, waves that have the frequency between 3-7 Hz are
, waves that have the frequency between 7-13 Hz are
, and waves have the frequency between 13-40
are 
[7].
Zhang YY et al / Acta Pharmacol Sin 2003 Feb; 24 (2): 157-162
ZHANG Yun-Yi1, LI Pei-Fen, LI Duan
Department of Pharmacology, School of Pharmacy, Fudan University, Shanghai 200032, China
1 Correspondence to Assoc Prof ZHANG Yun-Yi. Phn 86-21-6404-1900, ext 2465. E-mail zhcheng@shmu.edu.cn
Received 2002-06-14 Accepted 2002-11-19
KEY WORDS Ginkgo biloba; fast Fourier transform; electroencephalography; brain ischemia; reperfusion injury
ABSTRACT
AIM: To test the effect of Ginkgo biloba leaf extract on electroencephalography (EEG) during cerebral ischemia and reperfusion. METHODS: Based on the quantitative analysis of EEG using the fast Fourier transform (FFT), the effect of Ginkgo biloba extract (GbE) on rat EEG was surveyed in the model of middle cerebral artery (MCA) occlusion and global cerebral ischemia. RESULTS: In the global cerebral ischemia, GbE 8 and 16 mg/kg could accelerate the recovery of EEG after reperfusion, and GbE 4 mg/kg had the same effect but much weaker. In the MCA occlusion model, GbE 16 and 32 mg/kg greatly suppressed the drop of power spectrum of EEG. CONCLUSION: GbE could mitigate the cerebral damage caused by ischemia.
INTRODUCTION
The leaves of Ginkgo biloba are used to treat many diseases by the traditional Chinese medicine. Much research have been carried out since 1960s and it was found that Ginkgo biloba extract (GbE) could increase the rat cerebral blood flow[1], improve the mice memory impaired by ischemia[2,3], and protect the cerebral function[4,5]. But few research was focused on the effect of GbE on cerebral function following ischemic insult using electroencephalography (EEG), which reflects cerebral excitation. By using digital signal processing technique, this paper aimed to test the effect of GbE on EEG during cerebral ischemia and reperfusion.
MATERIALS AND METHODS
Materials GbE, brown powder, produced by Shanghai Zhongwei Biochemistry Co, Ltd, patch number 960706, was diluted by distilled water after drops of alcohol and sorbitol (3:4) mixture was added. The pH of solution was adjusted to 7.4 by adding NaOH 1 mol/L. Nimodipine, white powder, gifted by Dr ZHANG Rui-Wen, Shanghai New Drug Research & Developing Center, was diluted in 0.5 % carboxycellulose just before use. The solution was kept on sting to prevent precipitating.
Effect of GbE on rat after middle cerebral artery (MCA) occlusion[6] SD rats, weighing (180± 20)g, were anesthetized by ip urethane 1 g/kg. The skulk was exposed then cleaned using H2O2 . Drilling was performed at 4 points: 2-mm left and right to the sagittal suture and 2-mm in front of and behind coronal suture. Four electrodes were inserted into the holes and fixed. The left 2 electrodes formed one recording pare, the right 2 formed the other. An 0.2-mm nylon thread was inserted into right internal carotid artery to about 1.2 cm. The temperature of rectum was maintained at (37.0±0.5) ºC.
Drugs were given as follows: GbE 8, 16, and 32 mg×kg-1 iv, NS iv as negative control, nimodipine 20 mg/kg as positive control. Ten minutes later (20 min for nimodipine), EEG was recorded. Then the MCA was blocked by inserting the nylon thread to 2.2 cm. Meanwhile the left common carotid artery was blocked. EEG was recorded at 1, 5, 10, 15, 30, 45, and 60 min after occlusion.
Effect of GbE on rat global cerebral ischemia Spraw-Dawley rats, weighing (260±54) g, were inserted 2 electrodes at the points in front of the coronal suture for EEG recording. The 2 femoral arteries were used separately for blood pressure measuring and for bleeding. The rectum temperature was monitored and maintained at (37.0±0.5) ºC. The animals were left to rest for 30 min.
After blood pressure and EEG were recorded, the rats were given drugs as follows: NS, GbE 4, 8, and 16 mg×kg-1 iv and nimodipine 10 mg×kg-1 ig. Fifteen minutes later, blood was bleeding from femoral artery until the systolic blood pressure dropped to about 40 mmHg. Then the two aortors were clamped. The systolic blood pressure was maintained to 38-48 mmHg by bleeding or refilling the blood during the occlusion. The BP and EEG were recorded at 1, 5, 10, and 15 min after the occlusion. The reperfusion was established by refilling all blood and re-opening the aorta. The BP and EEG were recorded at 1, 5, 10, 15, and 30 min after reperfusion.
Data process and statistics The subgroup of
EEG waves are defined as follows: waves that have frequency between 1-3 Hz are defined as
, waves that have the frequency between 3-7 Hz are
, waves that have the frequency between 7-13 Hz are
, and waves have the frequency between 13-40
are [7].
The EEG signals were analyzed with computer after A/D convert (sample rate 200 points per second). By means of 1024-point fast Fourer transform (FFT), the power spectrum was calculated[8]. Then it was summed according to the EEG wave definition above. The data were expressed as the percentage of the value before drug. For it was not in normal distribution so the significance was evaluated using Mann-Whitney test. The software used for statistic calculation was StatView ver 5.0.1, SAS Institute Inc (Cary, NC27513, USA).
