Brain Rrwxch
Bullrrin, Vol.
26.
pp.
241-250.
Pergamon
Press
plc,
1991.
KS-Y23O/Yl
Printed in the U.S.A
$3.00
t
.OlJ
Induction of Heat Shock Protein 72-Like Immunoreactivity in the Hippocampal Formation Following Transient Global Ischemia MANUEL F. GONZALEZ,** DANIEL LOWENSTEIN,*Q SUSAN FERNYAK,*$ KINYA HISANAGA,*f ROGER SIMON*9 AND FRANK R. SHARP*++’
Departments
of *Neurology and i-Physiology, University of California, San Francisco the ifveterans Affuirs Medical Center and the &Tan Francisco General Hospital, San Francisco, CA 94121 Received 30 July 1990
GONZALEZ,
R. SIMON AND F. R. SHARP. Induction ofhetrr shock global ischemicc. BRAIN RES BULL 26(2) 24l250. 1991. -Global ischemia was produced in adult rats by combining bilateral carotid artery occlusions with systemic hypotension for 5 or 10 minutes. Induction of the 72 kD heat shock protein (HSP72) in the hippocampus was examined immunocytochemically 18-24 hours later. Several patterns of HSP72-like immunoreactivity (HSP72LI) were observed. Five minutes of ischemia induced HSP72 In isolated columns of CAla pyramidal neurons, or throughout CA1 pyramidal neurons and dentate hilar neurons. Ten mmutes of ischemia induced marked HSP72LI in CA3 pyramidal neurons, moderate HSP72Ll in dentate granule cells. and minimal HSP72L.I in CAI pyramidal, dentate hilar neurons, and hippocampal glia. Two hippocampi subjected to 10 minutes of &hernia exhibited marked HSP72LI in capillary endothelial cells but no neuronal or glial HSP72LI. It is proposed that (a) the induction of HSP72 In hippocampal sectors correlates with their vulnerability to global ischemia (CA1 > hilus > CA3 :> dentate gyms): (b) the induction of HSP72 in hippocampal cells correlates with their vulnerability to global ischemia in that mild ischemia induced HSP72 only in neurons, moderate &hernia in neurons and glia, and severe ischemia only in capillary endothelial cells; (c) the failure to induce HSP72 in hippocampal neurons in 2 cases of IO min ischemia may be related to severe injury causing disruption of protein synthesis in these cells.
protein
M. F., D. LOWENSTEIN.
72-like immunoreacrivir~
Global lschemia
S. FERNYAK,
in the hippocampal
Selective vulnerability
K. HISANAGA.
formation
Stroke
following
Calcium
expressed in response to a variety of harmful stimuli besides heat, including exposure to hypoxia. hypoglycemia, alcohol, calcium, heavy metals. and metabolic poisons (28,43), they are also referred to as “stress” proteins. Although HSPs are expressed by organisms ranging from bacteria to man, their physiological function is not completely understood. HSPs synthesized by cells exposed to sublethal levels of heat increase the resistance of these cells to subsequent lethal or harmful stimuli (10, 28, 43). This protective role of HSPs has been demonstrated in the mammalian nervous system. HSP induction in hyperthermic rats partially protects retinal cells from subsequent injurious light stimulation (1). Recently, Pelham (35) has hypothesized that HSPs may exert this protective function by to Dr. Frank R. Sharp,
Heat shock protein
Hippocampus
Rat
maintaining the tertiary structure of normal or partially denatured proteins. This hypothesis has been supported by studies showing that 70 kDa HSPs facilitate the movement of large proteins or peptides across mitochondrial and ribosomal membranes (13.14). In mammals, 70 kDa HSPs are among the most abundant and inducible (35). Some of these proteins are not present in resting cells. but are highly induced following heat shock or hypoglycemia. Others are expressed constitutively in resting cells and are induced to high levels following stress (35). The present study examined the induction of HSP72-like immunoreactivity (HSP72LI) in the rat’s hippocampal formation following global ischemia. The hippocampus is selectively vulnerable to this procedure (4. 20, 23-26. 36-39, 44, 49. 51, 52) and has populations of cells relatively susceptible or resistant to ischemia (37-39, 49). Our goal was to determine if HSP72 expression in the hippocampal formation is correlated with cellular vulnerability to ischemic injury.
EXPOSING cells to elevated temperatures induces the synthesis of a family of proteins, collectively known as the heat shock proteins (HSPs) (10, 28, 43). However, since these proteins are
‘Requests for reprints should be addressed cisco, CA 94121.
rrunsient
Neurology
241
Service (V127). VA Medical Center. 4150 Clement Street. San Fran-
FIG. I. Coronal sections of rat brains illustrating the patterns of HSP724ike immunoreactivity 18 hours after transient ischemia. Rats were subjected to global &hernia for 5 (A, C, D) or 10 (E, F) minutes by combining bilateral carotid occlusions with systemic hypotension. Panel B demonstrates control background staining in a hippocampus of a globaliy ischemic subject (10 min) in which primary antiserum was omitted during performance of the i~unoc~~hemis~. IX=dentate gyms; H = hip~campus: PO = posterior thalamus; H = hitus (in panels B-F). Calibration bars: Iz = 2 mm: BF= t mm.
METHOD Global ischemia was produced in 11 male Sprague-Dawley rats (Bantin and Kingman, 300-400 g) by a combination of carotid clamping and hypotension using the method of Smith et al.
