Brain Research, 510 (1990) 17-25
17
Elsevier BRES 15213
Chronic maintenance of presynaptic terminals in gliotic hippocampus following ischemia Takaaki Kirino, Akira Tamura and Keiji Sano Department of Neurosurgery, Teikyo University School of Medicine, Tokyo (Japan) (Accepted 25 July 1989)
Key words: Synapse; Dendrite; Gliosis; Ischemia; Hippocampus; Degeneration
Following brief cerebral ischemia, neurons are selectively damaged and die~ whereas glial cells and blood vessels survive. This phenomenon of selective vulnerability is well illustrated in the hippocampal CAI region. Five min of forebrain ischemia in the Mongolian gerbil produced selective neuronal necrosis in the hippocampal CA1 sector. After destruction and loss of CA1 neurons, a remarkable glial reaction (gliosis) was seen. The thickness of the CA1 subfield remained unchanged until 1 month after ischemia and then gradually shrank over several months. Ultrastructural observation of this region revealed persistent maintenance of presynaptic structures. Numerous presynaptic terminals containing synaptic vesicles were scattered throughout the gliotic scar tissue. These presynaptic terminals were apposed to degenerative structures which seemed most likely to be remnants of dendrites. In another group of animals, at one month following ischemic damage in the CA1 sector, the CA3 neurons were destroyed by kainic acid injection. In these animals, numerous degenerating presynaptic boutons were seen in the CAI sector when fixed 4 days following kainate injection. These results indicate that even in gliotic tissue, presynaptic terminals can survive and maintain their structural characteristics although neuronal cell bodies are almost absent.
INTRODUCTION
damaged following ischemic insult. O n the other hand,
N e u r o n s are usually the most vulnerable cell type of nervous tissue and are easily destroyed when exposed to
n u m e r o u s presynaptic terminals maintain their structure up to at least 7 days following ischemia when most of the postsynaptic structures have been destroyed and removed 7,9.
detrimental insults such as ischemia, hypoglycemia, status epilepticus, or excitotoxins. Following selective n e u r o n a l necrosis and loss, cell debris is cleared away and the glial reaction takes over. Neural structures are not likely to play any major functional roles in gliotic tissue. A n y n e u r o n a l elements including neural processes coming from outside n e u r o n s are believed to disappear in the gliotic region. A typical example of gliosis is glial 'sclerosis' of the hippocampal CA1 sector, which is frequently e n c o u n t e r e d in epileptic patients 14 and in patients who had suffered transient but profound cerebral ischemia 3. This pathological alteration of the CA1 has been k n o w n since the 19th century 26, and has attracted wide attention since it seems to be related to temporal lobe epilepsy 23 and amnestic syndrome following cardiac arrest 2~. Recently, it has become possible to reproduce ischemic hippocampal damage and ensuing gliosis easily in rodents. Brief transient ischemia is now known to kill hippocampal CA1 pyramidal cells in the Mongolian gerbil 9 and in the rat 2°'2s. In CA1, neurons are selectively
Ischemic damage in the hippocampal CA1 region, therefore, removes postsynaptic structures, but it leaves presynaptic terminals alive. The purpose of the present experiment is to see if the presynaptic terminals remain for an extended period of time after ischemia and to examine whether these terminals are truly presynaptic elements which come from intact regions lying outside the CA1 sector. To our knowledge, the fact that neural elements remain in a morphologically intact form in gliotic scar tissue for many months has not been described. MATERIALS AND METHODS Male Mongolian gerbils (Meriones unguiculatus, 60-911 g) were subjected to transient forebrain ischemia for 5 rain. The animals were anesthetized with 2% halothane in 30% O z and 70% N 2. Immediately after exposure of the bilateral common carotid arteries, the halothane vaporizer was turned off. Then, the carotid arteries were occluded with aneurysm clips. Following 5 min of bilateral carotid artery occlusion, the gerbils were returned to heated cages.
Correspondence: T. Kirino, Department of Neurosurgery, Teikyo University School of Medicine, 2-11-1 Kaga, Itabashi-ku, Tokyo 173, Japan. 0006-8993/90/$(}3.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
18 Light microscopy One month (n = 10), 3 months (n = 10), 6 months (n = 13), and 12 months (n = 7) following ischemia, the animals were fixed by transcardiac perfusion with 3.5% formaldehyde in 0.1 M phosphate buffer (pH = 7.4). The brain was kept in a refrigerator overnight and then removed. Specimens were embedded in paraffin and used for light microscopy and immunocytochemical studies. Five-/~mthick sections containing the dorsal hippocampi were cut and stained with hematoxylin-eosin or Luxol fast blue-Cresyl violet. Unoperated gerbils (n = 11) were fixed in the same way and used as normal controls. One section which included the dorsal hippocampus 0.5-1.0 mm posterior to its most rostral tip j2 was selected from each animal. Polaroid photographs of the right and left dorsal hippocampi were taken. The area of the CAI subfield of each hippocampal region and the length of the CA1 stratum pyramidale were measured on enlarged photographs using a digitizer (Graphtek KD4030). The area of the CA1 sector included the stratum oriens, stratum pyramidale, stratum radiatum, and stratum lacunosummoleculare. The width of the CA1 sector was calculated by dividing the area of the CA1 sector by the length of the CA1 stratum pyramidale. The average of the values of right and left sides was used as the value for each animal. Statistical analysis was done using Student's t-test. To confirm reactive gliosis, paraffin sections prepared for light microscopy were stained immunocytochemicaUy using antibody raised against glial fibrillary acidic protein (GFAP, Polysciences) and Vectastain ABC kit.
Electron microscopy Gerbils were subjected to 5 min of forebrain ischemia as described above. One month (n = 4), 3 months (n = 4), 6 months (n = 5), and 12 months (n = 7) following the operation, the animals were perfusion-fixed. The perfusate consisted of 2.0% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH = 7.4). After being kept in a refrigerator overnight, the brain was removed and specimens containing the dorsal hippocampus as described above were postfixed in 2% OsO 4. Tissue samples were embedded in Epon. Thin sections were cut on a diamond knife, stained with uranyl acetate and lead citrate, and observed using a Hitachi HS-9 electron microscope. The main area of observation was the neuropil in the stratum radiatum of the CA1 sector. Unoperated normal gerbils (n = 5), processed similarly, served as normal controls.
