- Email: [email protected]
Astroglial and microglial reactions in the gerbil hippocampus with induced ischemic tolerance
BRAIN RESEARCH ELSEVIER
Brain Research 664 (1994) 69,76
Research report
Astroglial and microglial reactions in the gerbil hippocampus with induced ischemic tolerance Hiroyuki Kato
a,*,
Kyuya Kogure b, Tsutomu Araki a, Yasuto Itoyama
a
a Department of Neurology, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980, Japan b Foundation for Brain and Nerve Diseases and the Institute ofNeuropathology, Kumagaya, Japan Accepted 23 August 1994
Abstract
Preconditioning of the brain with sublethal ischemia protects against neuronal damage following subsequent longer periods of ischemia (ischemic tolerance). In this study, we investigated astroglial and microglial reactions in the hippocampus following ischemia in a gerbil model of ischemic tolerance. Two minutes of forebrain ischemia (preconditioning ischemia) or sham operation was followed by 3 min of ischemia (second ischemia) 3 days later. The brains were perfusion-fixed after 4 h, 1 day, 2 days, and 7 days. Paraffin sections were used for the visualization of astrocytes by immunostaining against glia! fibrillary acidic protein (GFAP) and for the visualization of microglia by histochemical staining with isolectin-B4 from Griffonia simplicifolia. The preconditioning ischemia induced a moderate increase in astroglial and microglial staining. Two days after the second ischemia, GFAP staining further increased in astrocytes in the hippocampus with ischemic tolerance. In the CA1 region of the hippocampus without ischemic tolerance, in contrast, microglial activation with increased staining and morphological changes was pronounced. After 7 days, neuronal destruction resulted in the CA1 region without tolerance, where hypertrophic reactive astroglia and reactive microglia with phagocytic transformation accumulated intensely. However, the ischemic preconditioning prevented the CA1 neuronal damage and the activation of microglia was subsiding after 7 days. Thus, activation of glial cells occurred in a graded fashion in response to different degrees of neuronal injury. Astroglial but not microglial activation may have implications in neuronal survival in ischemic tolerance. The findings also suggest that neuron-glial and glia-glial interactions are under strict control and play a critical role in neuronal survival and death after ischemia.
Keywords: Cerebral ischemia; Ischemic tolerance; Astroglia; Microglia; Immunohistochemistry; Neuron-glia interaction; Hippocampus; Gerbil
I. Introduction
Neurons in specific brain regions are selectively susceptible to ischemic insults [2,20,39]. A brief period of cerebral ischemia, as short as 3 min in gerbils, destroys neurons in specific neuronal populations, such as pyramidal neurons in the CA1 subfield of the hippocampus [19]. The neuronal destruction occurs after 3 - 4 days of reperfusion and has been known as delayed neuronal death [20]. However, this selective vulnerability can be modified by preconditioning of the brain with a sublethal period of ischemia [19,21,22]. A
* Corresponding author. Fax: (81) (22) 272-5818. 0006-8993/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0006 -8993 ( 9 4 ) 0 1 0 4 7 - 1
2-min period of forebrain ischemia in gerbils produces no appreciable neuronal damage in the brain. However, preconditioning of the brain with this period of ischemia followed by 1-7 days of reperfusion protects against neuronal damage following a longer period of ischemia which normally damages the CA1 neurons [19]. This phenomenon has been termed ischemic tolerance and has received attention because the elucidation of its mechanism may afford a clue for protection against ischemic brain damage [19,21,22,23]. However, the mechanism of ischemic tolerance remains elusive although the role of altered gene expression in neurons, such as that of stress proteins, has been suggested [21,23]. On the other hand, glial cells in the brain greatly influence the maintenance and survival of neurons
7{I
H. Kato et aL /Brain Research 664 (1994) 69-76
[3,9,12,14,15,36]. Astroglia and microglia are rapidly activated after a wide range of brain injury, including cerebral ischemia [10,11,13,17,26,32]. Earlier studies, especially those in vitro, have shown that these two types of glial cells have opposing actions in regulating neuronal survival [12,14]. These studies have shown that astroglia support neuronal growth and survival. In contrast, microglia, when stimulated, release a variety of cytotoxic agents which may be important mediators of neuronal injury [3,12,14,34,36]. It is very important to clarify how these glial cells influence neuronal survival and death following cerebral ischemia in vivo. The purpose of this study was therefore to investigate the activation of astroglia and microglia in the hippocampus following ischemia in a gerbil model of ischemic tolerance.
2. Materials and methods 2.1. Induction of ischemia We used male Mongolian gerbils, 12-13 weeks old and weighing 60-80 g (Seiwa Experimental Animals, Fukuoka, Japan). Forebrain ischemia and ischemic tolerance were induced as reported previously [19]. Briefly, the gerbils were anesthetized with 2% halothane in a mixture of 30% oxygen and 70% nitrous oxide and both common carotid arteries were gently exposed. One minute after discontinuation of anesthesia, the carotid arteries were occluded with aneurysm clips. This procedure produces a severe reduction of blood flow in the forebrain [18]. A 2-min period of occlusion (ischemic preconditioning) or sham operation was followed by 3 days of reperfusion. Then 3 min of occlusion (the second ischemia) was again induced. Rectal temperatures of the animals were maintained at 37°C during surgery and ischemia with a heating pad and a lamp, and were also monitored after ischemia for 2 h to confirm lhe occurrence of
Fig. 1. Immunostaining against glial fibrillary acidic protein (GFAP) for the visualization of astrocytes (a,b,c) and isolectin staining for the visualization of microglia (d,e,f) in the hippocampus (CA1 subfield and dentate gyrus), a,d: sham-operated. Normal appearance, b,e: 3 days after 2-min ischemia. GFAP-stained astroglia and iectin-stained microglia increased in number, especially in the dentate hilus, c,f: 7 days after 3-min ischemia without ischemic preconditioning. Note accumulation of reactive astroglia and microglia in the CA1 subfield. Bar = 0.2 mm.
H. Kato et al. / Brain Research 664 (1994) 69-76 postischemic hyperthermia [19]. The animals were sacrificed 3 days after the preconditioning ischemia or sham operation, and at 4 h, 1 day, 2 days, and 7 days after the second ischemia. Under anesthesia with pentobarbital (50 mg/kg, i.p.), the brains were perfusion-fixed with 4% paraformaldehyde in 0.1 M phosphate buffer. The brains were removed 2 h later and postfixed in the same fixative overnight at 4°C. The brains were then routinely embedded in paraffin. Sections at a thickness of 5/xm were prepared. Each group contained 5 animals.
71
Laboratories). Lastly, the sections were reacted with diaminobenzidine and H 2 0 2 for color development. The sections were also histochemically stained with peroxidaselabeled isolectin-B4 from Griffonia simplicifolia seeds (Sigma) for the visualization of microglia as described previously [35]. Briefly, the sections were incubated with the isolectin (20 p,g/ml) in phosphate buffered saline with cations overnight at 4°C. Then, the sections were reacted with diaminobenzidine and H20 2 for color development. The sections were counterstained with hematoxylin.
