Neurophysiological changes of spiny neurons in rat neostriatum after transient forebrain ischemia: An in vivo intracellular recording and staining study

Pergamon

NeuroscienceVol. 67, No. 4, pp. 823-836, 1995 Elsevier Science Ltd Copyright 0 1995 IBRO Printed in Great Britain. All rightsreserved

0306-4522(95)00096-8

0306-4522/95 $9.50+ 0.00

NEUROPHYSIOLOGICAL CHANGES OF SPINY NEURONS IN RAT NEOSTRIATUM AFTER TRANSIENT FOREBRAIN ISCHEMIA: AN IN I/IV0 INTRACELLULAR RECORDING AND STAINING STUDY z. c. xu Department of Neurology, College of Medicine, University of Tennessee, Memphis, TN 38163, U.S.A. Abstract-The spontaneous activities, evoked postsynaptic potentials and membrane properties of spiny neurons in rat neostriatum were compared before, during and after 5-8 min ischemia using intracellular recording and staining techniques in ho. Severe forebrain ischemia was induced with the four-vessel occlusion method. Approximately 2.5 min after the onset of ischemia the baseline membrane potential quickly depolarized to - 20 mV and remained at this level during ischemia. Repolarization began within 2 min after recirculation. The onset of ischemic depolarization was directly related to the severity of ischemia and its latency was inversely related to brain temperature. Spontaneous firing and membrane potential fluctuation of spiny neurons ceased immediately after ischemia and slowly recovered several hours after recirculation. No neuronal hyperactivity was observed up to 7 h after recirculation. Cortically evoked inhibitory postsynaptic potentials and late depolarizations disappeared earlier after &hernia and recovered later following recirculation than the initial excitatory postsynaptic potentials. Membrane input resistance of spiny neurons was significantly increased but the time constant remained the same following recirculation. The rheobase and spike threshold of spiny neurons were significantly increased and the repetitive firing evoked by depolarizing current pulse was suppressed shortly after recirculation. The results of the present study indicated that the spontaneous activity and evoked postsynaptic responses of spiny neurons are suppressed and the excitability of spiny neurons is decreased after transient ischemia. The polysynaptic responses are more sensitive to ischemia than the monosynaptic ones.

Small to medium-sized neurons in the neostriatum are highly vulnerable to transient forebrain ischemia.2,46 Because spiny neurons are the efferent neurons of neostriatum6.24.44 and since studies have shown that most of the interneurons in the neostriatum are relatively spared after transient ischemia,‘,” the spiny neurons may be considered to be the main ischemia-vulnerable population in the neostriatum. The pathogenesis of this selective neuronal injury is not clear. It has been hypothesized that cerebral ischemia disturbs the release/reuptake of excitatory neurotransmitters or increases the sensitivity of postsynaptic receptors to exictatory transmitters. Excessive excitation of postsynaptic neurons, which causes a lethal influx of calcium, would then trigger the process of neuronal injury.‘x5’ Neostriatum receives massive glutamatergic projections from cerebral cortex.‘6~32~55Microdialysis studies have indicated that the extracellular levels of glutamate and aspartate are dramatically increased in hippocampus’ and in neostriatum” during ischemia. _____ Abbreviations:

AHP, after hyperpolarization; DAB, diaminobenzidine; EPSP, excitatory postsynaptic potential; ID, ischemic depolarization; IPSP, inhibitory postsynaptic potential; ISI, inter-stimulus interval, interspike interval; I-V, current-voltage; KPBS, potassium phosphate-buffered saline; PBS, phosphate-buffered saline.

The excitotoxic hypothesis would predict that activities of vulnerable neurons should increase following ischemia. However, results of electrophysiological studies on neuronal activity in hippocampus following transient ischemia are controversial. Using extracellular recording in ho, some studies have shown that the spontaneous firing rate in CA1 region of hippocampus is increased following ischemia7,56while others report that the spontaneous firing rate and evoked responses are suppressed.3,27,42Intracellular recording studies in vitro have indicated that the neuronal activities and evoked synaptic potentials of CA1 neurons are suppressed after ischemia/ hypoxia,

19X37.58

The bulk of the information about neurophysiological changes in ischemia-vulnerable neurons comes from studies on CA1 hippocampal neurons. Little is known about neurophysiological changes of spiny neurons in neostriatum after transient ischemia. While the spiny neurons in neostriatum and CA1 neurons in hippocampus are both vulnerable to transient ischemia, the threshold for injury and the time course of cell loss differ dramatically for these neuronal populations. In unanesthetized rats, 10 min ischemia produces CA 1 neuron degeneration while 30min ischemia is required to induce cell death in neostriatum.46 Neostriatal neurons die by 24 h after the insult whereas the CA1 neurons attain maximal 823

z. c. AU =’

824

damage 72 h following ischemia.46 Recent studies suggest that dopamine is another important factor affecting the outcome of postischemic cell injury in neostriatum.20~21 The mechanism of neuronal death of spiny neurons in neostriatum may therefore differ from that of CA1 neurons in hippocampus. To characterize the neurophysiological changes of spiny neurons after ischemia and reveal the mechanisms associated with selective neuronal damage following transient ischemia, the present study compares the spontaneous activities, evoked postsynaptic potentials and membrane properties of spiny neurons in rat neostriatum before, during and after transient forebrain ischemia. EXPERIMENTAL PROCEDURES

Preparation for transient forebrain &hernia

Male adult Wistar rats (220-32Og, Sasco) were used in the present study. Transient forebrain ischemia (5-8 min) was induced using the four-vessel occlusion method45 with some modification. The animals were fasted overnight to provide uniform blood glucose levels. For surgical preparation the animals were anesthetized with 2-3% halothane. An occluding device (a loop of silicone tubing) was placed loosely around each carotid artery to allow subsequent occlusion of these vessels with minimal mechanical disturbances of the animal. A femoral artery was cannulated to monitor blood pressure and for withdrawal of samples for blood gas measurements. The animal was then placed on a stereotaxic frame and the core body temperature was maintained with a heating pad through a temperature control unit (TC-120, Medical System). Mechanical ventilation was initiated via a nasal tube using a mixture of 0.882% halothane, 33% 0, and 66% N,, with stroke rate and volume adjusted to maintain the PCO, and PO, at approximately 40 and 120 mmHg, respectively. The vertebral arteries were electrocauterized. A very small temperature probe (0.024” D, YSI 511) was inserted beneath the skull in the extradural space after which brain temperature was maintained at 37°C with a heating lamp via a feedback temperature regulated system. The fiber optic probe of a laser-Doppler flowmeter (LASERFLO, Vasamedics) was placed on the surface of the cerebral cortexI to monitor local cortical blood flow. Severe forebrain ischemia was induced by occluding both common carotid arteries for 5-8 min. Upon release of the carotid artery clasps, cerebral blood flow resumed immediately. Intracelhd~r recording and staining

Preparation for intracellular recording in uioo was performed as described in previous studies.62,63The skull was opened to expose the recording site and for placement of stimulus electrodes. One pair of bipolar stimulating electrodes were placed into the ipsilateral medial agranular cortical field at a 30” angle to the vertical, 2 mm from the dural surface. Stimuli were constant current pulses of 0.1 ms duration ranging in amplitude from 0.1 to 3 mA. Recording electrodes were pulled from glass capillaries with filament (A-M systems) using a vertical electrode puller (Kopf 750). The tips of the electrodes were broken to produce a tip