(49). Briefly, subjects were anesthetized with 3.5% halothane using a nose cone and intubated. Anesthesia was maintained using a halothane, oxygen, nitrous oxide mixture so that no pain or cornea1 responses were elicitable. Both common carotid arteries
STRESS PROTEINS INDUCED BY ISCHEMIA
. /
/ .
*
. SLM
-
FIG. 2. The first pattern of HSP’IZ-like immunoreactivity observed following 5 minutes of global ischemia. Immunostained pyramidal CAla neurons were clustered in double (A) or single (B) columns. Single neurons occasionally occurred adjacent to these columns CC); their pyramidal morphology i< seen at higher magnification in (D). The asterisk in (D) indicates an axon of a hippocampal pyramidal cell. P= stratum pyramidale: SO = stratum miens; SR = stratum radiatum: SLM = stratum lacunosumimoleculare. Calibration bars: A-C = 100 pm; D = 25 pm.
were isolated via neck incisions and catheters were inserted into the right femoral artery and vein. EEG was monitored with needle electrodes inserted in muscles lateral to the skull, blood gases were monitored from samples taken from the arterial catheters. and body temperature was maintained at 37°C. Once these preparations were completed, halothane administration was discontinued and rats were maintained in a steady state for 30 min. Then. rats were administered heparin (50 U, IV) and ischemia was induced by clamping both carotid arteries with vascular clamps, infusing Arfonad (5 mgiml IV), and withdrawing venous blood to rapidly lower systemic blood pressure to approximately 50 mmHg. Six rats were rendered ischemic for 5 min, and 5 for IO min. Cessation of EEG activity was confirmed in all subjects. Following the desired period of ischemia the carotid clamps were removed, and blood pressure was restored by rapidly infusing the subject’s blood and stopping the Arfonad. Sodium bicarbonate solution (0.5 ml of 0.6M IV) was injected to
reverse systemic acidosis. When blood pressure and EEG activity were restored. rats were extubated, all catheters were removed. and incisions were sutured. Once fully recovered, subjects were returned to their cages and given free access to food and water. Two control rats were subjected to the same surgical procedures. but not rendered ischemic. Eighteen to 24 h after carotid occlusions or sham surgeries, all subjects were deeply anesthetized with ketamine (80 mg/kg) and xylazine (20 mg/kg IP). given heparin (100 U/kg IP) and perfused through the aorta with 100 ml of 0.9% saline, followed by 500 ml of 4% paraformaldehyde dissolved in 0.1 M pH 7.4 phosphate buffer (PB). Their brains were removed, and postfixed in the same fixative for 2L4 h. One hundredqm thick coronal sections were cut on a vibratome and collected in PB. The sections were then immersed in the following solutions at room temperature: 1% avidin in PB for 5 min: 1% biotin in PB for 5 min; 3 ml of PB with 2 drops of 0.001% hydrogen peroxide for 5 min: PB for
!;OiX\;%AL,E/
244
El
41
PIG. 3. The second pattern of HSP72-like immunoreactivity observed following 5 minutes of global ischemia. Immunostained pyramtdal neurons were evident throughout sector CA1 (A, B). HSP72-like immunoreactivity was also present in pyramidal neurons of sector CA4 (hilust (C). and in granule neurons of the dentate gyrus (D, upper border). Asterisks in A and C point to a pyramidal cell located in the border zone between CA4 and CA?c. DC = dentate gyms; G = granule layer: H = hilus; ibl= inner blade of hippocampal hilus; obl= outer blade of hippocampal hilua; F =~\traturn pyramidale: SO = stratum oriens; SR = stratum radiatum. Calibration bars: A = I mm: B-D = 200 km.
and PB containing 2% horse serum, 0.2% Triton X100. and 0.1% bovine serum albumin (BSA) for 2 h. This was followed by a 48 h incubation at 5°C in monoclonal antibody to HSP72 (Amersham) diluted 1:50-2ooO in the above horse serumtriton X/BSA PB mixture. Lastly, the sections were washed in PB and reacted at room temperature by the avidin-biotin-peroxidase (ABC) method using a Vectastain Kit (Vector Laboratories, Burlingame. CA). Sections were incubated in a biotinylated horse anti-mouse second antibody, sometimes preabsorbed with 2% rat serum, for 2-3 h. Sections were then placed in a avidin-horseradish peroxidase solution for 3 h, and reacted in 15 mg of DAB (Sigma) dissolved in 100 ml of PBS to which 0.001% hydrogen peroxide was added a drop at a time. Reacted sections were washed, mounted, dehydrated, and coverslipped. Representative control sections were processed as described above except that first antibody was deleted. The regions of the hippocampal formation were named according to the terminology of Lorente de N6 (29). The primary antibody used in this study is a mouse monoclonal IgG produced against purified human HeLa cell HSP72 5 min;
protein that appears to specifically recognize stress inducible HSP72 protein (56,58). On Western blots this antibody recognizes one major protein, and either a degradation product or a second protein (56). RESULTS A robust induction of HSP72LI occurred in individual brain cells following global ischemia (Figs. l-4). Golgi-like HSP72 immunostained neurons were found in the cortex, striatum, septum, hippocampal formation and thalamus of ischemic brains. However, the distribution of the immunostained cells in these regions varied between subjects (Figs. l-4), and between the left and right sides of individual brains (Fig. 1A). Because of these side-to-side differences, results on the hippocampal formation will be described in terms of individual hippocampi (12 subjected to 5 min ischemia, and 10 to 10 min ischemia). Sections from brains subjected to global ischemia not incubated in primary antiserum did not show any cellular immuno-
STRESS PROTEINS INDUCED BY ISCHEMIA
FIG. 4. The third pattern of HSP72-like immunoreactivity observed following 5 or IO minutes carotid occlusions. Intense HSP72like nnmunostaining was evident in most pyramidal neurons of CA3 (A, B, D), and in many granule cells of the dentate gyrus (C). Scattered @ial cells m the mnrr blade 01 the dentate gyms (A) and pyramidal cells throughout CA1 (B) also exhibited HSP72 immunostaining. There uas little tmmunostaining m the CA2 wctor (A. B). G = granule layer; H = hilus: HF= hippocampal fissure; ibl = inner blade of hippocampal hilus; obl= outer blade of hippocampal hilus: f’= stratum pyramidale; SO = stratum oriens: SR= stratum radiatum. Calibration bars: A, B = 500 pm; C. D= SO pm.