2.2. Visualization of glial cells 3. Results The sections were immunohistochemically stained with an antibody against glial fibrillary acidic protein (GFAP; Chemicon) for the visualization of astrocytes. The paraffin sections were deparaffinized and preincubated with 10% normal serum. The sections were then incubated with the anti-GFAP antibody (1:200) overnight at 4°C. Then the sections were incubated with biotinylated second antibody for 1 h at room temperature, followed by incubation with avidin-biotin-peroxidase complex for 30 min at room temperature according to the supplier's recommendations (Vectastain elite ABC kit, Vector
3.1. Astroglia I n s h a m - o p e r a t e d gerbils, a s t r o g l i a w e r e o n l y w e a k l y i m m u n o s t a i n e d f o r G F A P (Fig. l a , Fig. 2a). T h e y w e r e r e l a t i v e l y a b u n d a n t in s t r a t u m l a c u n o s u m - m o l e c u l a r e o f t h e C A 1 s u b f i e l d . T h e y w e r e also s e e n in s t r a t a radiatum and oriens of the CA1, but were only rarely
Fig. 2. Immunostaining against glial fibrillary acidic protein for the visualization of astrocytes in the CA1 subfield of the hippocampus, a: sham-operated. Normal appearance, b,c: 2 days and 7 days after 3-min ischemia without ischemic preconditioning. Note hypertrophic reactive astrocytes in c. d: 3 days after 2-rain ischemia, e,f: 2 days and 7 days after 3-min ischemia with ischemic preconditioning. Increased astroglial staining in d, e and f. Bar = 0.1 mm.
72
H. Kato et al. / Brain Research 004 (1994) 09-70
seen in other regions. G F A P immunostaining was seen in the cytoplasm and thin processes of astrocytes (Fig. 2a). Three days after 2 min of preconditioning ischemia, G F A P immunostaining increased in astrocytes in the entire hippocampus as compared with those in shamoperated gerbils (Fig. lb). The increase was most pronounced in the dentate hilus, and was moderate in stratum radiatum of the CA1 (Fig. 2d). Astroglia in the dentate hilus were slightly hypertrophic (Fig. 4a). Four hours and 1 day after 3-min ischemia without ischemic preconditioning, G F A P staining appeared similar to that of sham-operation. In animals with preconditioning, the G F A P staining was also appeared similar to that 3 days after 2 min of ischemia. Thus, GFAP-positive astroglia were seen in the entire hippocampus, especially in the dentate hilus and the CA1.
After 2 days, G F A P staining largely unchanged in animals without preconditioning (Fig. 2b). However, G F A P staining increased in the CA1 region of the animals with preconditioning (Fig. 2e) and the astroglia in the CA1 became slightly hypertrophic. After 7 days, CA1 pyramidal neurons were destroyed in animals without preconditioning. Astroglia in the CA1 region and the dentate hilus became more immunoreactive to G F A P and became hypertrophic, showing a typical morphology of reactive astrocytes (Fig. lc). The accumulation of reactive astrocytes in the CA1 was seen predominantly in strata oriens, radiaturn, and lacunosum-moleculare, but was sparse in stratum pyramidale (Fig. 2c). G F A P staining also slightly increased in the CA3 and the dentate gyrus. In the hippocampus of animals with preconditioning where CA1 neurons survived, astroglia with increased G F A P
Fig. 3. Histochemical staining with isolectin-B4 from Griffonia simplicifolia for the visualization of microglia in the CAI subfield of the hippocampus, a: sham-operated. Quiescent microglia are only rarely seen (arrow). b,c: 2 days and 7 days after 3-min ischemia without ischemic preconditioning. Note an intense activation and accumulation of reactive microglia in b and c. d: 3 days after 2-min ischemia, e,f: 2 days and 7 days after 3-min ischemia with ischemic preconditioning. Microglia are moderately activated in d and e but are subsiding in f. Bar = 0.l mm.
H. Kato et al. / Brain Research 664 (1994) 69-76
73
Four hours after 3-min ischemia in animals without preconditioning, some activated microglia were already
seen in the hipp0campus, especially in the CA1 sub-
T ~
~.~
Fig. 4. Immunostaining against glial fibrillary acidic protein for the visualization of astrocytes(a) and isolectin staining for the visualization of microglia(b) in the dentate hilus 3 days after 2-min ischemia. Reactive astroglia (a) and microglia(b) are seen in the dentate hilus. Bar = 0.1 mm.
staining were seen but they were not so hypertrophic (Fig. 2D. G F A P staining in the CA3 and the dentate gyrus decreased but the reactive astrocytes in the dentate hilus were still hypertrophic. 3.2. Microglia
In sham-operated animals, microglia were only rarely and weakly stained with the isolectin in the hippocampus (Fig. ld, Fig. 3a). When stained, the microglia had the morphology of quiescent microglia, i.e. small cell body and ramified thin processes (Fig. 3a). As reported earlier, microglia stained intensely with this concentration of isolectin could be considered as activated microglia [24]. Three days after 2 rain of preconditioning ischemia, microglial cells increased throughout the hippocampus, particularly in the dentate hilus and the CA1 subfield (Fig. le, Fig. 3d and Fig. 4b). The microglia had the morphology of activated microglia with increased isolectin staining and ramified, thorny processes.
field. In animals with preconditioning, isolectin staining was largely similar to that before the second ischemia. After 1 day, activated microglia were scattered in the hippocampus, especially in the CA1, in animals both with and without preconditioning. After 2 days, isolectin staining appeared unchanged in animals with preconditioning (Fig. 3e). However, activated microglia increased strikingly in the CA1 of animals without preconditioning and the microglia became larger and stained stronger (Fig. 3b). After 7 days, an intense accumulation of reactive microglia was seen in the CA1 of animals without preconditioning, especially in stratum pyramidale, where reactive microglia had the morphology of ameboid-like appearance (Fig. If and Fig. 3c). In stratum radiatum of the CA1, many microglia had the morphology of rod cells. In the CA3, dentate gyrus, and the dentate hilus, the number of microglia decreased. The hippocampus of the animals with preconditioning returned to near control levels and only scattered microglia with less activated morphology were seen in the CA1 (Fig. 3f).