Fig. 1. A photomicrograph of parasagittal section showing the stimulus site. The section was stained with Cresyl Violet. The track of stimulus electrode (arrow) is in the medial agranular cortical field.

resistance of 40-70 MD. The electrodes were filled with a solution of 2-4% neurobiotin (Vector)34 in 2 M potassium acetate. Cerebral spinal fluid was drained via a cisternal puncture to reduce brain pulsation and the animal was suspended by a clamp applied to the tail. After placement of a microelectrode in the cortex above the neostriatum for recording, the exposed surface of the brain was covered with soft paraffin wax. Only neurons with stable membrane potentials after impalement were selected for study. Recording of membrane potential, intracellularly applied current, cerebral flood flow, and arterial blood pressure were digitized with data acquisition programs Axodata (Axon Instruments) and Superscope II (GW instruments) and stored on Macintosh computers for off-line analysis. The statistical analysis in the present study consisted of an unpaired r-test using Statview 3.0 (Abacus Concepts). After each successful recording, neurobiotin was iontophoresed into the cell by passing depolarizing current pulses (2 Hz, 300 ms, I-1.5 nA) for 15-30 min. At the end of the experiment, the animal was deeply anesthetized, perfused transcardially with 0.01 M phosphate-buffered saline (PBS) followed by 4% paraformaldehyde in the same buffer. The brain was removed and stored in fixative overnight. Parasagittal sections were cut at 50 pm thickness using a vibratome, incubated in 0.1% horseradish peroxidase-conjugated avidin-D (Vector) in 0.01 M potassium phosphate-buffered saline (KPBS, pH 7.4) with 0.5% Triton X-100 overnight at room temperature. After detection of peroxidase activity with 3’,3’diaminobenzidine (DAB), sections were examined in KPBS. Those sections containing labeled neurons and stimulation sites were mounted on gelatin coated slides and counterstained with Cresyl Violet for light microscopy. Figure 1 is a photomicrograph showing the location of the stimulus electrode track. RESULTS

The present study involved 19 animals. The arterial gas tension and general physiological condition of

Table I. Physiological parameters and arterial gas tension during experiment Mean S.D. n

BP (mmHg)

Rectal T (“C)

Brain t (“C)

pH

PCO, (mmHg)

POz (mmHg)

Hgb (%)

Glucose

82 10 14

37.7 0.71 14

36.8 0.48 14

7.35 0.04 14

35.5 3.91 14

127.0 19.78 14

38 4.1 14

88.4 32.74 14

Responses of neostriatal neurons after ischemia

825

cases, the baseline membrane potential depolarized approximately 5 mV immediately after the onset of ischemia for ~20 s, and this was followed by a hyperpolarization to approximately 10 mV more negative than the preischemic level. The membrane potential then gradually depolarized until a large rapid depolarization (ischemic depolarization, ID) occurred. The membrane potential depolarized to approximately -2OmV (-21.88 +4.30mV, mean f SD.) within tens of seconds and was maintained at this level throughout the course of ischemia. In some cases (n = 9) the recording was terminated shortly after the onset of recirculation, presumably because the recording electrode came out of the cell due to the brain volume expansion, the membrane potential shifted toward 0 mV (- 2.56 + 7.06 mV). In neurons in which the intracellular recording was maintained (n = 6), the membrane potential began to repolarize within 2 min after recirculation. The membrane potential quickly reached the preischemic level and continued to hyperpolarize to approximately - 90 mV. The membrane potential did not recover to control levels until 1-2 h after recirculation. The membrane potential change of a spiny neuron during Fig. 2. An example of spiny neurons intracellularly stained ischemia and shortly after recirculation is presented with neurobiotin after recording. The soma of this neuron is oval and the dendrites are loaded with dense spines except in Fig. 3. the most proximal portions. The onset of ID was directly related to the severity of ischemia. ID occurred only when blood flow of cortex, and presumably also neostriatum, was rethese animals were maintained in the normal range duced to < 10% of the preischemic level. In two cases (Table 1). Intracellular recording was performed on in which blood flow was reduced to only about 40% 73 neurons, of which three were excluded from the of the preischemic level, no baseline membrane present study because one was identified histologipotential change occurred during ischemia (data not cally as an aspiny neuron and two were cortical shown). In one of these two neurons, ID did not neurons. Neurons with resting membrane potentials occur even when the moderate ischemia lasted for of -60 mV or greater and action potential ampli20 min. The amplitude of ID varied from 40 to 60 mV tudes of at least 60mV (measured from the resting depending on the preischemic baseline potential of membrane potential) were selected for further electrothe neuron so that the membrane potentials depolarphysiological analysis. A total of 66 neurons were ized to approximately -20 mV regardless of the selected, of which 56 were successfully stained and preischemic baseline potential. The latency of ID identified as spiny neurons (Fig. 2). They all had onset was associated with brain temperature. In round or oval somata approximately 15 pm in diamanimals in which the brain temperature was not eter, with dendrites radiating in all directions that regulated (n = 8), the brain temperature dropped to were loaded with dense spines except in their most 32-35°C during ischemia. In such circumstances the proximal portions. These features were the characteristics of medium spiny neurons in neostriatum.‘2*3’,59 latency of ID onset varied from 4 to 6min. If the brain temperature was maintained at 37°C with a Ten unidentified cells were considered as spiny neurheating lamp (n = 9), the ID reached -20 mV at ons and included in the present study based on the approximately 2.5 min after the onset of ischemia. electrode track, stereotaxic coordinates and neuroThe relationship between brain temperature and the physiological characteristics. In some animals latency of ID onset is illustrated in Fig. 4. (n = 5), a second ischemia was induced after the spontaneous membrane potential fluctuation of spiny Changes in spontaneous activities and evoked neurons had recovered from the first insult. Since no potentials differences were found between the neurophysiologiThe spontaneous activity of spiny neurons in neocal responses of spiny neurons after these two insults, striatum is well known for its low spontaneous firing the collected data were pooled for the final analysis. rate and membrane potential fluctuation between depolarizing and hyperpolarizing states6’ SponChanges in baseline membrane potential taneous firing was abolished immediately after the Intracellular recording was maintained throughout onset of ischemia and slowly recovered after recircuthe whole course of ischemia in 17 neurons. In these lation. The firing rate was 0.55 _t 1.24 spike/s before

z. c. xu

826

Membrane potential

I 1 nA

Current

Blood flow 100

lmin

0 E-----L

4

4

Fig. 3. Baseline membrane potential change of a spiny neuron during transient ischemia. The upper panel is the baseline membrane potential of the neuron; the middle panel is the intracellularly applied current; the lower panel is the blood flow monitored from cerebral cortex. Two arrows indicate the onset of ischemia and recirculation, respectively. Constant hyperpolarizing current pulses (1 Hz, - nA, 200 ms) were applied to monitor the membrane input resistance except four intervals from which membrane potential fluctuation before and after ischemia was compared. Cortical stimuli with constant subthreshold intensity were applied during the recording. The brain temperature of this animal was maintained at 37°C. Immediately after the onset of ischemia, the membrane potential slightly depolarized for tens of seconds followed by a hyperpolarization and then a slow depolarization. Membrane potential rapidly depolarized to - 20 mV approximately 2 min after ischemia and began to repolarize approximately 1.5 min after recirculation. Membrane potential fluctuation ceased immediately after ischemia. The changes in amplitude of membrane potential deflection caused by the hyperpolarizing pulses indicated that the input resistance of this neuron remained about the same during the early phase of ischemia but significantly decreased during ID.

