Selective neuronal vulnerability following transient cerebral ischemia in the gerbil: distribution and time course

An important feature of ischemic brain damage is the selective vulnerability of specific neuronal populations. We studied the distribution and time course of neuronal damage following transient cerebral ischemia in the gerbil, using light microscopy and 45Ca autoradiography. Following 5 min of ischemia, selective neuronal damage determined by abnormal 45Ca accumulation was recognized only in the hippocampal CA1 subfield and part of the inferior colliculus. Ischemia for 10 to 15 min caused extensive neuronal injury in the 3rd and 5th layers of neocortex, the striatum, the septum, the whole hippocampus, the thalamus, the medial geniculate body, the substantia nigra, and the inferior colliculus. Progression of the damage was rapid in the medial geniculate body and the inferior colliculus, moderate in the neocortex, striatum, septum, thalamus, and the substantia nigra, and was delayed in the hippocampal CA1 sector. However, the delayed damage of the hippocampus occurred earlier when the ischemia period was prolonged. Histological observation revealed neuronal loss in the identical sites of the 45Ca accumulation. This study revealed that the distribution and time course of selective neuronal damage by ischemia proceeded with different order of susceptibility and different speed of progression.     

Neuronal damage and calcium accumulation following transient cerebral ischemia in the rat

The purpose of this study was to examine the distribution of neuronal damage following transient cerebral ischemia in the rat model of four-vessel occlusion utilizing light microscopy as well as⁴⁵Ca-autoradiography. Transient ischemia was induced for 30 min. The animals were allowed to survive for 7 d after ischemia. In the animals subjected to ischemia, the most frequently and seriously damaged areas were the paramedian region of hippocampus, the hippocampal CA1 sector, and the dorsolateral part of striatum, followed by the inferior colliculus, the substantia nigra, the frontal cortex, and the thalamus, which were moderate damaged. Furthermore, the cerebellar Purkinje neurons, the hippocampal CA4 sector, the medial geniculate body, and the hippocampal CA3 sector were slightly affected.⁴⁵Ca-autoradiographyic study also revealed calcium accumulation in the identical sites of ischemic neuronal damage, except for the frontal cortex. Regional cerebral blood flow during 10 min of ischemia was severely decreased in selectively vulnerable areas. The blood flow in the medial geniculate body, the substantia nigra, the inferior colliculus, and the cerebellum was less pronounced than that in the selectively vulnerable areas. The present study demonstrates that transient cerebral ischemia can produce significant neuronal damage not only in the selectively vulnerable regions, but also in the brainstem.     

Emphasized selective vulnerability after repeated nonlethal cerebral ischemic insults in rats.

BACKGROUND AND PURPOSE: We examined the density and distribution of brain damage after repeated periods of nonlethal ischemic insult in rats in comparison with damage after single lethal periods of ischemic insult. METHODS: Transient cerebral ischemia was induced by four-vessel occlusion for 3, 10, 20, and 30 minutes, and 3-minute periods of ischemia were repeated two, three, or five times at 1-hour intervals, followed by 7 days of survival. RESULTS: Three minutes of ischemia produced no brain damage, but 10-30 minutes of ischemia produced neuronal damage, depending on the length of ischemia, to the selectively vulnerable forebrain regions such as hippocampal CA1 and CA4 subfields, neocortex, striatum, and ventral thalamus, as well as to the brain stem structures (medial geniculate body, substantia nigra, and inferior colliculus) and cerebellar Purkinje cells. Two 3-minute periods of ischemic insult produced neuronal damage to the hippocampal CA1 subfield. Three and five 3-minute insults produced neuronal damage extensively to the selectively vulnerable forebrain areas. An intense cumulative effect of damage was observed in the ventral thalamus, whereas the substantia nigra and the inferior colliculus were resistant to repeated ischemic insults. CONCLUSIONS: Our data indicate that the density and distribution of neuronal damage after repeated ischemic insults are altered as compared with after single ischemia.     