4. Discussion
The present study showed that both astroglia and microglia were activated after ischemia in a graded fashion in response to different degrees of neuronal injury. The first step of glial activation, which occurred without neuronal destruction, was an increased G F A P staining in astrocytes and an increased isolectin staining in microglia. This type of glial activation was generally seen after ischemia including ischemia-resistant regions such as the CA3 and the dentate gyrus although stronger in the vulnerable CA1 region. Such glial activation was also seen in the hippocampus that acquired ischemic tolerance. This early activation of glial cells was more rapid and striking in microglia and was subsiding after 7 days when ischemia was sublethal. In contrast, astroglial activation occurred relatively slowly but lasted longer even when ischemia was sublethal. Conspicuous differences in glial activation between animals with and without ischemic tolerance was observed 2 days after the second ischemia. In the hippocampus without ischemic tolerance, microglial activation with increases in isolectin staining and cell size was seen in the CA1 subfield although neuronal destruction was not yet evident. However, astroglial reactions were not obvious. In the CA1 subfield of the hippocampus with ischemic tolerance, in contrast, as-
74
H. Kato et al. / Brain Research 004 (1994) 09-70
troglial activation was evident but microglial activation was milder than that in animals without tolerance. These observations suggest that the balance between astroglial and microglial reactions may be related to the final outcome. The astroglial and microglial reactions became far intense when neuronal destruction occurred as seen in the CA1 of animals without tolerance and in the dentate hilus both with and without preconditioning. In the animal models used in this study, ischemia-induced neuronal damage is produced in these two regions. Even a 2-min period of preconditioning ischemia destroys a subpopulation of dentate hilar neurons [25]. Reactive astrocytes intensely immunoreactive to GFAP became hypertrophic. Many reactive microglia had the ameboid-like morphology, suggesting an increase in phagocytic activities. Thus, full-blown activation of glial cells was seen when neurons were destroyed. The function of astrocytes is to maintain local microenvironment under physiological conditions [1,6,29, 38]. The first is a rapid reuptake and inactivation of released neurotransmitters, such as excitatory amino acid glutamate, from synaptic spaces. Secondly, astrocytes contribute to form blood-brain barrier (BBB), and thirdly they maintain water and ionic homeostasis. It has been generally accepted that a massive release of glutamate during ischemia triggers the chain reactions leading to ischemia-induced neuronal injury by excitotoxic mechanisms [4,8]. Ischemia also induces water and ionic imbalance and this derangement becomes severe when BBB is broken down [16]. Therefore, if astroglial activation implies enhanced astroglial functions, reactive astroglia in the hippocampus with induced tolerance may have an increased capacity of glutamate reuptake and maintenance of water and ionic homeostasis. In fact, glutamine synthetase, which inactivates glutamate into glutamine and is present exclusively in astrocytes, increases after cerebral ischemia [30]. Therefore, ischemic insult rendered in the presence of astroglial activation may facilitate recovery from ischemia-induced environmental derangements when blood flow is restored. Thus, an increased capacity of astroglial function may be one of the mechanisms of ischemic tolerance. Furthermore, reactive astrocytes upregulate the expression of a large number of molecules that may benefit the injured neurons [9]. In vitro studies have shown that astrocytes actually produce a number of agents that support neuronal growth and survival [12,14]. In addition, astroglia-derived growth factors attenuate the toxic effects of microglia and may help to preserve neurons under attack by phagocytic cells [12,14]. Nerve growth factor and basic fibroblast growth factor, which are produced by astrocytes, have been shown to prevent ischemia-induced neuronal damage in vivo when administered exogenously before and
after the onset of ischemia [28,33]. Thus. it is assumed that the neuronal injury following the second ischemia is also reduced by the substances produced by the activated astrocytes in the hippocampus with ischemic tolerance. Growth factors released from astrocytes may also function at a later phase for wound repair. In vitro studies have shown that microglial cells release toxic agents [12,14,34]. When stimulated, microglia release a variety of cytotoxic agents which may be important mediators of neuronal injury, such as certain kinds of cytokines, reactive oxygen radicals, proteases, and glutamate. In this study, a strong activation of microglia was seen in the CA I subfield 2 days after 3 min of ischemia without preconditioning before the CA1 neurons were destroyed. These activated cells may have a neuron-killing effect. The fate of ischemic neurons may therefore depend on the reactive astroglia and microglia as they compete to govern the survival of neurons. However, considerable microglial activation, though not full-blown, was seen when ischemic insult was sublethal. Microglial ceils were frequently seen in close proximity to viable neurons. There was a rapid activation of microglia next to neurons that do not die. Of particular interest in this regard is the phenomenon termed synaptic stripping. Blinzinger and Kreutzberg [5] reported that activated microglia after facial nerve lesioning displace afferent terminals from the facial nerve cell bodies causing deafferentation. Whether similar deafferentation occurs in the hippocampus following ischemia is unknown and electron microscopic studies are necessary. However, if microglia cause similar deafferentation after ischemic preconditioning, it may protect the CA1 neurons from ischemia because excitotoxicity caused by exposure to glutamate is blocked. Actually, deafferentation has been reported to protect CA1 neurons from ischemic injury [7,40]. Just the presence of microglia therefore does not mean neuronal death and possibly is neuroprotective. There are various levels of microglial activation and this graded activation may explain these conflicting observations. When neurons were destroyed after ischemia, far intense activation of astrocytes and microglia was seen. They also changed morphology, i.e. hypertrophy of astrocytes and macrophage-like transformation of microglia. Elimination of damaged circuits would prevent continuing disruption and, perhaps, dysfunction of the entire neural system. Controlled cell death is a well-recognized process during wound healing and tissue renewal and arises when inflammatory cells release cytotoxins. Microglia may protect healthy neurons by removal of diseased cells because necrotic tissue elicits aggressive phagocytic microglia. In conclusion, the findings of this study suggest that activation of astroglia and microglia following ischemia
H. Kato et al. / Brain Research 664 (1994) 69-76
is under strict control, and that they are activated in a graded fashion in response to different degrees of neuronal injury. The graded response of microglia have been shown by the differential expression of immunomolecules and receptors on microglia [11,26], and therefore this kind of investigation may further clarify the differential activation of microglia in ischemic tolerance. It is quite possible that neuron-glia interactions, especially astroglial activation, have implications in promoting neuronal survival in ischemic tolerance. The regulation of glial effects upon neurons is further complicated by microglia-astroglial interactions because fully activated microglia may have neuron-killing effects while moderately activated microglia may not. Thus, the balance between neuron-glial and glia-glial interactions may play a critical role in neuronal survival and death after ischemia and in the manifestation of ischemic tolerance.
Acknowledgements This study was supported in part by Grant-in-Aid for Scientific Research 03404028 and Grant-in-Aid for Scientific Research on priority areas from the Ministry of Education, Science, and Culture of Japan, and by Kanae Fund of Research for New Medicine.