Kainic acid injection Gerbils were subjected to 5 min of ischemia as described above. One month following surgery, the animals were deprived of food overnight. On the following morning, they were injected with atropine sulphate (5/~g/animal) and anesthetized with pentobarbital (40 mg/kg, i.p.). They were then fixed on a stereotaxic apparatus. A small burr hole was made in the skull 1.6 mm posterior to the bregma and 3.0 mm right of the midline. An unbevelled stainless steel needle (external caliber = 100/~m) was lowered 1.5 mm from the dural contact. The tip of the needle was thus located in the CA3 sector of the right hippocampus. Kainic acid (Sigma) was dissolved in saline and its pH adjusted to 7.4. Using a microinfusion pump, 0.2 ,ul of kainate solution (1/~g//~l) was injected over a period of 4 rain. The needle was left undisturbed for 10 min and then
Fig. 1. Gerbil hippocampi showing gradual shrinkage of the CA1 sector. The CA1 subfield and the dentate gyrus (DG) are divided by the fused hippocampal fissure (arrowheads), At 1 month following ischemia (b), the thickness of the CA1 sector is similar to the normal (a). At 3 months after ischemia (c) and thereafter (d,e), the thickness progressively decreased. At 12 months after ischemia (e), however, the thickness is still maintained at 70% of the normal. (Paraffin, Cresyl violet and Luxol fast blue, bar in a = 2001zm, same magnification throughout.)
withdrawn. Four days following this procedure, ammals we~c perfusion-fixed. Four animals were fixed with ?,.5c~4 formaldehyde in 0.1 M phosphate buffer. These specimens were used for the silver
19
Fig. 2. At 1 month following ischemia, immunostaining with anti-glial fibrillary acidic protein (GFAP) demonstrates remarkable gliosis in the CA1 sector (left, bar = 200/am). A magnified picture (right) shows abundant reactive astrocytes in the CA1 sector. (Bar = 10 pm.)
impregnation method a for localization of degenerating presynaptic terminals. The rest (n = 4) of the animals were perfusion-fixed and processed for electron microscopy as described above. For comparison, gerbils (n = 4) were perfusion-fixed 1 month following ischemia without kainate injection and the specimens were processed for the silver impregnation method.
RESULTS
Light microscopy All of the gerbils (n = 40) w h i c h had b e e n s u b j e c t e d to 5 min of i s c h e m i a d e v e l o p e d n e u r o n a l necrosis in the
Fig. 3. At 1 month following ischemia, presynaptic terminals which contain many vesicles are seen. These terminals are apposed to degenerative dendritic processes (*). Thickned and specialized plasma membrane (arrowheads) is noticed between these structures. (Stratum radiatum of the CA1 sector, EM, bar = 1 #m.)
21
Fig. 6. At 12 months after ischemia, the neuropil consists of astrocytes (A), astrocytic processes (a) and many presynaptic terminals. Astrocytic processes (a) are filled with intermediate filaments. Presynaptic boutons are not usually apposed to postsynaptic structures. One specialized apposition is seen in this picture (arrow). (Stratum radiatum of the CA1 sector, EM, bar = 1 ~m.)
CA1 sector. The pattern of hippocampal damage was identical to that already described 9"~°. In 37 out of 40 gerbils, the CA1 damage was severe on both sides, but an asymmetrical tissue injury between left and right hippocampi was found in 3 animals. The average thickness of the CA1 sector in the unoperated, normal gerbils was 501 + 19 ( S . D . ) ~ m (Fig. la). At 1 month following ischemia (Fig. lb), the thickness was 505 + 30 ktm and there was no significant difference from the normal animals. At 3 months after ischemia (Fig. lc), the thickness was 425 + 23 ~m. At 6 months after ischemia (Fig. ld), the thickness was 387 + 39 ~m and at 12 months after operation (Fig. ld), the thickness was reduced to 359 +_ 14/~m. At 3 months or longer, the thickness of the CA1 sector was significantly lower (P < 0.001) than that of unoperated, normal gerbils. Immunocytochemical staining using antiserum against glial fibrillary acidic protein (GFAP) was observed only
sparsely in the normal hippocampus. There was no noticable difference between the hippocampus and other areas of the brain. On the other hand, G F A P antiserum produced intense staining of glial cells in the hippocampus following ischemia. The G F A P positive cells appeared in large numbers selectively in the CA1 sector at 1 month following ischemia and later (Fig. 2).
Electron microscopy The apical dendrites of pyramidal cells, their dendritic spines, and presynaptic terminals were seen in the stratum radiatum of the CA1 sector of the normal controls. These structures were not fundamentally different from the description given in standard textbooks (e.g. ref. 19). At 1 month following ischemia (Fig. 3), most of the pyramidal cells in the C A I sector had disappeared and only sparsely scattered neurons were found. In the
22
Fig. 7. a: the CA1 sector in an unoperated, normal gerbil. The stratum pyramidale, radiating dendritic shafts, and the intact ncuropil are seen. (Epon, Toluidine blue, bar - 10/~m.) b: at 1 month following ischemia, most of the pyramidal cells have been destroyed and removed. The stratum pyramidale is replaced with a cluster of glial cells. No radiating pattern of dendritic shafts is seen. The neuropil is not spongy at this stage. (Epon, Toluidine blue, bar = 10/~m.) c: at 1 month following ischemia, the CA3 neurons have been destroyed with kainate injection. In the stratum radiatum of the CA1 sector, the spongy state in the neuropil is remarkable compared to b. (Epon, Toluidinc blue, bar = 10 k~m.) d: an EM picture of the CA 1 stratum radiatum after kainate injection with preceding ischemic CA 1 injury (taken from the same specimen as c). The spongy state in c is due partially to extremely swollen presynaptic terminals. In this swollen terminal, synaptic vesicles and specialized membrane thickening are observed (arrow). (Bar = 1 urn.) stratum radiatum,
t h e typical r a d i a t i n g p a t t e r n o f d e n -
n e u r o p i l , T h e s e t e r m i n a l s w e r e f r e q u e n t l y a p p o s e d to
d r i t i c s h a f t s h a d b e e n lost a n d t h e n o r m a l d e n d r i t i c s p i n e
membranous
apparatus could not be found. On the other hand, many
flocculent material. In the interface between the mem-
structures which contained electron-dense,
t e r m i n a l s c o n t a i n i n g a b u n d a n t vesicles w e r e s e e n in t h e
branous structure and the vesicle-containing terminals,
23 thickening or specialization of the plasma membrane was noticed. Astroglial cells were frequently encountered, which contained numerous intermediate fibers in their cytoplasm. At 3 months following ischemia (Fig. 4), the vesiclecontaining terminals were seen commonly in the stratum radiatum. They were associated with a membranous structure, the inside of which was packed with electrondense, amorphous material. The fine structure of the neuropil was basically similar to that seen 1 month following ischemia (Fig. 3). At 6 months after ischemic insult (Fig. 5), the vesicle-containing terminals were still abundant, and were often localized in clusters of 5-10 terminals. They were usually not apposed to membranous structures. When membrane specialization was noticed in the vesicle-containing terminals, degenerative dendritic structures were seen opposite the specialized plasma membrane. At 12 months after ischemia, many presynaptic terminals were still seen and their fine structure was similar to that observed 6 months after ischemia (Fig. 6). Presynaptic terminals were not evenly distributed. While they were numerous and found as clusters in some regions, a complex network of astrocytic processes was dominant and presynaptic terminals were sparse in other areas. They usually did not show specialized apposition to postsynaptic structures. In the CA1 sector, neurons were only occasionally observed following ischemia. They were usually degenerating but appeared to be neurons because presynaptic boutons were found in contact with the surface of their perikarya. Even at 12 months following ischemia, highly degenerated but structurally recognizable neurons were sparsely scattered in the CA1 subfield.