&hernia and decreased to 0.23 f 0.38/s shortly after recirculation. The firing rate was further reduced to 0.02 + 0.04/s l-5 h a after recirculation. No hyperactivity was observed up to 7 h after recirculation. In

the present study, spiny neurons exhibited spontaneous membrane potential fluctuation and the pattern of this fluctuation was affected by the extent of anesthesia. As shown in Fig. 5, when the animal was lightly anesthetized with 0.8% halothane, the mem-

bef

40

80

120 160 200

240 280 320

360

400

Tie after occlusion (set) Fig. 4. Plot showing the relationship of brain temperature and the onset of ischemic depolarization. The abscissa is the time interval at which measurements being taken. The ordinate is the membrane potential. The value of each plot is the mean and standard deviation. In unregulated animals, the brain temperature dropped to 32 N 35°C during ischemia. The ID in these cases reached - 20 mV approximately 5.5 min after the onset of ischemia. When brain temperature was maintained at 37”C, ID reached its plateau at 2.5 min after ischemia.

brane potential was at the depolarizing state most of the time; when the halothane concentration was increased to 3% for about 7 min, the membrane potential of the same neuron shifted to the hyperpolarizing state most of the time. The spontaneous activity of spiny neurons ceased immediately after the onset of ischemia resulting in the disappearance of depolarizing state of spontaneous membrane potential fluctuation. The baseline potential was at the hyperpolarizing state at approximately - 85 mV. The spontaneous activity slowly recovered after recirculation. Isolated spontaneous depolarizations started to appear within 30 min. The initial amplitude of these depolarizations was smaller than 10 mV, but large depolarizations with amplitudes >20 mV gradually occurred and intermingled with small depolarizations. Action potentials were sometimes generated from these large deplarizations. Approximately l-2 h after recirculation, the membrane potential of spiny neurons resumed to spontaneous fluctuation between depolarizing and hyperpolarizing states, with full recovery to preischemic conditions approximately 5 h after recirculation. Examples of spontaneous activity changes of spiny neurons at different intervals after ischemia are presented in Fig. 6. The evoked postsynaptic potentials of spiny neurons were less sensitive to an ischemic insult than spontaneous activities. Before ischemia, stimulation of media1 agranular cortex elicited excitatory postsynaptic potentials (EPSPs) from spiny neurons. The initial EPSP was followed by a long-lasting hyperpolarization that was in turn followed by a

Responses of neostriatal neurons after ischemia A

827

before ischemia

0.8% halothane

2 mm after ischemia

-70 mV

-80

-15

-70

-65

-

C

25 min after recirculation

D

2.5 h after recirculation

-60

Membrane potential (mV) B

3% halo&me

I 20 mV -90 mV

E

5 h after recirculation

I-

20 mV

1 set

-90

-85

-80

-75

-70

-65

-60

-55

-62 mV

Membrane potential (rnv) Fig. 5. An example of anesthesia effects on spontaneous membrane potential fluctuation of spiny neurons. The traces above histograms are intracellular recordings from the same neuron at different conditions. The abscissa is the membrane potential. The ordinate is the counts of events representing the time that the membrane potential was maintained. (A) When animal was anesthetized with 0.8% halothane, the membrane potential of this neuron was maintained at depolarizing state most of the time. Histogram shows a peak at about - 72 mV. (B) When halothane concentration was increased to 3% for 7min, the membrane potential of the same neuron as in A stayed longer in hyperpolarizing state. As a result, the histogram shifted to the left with a peak at about -82 mV. The scales in B also apply to A.

depolarization. Shortly after the onset of ischemia, despite the loss of spontaneous activity, initial EPSPs could be elicited by cortical stimulation, but the following hyperpolarization and late depolarization disappeared. The amplitude of EPSPs gradually decreased and finally disappeared shortly before the

Fig. 6. Spontaneous activities of spiny neurons before and at different intervals after transient ischemia and recirculation. All recordings are from the same animal. (A) Spontaneous activity of a spiny neuron before ischemia. The neuron exhibited spontaneous membrane potential fluctuation with some action potentials generated from the depolarizing state. (B) Recording from the same neuron as in A 2 min after the onset of ischemia. Spontaneous membrane potential fluctuation disappeared. The membrane potential maintained at hyperpolarizing state. (C) Recording from another neuron 25 min after recirculation. The membrane potential was at hyperpolarizing state most of the time. Depolarizations with different amplitudes and various durations occasionally occurred. Action potentials were sometimes generated from large depolarizations. (D) Recording from a neuron 2.5 h after recirculation. Membrane potential was fluctuated between depolarizing and hyperpolarizing states. (E) Recording from a neurons 5 h after recirculation. Spontaneous activity of this neuron almost recovered to preischemic level. The scales in E apply to all traces. The top of the action potentials in these figures is truncated. onset of ID. After recirculation, the recovery of initial EPSPs occurred earlier than that of hyperpolarizations and late depolarizations. Cortical stimulation

z. c. xu

828

A

before &hernia

D

before ischemia

1 min after ischemia

E zo min aftexrecirculation

4 min after ischemia

F 60 min after recirculation

B

-11 mV C

Fig. 7. Cortically evoked postsynaptic potentials of spiny neurons before and at different intervals after transient ischemia and recirculation. (A-C) are recordings from the same neuron with constant stimulus intensity. (D-F) are recordings of different neurons from the same animal. (A) Before ischemia, cortical stimulation elicited initial EPSP followed by a long lasting hyperpolarization and then a late depolarization. (B) Approximately 1 min after the onset of ischemia, the hyperpolarization and late depolarization disappeared while initial EPSP persisted. The duration of initial EPSP was increased suggesting the elimination of IPSP that was masked by the EPSP. (C) Four min after ischemia, All evoked responses were abolished. (D) Cortically evoked responses of a spiny neuron before ischemia. (E) Recording from another neuron 20min after recirculation. Cortical stimulation elicited initial EPSPs but no hyperpolarizations and late depolarizations. (F) Recording from a neuron 1 h after recirculation. With the same stimuli as D and E, the hyperpolarizations and late depolarizations following initial EPSPs were sometime elicited. The scales in A apply to B and C. The scales in D apply to E and F.