References [1] Abbott, N.J., Revest, P.A. and Romero, I.A., Astrocyte-endothelial interaction: physiology and pathology, Neuropathol. Appl. Neurobiol., 18 (1992) 424-433. [2] Araki, T., Kato, H. and Kogure, K., Selective neuronal vulnerability following transient cerebral ischemia in the gerbil: distribution and time course, Acta Neurol. Scand., 80 (1989) 548-553. [3] Banati, R.B., Gehrmann, J., Schubert, P. and Kreutzberg, G.W., Cytotoxicity of microglia, Glia, 7 (1993) 111-118. [4] Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H., Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis, J. Neurochem., 43 (1984) 1369-1374. [5] Blinzinger, K. and Kreutzberg, G.W., Displacement of synaptic terminals from regenerating motoneurons by microglial cells, Z. Zellforsch., 85 (1968) 145-157. [6] Brightman, M., Implication of astroglia in the blood-brain barrier, Ann. NYAcad. Sci., 633 (1991) 343-347. [7] Buchan, A.M. and Pulsinelli, W.A., Septo-hippocampal deafferentation protects CA1 neurons against ischemic injury, Brain Res., 512 (1990) 7-14. [8] Choi, D.W., Cerebral hypoxia: some new approaches and unanswered questions, J. Neurosci., 10 (1990) 2493-2501. [9] Eddleston, E. and Mucke, L., Molecular profile of reactive astrocytes. Implications for their role in neurologic disease, Neuroscience, 54 (1993) 15-36. [10] Finsen, B.R., Jorgensen M.B., Diemer N.H. and Jimmer, J., Microglial MHC antigen expression after ischemic and kainic acid lesions of the adult rat hippocampus, Glia, 7 (1993) 41-49. [11] Gehrmann, J., Bonnekoh, P., Miyazawa, T., Hossmann, K.-A.
75
and Kreutzberg, G.W., Immunocytochemical study of an early microglial activation in ischemia, J. Cereb. Blood Flow Metab., 12 (1992) 257-269. [12] Giulian, D., Reactive glia as rivals in regulating neuronal survival, Glia, 7 (1993) 102-110. [13] Giulian, D. and Vaca, K., Inflammatory glia mediate delayed neuronal damage after ischemia in the central nervous system, Stroke, 24 (Suppl. I) (1993) 1-84-1-90. [14] Giulian, D., Vaca, K. and Corpuz, M., Brain glia release factors with opposing actions upon neuronal survival, J. Neurosci., 13 (1993) 29-37. [15] Hatten, M.E., Liem, R.K.H., Shelanski, M.L. and Mason, C.A., Astroglia in CNS injury, Glia, 4 (1991) 233-243. [16] Hossmann, K.-A., Development and resolution of ischemic brain swelling. In Papius, H.M. and Feindel, W. (Eds.), Dynamics of Brain Edema, Springer, Berlin, 1976, pp. 219-227. [17] J0rgensen, M.B., Finsen, B.R., Jensen, M.B., Castellano, B., Diemer, N.H. and Zimmer, J., Microglial and astroglial reactions to ischemic and kainic acid-induced lesions of the adult rat hippocampus, Exp. Neurol., 120 (1993) 70-88. [18] Kato, H., Araki, T., Kogure, K., Murakami, M. and Uemura, K., Sequential cerebral blood flow changes in short-term cerebral ischemia in the gerbil, Stroke, 21 (1990) 1346-1349. [19] Kato, H., Liu, Y., Araki, T. and Kogure K., 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 (1991) 238-242. [20] Kirino, T., Delayed neuronal death in the gerbil hippocampus following ischemia, Brain Res., 239 (1982) 57-69. [21] Kirino, T., Tsujita, Y. and Tamura, A., Induced tolerance to ischemia in gerbil hippocampal neurons. J. Cereb. Blood Flow Metab. 11 (1991) 299-307. [22] Kitagawa, K., Matsumoto, M., Tagaya, M., Hata, R., Ueda, H., Niinobe, M., Handa, N., Fukunaga, R., Kimura, K,. Mikoshiba, K. and Kamata, T., 'Ischemic tolerance' phenomenon found in the brain, Brain Res., 528 (1990) 21-24. [23] Liu, Y., Kato, H., Nakata, N. and Kogure, K., Temporal profile of heat shock protein 70 synthesis in ischemic tolerance induced by preconditioning ischemia in rat hippocampus, Neuroscience, 56 (1993) 921-927. [24] Marty, S., Dusart, I. and Peschanski, M., Glial changes following an excitotoxic lesion in the CNS. I. Microglia/macrophages, Neuroscience, 45 (1991) 529-539. [25] Matsuyama, T., Tsuchiyama, M., Nakamura, H., Matsumoto, M. and Sugita M., Hilar somatostatin neurons are more vulnerable to an ischemic insult than CA1 pyramidal neurons, J. Cereb. Blood Flow Metab., 13 (1993) 229-234. [26] Morioka, T., Kalehua, A.N. and Streit, W.J., The microglial reaction in the rat dorsal hippocampus following transient forebrain ischemia, J. Cereb. Blood Flow Metab., 11 (1991)966-973. [27] Morioka, T., Kalehua, A.N. and Streit, W.J., Progressive expression of immunomolecules on microglial cells in rat dorsal hippocampus following transient forebrain ischemia, Acta Neuropathol., 83 (1992) 149-157. [28] Nakata, N., Kato, H. and Kogure, K., Protective effects of basic fibroblast growth factor against hippocampal neuronal damage following cerebral ischemia in the gerbil, Brain Res., 605 (1993) 458-464. [29] Nicholls, D. and Attwell, D., The release and uptake of excitatory amino acids, Trends Pharmacol. Sci., 11 (1990) 462-468. [30] Petito, C.K., Chung, M.C., Verkhovsky, L.M. and Coopwe, A.J.L., Brain glutamine synthetase increases following cerebral ischemia in the rat, Brain Res., 569 (1992) 275-280. [31] Rothman, S.M. and Olney, J.W., Glutamate and the pathophysiology of hypoxic-ischemic brain damage, Ann. Neurol., 19 (1986) 105-111.
76
H. Kato et aL /Brain Research 604 (1994) 09-7(3
[32] Schmidt-Kastner, R., Szymas, J. and Hossmann, K.-A., lmmunohistochemical study of glial reaction and serum-protein extravasation in relation to neuronal damage in rat hippocampus after ischemia, Neuroscience, 38 (1990) 527-540. [33] Shigeno, T., Mima, T., Takakura, K., Graham, D.T., Kato, G., Hashimoto, Y. and Furukawa, S., Amelioration of delayed neuronal death in the hippocampus by nerve growth factor, J. Neurosci., 11 (1991) 2914-2919. [34] Streit, W.J., Graeber, M.B. and Kreutzberg, G.W., Functional plasticity of microglia: a review, Glia, 1 (1988) 301-307. [35] Streit, W.J., An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4), J. Histochem. Cytochem., 38 (1990) 1683-1686.
[36] Streit, W.J., Microglial-neuronal interactions, J. Chem. Neuroanat., 6 (1993) 261-266. [37] Vibulsreth, S., Hefti, F., Ginsberg, M.D., Dietrich, W.D. and Busto, R., Astrocytes protect cultured neurons from degeneration induced by anoxia, Brain Res., 422 (1987) 303-311. [38] Walz, W., Role of glial cells in the regulation of the brain ion microenvironment, Prog. Neurobiol., 33 (1989) 309-333. [39] Wieloch, T., Neurochemical correlates to selective neuronal vulnerability, Prog. Brain Res., 63 (1985) 69-85. [40] Wieloch, T., Lindvall, O., Blomqvist, P. and Gage, F.H., Evidence for amelioration of ischaemic neuronal damage in the hippocampal formation by lesions of the perforant path, Neurot Res., 7 (1985) 24-26.