Kainate injection At 1 month following ischemia, degenerating structures were selectively demonstrated by the Gallyas method in the CA1 sector and to a small extent in the CA4 subfield. Kainate injection in the CA3 sector destroyed CA3 pyramidal cells. Silver deposits were seen in the CA3 sector and in areas known to have synaptic connections with CA3 neurons, such as the inner 1/3 of the dentate molecular layer 1~. The silver impregnation method thus showed that kainate kills CA3 neurons even after isehemic CA1 damage. However, silver staining failed to show axon degeneration of CA3 neurons in the CA1 sector because there were already massive silver deposits in the C A I sector following ischemic insult. Semithin sections of Epon-embedded specimens revealed an extremely spongy state in the CA1 sector following kainate injection (Fig. 7c), which was never noticed at 1
month following ischemia without kainate injection (Fig. 7b). It was difficult, by electron microscopic observation, to tell the exact origin of every swelling seen in the CA1 sector following kainate injection. Many swollen structures, however, definitely contained synaptic vesicles in their periphery together with specialized membrane thickening (Fig. 7d). The spongy state in the CA1 sector seen in the kainate injected gerbils was caused, at least partially, by electron-lucent terminal degeneration ~ of presynaptic terminals. DISCUSSION Most of the neurons in the hippocampai CA1 sector were destroyed after brief forebrain ischemia in the Mongolian gerbil. The CA1 sector became almost devoid of neuronal structure, as demonstrated by immunostaining using antibody against neuron-specific enolase 11. A massive increase in GFAP-positive glial cells was shown by GFAP staining. These facts indicate that all of the morphological changes in the CA1 subfield are comparable to those of gliosis. The gliotic nature of the CA1 damage has been known since the early reports by Sommer 26 and others, and the change has often been called Ammon's horn sclerosis. The damaged hippocampus gradually shrank, which became obvious at 3 months following ischemia. The bulk of the CA1 sector, however, did not fall into total tissue atrophy during the observation period. The CA1 sector maintained its thickness at almost 70% of the normal value at 12 months following ischemia. Persistent presynaptic terminals were found abundantly throughout the observation period in the gliotic tissue of the hippocampal CA1 sector. The terminals contained numerous synaptic vesicles. In the early phase (within 3 months), the terminals were usually apposed to membranous structures with specialized membrane thickening. This specialization indicates that the membranous structure is dendritic in nature. The inside of this membrane structure, however, was highly degenerated. There were only a few neurons surviving in the CA1 sector at this stage. Did these degenerative postsynaptic processes have connections with neuronal cell bodies'? Neurons were too small in number to account for the existence of all of the postsynaptic structures. Why did the presynaptic terminals remain for such a long time? These questions remain unanswered by the present experiments. Westerberg et al. 27 studied the changes in excitatory amino acid receptor ligand binding in the rat hippocampus following ischemia. Surprisingly, although most of the postsynaptic structures were destroyed at 7 days after ischemia, only 25% of the binding of N M D A receptors was lost in CA1. This maintained binding
24 capacity may be accounted for by the preservation of postsynaptic m e m b r a n e remnants as shown in this experiment. These r e m n a n t s may play some role in the chronic m a i n t e n a n c e of presynaptic terminals since they are frequently associated with presynaptic structures even 6 - 1 2 months after ischemia. These findings show that presynaptic terminals maintained their structural integrity for at least 12 months, 1/3 of the gerbil's life span I in gliosis. Their integrity is further e m p h a s i z e d by the fact that, when C A 3 neurons were d e s t r o y e d by kainic acid injection, terminal degeneration was seen a b u n d a n t l y in the gliotic C A I region. Kainic acid is known to destroy neurons but not axons located a r o u n d the injection site ~6. Kainate selectively kills C A 3 neurons in the hippocampus 15. Presynaptic terminals in the gliotic CA1 sector fell into degeneration i n all likelihood because the kainate-sensitive C A 3 n e u r o n s died and their Schaffer collaterals d e g e n e r a t e d , resulting in d e g e n e r a t i o n of axon terminals in the CA1 sector. The m o r p h o l o g i c a l findings in this study seem to indicate that m a j o r pathological changes following ischemia are postsynaptic. Presynaptic processes in the vicinity of injured neurons survive the insult. Persistent presynaptic terminals and degenerative postsynaptic structures in the present study are very similar to those o b s e r v e d after kainate injury 4"17. These findings are in a g r e e m e n t with the excitotoxic hypothesis of ischemic h i p p o c a m p a l injury ~. This hypothesis is based on the facts that excitatory amino acid ( E A A ) neurotransmitters are harmful to neurons if their extracellular concentration exceeds a certain value 22 and that the E A A concentration is e l e v a t e d following ischemic insult 2"6. Excitotoxins are REFERENCES 1 Arrington, L.R., Beaty, T.C. Jr. and Kelley, K.C, Growth, longevity, and reproductive life of the Mongolian gerbil, Lab. Animal Sci., 23 (1973) 262-265. 2 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. 3 Brierley, J.B. and Graham, D.I., Hypoxia and vascular disorders of the central nervous system. In J.H. Adams, J.A.N. Corsellis, and L.W. Duchen (Eds.), Greenfield's Neuropathology, 4th edn., Edward Arnold, London, 1984, pp. 125-207. 4 Coyle, J.T., Molliver, M.E. and Kuhar, M.J., In situ injection of kainic acid: A new method for selectively lesioning neuronal cell bodies while sparing axons of passage, J. Comp. Neurol., 180 (1978) 301-324. 5 Gallyas, E, Wolff, J.R., Boettcher, H. and Zaborszky, L., A reliable and sensitive method to localize terminal degeneration and lysosomes in the central nervous system, Stain. Technol., 55 (1980) 299-306. 6 Hagberg, H., Lehmann, A., Sandberg, M., Nystrom, B., Jacobson, I. and Hamberger, A., Ischemia-induced shift of inhibitory and excitatory amino acids from intra- to extracellulat
believed to be detrimental to postsynaptic structures, causing 'axon-sparing d e n d r o s o m a t o t o x i c ' injury ~'. However, the time course of the neuronal injury in the C A I sector is exceptionally slow '~'2°. The extracellular E A A concentration is elevated but the increase is only transient 2'6 c o m p a r e d to delayed progression of the cell damage. The exact mechanism of delayed neuronal death in the CA1 neurons is still enigmatic. Gliosis is considered to be a tissue reaction which follows severe neuronal d e p o p u l a t i o n in the central nervous system. It is often described as 'scar' formation of the brain. W h e n a few neurons survive in a gliotic region, new and bizarre dendritic arborization often develops 24. These residual neurons are thought to play an important role in epileptogenesis. Presynaptic terminals have not attracted attention in gliotic tissue. The fine structure of these presynaptic terminals as seen in this e x p e r i m e n t is strikingly normal. They a p p e a r able to release neurotransmitters, which may be contained in the synaptic vesicles. R e l e a s e d transmitters may spread by diffusion and could well be a pathogenetic factor in local epileptogenesis. A l t h o u g h these hypotheses are made only on morphological grounds and the persistence of presynaptic terminals could be an u n i m p o r t a n t side p h e n o m e n o n of tissue shrinkage in the brain, further study of this topic seems to be warranted. Acknowledgements. The authors would like to thank Miss N. Tomukai and T. Iwasawa for their technical assistance and Dr. O. Gotoh for preliminary discussion. We also acknowledge the technical support of Dr. I. Takahashi, Laboratory of Electron Microscopy, Teikyo University School of Medicine. This work was supported in part by Grants-in-Aid, for Scientific Research 63570687 and 63440054 from the Ministry of Education, Science and Culture of Japan.
