous (s.c.) PTZ-induced seizures, the first “threshold seizure” (i.e., a repeated clonic seizure of forelimbs and/or hindlimbs with a seizure duration of at least 5 s, but without loss of righting reflexes) appears to be the most reliable end point, whereas in intravenous PTZ infusion test, the initial myoclonic twitch should be used as the end point.[77]

Today, there are still about 30–40% of all patients with epilepsy affected by intolerable toxicity, numerous adverse effects (such as sedation and myorelaxation), and seizure resistance to the current AEDs, which motivate researchers to develop novel approaches to treat epilepsy. Because some medicinal compounds reviewed herein, such as esculetin [68] and wogonin,[33] significantly decreased seizure response but did not cause sedative and myorelaxant effects, it is evident that medicinal plants have the potential to be a rich source for discovering safer and more effective antiepileptic reagents. Although there are insufficient data to support the clinical efficacy of these herbal components in patients with epilepsy—only some phytocannabinoids such as CBD [78, 79] and ∆9-THC [80] have individual case studies or have undergone a number of small-scale human trials—the in vitro and in vivo experimental data reviewed herein could supply the basic science evidence for research and development of novel AEDs from medicinal plants.


This study was supported by the National Natural Science Foundation of China (31160065), the Macao Science and Technology Development Fund (052/2012/A2) and the Research Committee of the University of Macau (SRG010-ICMS12-LP).


None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.


  • Image of creator

    Ms. Hui-Ling Zhu is a master__s student at University of Macau, whose major research focuses on discovery antiepileptic or anticonvulsant compounds from natural products including traditional Chinese medicine.