Brain Research 664 (1994) 69,76
Research report
Astroglial and microglial reactions in the gerbil hippocampus with induced ischemic tolerance Hiroyuki Kato
a,*,
Kyuya Kogure b, Tsutomu Araki a, Yasuto Itoyama
a
a Department of Neurology, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980, Japan b Foundation for Brain and Nerve Diseases and the Institute ofNeuropathology, Kumagaya, Japan Accepted 23 August 1994
Abstract
Preconditioning of the brain with sublethal ischemia protects against neuronal damage following subsequent longer periods of ischemia (ischemic tolerance). In this study, we investigated astroglial and microglial reactions in the hippocampus following ischemia in a gerbil model of ischemic tolerance. Two minutes of forebrain ischemia (preconditioning ischemia) or sham operation was followed by 3 min of ischemia (second ischemia) 3 days later. The brains were perfusion-fixed after 4 h, 1 day, 2 days, and 7 days. Paraffin sections were used for the visualization of astrocytes by immunostaining against glia! fibrillary acidic protein (GFAP) and for the visualization of microglia by histochemical staining with isolectin-B4 from Griffonia simplicifolia. The preconditioning ischemia induced a moderate increase in astroglial and microglial staining. Two days after the second ischemia, GFAP staining further increased in astrocytes in the hippocampus with ischemic tolerance. In the CA1 region of the hippocampus without ischemic tolerance, in contrast, microglial activation with increased staining and morphological changes was pronounced. After 7 days, neuronal destruction resulted in the CA1 region without tolerance, where hypertrophic reactive astroglia and reactive microglia with phagocytic transformation accumulated intensely. However, the ischemic preconditioning prevented the CA1 neuronal damage and the activation of microglia was subsiding after 7 days. Thus, activation of glial cells occurred in a graded fashion in response to different degrees of neuronal injury. Astroglial but not microglial activation may have implications in neuronal survival in ischemic tolerance. The findings also suggest that neuron-glial and glia-glial interactions are under strict control and play a critical role in neuronal survival and death after ischemia.
Keywords: Cerebral ischemia; Ischemic tolerance; Astroglia; Microglia; Immunohistochemistry; Neuron-glia interaction; Hippocampus; Gerbil
I. Introduction
Neurons in specific brain regions are selectively susceptible to ischemic insults [2,20,39]. A brief period of cerebral ischemia, as short as 3 min in gerbils, destroys neurons in specific neuronal populations, such as pyramidal neurons in the CA1 subfield of the hippocampus [19]. The neuronal destruction occurs after 3 - 4 days of reperfusion and has been known as delayed neuronal death [20]. However, this selective vulnerability can be modified by preconditioning of the brain with a sublethal period of ischemia [19,21,22]. A
* Corresponding author. Fax: (81) (22) 272-5818. 0006-8993/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0006 -8993 ( 9 4 ) 0 1 0 4 7 - 1
2-min period of forebrain ischemia in gerbils produces no appreciable neuronal damage in the brain. However, preconditioning of the brain with this period of ischemia followed by 1-7 days of reperfusion protects against neuronal damage following a longer period of ischemia which normally damages the CA1 neurons [19]. This phenomenon has been termed ischemic tolerance and has received attention because the elucidation of its mechanism may afford a clue for protection against ischemic brain damage [19,21,22,23]. However, the mechanism of ischemic tolerance remains elusive although the role of altered gene expression in neurons, such as that of stress proteins, has been suggested [21,23]. On the other hand, glial cells in the brain greatly influence the maintenance and survival of neurons
7{I
H. Kato et aL /Brain Research 664 (1994) 69-76
[3,9,12,14,15,36]. Astroglia and microglia are rapidly activated after a wide range of brain injury, including cerebral ischemia [10,11,13,17,26,32]. Earlier studies, especially those in vitro, have shown that these two types of glial cells have opposing actions in regulating neuronal survival [12,14]. These studies have shown that astroglia support neuronal growth and survival. In contrast, microglia, when stimulated, release a variety of cytotoxic agents which may be important mediators of neuronal injury [3,12,14,34,36]. It is very important to clarify how these glial cells influence neuronal survival and death following cerebral ischemia in vivo. The purpose of this study was therefore to investigate the activation of astroglia and microglia in the hippocampus following ischemia in a gerbil model of ischemic tolerance.
2. Materials and methods 2.1. Induction of ischemia We used male Mongolian gerbils, 12-13 weeks old and weighing 60-80 g (Seiwa Experimental Animals, Fukuoka, Japan). Forebrain ischemia and ischemic tolerance were induced as reported previously [19]. Briefly, the gerbils were anesthetized with 2% halothane in a mixture of 30% oxygen and 70% nitrous oxide and both common carotid arteries were gently exposed. One minute after discontinuation of anesthesia, the carotid arteries were occluded with aneurysm clips. This procedure produces a severe reduction of blood flow in the forebrain [18]. A 2-min period of occlusion (ischemic preconditioning) or sham operation was followed by 3 days of reperfusion. Then 3 min of occlusion (the second ischemia) was again induced. Rectal temperatures of the animals were maintained at 37°C during surgery and ischemia with a heating pad and a lamp, and were also monitored after ischemia for 2 h to confirm lhe occurrence of
Fig. 1. Immunostaining against glial fibrillary acidic protein (GFAP) for the visualization of astrocytes (a,b,c) and isolectin staining for the visualization of microglia (d,e,f) in the hippocampus (CA1 subfield and dentate gyrus), a,d: sham-operated. Normal appearance, b,e: 3 days after 2-min ischemia. GFAP-stained astroglia and iectin-stained microglia increased in number, especially in the dentate hilus, c,f: 7 days after 3-min ischemia without ischemic preconditioning. Note accumulation of reactive astroglia and microglia in the CA1 subfield. Bar = 0.2 mm.
H. Kato et al. / Brain Research 664 (1994) 69-76 postischemic hyperthermia [19]. The animals were sacrificed 3 days after the preconditioning ischemia or sham operation, and at 4 h, 1 day, 2 days, and 7 days after the second ischemia. Under anesthesia with pentobarbital (50 mg/kg, i.p.), the brains were perfusion-fixed with 4% paraformaldehyde in 0.1 M phosphate buffer. The brains were removed 2 h later and postfixed in the same fixative overnight at 4°C. The brains were then routinely embedded in paraffin. Sections at a thickness of 5/xm were prepared. Each group contained 5 animals.
71
Laboratories). Lastly, the sections were reacted with diaminobenzidine and H 2 0 2 for color development. The sections were also histochemically stained with peroxidaselabeled isolectin-B4 from Griffonia simplicifolia seeds (Sigma) for the visualization of microglia as described previously [35]. Briefly, the sections were incubated with the isolectin (20 p,g/ml) in phosphate buffered saline with cations overnight at 4°C. Then, the sections were reacted with diaminobenzidine and H20 2 for color development. The sections were counterstained with hematoxylin.