compartments, J. Cereb. Blood Flow Metab., 5 (1985) 413-419. 7 Johansen, F.F., Jorgensen, M.B., von Lubitz, D.K.J.E. and Diemer, N.H., Selective dendrite damage in hippocampal CA1 stratum radiatum with unchanged axon ultrastructure and glutamate uptake after transient cerebral ischaemia in the rat, Brain Research, 291 (1984) 373-377. 8 Jorgensen, M.B. and Diemer, N.H., Selective neuron loss after cerebral ischemia in the rat: possible rote of transmitter glutamate, Acta Neurol. Scand., 66 (1982) 536-546. 9 Kirino, T., Delayed neuronal death in the gerbil hippocampus following ischemia, Brain Research, 239 (1982) 57-69. 10 Kirino, T. and Sano, K., Selective vulnerability in the gerbil hippocampus following transient ischemia, Acta Neurophathol., 62 (1984) 201-208. 11 Kirino, T., Tamura, A. and Sano, K., Selective vulnerability of the hippocampus to ischemia - - reversible and irreversible type of ischemic cell damage, Prog. Brain Res., 63 (1985) 39-58. 12 Loskota, W.J., Lomax, P. and Verity, M.A., A Stereotaxic Atlas of the Mongolian Gerbil Brain (Meriones unguiculatus), Ann Arbor Science, Ann Arbor, 1974. 13 McWilliams, R. and Lynch, G., Terminal proliferation and synaptogenesis follwing partial deafferentation: the reinnervation of the inner molecular layer of the dentate gyrus following removal of its commissural afferents, J. Comp. Neurot., 180 (1978) 581-616.
25 14 Meldrum, B.S. and Corsellis, J.A.N., Epilepsy, In J.H. Adams, J.A.N. Corsellis, and L.W. Duchen (Eds.), Greenfield's Neuropathology, 4th edn., Edward Arnold, London, 1984, pp. 920-950. 15 Nadler, J.V., Perry, B.W. and Cotman, C.W., Intraventricular kainic acid preferentially destroys hippocampal pyramidal cells, Nature (Lond.), 271 (1978) 676-677. 16 Olney, J.W., Fuller, T. and de Gubareff, T., Acute dentrotoxic changes in the hippocampus of kainate treated rats, Brain Research, 176 (1979) 91-100. 17 Olney, J.W., Collins, R.C. and Sloviter, R.S., Excitotoxic mechanisms of epileptic brain damage. In A.V. DelgadoEscueta, A.A. Ward, Jr., D.M. Woodbury, and R.J. Porter (Eds.), Basic Mechanism of Epilepsies. Molecular and Cellular Approaches, Advances in Neurology, Vol. 44, Raven, New York, 1986, pp. 857-877. 18 O'Neal, J.T. and Westrum, L.E., The fine structural synaptic organization of the cat lateral cuneate nucleus. A study of sequential alteration in degeneration, Brain Research, 51 (1973) 97-124. 19 Peters, A., Palay, S.L. and Webster, H. de E, The Fine Structure of the Nervous System. The Neurons and Supporting Cells', Saunders, Philadelphia, 1976. 20 Pulsinelli, W.A., Brierley, J.B. and Plum, E, Temporal profile of neuronal damage in a model of transient forebrain ischemia,
Ann. Neurol., 11 (1982) 491-498. 21 Rothman, S.M., Synaptic activity mediates death of hypoxic neurons, Science, 220 (1983) 536-537. 22 Rothman, S., Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death, J. Neurosci., 4 (1984) 1884-1891. 23 Sano, K. and Malamud, N., Clinical significance of sclerosis of the cornu ammonis: ictal 'psychic phenomena', Arch. Neurol. Psychiat., 70 (1953) 40-53. 24 Scheibel, M.E. and Scheibel, A.B., Hippocampal pathology in temporal lobe epilepsy. A Golgi survey, U.C.L.A. Forum Med. Sci., 17 (1973) 311-337. 25 Smith, M.-L., Auer, R.N. and Siesjo, B.K., The density and distribution of ischemic brain injury in the rat following 2-10 rain of forebrain ischemia, Acta Neuropathol., 64 (1984) 319-332. 26 Sommer, W., Erkrankung des Ammonshorns als ~itiologisches Moment der Epilepsie, Arch. Psychiat., 10 (1880) 631-675. 27 Westerberg, E., Monaghan, D.T., Kalimo, H., Cotman, C.W. and Wieloch, T.W., dynamic changes of excitatory amino acid receptors in the rat hippocampus following transient cerebral ischemia, J. Neurosci., 9 (1989) 798-805. 28 Zola-Morgan, S., Squire, L.R. and Amaral, D.G., Human amnesia and the medial temporal region: Enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus, J. Neurosci., 6 (1986) 2950-2967.