to evoke initial EPSPs approxmately 10 min after recirculation while the late depolarizations following the initial EPSPs did not occur until approximately 1 h after recirculation. The changes of evoked synaptic potentials during ischemia and after recirculation are presented in Fig. 7. The initial EPSPs elicited by cortical stimulation were monosynaptic responses because of their constant latency with increasing stimulus intensities (Fig. 8). The latency of onset of the initial EPSP remained the same after recirculation, but the threshold, amplitude and duration were altered. As listed in Table 2, the threshold was significantly increased shortly after recirculation (P < 0.05) and slowly returned to preischemic level. Although the amplitude of EPSPs within 1 h after recirculation was close to that observed before ischemia, it was significantly increased l-5 h after recirculation (P < 0.01). The duration of EPSPs was increased by 54% within 1 h after recirculation (P < 0.01) and by 35% l-5 h after recirculation (P < 0.01). Cortical or local stimulation elicited short-lasting inhibitory postsynaptic potentials (IPSPs) from spiny neurons.38,4L@’Because the IPSPs were masked by large initial EPSPs, the inhibitory effect was usually detected with indirect approaches such as paired began

stimulation test. Evidence from the present study suggested that the IPSPs elicited by cortical stimulation disappeared earlier after ischemia and returned later after recirculation than initial EPSPs. Shortly after ischemia, the duration of initial EPSPs increased while its amplitude was slightly depressed (compare Fig. 7A and B), consistent with a suppression of IPSPs which normally limits the duration of the EPSP. Shortly after recirculation, when EPSPs had recovered, no hyperpolarizations following initial EPSPs were observed (Fig. 8). The duration of initial EPSPs was increased from 39.23 f 11.86 ms before ischemia to 60.53 + 23.07 ms shortly after recirculation (P < 0.01) indicating that the IPSPs were attenuated (compare Fig. 9A and B). The duration of EPSPs was slightly decreased from 60.53 &-23.07 to 53.13 f 12.36 ms l-5 h after recirculation indicating a slow recovery of IPSPs (Table 2). Additional support for this conclusion is provided by paired stimulation tests. Before ischemia, the amplitude of EPSPs following a test stimulus with a 20ms inter-stimulus interval (ISI) was approximately 40% of that elicited by conditioning stimulus while the EPSPs following stimuli at longer IS1 (> 40 ms) were at least as large as the responses to the conditioning stimuli, indicating the existence of short duration IPSPs upon cortical stimulation. This inhibitory effect was significantly reduced during early recirculation. The amplitude of EPSPs after test stimuli at 20 ms IS1 was increased from 40% to approximately 70% of that elicited by conditioning stimuli (P < 0.01) within 1 h after recirculation while the size of EPSPs after longer IS1 stimuli had returned to essentially the preischemic level. The inhibition fully recovered within l-5 h after recirculation (Fig. 9).

A.

before ischemia

-83 mV

B.

17 min after recirculation

-80 mV Fig. 8. Cortically evoked monosynaptic responses of a spiny neuron before ischemia and after recirculation. (A) Responses of a spiny neuron before ischemia. Cortical stimuli with increasing stimulus intensities elicited EPSPs with increasing amplitudes accordingly. The latencies of onset of EPSPs, however, were constant indicating that they are monosynaptic responses. (B) Responses of the same neuron as in A 17 min after recirculation. The cortical sstimuli elicited EPSPs with smaller amplitudes than those in A but the latencies remained the same. The hyperpolarizations following initial EPSPs disappeared.

829

Responses of neostriatal neurons after ischemia Table 2. Initial excitatory postsynaptic potential changes before and after ischemia Threshold

Latency

Amplitude

(mA)

(ms)

(mV)

Duration

(ms)

Before ischemia

1.24 f 0.23 (n = 17)

5.24 + 2.06 (n = 40)

16.16 + 6.19 (n = 40)

39.23 k 11.86 (n = 40)

0.5-l h after recirculation

1.45 & 0.12** (n = 6)

4.87 k 2.12 (n = 17)

16.36 k 4.02 (n = 17)

60.53 + 23.07* (n = 17)

l-5 h after recirculation

1.31 + 0.27 (n = 11)

4.02 + 1.69 (n = 16)

21.56 + 6.13* (n = 16)

53.13 + 15.36* (n = 16)

Values are means + SD.; *Unpaired f-test P -C0.01; **Unpaired t-test P < 0.05

Changes

in membrane

properties

The membrane properties of spiny neurons were ischemia transient altered after dramatically (Table 3). The height of action potentials was increased after recirculation, probably due to the hyperpolarization of the baseline membrane potential during this period. The spike threshold which was measured at the beginning of the upstroke of the action potential, was significantly increased from - 52 + 5.51 mV before ischemia to -45 & 7.53 mV within 1 h after recirculation (P < 0.01). The spike threshold fell slightly to -49 + 3.36 mV l-5 h after recirculation but remained persistently higher than the preischemic level (P < 0.05). The rheobase was also significantly increased from 0.47 f 0.17 to 0.76 + 0.30 nA shortly after recirculation (P < 0.01) but recovered l-5 h after recirculation. Membrane input resistance during ischemia was evaluated by monitoring the amplitude of membrane potential deflection in response to constant hyperpolarizing current pulses (1 Hz), -0.5 nA, 200 ms Fig. 3). To compare the changes of input resistance before and during ischemia, 10 pulses were averaged before ischemia and at baseline potential comparable to preischemic level during ischemia, respectively, and the input resistance was calculated. Using this method, the input resistance obtained before and during the early phase of ischemia before the onset of ID was almost identical being 33.77 + 7.61 and 32.94 f 11.36 MQ respectively (n = 17). The membrane input resistance decreased during ID and significantly increased after recirculation. The current-voltage (Z-V) relationship before ischemia and after recirculation was obtained by passing constant current pulses (200 ms, - 1.0 - 0.5 nA) and the current and voltage values were measured at the steady state of the transients (170 ms after the onset of the pulse). The input resistance was derived from the linear portion of the Z-V curve (-0.5 - 0 nA). The input resistance of spiny neurons increased from 31.12 + 13.25 Ma before ischemia to 39.55 + 11.17 MR shortly after recirculation (P < 0.05) and to 41.35 + 16.26 MR after l-5 h recirculation (P < 0.05). Figure 10 illustrates an example of the changes in Z-V relationship of a spiny neuron after

recirculation. spontaneous

Because of the disappearance of the membrane potential fluctuation, the

transients of membrane potential deflections evoked by the current pulses shortly after recirculation became smoother than those observed before ischemia, and was similar to those obtained in slice preparations3’ The strong rectification at the depolarizing extreme of the pulses was apparent but the inward rectification at the hyperpolarizing extreme was greatly attenuated. The input resistance of this neuron increased from 26.5 MR before ischemia to 38.2 Ma at approximately 6 min after recirculation. The time constant of spiny neurons after recirculation was similar to that determined before ischemia. Time constants were derived from membrane potential transients evoked by hyperpolarizing current pulses (100 ms, -0.1 - - 0.9 nA). The values were obtained by fitting a single exponential curve to the transients using data analysis program Axograph 2.0 (Axon Instruments). The time constant of spiny neurons before and at different intervals after recirculation is presented in Fig. 11. The average time constant (measured with a -0.2 nA hyperpolarizing pulse) before ischemia was 7.05 + 3.62 ms, which was not significantly different from 8.45 + 2.77 and 8.90 &-4.60 ms measured at different intervals after recirculation, respectively (Table 3). Repetitive firing of spiny neurons was investigated by application of depolarizing current pulses (400 ms, 0.1 - 1.5 nA). Action potentials evoked by these pulses were followed by afterhyperpolarizations (AHPs). There was no spike-frequency adaptation of the spike trains during application of constant current (up to 1.5 nA). The repetitive firing characteristics of spiny neurons dramatically changed shortly after recirculation, The responses of a spiny neuron to 0.5 and 1.OnA depolarizing pulses before ischemia are shown in Fig. 12A. Figure 12B shows the responses of the same neuron to the depolarizing pulses of the same intensities as Fig. 12A at 4.5 min after recirculation. No action potential was evoked with the 0.5 nA depolarizing pulse after recirculation, probably because of the increased spike threshold at this time. Action potentials were evoked with the 1.O nA depolarizing pulse, but the AHPs were larger than those of preischemia and the spike frequency was lower. Plots of spike frequency versus injected current amplitude for the first and second ISIS of spiny neurons are presented in Fig. 12C and D. Spike frequency increased linearly with the current intensity

z. c. XU

830

over the range shown. The spike frequency was dramatically depressed shortly after recirculation, especially at high current intensities, and slowly returned to the preischemic level after 2-5 h recirculation. DISCUSSION Methodological

considerations

Intracellular recording in vivo offers unique advantages for the study of neurophysiological mechanisms associated with neuronal injury following ischemia. It preserves the whole animal preparation and has the power to detect synaptic activity and membrane property changes. A technical difficulty of