2.2. Visualization of glial cells 3. Results The sections were immunohistochemically stained with an antibody against glial fibrillary acidic protein (GFAP; Chemicon) for the visualization of astrocytes. The paraffin sections were deparaffinized and preincubated with 10% normal serum. The sections were then incubated with the anti-GFAP antibody (1:200) overnight at 4°C. Then the sections were incubated with biotinylated second antibody for 1 h at room temperature, followed by incubation with avidin-biotin-peroxidase complex for 30 min at room temperature according to the supplier's recommendations (Vectastain elite ABC kit, Vector
3.1. Astroglia I n s h a m - o p e r a t e d gerbils, a s t r o g l i a w e r e o n l y w e a k l y i m m u n o s t a i n e d f o r G F A P (Fig. l a , Fig. 2a). T h e y w e r e r e l a t i v e l y a b u n d a n t in s t r a t u m l a c u n o s u m - m o l e c u l a r e o f t h e C A 1 s u b f i e l d . T h e y w e r e also s e e n in s t r a t a radiatum and oriens of the CA1, but were only rarely
Fig. 2. Immunostaining against glial fibrillary acidic protein for the visualization of astrocytes in the CA1 subfield of the hippocampus, a: sham-operated. Normal appearance, b,c: 2 days and 7 days after 3-min ischemia without ischemic preconditioning. Note hypertrophic reactive astrocytes in c. d: 3 days after 2-rain ischemia, e,f: 2 days and 7 days after 3-min ischemia with ischemic preconditioning. Increased astroglial staining in d, e and f. Bar = 0.1 mm.
72
H. Kato et al. / Brain Research 004 (1994) 09-70
seen in other regions. G F A P immunostaining was seen in the cytoplasm and thin processes of astrocytes (Fig. 2a). Three days after 2 min of preconditioning ischemia, G F A P immunostaining increased in astrocytes in the entire hippocampus as compared with those in shamoperated gerbils (Fig. lb). The increase was most pronounced in the dentate hilus, and was moderate in stratum radiatum of the CA1 (Fig. 2d). Astroglia in the dentate hilus were slightly hypertrophic (Fig. 4a). Four hours and 1 day after 3-min ischemia without ischemic preconditioning, G F A P staining appeared similar to that of sham-operation. In animals with preconditioning, the G F A P staining was also appeared similar to that 3 days after 2 min of ischemia. Thus, GFAP-positive astroglia were seen in the entire hippocampus, especially in the dentate hilus and the CA1.
After 2 days, G F A P staining largely unchanged in animals without preconditioning (Fig. 2b). However, G F A P staining increased in the CA1 region of the animals with preconditioning (Fig. 2e) and the astroglia in the CA1 became slightly hypertrophic. After 7 days, CA1 pyramidal neurons were destroyed in animals without preconditioning. Astroglia in the CA1 region and the dentate hilus became more immunoreactive to G F A P and became hypertrophic, showing a typical morphology of reactive astrocytes (Fig. lc). The accumulation of reactive astrocytes in the CA1 was seen predominantly in strata oriens, radiaturn, and lacunosum-moleculare, but was sparse in stratum pyramidale (Fig. 2c). G F A P staining also slightly increased in the CA3 and the dentate gyrus. In the hippocampus of animals with preconditioning where CA1 neurons survived, astroglia with increased G F A P
Fig. 3. Histochemical staining with isolectin-B4 from Griffonia simplicifolia for the visualization of microglia in the CAI subfield of the hippocampus, a: sham-operated. Quiescent microglia are only rarely seen (arrow). b,c: 2 days and 7 days after 3-min ischemia without ischemic preconditioning. Note an intense activation and accumulation of reactive microglia in b and c. d: 3 days after 2-min ischemia, e,f: 2 days and 7 days after 3-min ischemia with ischemic preconditioning. Microglia are moderately activated in d and e but are subsiding in f. Bar = 0.l mm.
H. Kato et al. / Brain Research 664 (1994) 69-76
73
Four hours after 3-min ischemia in animals without preconditioning, some activated microglia were already
seen in the hipp0campus, especially in the CA1 sub-
T ~
~.~
Fig. 4. Immunostaining against glial fibrillary acidic protein for the visualization of astrocytes(a) and isolectin staining for the visualization of microglia(b) in the dentate hilus 3 days after 2-min ischemia. Reactive astroglia (a) and microglia(b) are seen in the dentate hilus. Bar = 0.1 mm.
staining were seen but they were not so hypertrophic (Fig. 2D. G F A P staining in the CA3 and the dentate gyrus decreased but the reactive astrocytes in the dentate hilus were still hypertrophic. 3.2. Microglia
In sham-operated animals, microglia were only rarely and weakly stained with the isolectin in the hippocampus (Fig. ld, Fig. 3a). When stained, the microglia had the morphology of quiescent microglia, i.e. small cell body and ramified thin processes (Fig. 3a). As reported earlier, microglia stained intensely with this concentration of isolectin could be considered as activated microglia [24]. Three days after 2 rain of preconditioning ischemia, microglial cells increased throughout the hippocampus, particularly in the dentate hilus and the CA1 subfield (Fig. le, Fig. 3d and Fig. 4b). The microglia had the morphology of activated microglia with increased isolectin staining and ramified, thorny processes.
field. In animals with preconditioning, isolectin staining was largely similar to that before the second ischemia. After 1 day, activated microglia were scattered in the hippocampus, especially in the CA1, in animals both with and without preconditioning. After 2 days, isolectin staining appeared unchanged in animals with preconditioning (Fig. 3e). However, activated microglia increased strikingly in the CA1 of animals without preconditioning and the microglia became larger and stained stronger (Fig. 3b). After 7 days, an intense accumulation of reactive microglia was seen in the CA1 of animals without preconditioning, especially in stratum pyramidale, where reactive microglia had the morphology of ameboid-like appearance (Fig. If and Fig. 3c). In stratum radiatum of the CA1, many microglia had the morphology of rod cells. In the CA3, dentate gyrus, and the dentate hilus, the number of microglia decreased. The hippocampus of the animals with preconditioning returned to near control levels and only scattered microglia with less activated morphology were seen in the CA1 (Fig. 3f).