staining (Fig. I B ). This was true when immunocytochemistry was performed using horse anti-mouse second antisera that had been preabsorbed with rat serum. As previously noted by Vass et al. (57). some injured neurons appear to take up immunoglobulins (and other proteins), and probably because of cross-reactivity of rodent immunoglobulins, incubation with anti-mouse second antisera which have not been preabsorbed can produce nonspecific cellular staining in injured neurons and other cells (57). A diffuse brown staining was evident in some regions where there was frank cell necrosis. The light staining in the hippocampus in Fig. IB from a section not exposed to primary antibody (particularly CA I and CA3) is an example of this nonspecific artifact. This diffuse reaction product is likely produced by endogenous peroxidase activity present in injured tissue (18), which can be decreased by pretreating sections with dilute solutions of hydrogen peroxide prior to immunocytochemistry. Finally, sections of brains of control subjects not subjected to ischemia did not exhibit either background staining or specific HSP72 cellular staining. Following global ischemia, HSP72-like immunostaining ranged from virtually no staining (left hippocampus, Fig. IA) to intense staining of all hippocampal regions (right hippocampus. Fig. IA; E. F). Aside from the absence of staining observed in 4 hippo-
campi following 5 min bilateral carotid occlusions, HSP72-like immunostaining occurred in 4 patterns that roughly correlated with the duration of ischemia. The first pattern. with the least HSP72 induction, was observed in 4 hippocampi following 5 min occlusions. HSP72LI was induced only in CAla pyramidal neurons (Figs. IC: IA-D). These neurons were easily identifiable because of their stained dendrites, axons. and perikarya (Fig. ?A-D). They were clustered into columns approximately 0.2 mm wide separated by columns of similar size in which HSP72 was not induced (Figs. IC; 2A). A single column of CAla pyramidal cells was induced in one subject (Fig. 2B). Isolated immunostained pyramidal neurons were occasionally found adjacent to these columns of HSP72 induction (Fig. 2C. D). HSP72 was not induced in pyramidal neurons of regions CA1 b and CAlc or in hippocampal plia in these subjects. The second pattern was observed in two hippocampi of rats subjected to a S-min carotid occlusion. HSP72LI was induced in most, if not all. pyramidal neurons of CA1 (Figs. ID; 3A. B). Marked induction was also observed in some cells within CA4 (Fig. ID: 3C. D). This hippocampal region, which is also referred to as the hilus of the dentate gyrus. contains neurons of different morphologies (2). HSP72 was induced in the modified
FIG. 5. The fourth pattern of HSP724ike immunor~activity crbserved foliowing 10 minutes carotid occlusions. Immunostaining was evident in many blood vessels of both the CAI sector (upper portion) of the hippocampuv and the dentnre gyros (lower portion). The open arrow points to one of these vessels crossing the stratum radiatum (SR), and the filled array point to necrotic pyramidal cells. DG = dentate gyrus; ibl = inner blade of hilus; P = stratum pyramidale: SLM = stratum lacunosurn!moleculare; SO = stratum oriens. Calibration bar= 200 km.
