- Top of page
Summary: Purpose: To test the anticonvulsant activity of three preparations of American ginseng: whole root extract, whole leaves/stems extract, and a partially purified extract that concentrates the Rb ginsenosides (Rb extract).
Methods: One hour after treatment with normal saline, or one of the three ginseng preparations, seizures were induced in adult, male, Sprague–Dawley rats with kainic acid (KA; 10 mg/kg), pilocarpine (300 mg/kg, preceded by methylscopolamine, 1 mg/kg, s.c.), or pentylenetetrazol (PTZ, 50 mg/kg). Time to onset of seizure activity, duration of seizure activity for PTZ, seizure severity, and weight change for KA and pilocarpine were determined for each animal. The brains from animals who had received KA or pilocarpine were examined for severe neuronal stress, by using immunoreactivity for heat-shock protein (HSP)72.
Results: The Rb extract had a dose-dependent anticonvulsant effect in all three models of chemically induced seizures: increasing the latency to the seizures; decreasing the seizure score, weight loss, and subsequent neuronal damage after pilocarpine; and shortening the seizure duration and reducing mortality after PTZ. The Rb extract also significantly reduced the effects of KA, including completely blocking behavioral seizures. The root preparation increased the mortality rate after administration of pilocarpine, but had no other significant effects. The leaves/stems preparation, at 120 mg/kg, reduced the weight loss after pilocarpine, but had no other significant effects.
Conclusions: Ginseng extract made from either the root or leaves/stems is ineffective against chemically induced seizures. A partial purification of the whole extract that concentrates the Rb1 and Rb3 ginsenosides has significant anticonvulsant properties.
Used medicinally by Asians for 2,000 years, ginseng is now used regularly by ∼6 million Americans (1). Seven major species of ginseng are known, but the three most commonly used are Panax ginseng (Asian), Panax quinquefolius (American), and Panax japonicus (Japanese). Although some other constituents of the plant extract may have some activity, the ginsenosides are considered to have the most activity (2). Over 28 ginsenosides have been isolated from American ginseng, including: the panaxadiols (Rb1, Rb2, Rc, Rd, Rg3, Rh2, and Rh3) and the panaxatriols (Re, Rf, Rg1, Rg2, and Rh1). Pharmacologic effects attributed to ginsenosides have been shown in the central nervous system, the cardiovascular system, the endocrine system, and the immune system. They are thought to have antineoplastic, antistress, and antioxidant properties. The possibility that ginseng may have anticonvulsant activity was previously suggested by Lee et al. (3). It was reported that the seizures after administration of kainic acid (KA) were shorter when the animals were pretreated with ginseng, but the details of this effect were not provided, and the experiments were done by using a mixture of ginsenosides. It is not known which of the ginsenosides in the mixture contribute to the antiepileptic effect, but another study reported that one of the ginsenosides, Rd, has no effect on KA–induced neurotoxicity in the hippocampus (4).
The root, leaves, and stems of the ginseng plant all contain ginsenosides. Interestingly, American ginseng generally contains a lower Rg1/Rb1 ratio than does Asian ginseng. Given the reports that Rb1 is a central nervous system depressant and Rg1 is a stimulant (5), a lower Rg1/Rb1 ratio would be predicted to be more effective as a potential anticonvulsant agent. Therefore, for this project, American ginseng was used. The leaves, stems, and roots of both American and Asian ginseng are rich in the Rb ginsenosides, and an extract can be prepared that is enriched in the Rb ginsenosides (termed Rbextract) relative to the whole extract. This study tested whether ginseng, including an extract enriched in panaxadiols, may have activity as an anticonvulsant.
- Top of page
Ginseng from the root and from the leaves/stems of American ginseng were purchased from Jilin Ginseng in China (http://www.ginseng99.com). The levels of total ginsenosides in each were provided in an analysis from the company (Table 1). The levels of the Rb1, Rb3, and Rd ginsenosides (5) were determined by high-performance liquid chromatography (Column: Adsorbosphere XL- C-18B 90A 5μ, 250 × 4.6 mm id; flow phase, CH3CN: 0.06% TFA in H2O; flow rate, 0.3 ml/min; detection, UV at 203 nm). As can be seen from the data in Table 1, the levels of ginsenosides are higher in the leaves/stems preparation for both American and Asian ginseng. In addition, the levels of ginsenosides in the Rb group are highest in the leaves/stems from American ginseng. Therefore for this study, the partial purification to concentrate the Rb ginsenosides was done with leaves/stems from American ginseng. The results with this partial purification were then compared with those of the root preparation (total ginsenosides from roots: TG – Rt) and leaves/stems preparation (total ginsenosides from leaves and stems: TG – LS) also from American ginseng.
