Postischemic alterations of complex spike cell discharges and evoked potentials in rat hippocampal CA1 region Furukawa K, Yamana K, Kogure K. Postischemic alterations of complex spike cell discharges and evoked potentials in rat hippocampal CA1 region. Acta Neurol Scand 1992: 86: 142-147. Postischemic alterations of spontaneous discharges of complex spike cells (CS cells) and evoked potential in the rat hippocampal CA1 region were studied. Following 5 min of ischemia, CS cell discharge reappeared approximately 5 rnin after reperfusion and the frequency remained low, reaching a final value of 66.1 & 16.0% (n = 11) of preischemic frequency 2 h later. However, only one of 7 CS cells subjected to 20 min of ischemia exhibited discharges 2 h later. In the group with 5 rnin of ischemia, we obtained CS cell discharges from all rats at both 1 and 2 days after ischemia, with cluster frequencies indistinguishable from preischemic levels. In the group with 20 min of ischemia, discharges were noted in 7 neurons of 11 rats after 1 day, and in only 2 neurons of 8 rats after 2 days: their mean frequencies were lower than preischemic levels. In experiments of evoked potentials, the mean percentages of amplitudes of the post-synaptic potential (psp) 2 h after 3, 5 and 20 min of ischemia were 98.0 & 10.7 (n = 8), 70.7 f 8.22 (n = 9) and 45.1 6.34% (n = 7) of preischemic amplitudes, respectively. These results suggest that the functional deterioration of spike generation, as well as synaptic transmission, starts during transient ischemia and/or at the early stage of reperfusion.
Pyramidal neurons in the hippocampal CA1 region of rats and gerbils are reported to be selectively vulnerable to transient ischemia (1-3). There have been some reports on postischemic morphological alterations of CAI pyramidal neurons (4-6), and the postischemic changes in neuronal function (7- 10). Suzuki et al. (7) examined the postischemic changes in neuronal function by measuring the spontaneous impulse frequency of CA1 neurons in gerbils, and noted prolonged hyperexcitation of CA 1 neurons after transient ischemia. Hippocampal CA1 neurons can be electrophysiologically divided into two groups based on the impulse discharge pattern (1 1- 14). One group, complex-spike cells (CS cells), has impulses which form a cluster of 2-8 action potentials (12, 13) with less than 5 ms interspike intervals. And the other group, theta cells, has an impulse which occurs in the shape of a single action potential (12, 13). Complex spike cells may be regarded as pyramidal neurons because they can only be activated by antidromic stimulations, and they can discharge one or two action potentials to orthodromic stimulation at the same time as a population spike (13, 15). However, some theta cells are considered as pyramidal neu142
K. Furukawa, K. Yamana, K. Kogure Department of Neurology, Institute of Brain Diseases, Tohoku University School of Medicine, Sendai, Japan.
Key words: complex spike cell; delayed neuronal death; evoked potential; hippocampus; ischemia; rat K. Furukawa, Department of Neurology, Institute of Brain Diseases, Tohoku University School of Medicine, 1- 1 Seiryo-machi, Aoba-ku, Sendai, 980, Japan. Accepted for publication December 16, 199 1
rons and some of them are considered to be interneurons (13, 16, 17). Therefore, to investigate CS cells reflects the function of pyramidal cells more directly than theta cells. To study the postischemic changes of neuronal function, it is important to clarify changes in neuronal activity after transient ischemia of both types of neurons, CS and theta cells. Previous reports about postischemic change of spontaneous activities, including our reports (18), focused only on theta cells (7,9). Nobody has reported the change of CS cells’ discharge by ischemia so far. Consequently, our report is the first one which has investigated the discharge of CS cells after ischemia. The first object of this paper is to clarify the functional change of CS cells. In addition, we studied ischemia-induced change of evoked potential of the CA1 area. There are a lot of in vitro reports about the change of evoked potentials during and after hypoxia and/or hypoglycemia using hippocampal slices (19-25). However, little is known about ischemia-induced changes in synaptic transmission. The second object of this paper is to study ischemia-induced changes of synaptic functions in vivo.