STRESS PROTEINS INDUCED BY ISCHEMIA
pyramidal (Fig. 3C) and in hilar neurons. A few immunostained pyramidal neurons extended into sector CA3c (Fig. 3A. C, indicated by asterisks). Some HSP72 induction also occurred in granule cells in the inner blade of the dentate gyrus (Fig. 3A, D. upper border). The third pattern occurred in 10 hippocampi. 2 after 5 min occlusions. and 8 after IO min occlusions. Intense HSP72 immunostaining was present in most pyramidal neurons of sector CA3 (Figs. IE. F: 4). and in granule cells of the dentate gyrus (Figs. IE, F: 3A. C). particularly those of its inner blade (Fig. 4A). HSP72 immunostaining was evident in individual astrocytes of the inner blade of the dentate gyrus. and in scattered pyramidal neurons of all regions of CAI as well (Fig. 4B, C). However. these pyramidal neurons were not organized in columns, and their numbers were significantly less than in the first pattern. The CAI region also exhibited moderate to intense fragmentation of the neuropil (unstained holes. Figs. IE: JA), suggesting necrosis of cells in this region (18). Finally, between areas CA1 and CA3 there was a region about 0.5 mm wide, corresponding to sector pyramidal neurons (Fig. CA2. that had few immunostained 4A. B). The fourth pattern was observed in 2 hippocampi following 10 minutes of ischemia. No HSP72-positive neurons or astrocytes were observed but many immunostained blood vessels were present through all the layers of the hippocampus and dentate gyrus (Fig. 5). The spacing and radial orientation of most stained hippocampal vessels was reminiscent of the columns of pyramidal cells observed in pattern I. There were also necrotic pyramidal cells with apparent cystic lesions in all sectors of the hippocampus proper (solid arrows. Fig. 5). The distribution of HSP77 immunostaining in neocortex, striatum and other forebrain regions will be the subject of a separate report. DISCUSSION
This study shows that ischemia induces HSP72LI in pyramidal neurons of the CA1 . CA3 and CA4 sectors of the hippocampus. as well as in hilar intemeurons. dentate gyrus granule cells, and occasionally in glia. These data confirm previous results showing that HSP synthesis in the brain is markedly increased following global ischemia (16, 32, 33. 56). This increase takes place even though global ischemia causes a decrease in total protein synthesis in the brain (5. 15. 22. 55). Vass et al. (56) first examined cellular HSP72 induction in the brain following global ischemia with the same antibody used in this study. They reported that following 10 min bilateral carotid occlusions in the gerbil, HSP72LI was first detected in some dentate granule cells 4 h after ischemia. At 24 h postischemia, HSP72 induction peaked in these cells and a few scattered immunostained pyramidal cells were evident in CA1 and CA3. At 48 h. while a few stained cells were still present in the dentate gyms and CA I, immunoreactivity in CA3 was at its peak, where it continued to be evident for up to 96 h after ischemia. A comparison of their results and ours is difficult because we used a different species and model of global ischemia. did not study the time course of HSP72 induction, and examined the effects of carotid occlusions of different duration. However, while differences between the two studies exist. both essentially agree on the hippocampal regions and the types of cells that express HSP72LI 18-24 h after ischemia. Localtarion pul
oj’ HSP72-Like
Immunoreactivity
in the Hippocan-
Formation CA1
.vector.
We observed
that HSP72 can be induced in col-
241
umns of CAla pyramidal neurons when induction is absent in other hippocampal regions (Fig. 2). In addition. HSP72 can be induced in most pyramidal neurons throughout CA1 with sparing of most other regions (Fig. 3). Vass et al. (56) did not observe either phenomenon following global ischemia in the gerbil. This could be due to differences in the duration of ischemia used in the two studies or to species differences. Vass et al. (56) did observe scattered immunostained pyramidal neurons throughout CA I similar to those seen in the third pattern of this study. Silver impregnation studies of degenerating neurons following 5 min carotid occlusions in the gerbil have shown differences between region CAla and regions CAIb and c. While argyrophilic neurons appear sooner in region a, regions b and c exhibit the most intense neuronal degeneration in the hippocampal formation ( I I ). The columnar induction of HSP72 in neurons of CAla (Figs. IC, and 2) cannot be explained on a neuroanatomicai or functional basis (53.54). However. traditional staining techniques have demonstrated selective survival of islands of CAI pyramidal neurons following global ischemia using a four vessel occlusion model (5) and following cardiac arrest (4). It is possible that during episodes of moderate ischemia blood flow differences in the radially oriented individual blood vessels of CA1 a (9) cause HSP72 induction in columns of pyramidal cells. The fact that CA1 sustains a selective and marked decrease of blood flow compared to other hippocampal sectors during diffuse ischemia supports this po\sibility (49). CA3 sec’tor. HSP72 was induced in the CA3 sector in most subjects. In agreement with results of Vass et al. (56) CA3 neurons exhibited the greatest degree of HSP72 induction in the hippocampal formation (Fig. I). In the gerbil CA3 neurons sure ive IO min ischemia. (23) and in the rat are “resistant’* to global ischemia following cardiac arrest (4) or 20-30 nun four vessel occlusion (26. 36. 37, 39). Because CA3 cells are resistant and exhibit maximal expression of HSP72. Vass et al. (56) hypothesized that transient ischemic insults induce HSP72 primarily in neurons destined to survive. However, Smith et al. (49) using the same model used in this study showed that though CAI and CA4 neurons are more vulnerable. many CA3 pyramidal neurons are destroyed following IO min of systemic hypotension and bilateral carotid occlusion. Therefore. it is possible that some immunostained CA3 neurons of this study were destined to die. Several hypotheses may explain why HSP72 is markedly induced in CA3 pyramidal neurons. (a) CA3 pyramidal neurons survive the ischemic insult better than other cells (25.37): therefore they may be able to synthesize proteins for a longer period than other cells (5. 22. 55). (b) Since it has been shown in vitro that cells differ in their capability to synthesize HSP72 (10,28), CA3 cells may have a greater intrinsic ability to synthesize HSP72. (c) Finally. if intracellular Ca’ ’ concentration is one of the signals that initiates HSP72 synthesis, the fact that CA3 pyramidal cells may have a greater influx of calcium or less binding of intracellular calcium than other neurons (27.47 1may enhance HSP72 synthesis. This preferential CA3 HSP72 induction is not due to greater vulnerability. since CA I and CA4 are more vulnerable to ischemic injury than CA3 (11. 37. 38. 44, 39). or to regional distribution of excitatory amino acid receptors. since these are distributed uniformly among hippocampal pyramidal neurons ( 19, 30. 31). CA4 sector (hilus) and dentate gvus. HSP77 was also induced in sector CA4 neurons in most subjects. The rat hilus contains several types of cells, including modified pyramidal. basket and mossy neurons (2). Some of the immunostained neurons show these morphologies (Fig. 3C) and could be GABA (I 7.62) or cholecystokinin (20) containing cells. However. it is unlikely that they are xomatostatin containing neurons since hilar somatostatin cells are particularly vulnerable to global ischemia and may be