17
Elsevier BRES 15213
Chronic maintenance of presynaptic terminals in gliotic hippocampus following ischemia Takaaki Kirino, Akira Tamura and Keiji Sano Department of Neurosurgery, Teikyo University School of Medicine, Tokyo (Japan) (Accepted 25 July 1989)
Key words: Synapse; Dendrite; Gliosis; Ischemia; Hippocampus; Degeneration
Following brief cerebral ischemia, neurons are selectively damaged and die~ whereas glial cells and blood vessels survive. This phenomenon of selective vulnerability is well illustrated in the hippocampal CAI region. Five min of forebrain ischemia in the Mongolian gerbil produced selective neuronal necrosis in the hippocampal CA1 sector. After destruction and loss of CA1 neurons, a remarkable glial reaction (gliosis) was seen. The thickness of the CA1 subfield remained unchanged until 1 month after ischemia and then gradually shrank over several months. Ultrastructural observation of this region revealed persistent maintenance of presynaptic structures. Numerous presynaptic terminals containing synaptic vesicles were scattered throughout the gliotic scar tissue. These presynaptic terminals were apposed to degenerative structures which seemed most likely to be remnants of dendrites. In another group of animals, at one month following ischemic damage in the CA1 sector, the CA3 neurons were destroyed by kainic acid injection. In these animals, numerous degenerating presynaptic boutons were seen in the CAI sector when fixed 4 days following kainate injection. These results indicate that even in gliotic tissue, presynaptic terminals can survive and maintain their structural characteristics although neuronal cell bodies are almost absent.
INTRODUCTION
damaged following ischemic insult. O n the other hand,
N e u r o n s are usually the most vulnerable cell type of nervous tissue and are easily destroyed when exposed to
n u m e r o u s presynaptic terminals maintain their structure up to at least 7 days following ischemia when most of the postsynaptic structures have been destroyed and removed 7,9.
detrimental insults such as ischemia, hypoglycemia, status epilepticus, or excitotoxins. Following selective n e u r o n a l necrosis and loss, cell debris is cleared away and the glial reaction takes over. Neural structures are not likely to play any major functional roles in gliotic tissue. A n y n e u r o n a l elements including neural processes coming from outside n e u r o n s are believed to disappear in the gliotic region. A typical example of gliosis is glial 'sclerosis' of the hippocampal CA1 sector, which is frequently e n c o u n t e r e d in epileptic patients 14 and in patients who had suffered transient but profound cerebral ischemia 3. This pathological alteration of the CA1 has been k n o w n since the 19th century 26, and has attracted wide attention since it seems to be related to temporal lobe epilepsy 23 and amnestic syndrome following cardiac arrest 2~. Recently, it has become possible to reproduce ischemic hippocampal damage and ensuing gliosis easily in rodents. Brief transient ischemia is now known to kill hippocampal CA1 pyramidal cells in the Mongolian gerbil 9 and in the rat 2°'2s. In CA1, neurons are selectively
Ischemic damage in the hippocampal CA1 region, therefore, removes postsynaptic structures, but it leaves presynaptic terminals alive. The purpose of the present experiment is to see if the presynaptic terminals remain for an extended period of time after ischemia and to examine whether these terminals are truly presynaptic elements which come from intact regions lying outside the CA1 sector. To our knowledge, the fact that neural elements remain in a morphologically intact form in gliotic scar tissue for many months has not been described. MATERIALS AND METHODS Male Mongolian gerbils (Meriones unguiculatus, 60-911 g) were subjected to transient forebrain ischemia for 5 rain. The animals were anesthetized with 2% halothane in 30% O z and 70% N 2. Immediately after exposure of the bilateral common carotid arteries, the halothane vaporizer was turned off. Then, the carotid arteries were occluded with aneurysm clips. Following 5 min of bilateral carotid artery occlusion, the gerbils were returned to heated cages.
Correspondence: T. Kirino, Department of Neurosurgery, Teikyo University School of Medicine, 2-11-1 Kaga, Itabashi-ku, Tokyo 173, Japan. 0006-8993/90/$(}3.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
18 Light microscopy One month (n = 10), 3 months (n = 10), 6 months (n = 13), and 12 months (n = 7) following ischemia, the animals were fixed by transcardiac perfusion with 3.5% formaldehyde in 0.1 M phosphate buffer (pH = 7.4). The brain was kept in a refrigerator overnight and then removed. Specimens were embedded in paraffin and used for light microscopy and immunocytochemical studies. Five-/~mthick sections containing the dorsal hippocampi were cut and stained with hematoxylin-eosin or Luxol fast blue-Cresyl violet. Unoperated gerbils (n = 11) were fixed in the same way and used as normal controls. One section which included the dorsal hippocampus 0.5-1.0 mm posterior to its most rostral tip j2 was selected from each animal. Polaroid photographs of the right and left dorsal hippocampi were taken. The area of the CAI subfield of each hippocampal region and the length of the CA1 stratum pyramidale were measured on enlarged photographs using a digitizer (Graphtek KD4030). The area of the CA1 sector included the stratum oriens, stratum pyramidale, stratum radiatum, and stratum lacunosummoleculare. The width of the CA1 sector was calculated by dividing the area of the CA1 sector by the length of the CA1 stratum pyramidale. The average of the values of right and left sides was used as the value for each animal. Statistical analysis was done using Student's t-test. To confirm reactive gliosis, paraffin sections prepared for light microscopy were stained immunocytochemicaUy using antibody raised against glial fibrillary acidic protein (GFAP, Polysciences) and Vectastain ABC kit.
Electron microscopy Gerbils were subjected to 5 min of forebrain ischemia as described above. One month (n = 4), 3 months (n = 4), 6 months (n = 5), and 12 months (n = 7) following the operation, the animals were perfusion-fixed. The perfusate consisted of 2.0% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH = 7.4). After being kept in a refrigerator overnight, the brain was removed and specimens containing the dorsal hippocampus as described above were postfixed in 2% OsO 4. Tissue samples were embedded in Epon. Thin sections were cut on a diamond knife, stained with uranyl acetate and lead citrate, and observed using a Hitachi HS-9 electron microscope. The main area of observation was the neuropil in the stratum radiatum of the CA1 sector. Unoperated normal gerbils (n = 5), processed similarly, served as normal controls.