Calcium deposits in the thalamus following repeated cerebral ischemia and long-term survival in the gerbil

We investigated the long-term changes in the gerbil brain following three episodes of 2-min forebrain ischemia at 1-h intervals in comparison with a 6-min period of ischemia. The animals were sacrificed after 1 month and 6 months. Following either ischemic insult, the hippocampal CA1 region showed a loss of pyramidal neurons together with a diffuse calcium accumulation as shown by alizarin red S staining. Three 2-min ischemic insults additionally produced neuronal damage in the striatum and thalamus. The thalamic damage was accompanied by an accumulation of small calcium granules after 1 month and large calcium concretions after 6 months. Calcium staining in the striatum was weak. Thus, the thalamic neuronal damage was accompanied by an active process of calcification, which has not been described in experimental cerebral ischemia models. The observations show that repeated ischemic insults produce different long-term effects in different brain regions.     

Neuronal damage following non-lethal but repeated cerebral ischemia in the gerbil

Summary Brief, non-lethal transient forebrain ischemia in the gerbil can injure selectively vulnerable neurons when such ischemia is induced repeatedly. The influence of the number and interval of the ischemic insults on neuronal damage, as well as the time course of damage, following repeated 2-min forebrain ischemia were examined. A single 2-min forebrain ischemia were examined. A single 2-min ischemic insult caused no morphological neuronal damage. A moderate number of hippocampal CA1 neurons were destroyed following two ischemic insults with a 1-h interval, and destruction of almost all CA1 neurons resulted from three or five insults at 1-h intervals. Three and five insults also resulted in moderate to severe damage to the striatum and thalamus, depending on the number of episodes. Although three ischemic insults at 1-h intervals caused severe neuronal damage, this number of insults at 5-min and 4-h intervals caused destruction of relatively few neurons, and non neurons were destroyed at 12-h intervals. Following three ischemic insults at 1-h intervals, damage to the striatum, neocortex, hippocampal CA4 subfield and thalamus was observed at 6–24 h of survival, whereas damage to the hippocampal CA1 subfield appeared at 2–4 days. The results indicate that even a brief non-lethal ischemic insult can produce severe neuronal damage in selectively vulnerable regions when it is induced repeatedly at a certain interval. The severity of neuronal damage was dependent on the number and interval of ischemic episodes.     

Regional neuroprotective effects of pentobarbital on ischemia-induced brain damage

We investigated the neuroprotective effect of pentobarbital, a GABAA receptor-effector, on ischemic neuronal damage in the gerbils. The animals were allowed to survive for 7 days after 10-min ischemia induced by bilateral occlusion of the common carotid arteries. Morphological changes and abnormal calcium accumulation were evaluated in selectively vulnerable areas after ischemia. Pentobarbital (40 mg/kg, IP), administered 30 min prior to ischemia, significantly reduced neuronal cell loss in the neocortex, the striatum, and the hippocampal CA3 sector. However, pentobarbital failed to prevent the damage to the hippocampal CA1 sector and the thalamus. 45Ca autoradiographic study also revealed that a marked calcium accumulation was found in the selectively vulnerable regions after ischemia, which was consistent with the extent of histological neuronal damage. The abnormal calcium accumulation was reduced in the sites corresponding to most of the regions in which the protective effect of pentobarbital was found. The results suggest that ischemia-induced neuronal damage may be partly caused by an imbalance between excitatory and inhibitory input.     

Prevention of abnormal calcium accumulation in postischemic gerbil brain by vinconate

We investigated the effect of vinconate on ischemia-induced calcium accumulation in the gerbil brain. The animals were allowed to survive for 7 days after 10 min of ischemia. Abnormal calcium accumulation was evaluated in the gerbil brain after ischemia. Following 10 min ischemia, abnormal calcium accumulation was found in the neocortex, the striatum, the hippocampus, the thalamus, the substantia nigra and the inferior colliculus. Intraperitoneal administration of vinconate (100 mg/kg) immediately after 10 min of ischemia significantly reduced the areas of abnormal calcium accumulation only in the striatum. However, the application of vinconate (100 and 300 mg/kg) 10 min before ischemia dose-dependently decreased the areas of abnormal calcium accumulation in the striatum, the thalamus and the substantia nigra. Morphological observation revealed neuronal damage in the identical sites of the abnormal calcium accumulation. Furthermore, a autoradiographic study using 14C-vinconate showed that this drug easily penetrates the blood-brain barrier and especially localizes in the striatum and the thalamus after 5 min ischemia. The result suggests that vinconate reduces the areas of abnormal calcium accumulation in the postischemic gerbil brain. This effect seems to be mediated via the height distribution in the brain following ischemia.     