24x
TABLE 1 SUMMARY OF HSP72 IMMUNOSTAINING IN THE HIPPOCAMPI OF ADULT RATS SUBJECTED TO GLOBAL ISCHEMIA PRODUCED BY A COMBINATION OF TEMPORARY HYPOTENSION COMBINED WITH TEMPORARY CAROTID OCCLUSION ___. _~I_.__._.._.._. --
Pattern of HSP72 Staining No Staining CA 1a neurons CAla, b, c; CA4 neurons CA3, CAla-c, CA4, dentate granule neurons CA3, CAla-c, CA4, dentate granule neurons Hippocampai and dentate vessels
lost prior to CA1 rteurons (20). HSP72 induction in this region is not described by Vass et al. (56). This discrepancy may be due to a species difference since many, but not all, hilar basket cells in the gerbil die rapidly following carotid occlusion (11) and may not express HSP72. HSP72 was induced in granule neurons of the dentate gyrus in subjects exhibiting patterns 2 and 3 in this study (Figs. 1, 2, 3) and in all gerbils of Vass et al.‘s study (56). The fact that HSP72 i~unos~ning in granule cells occurred only after the protein was expressed in virtually all other hippocampal regions and was absent after moderate &hernia suggests that these cells require severe degrees of ischemia to induce the HSP72 protein. Dentate granule neurons are among the hippocampal neurons most resistant to global ischemia (11, 37, 38, 44, 49, 60, 61). Hippocampal blood vessels. HSW2 was induced in the blood vessels but not in neurons or glia of two hippocampi and dentate gyri following 10 min occlusions (Fig. 5). This induction is compatible with previous results showing ind~tion of HSPs in cerebral vessels in vivo and in vitro ( 1259). We hypothesize that the neurons and glia of these hippocampi were lethally injured shortly after global ischemia and rendered unable to synthesize HSP72. The fact that CA1 neurons appeared cystic (solid arrows, Fig 5) supports this assumption. We also hypothesize that endothelial cells were more resistant to ischemia and were able to synthesize HSP72 after neurons and glia were unable to do so.
Most studies of global ischemia have shown that CA1 pyramidal neurons and perhaps certain subpopulations of neurons in CA4 are the most vulnerable, CA3 neurons have intermediate resistance, and dentate granule cells are the most resistant to global ischemia (4, 20, 23, 26, 37-39, 43, 44, 51, 52). Significantly, Smith et al. (49), who developed the model used in this study, agree with this ranking of the vuInerabili~ of hippocampa1 cells. Our results suggest that induction of HSP72 in hip~ampus occurs in this order. However, the relationship is imperfect possibly because of the variability of the model used, and because severely injured cells may not be able to synthesize stress proteins. HSP72 as a Marker of Cell Injury Our suggestion that HSP72 immunostaining labels stressed cells is not supported by the fact that more CA1 neurons immunostained following 5 min compared to 10 min occlusions, or by the two instances that CA1 cells did not stain at dl foilowing 10 min occlusions. Nonetheless, the fact that cells exhibiting HSWZLI occur in areas selectively vulnerable to ischemia, as discussed above, supports the idea that HSP72 expression occurs in stressed
No. of ~ippocampi
Duration of Ischemia
4 4 2 2 8 2
5 min 5 min 5 min 5 min 10 min 10 min
or injured cells. The exception to this may occur when cells are so severely injured that transcription or translation of heat shock genes is impaired (15, 22, 55). Since HSPs enhance the resistance of cells to lethal stimuli (1,28), it could be maintained that cells expressing the protein have a greater probability of surviving. This study and previous work of Vass et al. (56) show that HSP72 expression was greatest in CA3 pyramidal neurons destined to survive. On the other hand, we have shown that neurons within brain areas infarcted by complete middle cerebral artery occlusions clearly destined to die may or may not express HSP72 ( 18). The fact that HSP72 can be induced both in cells destined to die or survive has also been documented in cultured cells (28.57). Therefore, it is not possible to answer the question of whether cells that express HSP72Ll are destined to survive or to die. Selective Vu~nerabil~~ in Relation to the Excitutox~c Hypothesis of ~sc~~i~ Damage
It has been hypothesized that excitatory ammo acids (EAAs) may mediate cell death in the hippocampus (7, 21, 34, 45, 60, 61) and other regions and may be the cause of selective cellular vulnerability to ischemia (3, 8, 2 1, 34, 40, 41, 45, 60). Glutamate and aspartate (1950) may mediate cell death via intracellular accumulation of calcium and/or chloride (6, 40-42). While the EAA hypothesis may explain the vulnerability of the whole hip~ampal formation to global isehemia, it does not explain why ischemia induces HSP72 selectively in CA1 p~~dal neurons (Table I), why CA1 pyramidal neurons are more vulnerable than CA3 pyramidal neurons to ischemia (4, 20, 23-26, 37, 39, 43, 44, 51, 52). or why CA3 pyramidal neurons are more vulnerable than CA1 pyramidal neurons to damage caused by stimulation of the perforant path (46,48). All pyramidal neurons receive excitatory amino acid inputs (46, 60-62). Additional factors may determine the vulnerability of particular subpopulations of cells. For example, the induction of HSP72 during ischemia and sus~ptibility of CA1 cells to global ischemia may be related to regional differences of blood flow in the hip~c~pus (9); and the susceptibility of CA4 and CA3 neurons to damage caused by perforant path stimulation may be due to their greater electrical excitability produced by differences in calcium entry or binding (46-48) or to other unknown factors. ACKNOWLEDGEMENTS
We wish to thank Dr. Stephen M. Sagar for his helpfu1 ccunmettts on this manuscript, and Pat Jasper. Theresa Marsh, and Matt Morton fur excellent technical assistance. This work was supported by NIH grants NS24666 and NS28167 tu F.R.S., NS01424 to D.L., and NS24728 to R.S., and by the VA Medical Research Service (F.R.S.) and the WCSF Academic Senate (D.L.).