Table 1. Analysis of ginseng products
| ||Total ginsenosides||Rb1||Rb3||Rd||Total Rb1, Rb3, Rd||% that is in Rb group (by weight)|
|American root||30.2%|| 2.4%||14.6%|| 9.9%||26.9%|| 8.1%|
|American leaves/stems||85.2%|| 1.9%||19.2%||14.7%||35.8%||30.5%|
|Asian root||30.2%|| 1.9%|| 3.2%|| 7.1%||12.1%|| 3.7%|
|Asian leaves/stems||85.1%|| 3.9%|| 6.8%||11.9%||22.6%||19.2%|
To purify the leaves/stems partially, the powdered product was extracted with 95% ethanol 3 times under reflux. After removal of solvent, the extract was suspended in water and then partitioned in ethyl acetate 5 times. The water partition was then freeze-dried to give a dry powder. This powder was dissolved in water and passed over an open octadecal silica column eluting with 30%, then with 55%, and finally with 75% methanol. The 75% methanol elution was evaporated and freeze-dried to give the final Rb extract used in this study. High-performance liquid chromatography analyses indicate that this partial purification contains three major ginsenosides (Fig. 1): Rb1 (24.8%), Rb3 (46.4%), and Rd (13.1%). No other single component was >10%. For reference, this partial purification will be termed Rbextract.
Figure 1. Chemical analysis of the ginseng extract. The chemical structures of the three component ginsenosides in the Rb extract are shown.
Download figure to PowerPoint
All animal experiments were carried out in accordance with the National Institutes of Health guide for the care and use of laboratory animals (NIH publication 8023, revised 1996) and with the approval of the local Animal Use Committee. Adult male Sprague–Dawley rats weighing 190–230 g were used throughout this study. These animals were housed in a room with controlled temperature (22 ± 1°C) and humidity (50 ± 5%) under a 12:12-h light/dark cycle.
KA (Ocean Produce Int., Shelburne, Nova Scotia, Canada), pentylenetetrazol (PTZ; Sigma Chemical Co., St. Louis, MO, U.S.A.) and pilocarpine (Sigma Chemical) were dissolved in normal saline, and the pH adjusted to 7.4. Convulsants were administered intraperitoneally. Methylscopolamine (Sigma Chemical) was administrated to animals (1 mg/kg, subcutaneously) 10 min before the pilocarpine. Preliminary experiments demonstrated that 10 mg/kg KA, 50 mg/kg PTZ, or 300 mg/kg pilocarpine was the minimal dose that resulted in behavioral seizure activity in all naive animals. Therefore these doses of the convulsants were used to test the anticonvulsant activity of the ginseng products. The time from injection of each chemical convulsant to the first appearance of seizure activity was measured for each animal and is referred to as the seizure latency. After administration of KA, this was the time to the first wet-dog shake. After pilocarpine, this was defined as the time to the first appearance of forelimb clonus.
For the PTZ-induced seizures, the total duration of the behavioral seizure activity also was measured for each animal. For pilocarpine and KA-induced seizures, the behavioral seizures were scored according to the scale of Racine et al. (6), and each animal was assigned the score of the most severe seizure observed. Animals were weighed before administration of the convulsants and again 24 h later. The change in body weight was then determined for each animal after pilocarpine and KA.
Doses of the whole root and leaves/stems preparations were calculated to provide equivalent doses of the ginsenosides in the Rb extract. Each gram of the Rb extract contains 800 mg of the ginsenosides of interest (Rb1, Rb3, and Rd). To determine the doses of root and leaves/stems preparations for use in this study, the relative concentrations of total ginsenosides and ginsenosides in the Rb group were taken into account (Table 1). For the ginseng root preparation (TG – Rt), doses of 450 and 600 mg/kg include Rb ginsenosides at 36 and 48 mg/kg, respectively. For the leaves/stems preparation (TG – LS), doses of 120 and 160 mg/kg include Rb ginsenosides at 36 and 48 mg/kg, respectively. Ginseng products were dissolved in normal saline and administered intraperitoneally 60 min before the chemical convulsant.