Long-term observations on calcium accumulation in postischemic gerbil brain

We studied delayed postischemic calcium accumulation and neuronal damage in the gerbil brain, using 45Ca autoradiography as a marker for detection of injured tissue and light microscopy. Transient cerebral ischemia was induced for 15 min. Sham-operated gerbils showed no abnormal calcium accumulation and neuronal damage throughout the brain. At 2 and 7 days following 15 min of ischemia, marked calcium accumulation and mild to severe neuronal damage were found in the selectively vulnerable areas such as neocortex, striatum, hippocampus and thalamus, and brainstem such as medial geniculate body, substantia nigra and inferior colliculus. After 1-2 months of recirculation, the calcium accumulation was not recognized in the brainstem. But, the accumulation was still detectable in the striatum, the hippocampus and the thalamus. Morphological study showed that marked proliferation of glia cells was rapid in the inferior colliculus and was relatively slow in the striatum and the hippocampus, although these structures were severely damaged after ischemia. The result suggests that the speed of restoration of injured tissue and the mechanisms for the damage after cerebral ischemia may be different between the selectively vulnerable areas and the brainstem. Furthermore, they suggest that 45Ca autoradiographic technique may provide a useful approach for diagnosis of the restoration of injured tissue at chronic stage following cerebral ischemia.     

Repeated focal cerebral ischemia in gerbils is associated with development of infarction.

We induced repeated focal cerebral ischemia in gerbils. Single 5-min ischemia produced neuronal damage limited to the ipsilateral CA1 and CA4 hippocampus. Two 5-min ischemic insults spaced at a 1-h interval caused selective neuronal damage to the CA1, CA3 and CA4 hippocampus, striatum, neocortex, and thalamus. Three 5-min ischemic insults at 1-h intervals produced infarction. Thus, repeated focal ischemia produced cumulative brain damage by conversion of sublethal damage into selective neuronal damage and of the neuronal damage into infarction.     

Regional changes in [H]inositol 1,4,5-triphosphate binding in the gerbil brain following repeated sublethal ischemia

We investigated the regional changes in [³H]inositol 1,4,5-triphosphate (IP₃) binding in the brain following ischemia using in vitro autoradiography. Three 2-min ischemic insults at 1-hr intervals and a 6-min period of ischemia were induced in gerbils and they were killed after 1, 4, and 28 days. Normal animals had high [³H]IP₃ binding in the CA1 subfield of the hippocampus and the striatum. The binding in the CA1 decreased strikingly after both 6-min ischemia and three 2-min ischemic insults. The [³H]IP₃ binding also decreased in the lateral striatum after three 2-min ischemic insults but not after 6 min of ischemia. Histological observations confirmed neuronal damage to these areas of reduced binding. By contrast, we found a marked increase in [³H]IP₃ binding in the ventral thalamus 28 days after three 2-min ischemic insults. Histological observations with Nissl staining revealed an accumulation of fine granular deposits there. Thus, repeated ischemic insults produced more extensive neuronal damage and changes in [³H]IP₃ binding than a single equivalent period of ischemia. The increased [³H]IP₃ binding in the thalamus coincidentally with an accumulation of Nissl-positive granules at the chronic stage after repeated ischemia is of considerable interest.     

'Ischemic tolerance' phenomenon detected in various brain regions.

We investigated the effects of mild and non-lethal ischemic insult on neuronal death following subsequent lethal ischemic stress in various brain regions, using a gerbil model of bilateral cerebral ischemia. Single 10-min ischemia consistently caused neuronal damage in the hippocampal CA1, CA2, CA3 and CA4, layer III/IV of the cerebral cortex, dorsolateral part of the caudoputamen and ventrolateral part of the thalamus. On the other hand, in double ischemia groups, 2-min ischemic insult 2 days before 10-min ischemia exhibited significant protection in the CA1 and CA3 of the hippocampus, the cerebral cortex, the caudoputamen and the thalamus. Five-min ischemic insult 2 days before 10-min ischemia also showed protective effect in the same areas as those of 2-min ischemia except for the CA1 region of the hippocampus, while 1-min ischemic insult exhibited no protective effect in any brain regions. In the immunoblot analysis, both 2- and 5-min ischemia caused increased synthesis of heat shock protein 72 (HSP 72) in the hippocampus, but 1-min ischemia did not. The present study demonstrated that the 'ischemic tolerance' phenomenon was widely found in the brain and also suggested that ischemic treatment severe enough to cause HSP 72 synthesis might be needed for induction of 'ischemic tolerance'.     