Kainic acid injection Gerbils were subjected to 5 min of ischemia as described above. One month following surgery, the animals were deprived of food overnight. On the following morning, they were injected with atropine sulphate (5/~g/animal) and anesthetized with pentobarbital (40 mg/kg, i.p.). They were then fixed on a stereotaxic apparatus. A small burr hole was made in the skull 1.6 mm posterior to the bregma and 3.0 mm right of the midline. An unbevelled stainless steel needle (external caliber = 100/~m) was lowered 1.5 mm from the dural contact. The tip of the needle was thus located in the CA3 sector of the right hippocampus. Kainic acid (Sigma) was dissolved in saline and its pH adjusted to 7.4. Using a microinfusion pump, 0.2 ,ul of kainate solution (1/~g//~l) was injected over a period of 4 rain. The needle was left undisturbed for 10 min and then
Fig. 1. Gerbil hippocampi showing gradual shrinkage of the CA1 sector. The CA1 subfield and the dentate gyrus (DG) are divided by the fused hippocampal fissure (arrowheads), At 1 month following ischemia (b), the thickness of the CA1 sector is similar to the normal (a). At 3 months after ischemia (c) and thereafter (d,e), the thickness progressively decreased. At 12 months after ischemia (e), however, the thickness is still maintained at 70% of the normal. (Paraffin, Cresyl violet and Luxol fast blue, bar in a = 2001zm, same magnification throughout.)
withdrawn. Four days following this procedure, ammals we~c perfusion-fixed. Four animals were fixed with ?,.5c~4 formaldehyde in 0.1 M phosphate buffer. These specimens were used for the silver
19
Fig. 2. At 1 month following ischemia, immunostaining with anti-glial fibrillary acidic protein (GFAP) demonstrates remarkable gliosis in the CA1 sector (left, bar = 200/am). A magnified picture (right) shows abundant reactive astrocytes in the CA1 sector. (Bar = 10 pm.)
impregnation method a for localization of degenerating presynaptic terminals. The rest (n = 4) of the animals were perfusion-fixed and processed for electron microscopy as described above. For comparison, gerbils (n = 4) were perfusion-fixed 1 month following ischemia without kainate injection and the specimens were processed for the silver impregnation method.
RESULTS
Light microscopy All of the gerbils (n = 40) w h i c h had b e e n s u b j e c t e d to 5 min of i s c h e m i a d e v e l o p e d n e u r o n a l necrosis in the
Fig. 3. At 1 month following ischemia, presynaptic terminals which contain many vesicles are seen. These terminals are apposed to degenerative dendritic processes (*). Thickned and specialized plasma membrane (arrowheads) is noticed between these structures. (Stratum radiatum of the CA1 sector, EM, bar = 1 #m.)
21
Fig. 6. At 12 months after ischemia, the neuropil consists of astrocytes (A), astrocytic processes (a) and many presynaptic terminals. Astrocytic processes (a) are filled with intermediate filaments. Presynaptic boutons are not usually apposed to postsynaptic structures. One specialized apposition is seen in this picture (arrow). (Stratum radiatum of the CA1 sector, EM, bar = 1 ~m.)
CA1 sector. The pattern of hippocampal damage was identical to that already described 9"~°. In 37 out of 40 gerbils, the CA1 damage was severe on both sides, but an asymmetrical tissue injury between left and right hippocampi was found in 3 animals. The average thickness of the CA1 sector in the unoperated, normal gerbils was 501 + 19 ( S . D . ) ~ m (Fig. la). At 1 month following ischemia (Fig. lb), the thickness was 505 + 30 ktm and there was no significant difference from the normal animals. At 3 months after ischemia (Fig. lc), the thickness was 425 + 23 ~m. At 6 months after ischemia (Fig. ld), the thickness was 387 + 39 ~m and at 12 months after operation (Fig. ld), the thickness was reduced to 359 +_ 14/~m. At 3 months or longer, the thickness of the CA1 sector was significantly lower (P < 0.001) than that of unoperated, normal gerbils. Immunocytochemical staining using antiserum against glial fibrillary acidic protein (GFAP) was observed only
sparsely in the normal hippocampus. There was no noticable difference between the hippocampus and other areas of the brain. On the other hand, G F A P antiserum produced intense staining of glial cells in the hippocampus following ischemia. The G F A P positive cells appeared in large numbers selectively in the CA1 sector at 1 month following ischemia and later (Fig. 2).
Electron microscopy The apical dendrites of pyramidal cells, their dendritic spines, and presynaptic terminals were seen in the stratum radiatum of the CA1 sector of the normal controls. These structures were not fundamentally different from the description given in standard textbooks (e.g. ref. 19). At 1 month following ischemia (Fig. 3), most of the pyramidal cells in the C A I sector had disappeared and only sparsely scattered neurons were found. In the
22
Fig. 7. a: the CA1 sector in an unoperated, normal gerbil. The stratum pyramidale, radiating dendritic shafts, and the intact ncuropil are seen. (Epon, Toluidine blue, bar - 10/~m.) b: at 1 month following ischemia, most of the pyramidal cells have been destroyed and removed. The stratum pyramidale is replaced with a cluster of glial cells. No radiating pattern of dendritic shafts is seen. The neuropil is not spongy at this stage. (Epon, Toluidine blue, bar = 10/~m.) c: at 1 month following ischemia, the CA3 neurons have been destroyed with kainate injection. In the stratum radiatum of the CA1 sector, the spongy state in the neuropil is remarkable compared to b. (Epon, Toluidinc blue, bar = 10 k~m.) d: an EM picture of the CA 1 stratum radiatum after kainate injection with preceding ischemic CA 1 injury (taken from the same specimen as c). The spongy state in c is due partially to extremely swollen presynaptic terminals. In this swollen terminal, synaptic vesicles and specialized membrane thickening are observed (arrow). (Bar = 1 urn.) stratum radiatum,
t h e typical r a d i a t i n g p a t t e r n o f d e n -
n e u r o p i l , T h e s e t e r m i n a l s w e r e f r e q u e n t l y a p p o s e d to
d r i t i c s h a f t s h a d b e e n lost a n d t h e n o r m a l d e n d r i t i c s p i n e
membranous
apparatus could not be found. On the other hand, many
flocculent material. In the interface between the mem-
structures which contained electron-dense,
t e r m i n a l s c o n t a i n i n g a b u n d a n t vesicles w e r e s e e n in t h e
branous structure and the vesicle-containing terminals,
23 thickening or specialization of the plasma membrane was noticed. Astroglial cells were frequently encountered, which contained numerous intermediate fibers in their cytoplasm. At 3 months following ischemia (Fig. 4), the vesiclecontaining terminals were seen commonly in the stratum radiatum. They were associated with a membranous structure, the inside of which was packed with electrondense, amorphous material. The fine structure of the neuropil was basically similar to that seen 1 month following ischemia (Fig. 3). At 6 months after ischemic insult (Fig. 5), the vesicle-containing terminals were still abundant, and were often localized in clusters of 5-10 terminals. They were usually not apposed to membranous structures. When membrane specialization was noticed in the vesicle-containing terminals, degenerative dendritic structures were seen opposite the specialized plasma membrane. At 12 months after ischemia, many presynaptic terminals were still seen and their fine structure was similar to that observed 6 months after ischemia (Fig. 6). Presynaptic terminals were not evenly distributed. While they were numerous and found as clusters in some regions, a complex network of astrocytic processes was dominant and presynaptic terminals were sparse in other areas. They usually did not show specialized apposition to postsynaptic structures. In the CA1 sector, neurons were only occasionally observed following ischemia. They were usually degenerating but appeared to be neurons because presynaptic boutons were found in contact with the surface of their perikarya. Even at 12 months following ischemia, highly degenerated but structurally recognizable neurons were sparsely scattered in the CA1 subfield.