A

before ischemia

-64 mV

25 min after recirculation

-91 mV

2 h after recirculation

-83 mV

40

60

80

100

120

Inter-stimulus interval (ms) Fig. 9

140

maintaining stable intracellular recording in vivo arises from the movements caused by respiration and the vascular pulsation of the animal. The fluctuations in brain volume caused by ischemia and recirculation make the recording even more difficult. Although successful intracellular recordings have been continuously performed before, during, and after 5-8 min ischemia in the present study, the number of such recordings is relatively low. Of necessity, recordings from different cells were collected at various intervals after recirculation. The interval between the onset of recicrulation and each recording was noted and data from recordings in the same interval were grouped across the experiments. Taking advantage of intracellular staining techniques, identified neurons were sorted according to their morphological appearance. All data presented in this paper were collected from spiny neurons. Combining the data from occasional continuous recordings and those from the interval groups provided an overall view of temporal changes in spiny neuron physiology after transient ischemia. Some of the data in the present study were obtained during sequential periods of ischemia and reperfusion in the same animal. Repeated ischemia has been shown to have different effects on neuronal histopathology depending on the duration of and interval between the insults. It has been shown that sublethal ischemia induced tolerance to subsequent lethal ischemia in ischemia-vulnerable cells if the second insult was performed 24-48 h after the first

Fig. 9. Responses of spiny neurons to cortically evoked paired stimulation before ischemia and at different intervals after recirculation. (A-C) are recordings from the same animal. (A) Superimposed recordings from a spiny neuron showing the responses to paired stimulations with different ISIS before ischemia. Conditioning stimuli elicited initial EPSPs followed by hyperpolarizations. The amplitude of EPSP elicited by test stimulus of 20 ms IS1 was much smaller than that of control EPSP while the amplitude of EPSPs elicited by stimuli of ISI longer than 40 ms did not attenuate suggesting the existence of a short duration IPSP masked by initial EPSP. (B) Responses to paired stimulation from another neuron 25 min after recirculation. The duration of intial EPSPs elicited by conditioning stimuli was longer and the amplitude of EPSP elicited by test stimulus of 20 ms IS1 was larger than those shown in A. These results suggested the great suppression of IPSP after ischemia (C) Recordings from a neuron 2 h after recirculation. The size of initial EPSPs and the amplitude of EPSP elicited by stimulus of 20ms IS1 were comparable to those in A suggesting the recovery of IPSP. (D) Plot showing the relationship between IS1 and the amplitude of test stimuli in paired stimulation. The ordinate is the percentage of test EPSPs’ amplitude to conditioning EPSPs’ amplitude. The value of each plot is the mean and standard deviation. The amplitude of EPSPs elicited by test stimuli of 20 ms IS1 (asterisk) within 1h after recirculation was significantly larger than those before ischemia (P < 0.01) and long after recirculation (P i 0.05). No significant difference was found among the amplitude of EPSPs elicited by test stimuli with ISI longer than 40ms before and after recirculation. Data suggested a transient suppression of cortically evoked IPSPs in spiny neurons following 5-8 min forebrain ischemia.

Responses

of neostriatal

neurons

after ischemia

831

insult.29.33.35On the other hand, if the second ischemia was performed l-2 h after the first one, ischemic damage was worsen.29.57 Repeated ischemia had minimal effect if the interval of insults was between 6 and 24 h (I. Halaby and W. A. Pulsinelli, unpublished observation). Because ischemic tolerance is not the focus of the present study, efforts have been taken to reduce the effect of repeated ischemia. The duration of ischemia was relatively short (5-8 min) in comparison with the lethal ischemia to spiny neurons in neostriatum (30min). In addition, the second ischemia was induced approximately 5 h after the first insult, at which time the spontaneous membrane potential fluctuation, a sign of recovery from the first insult, had been restored. As a result, no difference was found between the electrophysiological responses of spiny neurons to the first and second ischemia. Ischemic depolarization The present study provides insights into the neurophysiological changes of spiny neurons in neostriatum after transient ischemia. The ID induced in spiny neurons of neostriatum by severe ischemia is similar to that induced in CA1 pyramidal neurons of hippocampus. In both cases when brain temperature is maintained at 37°C the onset of ID is approximately 3 min after ischemia and the membrane potential begins to repolarize within 2 min after recirculation. Whether the membrane potential of ID (approximately -20 mV) is the true value of the membrane potential during ischemia is open to discussion. Previous studies have shown that the membrane potential during hypoxic depolarization was approximately - 20 mV if referred to ground potential but was actually close to zero if extracellular potential was used as a reference because of the negative extracellular potential shift during this period. “J’ In the present study, using ground as a reference, we detected a potential shift from - - 20 mV toward 0 mV in recordings in which cells were lost shortly after recirculation, suggesting that the potential recorded during ID closely, if not exactly, reflects the true value of the membrane potential. The amplitude of ID (_ 50 mV) is consistent with previous reports on ID induced in cortical” and hippocampal neurons53.s4,63 in vivo. These observations suggest that these neurons may have similar mechanisms maintaining membrane potentials, some components of which are sensitive to ischemia, while other components may be relatively resistant and thereby contribute to the - 20 mV maintained during ID. This does not imply that all neurons in the CNS show identical changes in membrane potential after ischemia. An in vitro study has shown that 5 min after anoxia. CAI neurons in hippocampus depolarized to approximately -20 mV while neurons in dorsal vagal motor nucleus depolarized to approximately - 30 mV.14 In the present study, an aspiny neuron of neostriatum depolarized to approximately - 40 mV64

z. c. XU

832

in comparison with the - 20 mV ID of spiny neurons in the same region induced by the ischemia of same severity. It is not clear what factors determine these variations in ID or whether the differences in ID among neurons are related to selective neuronal damage following transient ischemia.

A

before ischemia

Ol

-0.8

-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 Hyperpolarizing current (nA)

-0.1

Fig. 11. Time constant of spiny neurons before ischemia and at different intervals after recirculation. The abscissa is the current range of hyperpolarizing pulses. The ordinate is the time constant derived from the membrane potential transients evoked by the current pulses. The value of each plot is the mean and standard deviation. No significant difference was found among time constants measured at different current pulses before and at different intervals after ischemia.

-15 mV

B

-0.9

6 mm after recirculation

-89 mV

DifSerentiul sensitivity of monosynaptic versus polysynaptic responses to ischemia

00.