STRESS PROTEINS
INDUCED
BY ISCHEMIA
249
REFERENCES I. Barbe. M. F.: Tytell, M.: Cower, D. J.; Welch, W. J. Hyperthermia protects against light damage in the rat retina. Science 241: 18171820; 1988. 2. Bayer. S. A. Hippocampal region. In: The rat nervous system, vol. I. In: Paxinos, G., ed. Sydney: Academic Press; 1985:335-352. 3. Benveniste, H.; Drejer, J.; Schousboe, A.; 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: 1369-1374; 1984. 4. Blomqvist, P.; Wieloch, T. Ischemic brain damage in rats following cardiac arrest using a long-term recovery model. J. Cereb. Blood Flow Metab. 5:42043 I; 1985. 5. Bodsch. W.; Takahashi, K.: Barbier. A.; Grosse Ophoff. B.: Hossmann. K.-A. Cerebral protein synthesis and ischemia. Prog. Brain Res. 63~237-259: 1985. 6. Choi. D. W. Ionic dependence of glutamate neurotoxicity. J. Neurosci. 7:369-379; 1987. 7. Clark, G. D.; Rothman. S. M. Blockade of excitatory amino acid receptors protects anoxic hippocampal slices. Neuroscience 21:665671; 1987. 8. Collins, R. C. Selective vulnerability of brain: new insights from the excitatory synapse. Metab. Brain Dis. I:23 l-240; 1986. 9. Coyle, P. Vascular patterns of the rat hippocampal formation. Exp. Neurol. 52:447i158: 1976. IO. Craig, E. A. The heat shock response. CRC Crit. Rev. Biochem. 18:239-280; 1985. I I. Crain. B. J.: Westerkam. W. D.; Harrison, A. H.; Nadler, J. V. Selective neuronal death after transient forebrain ischemia in the Mongolian gerbil: a silver impregnation study. Neuroscience 27:3871102: 1988. 12. Currie, R. W .; White, F. P. Trauma-induced protein in rat tissues: a physiological role for a “heat shock” protein. Science 214:72-73: 1981. 13. Deshaies. R. J.; Koch, B. D.; Schekman. R. The role of stress proteins in membrane biogenesis. Trends Biol. Sci. 13:384-388; 1988. 14. Deshaies, R. J.: Koch. B. D.: Werner-Washbume, M.: Craig, E. A.; Schekman. R. A subfamily of stress proteins facilitates translocation of secretory and mitochondrial precursor polypeptides. Nature 332: 800-805; 19X8. 15. Dienel. G. A.; Cruz. N. F.; Rosenfeld, S. J. Temporal profiles of proteins responsive to transient ischemia. J. Neurochem. 44600-610; 1985. 16. Dienel. G. A.; Kiessling, M.; Jacewicz, M.; Pulsinelli, W. A. Synthesis of heat shock proteins in rat brain cortex after transient ischemia. J. Cereb. Blood Flow Metab. 6:505-510; 1986. 17. Gamrani, H.; Onteniente. B.; Seguela, P.: Geffard, M.; Calas. A. Gamma-aminobutyric acid immunoreactivity in the rat hippocampus. A light and electron microscopic study with antiGABA antibodies. Brain Res. 364:30-38: 1986. 18. Gonzalez, M. F.; Shiraishi, K.; Kisanaga, K.; Sagar, S. M.; Mandabach. M.; Sharp, F. R. Heat shock proteins as markers of neural injury. Mol. Brain Res. 6:93-100; 1989. 19. Greenamayre. J. T.: Olson, J. M.; Penney, J. B.; Young, A. B. Autoradiographic characterization of N-methyl-D-aspartate-, quisqualateand kainate-sensitive glutamate binding sites. J. Pharmacol. Exp. Ther. 233:254-263: 1985. 20. Johansen, F. F.; Zimmer, J.; Diemer, N. H. Early loss of somatostatin neurons in dentate hilus after cerebra1 ischemia in the rat precedes CA-I pyramidal cell loss. Acta Neuropathol. (Berl.) 73:110-l 14; 1987. 21. Jorgensen. M. B.; Johansen, F. F.; Diemer, N. H. Removal of the entorhinal cortex protects hippocampal CA-1 neurons from ischemic damage. Acta Neuropathol. (Berl.) 73:189-194; 1987. 22. Kiessling, M.; Dienel, G. A.; Jacewicz, M.; Pulsinelli, W. A. Protein synthesi\ in postischemic rat brain: a two-dimensional electrophoretic analysis. J. Cereb. Blood Flow Metab. 6:642649; 1986. 23. Kirino, T. Delayed neuronal death in the gerbil hippocampus following ischemia. Brain Res. 239:5749; 1982. 24. Kirino. T.; Sano, K. Selective vulnerability in the gerbil hippocam-
pus
25.