Kainate injection At 1 month following ischemia, degenerating structures were selectively demonstrated by the Gallyas method in the CA1 sector and to a small extent in the CA4 subfield. Kainate injection in the CA3 sector destroyed CA3 pyramidal cells. Silver deposits were seen in the CA3 sector and in areas known to have synaptic connections with CA3 neurons, such as the inner 1/3 of the dentate molecular layer 1~. The silver impregnation method thus showed that kainate kills CA3 neurons even after isehemic CA1 damage. However, silver staining failed to show axon degeneration of CA3 neurons in the CA1 sector because there were already massive silver deposits in the C A I sector following ischemic insult. Semithin sections of Epon-embedded specimens revealed an extremely spongy state in the CA1 sector following kainate injection (Fig. 7c), which was never noticed at 1
month following ischemia without kainate injection (Fig. 7b). It was difficult, by electron microscopic observation, to tell the exact origin of every swelling seen in the CA1 sector following kainate injection. Many swollen structures, however, definitely contained synaptic vesicles in their periphery together with specialized membrane thickening (Fig. 7d). The spongy state in the CA1 sector seen in the kainate injected gerbils was caused, at least partially, by electron-lucent terminal degeneration ~ of presynaptic terminals. DISCUSSION Most of the neurons in the hippocampai CA1 sector were destroyed after brief forebrain ischemia in the Mongolian gerbil. The CA1 sector became almost devoid of neuronal structure, as demonstrated by immunostaining using antibody against neuron-specific enolase 11. A massive increase in GFAP-positive glial cells was shown by GFAP staining. These facts indicate that all of the morphological changes in the CA1 subfield are comparable to those of gliosis. The gliotic nature of the CA1 damage has been known since the early reports by Sommer 26 and others, and the change has often been called Ammon's horn sclerosis. The damaged hippocampus gradually shrank, which became obvious at 3 months following ischemia. The bulk of the CA1 sector, however, did not fall into total tissue atrophy during the observation period. The CA1 sector maintained its thickness at almost 70% of the normal value at 12 months following ischemia. Persistent presynaptic terminals were found abundantly throughout the observation period in the gliotic tissue of the hippocampal CA1 sector. The terminals contained numerous synaptic vesicles. In the early phase (within 3 months), the terminals were usually apposed to membranous structures with specialized membrane thickening. This specialization indicates that the membranous structure is dendritic in nature. The inside of this membrane structure, however, was highly degenerated. There were only a few neurons surviving in the CA1 sector at this stage. Did these degenerative postsynaptic processes have connections with neuronal cell bodies'? Neurons were too small in number to account for the existence of all of the postsynaptic structures. Why did the presynaptic terminals remain for such a long time? These questions remain unanswered by the present experiments. Westerberg et al. 27 studied the changes in excitatory amino acid receptor ligand binding in the rat hippocampus following ischemia. Surprisingly, although most of the postsynaptic structures were destroyed at 7 days after ischemia, only 25% of the binding of N M D A receptors was lost in CA1. This maintained binding
24 capacity may be accounted for by the preservation of postsynaptic m e m b r a n e remnants as shown in this experiment. These r e m n a n t s may play some role in the chronic m a i n t e n a n c e of presynaptic terminals since they are frequently associated with presynaptic structures even 6 - 1 2 months after ischemia. These findings show that presynaptic terminals maintained their structural integrity for at least 12 months, 1/3 of the gerbil's life span I in gliosis. Their integrity is further e m p h a s i z e d by the fact that, when C A 3 neurons were d e s t r o y e d by kainic acid injection, terminal degeneration was seen a b u n d a n t l y in the gliotic C A I region. Kainic acid is known to destroy neurons but not axons located a r o u n d the injection site ~6. Kainate selectively kills C A 3 neurons in the hippocampus 15. Presynaptic terminals in the gliotic CA1 sector fell into degeneration i n all likelihood because the kainate-sensitive C A 3 n e u r o n s died and their Schaffer collaterals d e g e n e r a t e d , resulting in d e g e n e r a t i o n of axon terminals in the CA1 sector. The m o r p h o l o g i c a l findings in this study seem to indicate that m a j o r pathological changes following ischemia are postsynaptic. Presynaptic processes in the vicinity of injured neurons survive the insult. Persistent presynaptic terminals and degenerative postsynaptic structures in the present study are very similar to those o b s e r v e d after kainate injury 4"17. These findings are in a g r e e m e n t with the excitotoxic hypothesis of ischemic h i p p o c a m p a l injury ~. This hypothesis is based on the facts that excitatory amino acid ( E A A ) neurotransmitters are harmful to neurons if their extracellular concentration exceeds a certain value 22 and that the E A A concentration is e l e v a t e d following ischemic insult 2"6. Excitotoxins are REFERENCES 1 Arrington, L.R., Beaty, T.C. Jr. and Kelley, K.C, Growth, longevity, and reproductive life of the Mongolian gerbil, Lab. Animal Sci., 23 (1973) 262-265. 2 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. 3 Brierley, J.B. and Graham, D.I., Hypoxia and vascular disorders of the central nervous system. In J.H. Adams, J.A.N. Corsellis, and L.W. Duchen (Eds.), Greenfield's Neuropathology, 4th edn., Edward Arnold, London, 1984, pp. 125-207. 4 Coyle, J.T., Molliver, M.E. and Kuhar, M.J., In situ injection of kainic acid: A new method for selectively lesioning neuronal cell bodies while sparing axons of passage, J. Comp. Neurol., 180 (1978) 301-324. 5 Gallyas, E, Wolff, J.R., Boettcher, H. and Zaborszky, L., A reliable and sensitive method to localize terminal degeneration and lysosomes in the central nervous system, Stain. Technol., 55 (1980) 299-306. 6 Hagberg, H., Lehmann, A., Sandberg, M., Nystrom, B., Jacobson, I. and Hamberger, A., Ischemia-induced shift of inhibitory and excitatory amino acids from intra- to extracellulat
believed to be detrimental to postsynaptic structures, causing 'axon-sparing d e n d r o s o m a t o t o x i c ' injury ~'. However, the time course of the neuronal injury in the C A I sector is exceptionally slow '~'2°. The extracellular E A A concentration is elevated but the increase is only transient 2'6 c o m p a r e d to delayed progression of the cell damage. The exact mechanism of delayed neuronal death in the CA1 neurons is still enigmatic. Gliosis is considered to be a tissue reaction which follows severe neuronal d e p o p u l a t i o n in the central nervous system. It is often described as 'scar' formation of the brain. W h e n a few neurons survive in a gliotic region, new and bizarre dendritic arborization often develops 24. These residual neurons are thought to play an important role in epileptogenesis. Presynaptic terminals have not attracted attention in gliotic tissue. The fine structure of these presynaptic terminals as seen in this e x p e r i m e n t is strikingly normal. They a p p e a r able to release neurotransmitters, which may be contained in the synaptic vesicles. R e l e a s e d transmitters may spread by diffusion and could well be a pathogenetic factor in local epileptogenesis. A l t h o u g h these hypotheses are made only on morphological grounds and the persistence of presynaptic terminals could be an u n i m p o r t a n t side p h e n o m e n o n of tissue shrinkage in the brain, further study of this topic seems to be warranted. Acknowledgements. The authors would like to thank Miss N. Tomukai and T. Iwasawa for their technical assistance and Dr. O. Gotoh for preliminary discussion. We also acknowledge the technical support of Dr. I. Takahashi, Laboratory of Electron Microscopy, Teikyo University School of Medicine. This work was supported in part by Grants-in-Aid, for Scientific Research 63570687 and 63440054 from the Ministry of Education, Science and Culture of Japan.