4. Discussion
The present study showed that both astroglia and microglia were activated after ischemia in a graded fashion in response to different degrees of neuronal injury. The first step of glial activation, which occurred without neuronal destruction, was an increased G F A P staining in astrocytes and an increased isolectin staining in microglia. This type of glial activation was generally seen after ischemia including ischemia-resistant regions such as the CA3 and the dentate gyrus although stronger in the vulnerable CA1 region. Such glial activation was also seen in the hippocampus that acquired ischemic tolerance. This early activation of glial cells was more rapid and striking in microglia and was subsiding after 7 days when ischemia was sublethal. In contrast, astroglial activation occurred relatively slowly but lasted longer even when ischemia was sublethal. Conspicuous differences in glial activation between animals with and without ischemic tolerance was observed 2 days after the second ischemia. In the hippocampus without ischemic tolerance, microglial activation with increases in isolectin staining and cell size was seen in the CA1 subfield although neuronal destruction was not yet evident. However, astroglial reactions were not obvious. In the CA1 subfield of the hippocampus with ischemic tolerance, in contrast, as-
74
H. Kato et al. / Brain Research 004 (1994) 09-70
troglial activation was evident but microglial activation was milder than that in animals without tolerance. These observations suggest that the balance between astroglial and microglial reactions may be related to the final outcome. The astroglial and microglial reactions became far intense when neuronal destruction occurred as seen in the CA1 of animals without tolerance and in the dentate hilus both with and without preconditioning. In the animal models used in this study, ischemia-induced neuronal damage is produced in these two regions. Even a 2-min period of preconditioning ischemia destroys a subpopulation of dentate hilar neurons [25]. Reactive astrocytes intensely immunoreactive to GFAP became hypertrophic. Many reactive microglia had the ameboid-like morphology, suggesting an increase in phagocytic activities. Thus, full-blown activation of glial cells was seen when neurons were destroyed. The function of astrocytes is to maintain local microenvironment under physiological conditions [1,6,29, 38]. The first is a rapid reuptake and inactivation of released neurotransmitters, such as excitatory amino acid glutamate, from synaptic spaces. Secondly, astrocytes contribute to form blood-brain barrier (BBB), and thirdly they maintain water and ionic homeostasis. It has been generally accepted that a massive release of glutamate during ischemia triggers the chain reactions leading to ischemia-induced neuronal injury by excitotoxic mechanisms [4,8]. Ischemia also induces water and ionic imbalance and this derangement becomes severe when BBB is broken down [16]. Therefore, if astroglial activation implies enhanced astroglial functions, reactive astroglia in the hippocampus with induced tolerance may have an increased capacity of glutamate reuptake and maintenance of water and ionic homeostasis. In fact, glutamine synthetase, which inactivates glutamate into glutamine and is present exclusively in astrocytes, increases after cerebral ischemia [30]. Therefore, ischemic insult rendered in the presence of astroglial activation may facilitate recovery from ischemia-induced environmental derangements when blood flow is restored. Thus, an increased capacity of astroglial function may be one of the mechanisms of ischemic tolerance. Furthermore, reactive astrocytes upregulate the expression of a large number of molecules that may benefit the injured neurons [9]. In vitro studies have shown that astrocytes actually produce a number of agents that support neuronal growth and survival [12,14]. In addition, astroglia-derived growth factors attenuate the toxic effects of microglia and may help to preserve neurons under attack by phagocytic cells [12,14]. Nerve growth factor and basic fibroblast growth factor, which are produced by astrocytes, have been shown to prevent ischemia-induced neuronal damage in vivo when administered exogenously before and
after the onset of ischemia [28,33]. Thus. it is assumed that the neuronal injury following the second ischemia is also reduced by the substances produced by the activated astrocytes in the hippocampus with ischemic tolerance. Growth factors released from astrocytes may also function at a later phase for wound repair. In vitro studies have shown that microglial cells release toxic agents [12,14,34]. When stimulated, microglia release a variety of cytotoxic agents which may be important mediators of neuronal injury, such as certain kinds of cytokines, reactive oxygen radicals, proteases, and glutamate. In this study, a strong activation of microglia was seen in the CA I subfield 2 days after 3 min of ischemia without preconditioning before the CA1 neurons were destroyed. These activated cells may have a neuron-killing effect. The fate of ischemic neurons may therefore depend on the reactive astroglia and microglia as they compete to govern the survival of neurons. However, considerable microglial activation, though not full-blown, was seen when ischemic insult was sublethal. Microglial ceils were frequently seen in close proximity to viable neurons. There was a rapid activation of microglia next to neurons that do not die. Of particular interest in this regard is the phenomenon termed synaptic stripping. Blinzinger and Kreutzberg [5] reported that activated microglia after facial nerve lesioning displace afferent terminals from the facial nerve cell bodies causing deafferentation. Whether similar deafferentation occurs in the hippocampus following ischemia is unknown and electron microscopic studies are necessary. However, if microglia cause similar deafferentation after ischemic preconditioning, it may protect the CA1 neurons from ischemia because excitotoxicity caused by exposure to glutamate is blocked. Actually, deafferentation has been reported to protect CA1 neurons from ischemic injury [7,40]. Just the presence of microglia therefore does not mean neuronal death and possibly is neuroprotective. There are various levels of microglial activation and this graded activation may explain these conflicting observations. When neurons were destroyed after ischemia, far intense activation of astrocytes and microglia was seen. They also changed morphology, i.e. hypertrophy of astrocytes and macrophage-like transformation of microglia. Elimination of damaged circuits would prevent continuing disruption and, perhaps, dysfunction of the entire neural system. Controlled cell death is a well-recognized process during wound healing and tissue renewal and arises when inflammatory cells release cytotoxins. Microglia may protect healthy neurons by removal of diseased cells because necrotic tissue elicits aggressive phagocytic microglia. In conclusion, the findings of this study suggest that activation of astroglia and microglia following ischemia
H. Kato et al. / Brain Research 664 (1994) 69-76
is under strict control, and that they are activated in a graded fashion in response to different degrees of neuronal injury. The graded response of microglia have been shown by the differential expression of immunomolecules and receptors on microglia [11,26], and therefore this kind of investigation may further clarify the differential activation of microglia in ischemic tolerance. It is quite possible that neuron-glia interactions, especially astroglial activation, have implications in promoting neuronal survival in ischemic tolerance. The regulation of glial effects upon neurons is further complicated by microglia-astroglial interactions because fully activated microglia may have neuron-killing effects while moderately activated microglia may not. Thus, the balance between neuron-glial and glia-glial interactions may play a critical role in neuronal survival and death after ischemia and in the manifestation of ischemic tolerance.
Acknowledgements This study was supported in part by Grant-in-Aid for Scientific Research 03404028 and Grant-in-Aid for Scientific Research on priority areas from the Ministry of Education, Science, and Culture of Japan, and by Kanae Fund of Research for New Medicine.