Severe neuronal stress in the brain was assessed by using immunohistochemistry for HSP72 (7). Twenty-four hours after administration of pilocarpine or KA, animals were deeply anesthetized and perfused through the heart with 4% buffered paraformaldehyde. The brains were removed and fixed overnight. Coronal sections (50 μm) were cut with a Vibratome (Technical Products, Int, Inc. O'Fallon, MO, U.S.A.). For immunostaining, free-floating sections were blocked in 4% normal horse serum, 1% bovine serum albumin, and 0.3% Triton-100 in phosphate-buffered saline (PBS; 2 h at room temperature). Then the sections were processed for immunolabeling with anti-HSP72 (1:2,000, mouse monoclonal antibody; Oncogene Research Products, Boston, MA, U.S.A.) overnight at 4°C, followed by biotinylated goat anti-mouse (1:200; Vector Laboratories, Burlingame, CA, U.S.A.). Sections were then incubated with avidin–peroxidase complex (ABC kit; 1 h at room temperature; Vector Laboratories), followed by 3-3′-diaminobenzidine for visualization (Vector Laboratories). Adjacent sections were stained with cresyl violet. Naïve animals do not express HSP72 in the brain, but after status epilepticus, widespread expression of HSP72 occurs, particularly in the cortex and hippocampus. For these experiments, the brains were scored as positive or negative for HSP72. In brains scored as negative, no neurons were positive for HSP72 in any stained section. In brains scored as positive, neurons positive for HSP72 were seen throughout the brain, and the stain was clearly positive without the need for magnification (Fig. 2).
Figure 2. Nissl stain and immunoreactivity for heat-shock protein (HSP)72. Each pair of photographs was obtained from two adjacent sections from a single brain. Left: Section stained with cresyl violet. Right: Section stained for HSP72 immunoreactivity. The pair of sections on the left is from a control, untreated, animal. The middle pair is from an animal that had received 300 mg/kg of pilocarpine 24 h earlier. The sections on the right (Rb/Pilocarpine) are from an animal that received 40 mg/kg of the Rb extract 1 h before the pilocarpine and was killed 24 h later. The control and Rb/Pilocarpine sections were scored as negative for HSP72, whereas the animal treated with pilocarpine alone (middle) was scored as positive.
Download figure to PowerPoint
The latency to seizure onset, seizure duration, change in body weight, and seizure score were averaged across animals in the same treatment group. Comparisons between groups were done with an analysis of variance (ANOVA) with nonparametric measures (Kruskal–Wallis) and post hoc test (Dunn's) with comparison with chemical convulsant alone. The mortality and HSP72 immunoreactivity were analyzed with a contingency table by using χ2 analysis. A statistical difference was determined by a value of p < 0.05.
- Top of page
This study compared the anticonvulsant effects of a preparation of root, a preparation of leaves and stems, and a partially purified preparation from American ginseng. The possibility that ginseng may have anticonvulsant activity was previously suggested by Lee et al. (3), but those experiments were done by using a mixture of ginsenosides, and the details of the anticonvulsant activity were not provided. In this study, the partially purified preparation (Rb extract), had significant anticonvulsant activity in all three models of chemically induced seizures, whereas the preparations from leaves/stems and root had little or no anticonvulsant activity. These results suggest that some ginsenosides are anticonvulsant, but that other components of the whole extract from roots or leaves/stems are proconvulsant, or alternatively contain components that neutralize the anticonvulsant effects of the active components. Whether the Rb extract or a purified ginsenoside would be clinically effective for patients with epilepsy will require further testing. In the current study, the root preparation had no effect on the seizures produced by PTZ or pilocarpine, but did significantly increase mortality after pilocarpine. The increased mortality rate after administration of the root preparation and the fact that most of the ginseng available in the American market is made from the root of American ginseng suggests that people with epilepsy, especially those with a history of status epilepticus, should not take ginseng.
The optimal dose of Rb extract appeared to be 40 mg/kg, with 60 mg/kg being roughly equivalent, but less effective on some of the measured parameters, including latency to seizures after pilocarpine and seizure duration after PTZ. This inverted U-shaped dose–response relation is relatively common with complex herbal products and also was reported by Lee et al. (3). It suggests that other constituents in the herbal complex have opposite or toxic effects. Thus the effect of the mixture is a sum of the relative doses and maximal effects of the individual components within the mixture.
The latency to seizure onset appeared to be the most sensitive to the effects of the Rb extract. The latency was increased in every model, and as the dose of the Rb extract was increased, it was the effect seen at the lowest doses. This suggests that the most potent effect of the active ginsenoside(s) is to slow, or block, synchronization or speed of spread of focal epileptiform activity. This is supported by the observation that the Rb extract could completely block seizures induced by KA, but not those induced by pilocarpine or PTZ. The seizures induced by KA begin slowly with hippocampal/limbic system seizures with a latency to the first wet-dog shake (a limbic seizure) of 35 min. These seizures then gradually spread to the cortex and other brain areas (9). In contrast, the seizures induced by pilocarpine, which are also partial seizures with secondary generalization, have a latency of <20 min to the first forelimb clonus (a cortical seizure). PTZ induces generalized seizures with a latency of <2 min. Thus the seizures induced by KA have a more gradual onset. The differential effect of the Rb extract in the different models also could due to the Rb extract being more effective in the limbic system compared with the cortex. With increasing doses, the next effect of the Rb extract was a decrease in seizure score and decrease in the weight loss after pilocarpine and KA. This suggests that the Rb extract is reducing the severity of the seizure activity and reducing the duration of the most severe seizures. The exact mechanism by which the ginsenosides produce the anticonvulsant action remains to be determined.