‘Ischemic tolerance’ phenomenon detected in various brain regions

We investigated the effects of mild and non-lethal ischemic insult on neuronal death following subsequent lethal ischemic stress in various brain regions, using a gerbil model of bilateral cerebral ischemia. Single 10-min ischemia consistently caused neuronal damage in the hippocampal CA1, CA2, CA3 and CA4, layer III/IV of the cerebral cortex, dorsolateral part of the caudoputamen and ventrolateral part of the thalamus. On the other hand, in double ischemia groups, 2-min ischemic insult 2 days before 10-min ischemia exhibited significant protection in the CA1 and CA3 of the hippocampus, the cerebral cortex, the caudoputamen and the thalamus. Five-min ischemic insult 2 days before 10-min ischemia also showed protective effect in the same areas as those of 2-min ischemia except for the CA1 region of the hippocampus, while 1-min ischemic insult exhibited no protective effect in any brain regions. In the immunoblot analysis, both 2- and 5-min ischemia caused increased synthesis of heat shock protein 72 (HSP 72) in the hippocampus, but 1-min ischemia did not. The present study demonstrated that the ‘ischemic tolerance’ phenomenon was widely found in the brain and also suggested that ischemic treatment severe enough to cause HSP 72 synthesis might be needed for induction of ‘ischemic tolerance’.     

Regional impairment of protein synthesis following brief cerebral ischemia in the gerbil

Summary Regional cerebral protein synthesis following brief ischemia was investigated in the Mongolian gerbil, utilizing l-[methyl-¹⁴C]methionine autoradiography. Transient ischemia was induced for 1,2 or 3 min. At various recirculation periods up to 48 h, animals received a single dose of l-[methyl-¹⁴C]-methionine and then were terminated 35 min later. Sham-operated animals showed a normal pattern of amino acid incorporation into the proteins of the brain. Following 1-min ischemia, the pattern of protein synthesis was similar to that in the sham-operated gerbils. Ischemia for 2 min, however, caused marked inhibition of protein synthesis in the neocortex, striatum, hippocampal CA1 sector and the thalamus at 1 h of recirculation. Extensive recovery of protein synthesis was found in the neocortex, the striatum, the hippocampal CA1 sector and the thalamus at 5–24 h of recirculation, but, a slight inhibition was detectable in the hippocampal CA1 sector in one of six animals. This inhibition had fully recovered at 48 h of recirculation. Following 3-min ischemia, severe impairment of protein synthesis was found in the neocortex, striatum, the whole hippocampus and the thalamus. After 5–24 h of recirculation, the protein synthesis in these regions had gradually recovered, except that complete lack of amino acid incorporation was seen in the hippocampal CA1 subfield. This impairment of protein synthesis in the hippocampal CA1 sector was not recovered at 48h of recirculation. Morphological study indicated that 2-min ischemia did not produce any significant neuronal damage in the brain, whereas gerbils subjected to 3-min ischemia revealed a mild neuronal damage in the hippocampal CA1 sector. The present study indicates that even non-lethal ischemia can produce a severe inhibition of protein synthesis in the selectively vulnerable regions during the early stage of recirculation.     

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

We examined the response of the gerbil brain to secondary ischemic insult following pretreatment with brief ischemia at intervals of 5 min, 1 and 6 h, 1, 2, 4, 7 and 14 days. Two minutes of bilateral carotid artery occlusion produced no histopathological brain damage, whereas 3 min of occlusion caused a moderate to severe reduction in the number of hippocampal CA1 pyramidal cells. Two-minute occlusion followed by 3-min occlusion at 5-min, 1- and 6-h intervals resulted in almost complete destruction of CA1 neurons. Additional neuronal damage was observed in the striatum at a 1-h interval and in the thalamus and the neocortex at 1- and 6-h intervals. The neuronal damage was most severe at a 1-h interval. Two-minute ischemia followed by 3-min ischemia at intervals of 1, 2, 4 and 7 days, however, caused a marked protective effect, and the hippocampal CA1 neurons were preserved. The protective effect was not observed at a 14-day interval and following pretreatment with 1-min ischemia. Thus, pretreatment with brief ischemia leads to complex responses of the brain to secondary ischemic insult; cumulative damage at intervals of 1-6 h and protective effects at intervals of 1-7 days.     