-60 s g

a ‘a si g

l

-70 E

00°

-80

00

ooo”

-90

B

-100

g

-110

00

E -120

-1.2

l0

ooo l

l I

o before l 6 min after

0

-0.9

0

l

-0.6

The present study indicates that cortically evoked polysynaptic responses from spiny neurons are more sensitive to ischemia than monosynaptic ones. Evidence supporting this conclusion comes from the changes in spontaneous membrane potential fluctuation after ischemia. Under control conditions the membrane potential of spiny neurons spontaneously shifts between depolarizing and hyperpolarizing states61 Studies have suggested that the hyperpolarizing state is the baseline potential of the neuron and the depolarizing state is the synchronized polysynaptic event originated from cerebral cortex and thalamus6’ The spontaneous depolarizations and the late depolarizations following initial EPSPs disappeared or were greatly reduced in amplitude in neostriatal grafts, in which cortical and thalamic afferents were only partially restored,62 and in neostriatal slices, in which most of the afferents were severed.5,39The fact that the baseline membrane potential shifts to the hyperpolarizing state with deep anesthesia and to the depolarizing state with light anesthesia also strongly supports the above notion. Immediately after the onset of ischemia, spontaneous membrane potential fluctuation ceases while cortically elicited monosynaptic EPSPs persist, indicating the polysynaptic responses are more easily suppressed by &hernia than the monosynaptic ones. Other evidence comes from the observation that IPSPs are more sensitive to ischemia than monosynaptic EPSPs. In the present study the IPSPs in spiny neurons disappeared earlier after ischemia and recovered later following recirculation than initial EPSPs. Similar results have been reported in CA1 pyramidal neurons after transient ischemia in vim”

-0.3

0

0.3

0.6

Current @A) Fig. 10. Current-voltage relationship of a spiny neuron before ischemia and shortly after recirculation. A and B arc recordings from the same neuron. Each trace is the average of four recordings. The upper panels are the membrane potential deflection caused by the current pulses. The lower panels are intracellularly applied constant current pulses. Scales in A apply to B. (A) Current-voltage relationship before ischemia. The traces of voltage deflection are noisy because of the spontaneous membrane potential fluctuation. (B) The I-V relationship of the same neuron as in A after 6 min recirculation. Because the membrane potential fluctuation has not recovered at this time, the traces are smooth. The membrane potential deflections are greater that those before ischemia suggesting the increase of input resistance. (C) The plot of I-V curve from A and B showing the input resistance change after recirculation. The slope of I&V curve 6 min after recirculation is steeper than that before ischemia suggesting the increase of membrane input resistance. The input resistance of this neuron measured at the linear portion (-0.5 - 0 nA) increased from 26.5 MB before ischemia to 38.2 MR 6 min after recirculation. The inward recitication at the hyperpolarization extreme was greatly attenuated.

Responses of neostriatal neurons after ischemia

C

before ischemia

200

First IS1

I

-82 mV../

0.3

0.4

0.3

0.6

0.;

0.8

0:9

1:o

1.1

1.2

1.3

Current (nA)

-84 mV _I

I

D

I

I

Second ISI

140 -I

B

4.5 min after recirculation

a

100 -

]POmV

Mmr

-80 mV J

m-

A!

40_ . 20 -

-73rnVfpL J

so-

l

cz

J I

G 8

.

OJ.

1

0.3

0.4

0.5

0.6

0.7

0.8

Current

1

0.9

1.0

1.1

1.2

1.3

(nA)

Fig. 12. Repetitive firing of spiny neurons before ischemia and at different intervals after recirculation. (A) Responses of a spiny neuron to depolarizing current pulses before ischemia. The upper panel is the repetitive firing evoked by 1 nA depolarizing current pulse. The lower panel is the firing evoked by 0.5 nA depolarizing current pulse. No spike frequency adaptation of the spike trains evoked by these depolarizing current pulses was observed. (B) Responses of the same neuron to the same current intensity as in A after 45 min recirculation. The repetitive firing of the neuron was suppressed during this period. No action potential was evoked by 0.5 nA depolarizing current pulse. The frequency of spike train evoked by I nA current pulse was lower and the AHPs were larger than those in A. Scales in B apply to A. (C) Plot showing the relationship between spike frequency and depolarizing currents of first interspike interval (ISI) of spiny neurons before and after ischemia. Value of each individual plot is the mean. The spike frequency increased linearly with increasing current intensities. In comparison with that before ischemia, the spike frequency, especially at high current intensities, was dramatically decreased within 2 h after recirculation. The spike frequency returned to preishemic level 2-5 h after recirculation. These data suggested the transient suppression of repetitive firing of spiny neurons after ischemia. (D). The relationship between spike frequency and depolarizing currents of second ISI of spiny neurons before and after ischemia. Similar to first ISI, the frequency of second IS1 was suppressed within 2 h after recirculation and returned to preischemic level 2-5 h following recirculation.

and in neurons in hippocampus,‘9,36,37 and neocortex40.50 after hypoxia in vitro. The above observations cannot be attributed to the greater sensitivity of inhibitory interneurons to ischemia because it has been shown that the interneurons in hippocampus and neostriatum are more resistant to ischemia than Since the IPSPs elicited the principal neurons. 8~‘7~‘8,23,43 by afferent stimulation are mediated through more than one synapse, it is probably the sensitivity of synaptic transmission that determines the loss of IPSPs after ischemia. This also provides an explanation for the high sensitivity of spontaneous membrane potential fluctuation, which is also a polysynaptic event, to transient ischemia. It is likely that the synaptic transmission is the key element of these phenomena. Failure of synaptic transmission, which is presumably directly related to the energy

depletion during ischemia, blocks the signals conducted along the pathway. That is why polysynaptic events, such as spontaneous membrane potential fluctuation and IPSPs, are more sensitive to ischemia than the monosynaptic responses that are directly driven by the stimuli. Ischemic

eflect on membrane

properties

The membrane properties of spiny neurons were dramatically altered after transient ischemia. In the present study, the input resistance of spiny neurons was unchanged during the initial phase of ischemia and increased after recirculation. This result is different from those obtained in vitro but is consistent with previous in vivo observations. It has been reported that the input resistance of hippocampal or cortical neurons in vitro was decreased after

z. c xu

834

hypoxia”~26~28~37~40~4y and returned to control level after reoxygenation.‘y,26 On the other hand, an early in uivo study has shown that the input resistance of spinal cord motor neurons did not change after 26 min hypoxia and slightly increased after resuming air ventilation.15 Increase in input resistance of medullary respiratory neurons after systemic hypoxia has also been reported in a recent study using in uivo preparations.48 The above conflicting results may stem from the differences between in vivo versus in vitro preparations. One possible explanation of increase in input resistance after ischemia in viva has been postulated as due to the decrease in electric shunt of synaptic currents caused by the abolishment of afferent bombardments following ischemia.63 However, the mechanisms underlying changes in input resistance after &hernia in viva are not clear and need further investigation. In the present study, the rheobase and spike threshold of spiny neurons are significantly increased shortly after recirculation and slowly returned to preischemic level. The increase in rheobase can be partially explained by the hyperpolarization of the membrane potential during this period. The increase in spike threshold, however, provides strong evidence indicating the suppression of neuronal excitability due to ischemia. Increased spike threshold accounts for some of the postischemic changes observed in the present study. For example, the sensitivity of polysynaptic events to ischemia can be explained by the increase in spike threshold. Increase in spike threshold may also contribute to the depression of

repetitive firing evoked by depolarizing pulses shortly after recirculation. It has been shown in vitro that dopamine inhibited the repetitive firing of neostriatal neurons by increasing the spike threshold5* and this inhibition was postsynaptically mediated by the activation of Dl dopamine receptors4 The extracellular dopamine concentration in neostriatum is increased during ischemia.22,47 Increases in spike threshold of spiny neurons after transient ischemia may be a consequence of such excessive dopamine release. CONCLUSIONS In summary, the spontaneous activity and evoked postsynaptic responses of spiny neurons in neostriatum are suppressed, and the excitability of spiny neurons is decreased following 5-8 min forebrain ischemia. The polysnaptic responses are more sensitive to ischemia than monosynaptic responses. No hyperactivity has been observed up to 7 h after cerebral recirculation. Although 5-8 min forebrain ischemia is not long enough to produce neuronal damage in neostriatum, the present study provides a foundation for further studies of neurophysiological changes in vulnerable spiny neurons after transient ischemia using intracellular recording in an in vivo preparation. Acknowledgements-1 thank Drs T. S. Nowak and C. J. Wilson for valuable comments on this manuscript, Miss M. O’Conor for secretarial assistance. This research was supported by grants UTMG R07-3280-74, AHA 94008130, NIH NS33103-01.