CS cell discharges, EPs Material and methods Animal preparation
Male Wistar rats weighing between 270-3 10 g were used in this study. The rats were anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and their bilateral vertebral arteries were occluded by electrocautery as described by Pulsinelli & Brierley (26). Acute experiment
Twenty-four hours after occluding the vertebral arteries, the rats were anesthetized with 1% halothane in 30% O,, 70% N,O. The bilateral common carotid arteries and the trachea were isolated through a ventral, midline cervical incision. A cotton thread was looped around each common carotid artery and it was passed through a polyethylene tube (10 cm in length, PE- 160, intramedic) (27). Transient forebrain ischemia and reperfusion was induced by drawing and releasing the thread manually. The head of the rat was fixed on a stereotaxic instrument and the rat was artificially ventilated to maintain physiological conditions (PaO,> 100 mmHg, PaCO, 3540 mmHg). The temperature was kept close to 37 O C by a heating mat. For recording spontaneous CS cell activity, a small bony opening was made 3.9 mm posterior and 2.1 mm lateral to the bregma. The recording electrode, which was a sharpened tungsten wire insulated with lacquer (tip diameter less than 1 pm) was inserted through the opening. The electrode was advanced carefully until stable recordings from CA1 neurons could be obtained. The indifferent electrode was a stainless steel screw placed on the frontal bone. For recording evoked potential, a stimulus electrode, made of bipolar enamel coated stainless steel (100 pm diameter), was advanced through an opening drilled at 3.9 mm posterior and 4.2 mm lateral to the bregma to stimulate CA3 neurons. Submaximal square pulse stimulations (0.1 ms, duration) were applied at 0.2 Hz. The recording electrode, which was an enamel-coated, stainless steel wire (100 pm diameter) cut at a right angle, was inserted though an opening 3.9 mm posterior, 2.1 mm lateral to the bregma. The depth of both the stimulus electrode and the recording electrode was varied until stable recordings were obtained (the depth of the stimulus electrode was 3.2-3.4 mm, recording electrode was 2.1-2.3 mm from brain surface). A stainless steel screw was mounted on a frontal bone and used as an indifferent electrode. Electrical events were pre-amplified (150 Hz10 kHz for spontaneous action potential, 5 Hz10 kHz for evoked potential) and stored in an F M tape recorder.
Twenty-four hours after vertebral artery cauterization, the rats were anesthetized with 1% halothane. Both common carotid arteries were isolated through the neck incision. Blood flow of the bilateral common carotid arteries was occluded by aneurysm clips. While ischemic insults of 5 min or 20 min were applied, the anesthetic gas was changed to room air in order to minimize its effect. After 1 or 2 days recovery from the ischemic insult, the rat was reanesthetized with 1% halothane and fixed on a stereotaxic instrument. Spontaneous complex cell firings were recorded as described above. Statistical comparisons were made using Wilcoxon rank-sum test. Morphology
At the end of the experiments, we applied DC current (30 pA, 20 s) through the recording and stimulating electrode to burn the recording and stimulating sites. Animals were perfusion-fixed with FAM (Formalin, Acetic acid and Methanol, 1:1:8). The brains were removed and immersed in fixative and then embedded in paraffin. Five pm paraffin sections were stained with cresyl violet and the burns at the recording and stimulating site together with the morphological changes of CA1 neurons were examined with a light microscope (Fig. 1). Results
The spontaneous discharge of CS cells formed a cluster of several spikes, with characteristic short inter-spike intervals (less than 5 ms) and decreasing its amplitude (Fig. 2). Each CS cell burst consists of 2-8 impulses (different from cell to cell), so we measured one cluster of spikes as one discharge in the
Fig. 1 . The recording (CA1) and stimulating (CA3) sites for evoked potential recording in the hippocampus are indicated by
burning with the electrodes. Creysyl violet staining.
Furukawa et al.
Fig. 2. A typical firing pattern of a complex spike cell. The horizontal scale indicates 5 ms and the vertical scale indicates 10 pV.
following experiments. Spontaneous cluster frequencies of CS cells of the CA1 region in the hippocampus from control, anesthetized rats varied from 0.084-1.83 Hz, with an average of 0.49 ~f:0.076 Hz (mean k S.E., n = 24). In the control group, the average spontaneous cluster frequency scarcely changed during 2 hours, that is, the frequency 2 h later was 0.52 & 0.11 Hz (n = 7). Sixteen out of 18 CS cells did not show any transient frequency increase just after the beginning of ischemia. The spontaneous discharge disappeared within 30 s after the beginning of ischemia. When the duration of the ischemia period was 5 min, the spontaneous discharge reappeared approximately 5 min after reperfusion (Fig. 3). The cluster frequency increased gradually, but did not show complete recovery during 2 h of recording period. Its average frequency 2 h after 5 min of ischemia was 66.1 k 16.0% ,. (n = 11) of the preischemic level. The number of the impulses in one cluster did not change compared to that of preis-
chemic state. When the duration of the ischemic period was prolonged to 20 min, the discharge was seen in only one out of 7 cells up to 2 h later, and the recovered cluster frequency was not higher than the preischemic level. However, the number of the impulses in one cluster also did not change in the case which discharge was recovered. When CS cell's spontaneous discharge did not recover until 2 h after ischemia, we moved the recording electrode along the long axis of electrode and searched the cell's discharge to eliminate the possibility that some movement of the CS cell occurred by brain edema or another cause. But we could not record CS cell's discharge in all of such cases. The mean amplitude of post-synaptic potential (psp) was 0.85 0.22 mV, (n = 32), disappearing approximately 1 rnin after the beginning of ischemia insult. But, the presynaptic volley remained until 23 min after ischemic insult. We caused 3, 5 and 20 min of ischemia, and a typical continuous recording of the evoked potential in the case of 5 rnin of ischemia is shown in Fig. 4(A). Typical forms of evoked potentials are also indicated in Fig. 4(B): preischemia, and in (C): 2 h after 5 rnin of ischemia. When the ischemic period was 3 min, the presynaptic volley endured at the end of ischemia in 4 of 8 rats, psp were completely suppressed, and evoked potential reappeared approximately 2.2 min after reperfusion (Fig. 5(A)). When the ischemic periods were prolonged to 5 and 20 min, the evoked potential reappeared at 3.7 and 20.4 min after reperfusion, respectively. After 3 rnin of ischemia the psp amplitude recovered to the preischemic level 2 h after reperfusion (98.7 k 10.7% compared to preischemic level, n = 8), (Fig. 4(A)). However, after 5 and 20 rnin of ischemia, there was not such full recovery (5 min; 70.71f:8.2%, n = 9 , 20min: 45.1?6.3%, n=7).