References [1] Abbott, N.J., Revest, P.A. and Romero, I.A., Astrocyte-endothelial interaction: physiology and pathology, Neuropathol. Appl. Neurobiol., 18 (1992) 424-433. [2] Araki, T., Kato, H. and Kogure, K., Selective neuronal vulnerability following transient cerebral ischemia in the gerbil: distribution and time course, Acta Neurol. Scand., 80 (1989) 548-553. [3] Banati, R.B., Gehrmann, J., Schubert, P. and Kreutzberg, G.W., Cytotoxicity of microglia, Glia, 7 (1993) 111-118. [4] Benveniste, H., Drejer, J., Schousboe, A. and Diemer, N.H., Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis, J. Neurochem., 43 (1984) 1369-1374. [5] Blinzinger, K. and Kreutzberg, G.W., Displacement of synaptic terminals from regenerating motoneurons by microglial cells, Z. Zellforsch., 85 (1968) 145-157. [6] Brightman, M., Implication of astroglia in the blood-brain barrier, Ann. NYAcad. Sci., 633 (1991) 343-347. [7] Buchan, A.M. and Pulsinelli, W.A., Septo-hippocampal deafferentation protects CA1 neurons against ischemic injury, Brain Res., 512 (1990) 7-14. [8] Choi, D.W., Cerebral hypoxia: some new approaches and unanswered questions, J. Neurosci., 10 (1990) 2493-2501. [9] Eddleston, E. and Mucke, L., Molecular profile of reactive astrocytes. Implications for their role in neurologic disease, Neuroscience, 54 (1993) 15-36. [10] Finsen, B.R., Jorgensen M.B., Diemer N.H. and Jimmer, J., Microglial MHC antigen expression after ischemic and kainic acid lesions of the adult rat hippocampus, Glia, 7 (1993) 41-49. [11] Gehrmann, J., Bonnekoh, P., Miyazawa, T., Hossmann, K.-A.
75
and Kreutzberg, G.W., Immunocytochemical study of an early microglial activation in ischemia, J. Cereb. Blood Flow Metab., 12 (1992) 257-269. [12] Giulian, D., Reactive glia as rivals in regulating neuronal survival, Glia, 7 (1993) 102-110. [13] Giulian, D. and Vaca, K., Inflammatory glia mediate delayed neuronal damage after ischemia in the central nervous system, Stroke, 24 (Suppl. I) (1993) 1-84-1-90. [14] Giulian, D., Vaca, K. and Corpuz, M., Brain glia release factors with opposing actions upon neuronal survival, J. Neurosci., 13 (1993) 29-37. [15] Hatten, M.E., Liem, R.K.H., Shelanski, M.L. and Mason, C.A., Astroglia in CNS injury, Glia, 4 (1991) 233-243. [16] Hossmann, K.-A., Development and resolution of ischemic brain swelling. In Papius, H.M. and Feindel, W. (Eds.), Dynamics of Brain Edema, Springer, Berlin, 1976, pp. 219-227. [17] J0rgensen, M.B., Finsen, B.R., Jensen, M.B., Castellano, B., Diemer, N.H. and Zimmer, J., Microglial and astroglial reactions to ischemic and kainic acid-induced lesions of the adult rat hippocampus, Exp. Neurol., 120 (1993) 70-88. [18] Kato, H., Araki, T., Kogure, K., Murakami, M. and Uemura, K., Sequential cerebral blood flow changes in short-term cerebral ischemia in the gerbil, Stroke, 21 (1990) 1346-1349. [19] Kato, H., Liu, Y., Araki, T. and Kogure K., 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 (1991) 238-242. [20] Kirino, T., Delayed neuronal death in the gerbil hippocampus following ischemia, Brain Res., 239 (1982) 57-69. [21] Kirino, T., Tsujita, Y. and Tamura, A., Induced tolerance to ischemia in gerbil hippocampal neurons. J. Cereb. Blood Flow Metab. 11 (1991) 299-307. [22] Kitagawa, K., Matsumoto, M., Tagaya, M., Hata, R., Ueda, H., Niinobe, M., Handa, N., Fukunaga, R., Kimura, K,. Mikoshiba, K. and Kamata, T., 'Ischemic tolerance' phenomenon found in the brain, Brain Res., 528 (1990) 21-24. [23] Liu, Y., Kato, H., Nakata, N. and Kogure, K., Temporal profile of heat shock protein 70 synthesis in ischemic tolerance induced by preconditioning ischemia in rat hippocampus, Neuroscience, 56 (1993) 921-927. [24] Marty, S., Dusart, I. and Peschanski, M., Glial changes following an excitotoxic lesion in the CNS. I. Microglia/macrophages, Neuroscience, 45 (1991) 529-539. [25] Matsuyama, T., Tsuchiyama, M., Nakamura, H., Matsumoto, M. and Sugita M., Hilar somatostatin neurons are more vulnerable to an ischemic insult than CA1 pyramidal neurons, J. Cereb. Blood Flow Metab., 13 (1993) 229-234. [26] Morioka, T., Kalehua, A.N. and Streit, W.J., The microglial reaction in the rat dorsal hippocampus following transient forebrain ischemia, J. Cereb. Blood Flow Metab., 11 (1991)966-973. [27] Morioka, T., Kalehua, A.N. and Streit, W.J., Progressive expression of immunomolecules on microglial cells in rat dorsal hippocampus following transient forebrain ischemia, Acta Neuropathol., 83 (1992) 149-157. [28] Nakata, N., Kato, H. and Kogure, K., Protective effects of basic fibroblast growth factor against hippocampal neuronal damage following cerebral ischemia in the gerbil, Brain Res., 605 (1993) 458-464. [29] Nicholls, D. and Attwell, D., The release and uptake of excitatory amino acids, Trends Pharmacol. Sci., 11 (1990) 462-468. [30] Petito, C.K., Chung, M.C., Verkhovsky, L.M. and Coopwe, A.J.L., Brain glutamine synthetase increases following cerebral ischemia in the rat, Brain Res., 569 (1992) 275-280. [31] Rothman, S.M. and Olney, J.W., Glutamate and the pathophysiology of hypoxic-ischemic brain damage, Ann. Neurol., 19 (1986) 105-111.
76
H. Kato et aL /Brain Research 604 (1994) 09-7(3
[32] Schmidt-Kastner, R., Szymas, J. and Hossmann, K.-A., lmmunohistochemical study of glial reaction and serum-protein extravasation in relation to neuronal damage in rat hippocampus after ischemia, Neuroscience, 38 (1990) 527-540. [33] Shigeno, T., Mima, T., Takakura, K., Graham, D.T., Kato, G., Hashimoto, Y. and Furukawa, S., Amelioration of delayed neuronal death in the hippocampus by nerve growth factor, J. Neurosci., 11 (1991) 2914-2919. [34] Streit, W.J., Graeber, M.B. and Kreutzberg, G.W., Functional plasticity of microglia: a review, Glia, 1 (1988) 301-307. [35] Streit, W.J., An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4), J. Histochem. Cytochem., 38 (1990) 1683-1686.
[36] Streit, W.J., Microglial-neuronal interactions, J. Chem. Neuroanat., 6 (1993) 261-266. [37] Vibulsreth, S., Hefti, F., Ginsberg, M.D., Dietrich, W.D. and Busto, R., Astrocytes protect cultured neurons from degeneration induced by anoxia, Brain Res., 422 (1987) 303-311. [38] Walz, W., Role of glial cells in the regulation of the brain ion microenvironment, Prog. Neurobiol., 33 (1989) 309-333. [39] Wieloch, T., Neurochemical correlates to selective neuronal vulnerability, Prog. Brain Res., 63 (1985) 69-85. [40] Wieloch, T., Lindvall, O., Blomqvist, P. and Gage, F.H., Evidence for amelioration of ischaemic neuronal damage in the hippocampal formation by lesions of the perforant path, Neurot Res., 7 (1985) 24-26.