Which of the components in the Rb extract is likely to be the active component(s)? The extract contains 24.8% Rb1, 46.4% Rb3, and 13.1% Rd. It is least likely that Rd had anticonvulsant activity. First, it is not concentrated to a significant degree in the Rb extract (from 14.7% in the leaves/stems starting material to 13.1% in the Rb extract), and there is little anticonvulsant activity of the starting material compared with the Rb extract. Second, it has been reported (Lee et al., 2003) that the isolated ginsenoside Rd has no effect on pyramidal cell death in CA1 or CA3 in the hippocampus of mice treated with KA. By contrast, evidence indicates that Rb1, and possibly Rb3, have anticonvulsant activity. The ginseng preparation used by Lee et al. (3) contained 18.3% Rb1 (levels of Rb3 were not reported). The effective doses in that study correspond to doses of 9.2–18.3 mg/kg of Rb1, which is similar to the doses of Rb ginsenosides that were active against KA-induced seizures in the present report. In addition, the relative activity of the ginseng preparations in the present study correlates with increasing concentrations of Rb1 and Rb3. Studies with individual ginsenosides are needed to determine the exact contribution of each component to the total effectiveness of the extract.
Although ginseng has been used by millions of people for thousands of years, the exact effects of the components within the product are not known. Some evidence exists of an effect in the central nervous system for both the whole extract and some of the individual ginsenosides, particularly a neuroprotective action. Pretreatment with whole ginseng extract has been shown to reduce the neuronal loss after prolonged seizures induced with KA (3). This attenuation of KA-induced damage was confirmed by Lee et al. (4). However, these studies do not separate the anticonvulsant effect from a neuroprotective effect. In the present study, no evidence was seen of an additional neuroprotective effect, beyond the reduction in seizure activity.
Additional evidence exists of a neuroprotective effect of the individual panaxadiol ginsenosides, most notably Rb1, which demonstrates that Rb1 has effects in the central nervous system and supports the hypothesis that Rb1 may be one of the active components in the Rb extract. Pretreatment with the isolated ginsenoside Rb1 reduces the loss of neurons in the hippocampus in gerbils after bilateral occlusion of the carotid arteries (10,11). This action is thought to be, in part, mediated by the scavenging of free radicals by Rb1 (12). More recent studies have examined the effects of panaxadiols, including Rb1, in some in vitro systems. Rb1 protects against glutamate-induced excitotoxicity in neuronal cultures (13,14). In additional work, Rb1 has been shown to have anxiolytic effects (15), but the mechanism behind this effect is not known. Rb1, injected into the lateral ventricle, has no effect on basic synaptic transmission (16). This suggests that Rb1 is not anxiolytic by suppression of excitatory transmission. Other measures of γ-aminobutyric acid (GABA)ergic function in vivo that might explain the reported anxiolytic and anticonvulsant actions have not been tested.
How do ginsenosides work? Ginsenosides are amphiphilic and have the ability to intercalate into the plasma membrane. This leads to changes in membrane fluidity, which can alter membrane function and secondarily alter the function of receptors in that membrane. Evidence exists of interactions with membrane-bound receptors and the possibility of binding to steroid receptors in the cytoplasm, leading to changes in gene expression (2), but this action has not been demonstrated in the central nervous system. The most accepted theory about the mechanism of action is that the ginsenosides (at least some of them) have antioxidant properties and the ability to scavenge free radicals (12,17). These actions might explain the neuroprotective effects of ginseng, but do not explain the anticonvulsant action reported here.
In the present experiments, the changes in body weight that were calculated for each animal appear to be a very sensitive and accurate measure of the total seizure severity and duration. In general, as young-adult animals, male Sprague–Dawley rats, fed ad lib, will gain 10–15 g/day (http://www.harlan.com/strain%20details/rats/sd.html). Animals that have status epilepticus will actually lose weight in the 24 h after convulsant administration. Over the years of using KA to induced prolonged seizures, a correlation was noted between the severity of the seizure activity and the amount of weight lost by the animals in the first 24 h. The weight loss is multifaceted, with contributions from the motor seizures, reduced food intake, and general distress. Thus the weight change gives an indication of the overall well-being of the animal. In contrast, the seizure score simply records the most severe seizure in each animal, with no indication of the duration of the most severe seizures. Therefore both seizure score and weight change were reported here.