compartments, J. Cereb. Blood Flow Metab., 5 (1985) 413-419. 7 Johansen, F.F., Jorgensen, M.B., von Lubitz, D.K.J.E. and Diemer, N.H., Selective dendrite damage in hippocampal CA1 stratum radiatum with unchanged axon ultrastructure and glutamate uptake after transient cerebral ischaemia in the rat, Brain Research, 291 (1984) 373-377. 8 Jorgensen, M.B. and Diemer, N.H., Selective neuron loss after cerebral ischemia in the rat: possible rote of transmitter glutamate, Acta Neurol. Scand., 66 (1982) 536-546. 9 Kirino, T., Delayed neuronal death in the gerbil hippocampus following ischemia, Brain Research, 239 (1982) 57-69. 10 Kirino, T. and Sano, K., Selective vulnerability in the gerbil hippocampus following transient ischemia, Acta Neurophathol., 62 (1984) 201-208. 11 Kirino, T., Tamura, A. and Sano, K., Selective vulnerability of the hippocampus to ischemia - - reversible and irreversible type of ischemic cell damage, Prog. Brain Res., 63 (1985) 39-58. 12 Loskota, W.J., Lomax, P. and Verity, M.A., A Stereotaxic Atlas of the Mongolian Gerbil Brain (Meriones unguiculatus), Ann Arbor Science, Ann Arbor, 1974. 13 McWilliams, R. and Lynch, G., Terminal proliferation and synaptogenesis follwing partial deafferentation: the reinnervation of the inner molecular layer of the dentate gyrus following removal of its commissural afferents, J. Comp. Neurot., 180 (1978) 581-616.
25 14 Meldrum, B.S. and Corsellis, J.A.N., Epilepsy, In J.H. Adams, J.A.N. Corsellis, and L.W. Duchen (Eds.), Greenfield's Neuropathology, 4th edn., Edward Arnold, London, 1984, pp. 920-950. 15 Nadler, J.V., Perry, B.W. and Cotman, C.W., Intraventricular kainic acid preferentially destroys hippocampal pyramidal cells, Nature (Lond.), 271 (1978) 676-677. 16 Olney, J.W., Fuller, T. and de Gubareff, T., Acute dentrotoxic changes in the hippocampus of kainate treated rats, Brain Research, 176 (1979) 91-100. 17 Olney, J.W., Collins, R.C. and Sloviter, R.S., Excitotoxic mechanisms of epileptic brain damage. In A.V. DelgadoEscueta, A.A. Ward, Jr., D.M. Woodbury, and R.J. Porter (Eds.), Basic Mechanism of Epilepsies. Molecular and Cellular Approaches, Advances in Neurology, Vol. 44, Raven, New York, 1986, pp. 857-877. 18 O'Neal, J.T. and Westrum, L.E., The fine structural synaptic organization of the cat lateral cuneate nucleus. A study of sequential alteration in degeneration, Brain Research, 51 (1973) 97-124. 19 Peters, A., Palay, S.L. and Webster, H. de E, The Fine Structure of the Nervous System. The Neurons and Supporting Cells', Saunders, Philadelphia, 1976. 20 Pulsinelli, W.A., Brierley, J.B. and Plum, E, Temporal profile of neuronal damage in a model of transient forebrain ischemia,
Ann. Neurol., 11 (1982) 491-498. 21 Rothman, S.M., Synaptic activity mediates death of hypoxic neurons, Science, 220 (1983) 536-537. 22 Rothman, S., Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death, J. Neurosci., 4 (1984) 1884-1891. 23 Sano, K. and Malamud, N., Clinical significance of sclerosis of the cornu ammonis: ictal 'psychic phenomena', Arch. Neurol. Psychiat., 70 (1953) 40-53. 24 Scheibel, M.E. and Scheibel, A.B., Hippocampal pathology in temporal lobe epilepsy. A Golgi survey, U.C.L.A. Forum Med. Sci., 17 (1973) 311-337. 25 Smith, M.-L., Auer, R.N. and Siesjo, B.K., The density and distribution of ischemic brain injury in the rat following 2-10 rain of forebrain ischemia, Acta Neuropathol., 64 (1984) 319-332. 26 Sommer, W., Erkrankung des Ammonshorns als ~itiologisches Moment der Epilepsie, Arch. Psychiat., 10 (1880) 631-675. 27 Westerberg, E., Monaghan, D.T., Kalimo, H., Cotman, C.W. and Wieloch, T.W., dynamic changes of excitatory amino acid receptors in the rat hippocampus following transient cerebral ischemia, J. Neurosci., 9 (1989) 798-805. 28 Zola-Morgan, S., Squire, L.R. and Amaral, D.G., Human amnesia and the medial temporal region: Enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus, J. Neurosci., 6 (1986) 2950-2967.