REFERENCES 1. Benveniste H., Drejer J., Schousboe A. and Diemer N. H. (1984) Elevation of the extracellular concentration of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. J. Neurochem. 43, 1369-l 374. 2. Brierley J. B. (1976) Cerebral hypoxia. In Greenfield’s Neuropathology (eds Blackwood W. and Corsellis A.), pp. 43-85. Arnold, London. 3. Buzsaki G., Freund T. F., Bayardo F. and Somogyi P. (1989) Ischemia-induced changes in the electrical activity of the hippocampus. Exp. Brain Res. 18, 2688278. 4. Calabresi P.; Mercuri N., Stanzione P., Stefani A. and Bernardi G. (1987) Intracellular studies on the dopamine-induced Neuroscience 20, 7577771. firing inhibition of neostriatal neurons in vitro: evidence for Dl receptor involvement. 5. Calabresi P., Mercuri N. B. and Bernardi G. (1990) Synaptic and intrinsic control of membrane excitability of neostriatal neurons-II. An in vitro analysis. J. Neurophysiol. 63, 663-675. 6. Chang H. T., Wilson C. J. and Kitai S. T. (1981) Single neostriatal efferent axons in the globus pallidus: a light and electron microscopic study. Science 213, 9155918. I. Chang H. S., Sasaki T. and Kassell N. F. (1989) Hippocampal unit activity after transient ischemia in rats. Stroke 20, 1051-1058. in the striatum of adult gerbils: 8. Chesselet M.-F., Gonzales C., Lin C. S., Polsky K. and Jin B. K. (1990) Ischemicdamage relative sparing of somatostatinergic and cholinergic interneurons contrasts with loss of efferent neurons. Exp. Neural. 110, 209-218. 9. Choi D. W. and Rothman S. M. (199oj The role of glutamate neurotoxicity in hypoxic-ischemic neuronal death. A. Rev. Neurosci. 13, 171-182. 10. Collewijn H. and Van Harreveld A. (1966) Membrane potential of cerebral cortical cells during spreading depression and asphyxia. Exp. Neural. 15, 425-436. 11. Czeh G., Aitken P. G. and Somjen G. G. (1993) Membrane currents in CA1 pyramidal cells during spreading depression (SD) and SD-like hypoxic depolarization. Brain Res. 632, 1955208. 12. DiFiglia M., Pasik P and Pasik T. (1976) A Golgi study of neuronal types in the neostriatum of Monkeys. Bruin Res.

114,245-256. 13. Dirnagl U., Kaplan B., Jacewicz M. and Pulsinelli W. (1989) Continuous measurement of cerebral blood flow by laser-Dropple flowmetry in a rat stroke model. J. cerebr. Blood Flow Metab. 9, 589-596. 14. Donnelly D. F., Jiang C. and Haddad G. G. (1992) Comparative responses of brain stem and hippocampal neurons to 0, deprivation: in vitro intracellular studies. Am. J. Physiol. 262, 549-554.