0 5 rnin isch.
Time (minutes) Fig. 3. The time course of the change of cluster frequency of CS cell discharge after 5 rnin ischemia. The spontaneous cluster frequency just before the ischemic insult is 100%. Each point indicates the mean frequency in 11 CS cells. Vertical lines represent S.E.
Fig. 4. The continuous recording of evoked potentials in rats which suffered 5 min ischemia (A) and typical recordings of the control (B) and experimental group 2 h after 5 rnin of ischemia
(C). The length of the arrow was measured as the amplitude of psp. Numerals under the continuous recording show minutes after reperfusion. Upper horizontal and vertical scales indicate 1 rnin and 1 mV, respectively. Lower scales indicate 10 ms and 1 mV.
CS cell discharges, EPs
0 3 min isch.
0 5 rnin isch.
20 min isch
al 0) c
Time (minutes) (B)
Time (minutes) (C)
Fig. 5. The time course of the change of amplitudes of psp ((A): 3 min, (B): 5 min, (C): 20 min). The amplitude just before the ischemic insult is 100%. The points indicate the mean amplitudes in 8 rats for 3 rnin of ischemia, 9 rats for 5 min of ischemia and 7 rats for 20 rnin of ischemia. Vertical lines represent S.E.
(Figs. 5, B, C). Furthermore, the amplitude of the population spike decreased (the percentage of the average amplitude was 20.1 k 10.2% compared to the preischemic level, n = 5) and the threshold stimulus for evoking a population spike increased in all animals examined 2 h after 5 min of ischemia. In a group with 20 min of ischemia, the population spike didn't reappear in 4 of 6 rats even if stimulus intensity was increased up to 3 times the preischemic intensity (approximately 150- 180 V). In the chronic experiments, we detected discharge of CS cells from all the rats subjected to 5 min of ischemia at 1 and 2 days later (Fig. 6). Their frequencies were very similar to the controls; the mean cluster frequency at one day after 5 min of ischemia was 0.58 & 0.12 Hz, n = 9 and that at 2 days after ischemia was 0.62 k 0.14 Hz, n = 7. After 20 rnin of ischemia, it was hard to detect any CS cell discharge. We detected 7 cells in 11 rats at 1 day after 20 rnin
20 min isch
of ischemia and only 2 cells in 8 rats at 2 days after 20 min of ischemia. The mean frequencies were 0.56k0.18 Hz for 1 day and 0.32k0.24Hz for 2 days. In the morphological experiments, approximately half of the pyramidal cells in CA1 region showed necrotic change at 2 days after 20 rnin of ischemia, but we couldn't detect morphological change at 2 days after 5 min of ischemia (data not shown). Discussion CS cell functional impairment post-transient ischemia
Spontaneous discharges of CS cells were depressed during transient ischemia. A 2-h recirculation period allowed 66.1% recovery of CS cell's spontaneous cluster frequencies after 5 rnin of ischemia. Neuronal activities 1 or 2 days after 5 rnin of ischemia were near preischemic levels. However, recovery of spontaneous discharge from 20 rnin of ischemia was observed in only 1 CS cell of 7 neurons tested. The other 6 neurons did not show spontaneous discharges during 2 h of recirculation. In addition, it was very difficult to record spontaneous discharges 1 or 2 days after 20 min of ischemia (7 neurons of 11 rats after 1 day, 2 neurons of 8 rats after 2 days) indicating that most CS cells were electrically silent at 2 days after 20 min of ischemia. These results suggest that CS neurons lose their neuronal function during ischemia and/or in the early stage of reperfusion. Comparison of theta and CS cell
Fig. 6. Mean cluster frequencies of spontaneous CS cell discharge at 1 and 2 days after ischemia. The closed, hatched and open columns represent control, 5 min of ischemia and 20 min of ischemia, respectively. N = 24 in the control experiments. N = 9 and 7 for 1 and 2 days after 5 min of ischemia, and n = 7 and 2 for 1 and 2 days after 20 rnin of ischemia, respectively. Vertical lines represent S.E. p