26.
27.
28. 29.
30.
31.
32. 33.
34.
35. 36.
37. 38.
39.
40. 41.
42. 43.
44.
45.
46.
47.
48.
following transient ischemia. Acta Neuropathol. (Berl.) 62:201208; 1984. Kirino, T.: Sane, K. Fine structural nature of delayed neuronal death following ischemia in the gerbil hippocampus. Acta Neuropathol. (Berl.) 62:209-218; 1984. Kirino. T.; Tamura. A.: Sano. K. Delayed neuronal death in the rat hippocampus following transient forebrain ischemia. Acta Neuropathol, (Berl.) 64:139-147; 1984. Kudo, Y.: Ogura, A. Glutamate-induced increase in intracellular Ca” concentration in isolated hippocampal neurones. Br. J. Pharmacol. 89:19l-198: 1986. Lindquist. S. The heat-shock response. Annu. Rev. Biochem 55: 1151-l 191; 1986. Lorente de No R. Studies on the structure of the cerebral cortex. II. Continuation of the study of the ammonic system. J. Psychol. Neuml. (Leipzig) 46:1 13-177: 1934. Monaghan. D. T.: Cotman, C. W. Distribution of ,Y-methyl-D-aspartate-sensitive L-[‘HIglutamate-binding \ites in rat main. J. Neut-osci. 5:2909-2919; 1985. Monaghan. D. T.; Holets. V. R.: Toy. D. W.; Cotman. C. W Anatomical distributions of four pharmacologically distinct ‘H-L.-glutamate binding sites. Nature 306: 176-l 79: 1983. Nowak, T. S. Synthesis of a stress protein following transient ischemia in the gerbil. J. Neurochem. 45:16351641: 19X5. Nowak. T. S.; Marini. A. M.; Kozuka, M. Differential localization of 70 kDa stress protein induction in brain after i\chemia and hyperthermia. J. Cereb. Blood Flow Metab. 9:S54: 1989. Onodera, H.; Sato. G.; Kogure, K. Le\iona to Schaffer collaterals prevent iachemic death of CAI pyramidal cells. Neurosci. Lett. 68: 169-17-l; 1986. Pelham. H. R. B. Speculations on the functions of the major heat shock and glucose-regulated proteins. Cell 46:959-96 I: 1986. Petito. C. K.: Pulsinelli. W. A. Delayed neuronal recovery and neuronal death in rat hippocampus following severe cerebral ischemia: possible relationship to abnormalities in neuronal processe\. J. Ccreb. Blood Flow Metab. 4:194-205: 1984. Pulsinelli. W. A. Selective neuronal vulnerability: morphological and molecular characteristics. Prog. Brain Res. 63:29-37; 1985. Pulsinelli. W. A.: Brierley, J. B.: Plum. F. Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann. Neurol. ll:491~98: 1982. Pulsinelli, W. A.; Levy, D. E. Duffy, T. E. Regional cerebral hlood flow and glucose metabolism following transient forehram ischemia. Ann. Neurol. Il:499-509: 1982. Rothman. S. M.: Olney, J. W. Glutamate and the pathophysiology of hypoxic-ischemic brain damage. Ann. Neurol. 19: 105-l II: 1986. Rothman. S. Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death. J. Neurosci. 4:1884-1891: 1984. Rothman. S. M. The neurotoxicity of excitatory amino acids is produced by passive chloride influx. J. Neurosci. 5: 1483-1489; 1985. Schlesinger, M. J.; Ashbumer. M.: Tissierea. A. Heat shock from bacteria to man. Cold Spring Harbor. NY: Cold Spring Harbor Laboratory; 1982. Schmidt-Kastner, R.; Hossmann, K.-A. Distribution of ischemic neuronal damage in the dorsal hippocampus of rat. Acta Neuropathol. (Berl.) 76:4ll-t21; 1988. Simon, R. P.; Swan. J. H.; Griffiths. T.: Meldrum, B. S. Blockade of N-methyl-D-aspartate receptors may protect against ischemic damage in the brain. Science 226:850-852; 1984. Sloviter. R. S. “Epileptic” brain damage in rats induced by sustained electrical stimulation of the perforant path. I. Acute electrophysiological and light microscopic studies. Brain Res. Bull. IO: 675-697; 1983. Sloviter. R. S. Calcium-binding protein (Calbindin-DZXI) and parvalbumin immunocytochemistry: Localization in the rat hippocampus with specific reference to the selective vulnerability of hippocampal neurons to seizure activity. J. Comp. Neurol. 280:183-196: 1989. Sloviter. R. S.; Damiano. B. P. Sustained electrical stimulation of
49.
50.
5 1.
52.
53.
54.