Postischemic changes of [3H]nimodipine and [3H]ryanodine binding in the gerbil striatum and hippocampus

Sequential alterations of [³H]nimodipine and [³H]ryanodine binding in gerbils were investigated in selectively vulnerable regions, such as the striatum and hippocampus, 1 h to 7 days after 10 min of transient cerebral ischemia. [³H]Nimodipine binding showed no significant changes in the striatum and hippocampus up to 48 h after ischemia. Seven days after ischemia, however, a severe reduction in [³H]nimodipine binding was observed in the dorsolateral striatum, hippocampal CA1 (stratum oriens, stratum pyramidale and stratum radiatum) and hippocampal CA3 sector. On the other hand, [³H]ryanodine binding showed a significant increase in the hippocampus 1 h after ischemia. Five hours after ischemia, a significant reduction in [³H]ryanodine binding was observed only in the hippocampal CA1 sector. Thereafter, the striatum and hippocampus showed no significant alterations in [³H]ryanodine binding up to 48 h after ischemia. After 7 days, a marked reduction in [³H]ryanodine binding was observed in the striatum and hippocampus which were particularly vulnerable to ischemia. These results demonstrate that postischemic alteration in [³H]nimodipine and [³H]ryanodine binding is produced with different processes in the hippocampus. They also suggest that the mechanism for striatal cell damage caused by transient cerebral ischemia may, at least in part, differ from that for hippocampal neuronal damage. Furthermore, our findings suggest that abnormal calcium release from intracellular stores may play a pivotal role in the development of hippocampal neuronal damage.     

Sequential changes in muscarinic acetylcholine, adenosine A1 and calcium antagonist binding sites in the gerbil hippocampus following repeated brief ischemia.

We performed quantitative autoradiography to determine sequential alterations in the binding of muscarinic cholinergic and adenosine A1 receptors and of an L-type calcium channel blocker in the gerbil hippocampus following repeated brief ischemic insults. [3H]Quinuclidinyl benzilate (QNB). [3H]cyclohexyladenosine (CHA) and [3H]PN200-110 were used to label muscarinic and adenosine A1 receptors and L-type calcium channels, respectively. Changes at 1 h, 6 h, 1 day, 4 days and 1 month after three 2-min ischemic insults were compared with changes after single 2- or 6-min ischemia. Two-minute ischemia, which causes no histopathological neuronal damage, produced no persistent alterations in binding sites. We observed a transient and mild increase in binding activities, especially in [3H]CHA binding, at 1 h of recirculation. Following 6-min ischemia and three 2-min ischemic insults. [3H]QNB and [3H]PN200-110 binding decreased by more than 50% in the CA1 subfield by 1 month, but [3H]CHA binding decreased transiently by 20-30% at 4 days when delayed neuronal death of hippocampal CA1 pyramidal cells took place. Reductions in binding, especially in [3H]QNB binding, following three 2-min ischemic insults were greater and appeared earlier than those after 6-min ischemia. Furthermore, alterations extended to the CA3 subfield and the dentate gyrus following repeated insults. Thus, alterations in receptor binding after repeated ischemic insults were greater than those after equivalent single period of ischemia.     

Time profile of calcium accumulation in hippocampus,striatum and frontoparietal cortex after transient forebrain ischemia in the gerbil

Summary The topical and temporal relationship between neuronal injury and calcium loading was investigated in gerbils following bilateral carotid artery occlusion for 5 or 10 min and recirculation times from 15 min to 7 days. The association of histochemically visible calcium deposits with neuronal death was assessed by combining two calcium stains, alizarin red and arsenazo III, with conventional histological techniques. Neuronal calcium accumulation was evaluated morphometrically in the striatum, the frontoparietal cortex and the CA1 and CA4 sectors of the hippocampus. After 5-min ischemia and 1–2 days of recirculation numerous calcium-containing neurons appeared in the CA4 sector but only a few were present in the CA1 sector. After 4 days of recirculation calcium accumulation was visible in the whole CA1 sector and the dorso-lateral part of striate nucleus. After 10-min ischemia calcium accumulation started in these regions, as well as in the cortex, already after 1 day. In the CA1 sector calcium accumulation followed a typical time course: on day 2 only the lateral parts were affected, while on day 4 the whole CA1 neuronal band was calcium positive. The regional distribution of histological lesions matched that of calcium loading and, furthermore, the lesions appeared after a corresponding delay in the respective regions. Morphometric evaluations of calcium staining and histological lesions in the CA1 sector revealed a high correlation, indicating that calcium accumulation and neuronal death are closely associated both topically and temporally. This suggests that disturbances of calcium homeostasis such as those measured by this histochemical technique are the consequence of and not the reason for ischemic cell death.     