Responses

of neostriatal

neurons

after &hernia

835

15. Eccles R. M., Loyning Y. and Oshima T. (1966) Effects of hypoxia on the monosynaptic reflex pathway in cat spinal cord. J. Neurophysiol29, 3 15-332. 16. Fonnum F. and Storm-Methisen J. (1977) High affinity uptake ofglutamate in terminals of corticostriatal axons. Nature 266, 3777378. 17. Francis A. and Pulsinelli W. (1982) The response of GABAergic and cholinergic neurons to transient cerebral ischemia. Brain Res. 243, 271-278. 18. Freund I. T., Buzaki G., Leon A., Bairnbridge K. G. and Somogyi P. (1990) Relationship of neuronal vulnerability and calcium binding protein immunoreactivity in ischemia. Exp. Brain Res. 83, 55-66. neurons in L’ilro 19. Fujiwara N., Higashi H., Shimoji K. and Yoshimura M. (1987) Effects of hypoxia on rat hippocampal J. Physiol. 384, 131~151. 20. Globus M. Y.-T., Ginsberg M. D., Dietrich D., Busto R. and Scheinberg P. (1987) Substantia nigra lesion protects against ischemic damage in the striatum. Neurosci. Lett. 80, 251-256. extracellular 21. Globus M. Y.-T., Busto D., Dietrich R., Martinez E., Valdes I. and Ginsberg M. D. (1988) Intra-ischemic release of dopamine and glutamate is associated with striatal vulnerability to ischemia. Neurosci. Left. 91, 36640. 22. Globus M. Y.-T., Busto D. R., Dietrich W. D., Martinez E. M., Valdes I. and Ginsberg M. D. (1988) Effect of ischemia on the in airo release of striatal dopamine, glutamate and g-aminobutyric acid studied by intracerebral microdialysis. J. Neurochem. 51, 145551464. 23. Gonzales C., Lin R. C-S. and Chesselet M.-F. (1992) Relative sparing of GABAergic interneurons in the striatum of gerbils with ischemia-induced lesions. Neurosci. Left. 135, 53-58. 24. Grofova I. (1975) The identification of striatal and pallidal neurons projecting to the substantia nigra. An experimental study by means of retrograde axonal transport of horseradish peroxidase. Brain Res. 91, 286-291. R. G. and Williams V. F. (1971) Electrical activity and ultrastructure of cortical neurons and synapses in 25. Grossman ischemia. In Brain Hypoxia (eds Brierley J. B. and Meldrum, B. S.), pp. 61-75. Heinemann, London, 26. Hasen A. J., Hounsgaard J. and Jahnsen H. (1982) Anoxia increases potassium conductance in hippocampal nerve cells. Acta Physiol. Stand. 115, 301-310. 27. Imon H., Mitani A., Andou Y.. Arai T. and Kataoka K. (1991) Delayed neuronal death is induced without postischemic hyperexcitability: continuous multiple-unit recording from ischemic CAI neurons. J. cerebr Blood Flow Meteb. 11, 819-823. in 28. Jiang C. and Haddad G. G. (1992) Differential responses of neocortical neurons to glucose and/or O2 deprivation the human and rat. J. Neurophysiol68, 216552173. 29. Kato H., Liu Y., Araki T. and Kogure K. (1991) Temporal profile of the effects of pretreatment with brief cerebral ischemia on the neuronal damage following secondary ischemic insult in the gerbil: cumulative damage and protective effects. Brain Res. 553, 238-242. 30. Kawaguchi Y., Wilson C. J. and Emson P. C. (1989) Intracellular recording of identified neostriatal patch and matrix spiny cells in a slice preparation preserving cortical inputs. J. Neurophysiol. 62, 1052-1068. 31. Kemp J. M. and Powell T. P. S. (1971) The structure of the caudate nucleus of the cat: light and electron microscopy. Phil. Trans. Sot. Lond. B. 262, 3833401. 32. Kemp J. M. and Powell T. P. S. (1971) The site of termination of afferent fibers in the caudate nucleus. Phil. Trans. Sot. Lond. B. 262, 413-427. 33. Kirino T., Tsujita Y. and Tamura A. (1991) Induced tolerance to ischemia in gerbil hippocampal neurons. J. cerehr. Blood Flout Metab. 11, 299-307. W. A. (1991) Biotin-containing compound N-(2-aminoethyl) biotinamide for intracellular 34. Kita H. and Armstrong labeling and neuronal tracing studies: comparison with biocytin. J. Neurosci. Mefh. 37, 1141-I 150. 35. Kitagawa K., Matsumoto M., Tagaya M., Hata R., Ueda H., Niinobe M., Handa N., Fukunaga R.. Kimura K., Mikoshiba K. and Kamada T. (1990) “Ischemic tolerance” phenomenon found in the brain. Brain Res. 528, 21-24. 36. Krnjevic K.. Xu Y. 2. and Zhang L. (1991) Anoxic block of GABAergic IPSPs. Neurochem. Res. 16, 2799284. 37. Leblond J. and Krnjevic K. (1989) Hypoxic changes in hippocampal neurons. J. Neurophysiol. 62, l-14. J. W., Park M. R. and Kitai S. T. (1981) Inhibition in slices of rat neostriatum. Brain Res. 212, 38. Lighthall 1822187. 39. Lighthall J. W. and Kitai S. T. (1983) A short duration GABAergic inhibition in identified neostriatal medium spiny neurons. In vitro slices study. Brain Res. Bull. 11, 103-110. 40. Luhmann H. J. and Heinemann U. (1992) Hypoxia-induced functional alterations in adult rat neocortex. J. Neurophysiol. 67, 798-8 1I. 41, Misgeld U., Wagner A. and Ohno T. (1982) Depolarizing IPSPs and depolarizations by GABA of rat neostriatum cells in vitro. Exp. Brain Res. 45, 108-l 14. 42. Mitani A., Imon H., Iga K., Kubo H. and Kataoka K. (1990) Gerbil hippocampal extracellular glutamate and neuronal activity after transient &hernia. Brain Res. Bull. 25, 319-324. 43. Nitsch C., Scotti A., Sommacal and Kalt G. (1989) GABAergic hippocampal neurons resistant to ischemia-induced Neurosci. Left. 105, 2633268. neuronal death contain the Ca”+-binding protein parvalbumin. 44. Preston R. J., Bishop G. A. and Kitai. (1980) Medium spiny neurons from rat striatum: an intracellular horseradish peroxidase study. Brain Res. 183, 2533263. 45. Pulsinelli W. A. and Brierley J. A. (1979) New model of bilateral hemispheric ischemia in the unanesthetized rat. Stroke 10, 2677272. 46. Pulsinelh W. A., Brierley J. B. and Plum F. (1982) Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann. Neural. 11, 491-498. 47. Richard D. A., Obrenovitch T. P., Symon L. and Curson G. (1993) Extracehular dopamine and serotonin in the rat striatum during transient ischemia of different severities: a microdialysis study. J. Neurochem. 60, 128-136. 48. Richter D. W., Bischoff A.. Anders K., Bellinaham M. and Windhorst U. (1991) Resoonse of the medullarv _. respiratory _ network of the cat to hypoxia. J. Physiol 44; 231-256. 49. Rosen A. S. and Morris M. E. (1991) Depolarizing effects of anoxia on pyramidal cells of rat neocortex. Neurosci. Lerr. 124, 1699173. 50. Rosen A. S. and Morris M. E. (1993) Anoxic depression of excitatory and inhibitory postsynaptic potentials in rat neocortical slices. J. Neurophysiol. 69, 1099117.

z. c. xu

836

51. Rothman S. M. and Olney J. W. (1986) Glutamate and the path physiology of hypoxic-ischemic brain damage. Ann. Neural. 19, 105-l 11. 52. Rutherford A., Garcia-Munoz M. and Arbuthnott G. W. (1987) An afterhyperpolarization recorded in striatal cells in vitro: effect of dopamine administration. Exp. Brain Res. 71, 3999405. 53. Silver I. A. and Erecinska M. (1990) Intracellular and extracelular changes of [Ca”] in hypoxia and ischemia in rat brain in vivo. J. gen. Physiol. 95, 837-866. 54. Silver I. A. and Erecinska M. (1992) Ion homeostasis in rat brain in Go: intra- and extracellular [Ca”] and [H+] in the hippocampus during recovery from short-term, transient ischemia. J. cerebr. Blood How Meteb. 12, 7599772. 55. Spenser H. J. (1976) Antagonism of cortical excitation of striatal neurons by glutamic acid diethy ester: evidence for glutamic acid as an excitatory transmitter in the rat striatum. Bruin Res. 102, 91-101. 56. Suzuki R., Yamaguchi T., Li C. L. and Klatzo I. (1983) The effects of 5-minute ischemia in mongolian gerbils-II. changes of spontaneous neuronal activity in cerebral cortex and CA1 sector of hippocampus. Acta Neuropath. Berlin 60, 217-222.

51. Tomida S., Nowak T. S., Vass K., Lohr J. M. and Klatzo I. (1987) Experimental model for repetitive ischemic attacks in the gerbil: the cumulative effect of repeated ischemic insults. J. cerebr. Blood Flow Metab. 7, 773-782. 58. Urban L., Neil1 K. H., Cram B. J., Nadler J. V. and Somjen G. G. (1989) Postischemic synaptic physiology in area CA1 of the gerbil hippocampus studied in vitro. J. Neurosci. 9, 3966-3975. 59. Wilson C. J. and Groves P. M. (1980) Fine structure and synaptic connection of the common spiny neuron of the rat striatum: a study employing intracellular injection of horseradish peroxidase. J. camp. Neural. 194, 599-615. 60. Wilson C. J., Chang H. T. and Kitai S. T. (1983) Disfacillitation and long-lasting inhibition of neostriatal neurons in the rat. Exp. Brain Res. 51, 227-235. 61. Wilson C. J. (1993) The generation of natural firing patterns in neostriatal neurons. In Progress in Brain Research (eds Arbuthnott G. W. and Emson P. C.) Vol. 99, pp 2777297. Elsevier, Amsterdam. 62. Xu Z. C., Wilson C. J. and Emson P. C. (1991) Synaptic potentials evoked in spiny neurons in rat neostriatal grafts by cortical and thalamic stimulation. J. Neurophysiol. 65, 477-493. 63. Xu Z. C. and Pulsinelli W. A. (1994) Responses of CA1 pyramidal neurons in rat hippocompus to transient forebrain ischemia: an in viuo intracellular recording study. Neurosci. Lett. 171, 187-191. 64. Xu Z. C. and Pulsinelli W. A. (1994) The responses of &hernia-resistant neurons after transient ischemia is different from those of ischemia-vulnerable neurons: an in uivo intracellular recording study. Sot. Neurosci. Abstr. 423, 3. (Accepted 13 January 1995)