RESULTS
Effect of GbE on MCA blocking model The EEG from left side did
not change greatly, but at the right electrode pair, the power spectrum dropped
greatly when MCA was blocked (Fig 1). All waves in GbE 32 mg/kg-treated
group were significantly higher than those of NS group at 1-30 min occlusion
except wave at 15 min and
wave at 30 min. When rats were given GbE 16 mg/kg, the power spectrum
of EEG was also higher than that of the NS group, but the improvement was weaker.
GbE 8 mg/kg did not have any effect. Nimodipine 20 mg/kg had the effect
on , ,
and wave sometimes but not on
wave (Tab 1). When right MCA was
occluded, the power spectrums of the waves from left electrode pair dropped
slightly also. The change was smaller than that of right pair and there was
no significant difference between NS and drug values.
Fig 1. EEG wave (up) and its power spectrum (down) before MCA occlusion (A) and at 10 min after MCA occlusion (B). The data were from a rat of NS group and contained 5-s period of EEG. Compared with A, EEG and power spectrum of B were much smaller.
Tab 1. Effect of Ginkgo biloba extract (GbE) on EEG power spectrum of rats with MCA occlusion. Mean±SD. aP>0.05, bP<0.05, cP<0.01 vs normal saline (NS) group.
Action of GbE on global cerebral ischemia During ischemia and
reperfusion, the power spectrums of NS group droped quickly. The higher the
wave frequency, the more the amplitude dropped. For the latent electrode pair,
the ,
and waves in GbE 16 mg/kg-treated
group and wave in GbE
8 mg/kg-treated group were significantly higher than those in control group
at 10 min after drug administration. During ischemia, the power spectrum of
all groups was reduced rapidly. Though they were slightly higher than those
of NS group, no statistic difference was found. After reperfusion, the EEG power
spectrum of all GbE groups recovered rapidly. In GbE 16 mg/kg-treated
group, the power spectrum was significantly higher than that of NS group in
all waves from 5 min after reperfusion and on. In GbE 8 mg/kg-treated
group, the change of wave was
aparent at 30 min after reperfusion, but the change of ,
and
waves were observed at 5 min after reperfusion. In GbE 4 mg/kg treated-group
only and
waves were higher at 5 min after reperfusion. These results showed that higher
frequency waves were more sensitive to the drug treatment. This phenomenon was
more prominent in nimodipine-treated group (Tab 2).
Tab 2. Effect of GbE on EEG power spectrum of rats with global cerebral ischemia. Mean±SD. aP>0.05, bP<0.05, cP<0.01 vs NS group.
The mean blood pressure was restored after reperfusion and reached the peak at 10 min. GbE 8 and 16 mg/kg had a significant effect occasionally but GbE 4 mg/kg and nimodipine did not (Tab 3).
Tab 3. Effect of GbE on mean blood pressure of rat (mmHg). Mean±SD. aP>0.05, bP<0.05, cP<0.01 vs NS group.
DISCUSSION
FFT is a tool that combines computer science and modern signal processing technique. It converts signals to a serious of different frequency sine waves. Counting these sine waves can quantitatively define the detected signals. Normally, EEG waves were divided into some groups according to its frequency. The altitude and duration, but not the shape, of individual wave were measured. That is to say the high frequency sub-wave was discarded. Besides, the individual EEG wave is repeatless and could not represent the outlook of EEG. And, even the same piece of EEG was studied, there would be a large deviation between data sets, different observers, and different time of study. Converting EEG, by FFT, to the sine waves and then counting them can be an easy and automatic step in EEG analysis. Because FFT uses a relatively long period of original EEG for calculation, the results would be more reliable and accurate. In this paper, the original EEG was used in FFT without any pre-processing. That is to say the original EEG was cut by a square window. The sudden truncation brings distortion to the results at the edge of high frequency[8]. Though the overload is very high which could be 10 % of the normal, it would be located in a narrow high frequency band on the condition that the number of sample data is large. Here, 1024-point FFT was used and sine waves below 40 Hz were summed, so the overload was excluded and could be neglected.
Results of this paper showed that the power spectrum of EEG dropped slowly when global cerebral ischemia was formed. In the model of MCA blocking, though the power spectrum decreased rapidly, it remained in a higher level, for GbE could increase cerebral blood flow[1]. And there may be blood flowing from unblocked side to the ischemia zone in an MCA model. But in the model of global cerebral ischemia, the blood supply was extremely low and may not be increased any more. It may explain the fact why GbE 16 mg/kg improved the EEG changes at the early stage of MCA blocking but not at the other one.
GbE suppressed the decrease in SOD and MDA accumulations that caused by ischemia[9,10]. These phenomena reflected that GbE had an anti-oxidation effect. GbE helps alleviate the subcellular damages of cerebral ischemia[11] and allows mitochondria to maintain their respiratory activity under ischemic conditions as long as some oxygen is present, thus delaying the onset of ischemia-induced damage[12]. The data of this paper showed that the GbE increased EEG power spectrum, it was more effective on high frequency waves than on low frequency ones. The more active the cerebral cells are, the higher the EEG frequency they elicite. We may conclude that GbE protects the cerebral cell function against loss during ischemia. Once the normal blood flow was restored, the cerebral cell function could recover rapidly. The fact that the blood pressure of animals treated with GbE was significantly increased when reperfusion was started following the global cerebral ischemia also suggested that the function of central nervous system was strengthened and regained the control over peripheral tissues.
REFERENCES