During repetitive forebrain ischemia, post-ischemic hypothermia protects neurons from damage.

In rodents damage from repetitive transient cerebral ischemia is more severe than that seen with a single ischemic insult of similar duration. Mild hypothermia has been shown to be very effective in protecting the brain during single ischemic insults. We tested the protective effects of hypothermia in repetitive ischemic insults. We used the gerbil model of repetitive ischemia (three minutes ischemia repeated at one hourly intervals three times) and histological evaluation was done using the silver staining technique. Our study reveals that a decrease in body and scalp temperature by 1-2 degrees Celsius can significantly reduce neuronal damage in the cerebral cortex, CA1 region of the hippocampus and substantia nigra reticulata during repetitive ischemia. As the hypothermia was induced after the initial insult, we believe this offers an opportunity for intervention in the clinical settings.     

Effects of normothermic versus mild hyperthermic forebrain ischemia in rats

We compared the neuropathological consequences of global forebrain ischemia under normothermia versus mild hyperthermia. Twenty-one rats underwent 20 minutes of four-vessel occlusion during which brain temperature was maintained at either 37 degrees C (normothermia, n = 9) or 39 degrees C (hyperthermia, n = 12). Quantitative neuropathological assessment was conducted 1 or 3 days later. At 1 day following the ischemic insult, normothermic rats demonstrated neuronal injury mainly confined to the most dorsolateral striatum. By 3 days, ischemic cells were present throughout the striatum and CA1 hippocampus in normothermic animals. Compared with normothermic rats, intraischemic hyperthermia significantly increased the extent and severity of brain damage at 1 day after the ischemic insult. Areas of severe neuronal necrosis and frank infarction included the cerebral cortex, CA1 hippocampus, striatum, and thalamus. Morphologic damage was also detected in the cerebellum and pars reticulata of the substantia nigra. An overall mortality rate of 83% was demonstrated at 3 days in the hyperthermic ischemic group. We conclude that intraischemic hyperthermia 1) markedly augments ischemic brain damage and mortality compared with normothermia, 2) transforms ischemic cell injury into frank infarction, and 3) accelerates the morphological appearance of ischemic brain injury in regions usually demonstrating delayed neuronal necrosis. These observations on mild hyperthermia may have important implications for patients undergoing cardiac or cerebrovascular surgery as well as patients following cardiac arrest or those with stroke-in-evolution.     

Ischemic tolerance to memory impairment associated with hippocampal neuronal damage after transient cerebral ischemia in rats

When rats were trained preoperatively with a three-panel runway task and were then exposed to 10-min ischemia by the method of 4-vessel occlusion, they showed no increase in the number of errors (attempts to pass through two incorrect panels of the three panel-gates at four choice points), having normal retention of memory performance learned before the ischemic insult. Next, we investigated the abilities of ischemic rats to acquire the three-panel runway task and to learn a subsequent reversal task, where the correct panel-gate locations were changed. Rats with 5-min ischemia exhibited performance as good as that of control rats, but rats exposed to 10- and 20-min ischemia showed more errors than control rats during 10 acquisition sessions and 5 subsequent reversal sessions, each of which (consisting of 6 trials) was given once a day. Marked neuronal degeneration was observed in the hippocampal CA1 sector from the rats with 10- and 20-min ischemia. Exposure to sublethal 5-min ischemia followed by 10-min ischemia at a 2-h interval had no effect on either the memory impairment during acquisition and reversal tests or the hippocampal CA1 damage. When rats were exposed to 5-min ischemia 2 days before lethal 10-min ischemia, they showed acquisition and subsequent reversal learning as good as that of control rats. Preconditioning with sublethal 5-min ischemia followed by 2 days of reperfusion also prevented the neuronal destruction of the hippocampal CA1 sector induced by 10-min ischemia. These findings suggest that postischemic hippocampal CA1 neuronal damage does not affect retention of spatial memory acquired before ischemia, but produces a significant impairment of acquisition and subsequent reversal learning. The present results also demonstrate that preconditioning with sublethal ischemia can develop tolerance to subsequent lethal ischemia to prevent the learning impairment related to the hippocampal CA1 neuronal damage.