55.
the perforant path duplicates kainate-induced electrophysiological effects and hippocampal damage in rats. Neurosci. Lett. 24:279-284; 1981. Smith, M.-L.; Auer, R. N.; Siesjo, B. K. The denstty and distribution of ischemic brain injury in the rat following 2-10 min of forebrain ischemia. Acta Neuropathol. (Berl.) 64:319-332: 1984. Storm-Mathisen, J. Localization of transmitter candidates in the brain: the hippocampal formation as a model. Prog. Neurobiol. 8:119-l8 1; 1977. Suzuki, R.; Yamaguchi, T.; Kirino, T.; Orzi, F.; Klatzo, 1. The effects of 5-minute ischemia in Mongolian gerbils: I. Blood-brain barrier, cerebral blood flow, and local cerebral glucose utilization changes. Acta Neuropathol. (Berl.) 60:207-216; 1983. Suzuki, R.; Yamaguchi, T.; Li, C.-L.; Klatzo. I. The effects of 5minute ischemia in Mongolian gerbils: II. Changes of spontaneous neuronal activity in cerebral cortex and CA1 sector of hippocampus. Acta Neuropathol. (Berl.) 60:217-222; 1983. Swanson, L. W.; Cowan, W. M. An autoradiographic study of the organization of the efferent connections of the hippocampal formation in the rat. J. Comp. Neurol. 172:49-84; 1977. Swanson, L. W.; Wyss, J. M.; Cowan, W. M. An autoradiographic study of the organization of intrahippocampal association pathways in the rat. J. Comp. Neurol. 181:681-716; 1978. Thilmann, R.; Xie. Y.; Kleihues. P.: Kiessling. M. Persistent inhi-
56.
57.
58.
59.
60. 61.
62.
bition of protein synthesis precedes delayed neuronal death m posttschemic gerbil hippocampus. Acta Neuropathol. (Berl. I 71:889~. 1986. Vass, K.; Welch, W. J.; Nowak, 7’. S. Localization of 70-kDa strebs protein induction in gerbil brain after ischemia. Acta Neuropathol. (Berl.) 77:128-135; 1988. Vass, K.; Berger, M. L.; Nowak, T. S., Jr.; Welch, W. J.; Lassmann. H. Induction of stress protein HSP70 in nerve cells after status epilepticus in the rat. Neurosci. Lett. 100:259-264; 1989. Welch, W. J.; Suhan, J. P. Cellular and biochemical events in mammalian cells during and after recovery from physiological stress. J. Cell Biol. 103:2035-2052; 1986. White, F. P. Differences in protein synthesized in viva and in vitro by cells associated with the cerebral vasculature. A protein synthesized in response to trauma. Neuroscience 5:1793-1799; 1980. Wieloch, T. Neurochemical correlates to selective neuronal vulnem bility. Prog. Brain Res. 63:69-85; 1985. Wieloch, T.; Lindvall, 0.; Blomqvist, P.; Gage, F. H. Evidence for amelioration of ischaemic neuronal damage in the hippocampal formation by lesions of the perforant path. Neurol. Res. 7:24-26; 1985. Woodson. W.; Nitecka, L.; Ben-Ari. Y. Organization of the GABAergic system in the rat hippocampal formation: a quantitative immunocytochemical study. J. Comp. Neurol. 280:254-271: 1989.
Bullrrin, Vol.
26.
pp.
241-250.
Pergamon
Press
plc,
1991.
KS-Y23O/Yl
Printed in the U.S.A
$3.00
t
.OlJ
Induction of Heat Shock Protein 72-Like Immunoreactivity in the Hippocampal Formation Following Transient Global Ischemia MANUEL F. GONZALEZ,** DANIEL LOWENSTEIN,*Q SUSAN FERNYAK,*$ KINYA HISANAGA,*f ROGER SIMON*9 AND FRANK R. SHARP*++’
Departments
of *Neurology and i-Physiology, University of California, San Francisco the ifveterans Affuirs Medical Center and the &Tan Francisco General Hospital, San Francisco, CA 94121 Received 30 July 1990
GONZALEZ,
R. SIMON AND F. R. SHARP. Induction ofhetrr shock global ischemicc. BRAIN RES BULL 26(2) 24l250. 1991. -Global ischemia was produced in adult rats by combining bilateral carotid artery occlusions with systemic hypotension for 5 or 10 minutes. Induction of the 72 kD heat shock protein (HSP72) in the hippocampus was examined immunocytochemically 18-24 hours later. Several patterns of HSP72-like immunoreactivity (HSP72LI) were observed. Five minutes of ischemia induced HSP72 In isolated columns of CAla pyramidal neurons, or throughout CA1 pyramidal neurons and dentate hilar neurons. Ten mmutes of ischemia induced marked HSP72LI in CA3 pyramidal neurons, moderate HSP72Ll in dentate granule cells. and minimal HSP72L.I in CAI pyramidal, dentate hilar neurons, and hippocampal glia. Two hippocampi subjected to 10 minutes of &hernia exhibited marked HSP72LI in capillary endothelial cells but no neuronal or glial HSP72LI. It is proposed that (a) the induction of HSP72 In hippocampal sectors correlates with their vulnerability to global ischemia (CA1 > hilus > CA3 :> dentate gyms): (b) the induction of HSP72 in hippocampal cells correlates with their vulnerability to global ischemia in that mild ischemia induced HSP72 only in neurons, moderate &hernia in neurons and glia, and severe ischemia only in capillary endothelial cells; (c) the failure to induce HSP72 in hippocampal neurons in 2 cases of IO min ischemia may be related to severe injury causing disruption of protein synthesis in these cells.
protein
M. F., D. LOWENSTEIN.
72-like immunoreacrivir~
Global lschemia
S. FERNYAK,
in the hippocampal
Selective vulnerability
K. HISANAGA.
formation
Stroke
following
Calcium
expressed in response to a variety of harmful stimuli besides heat, including exposure to hypoxia. hypoglycemia, alcohol, calcium, heavy metals. and metabolic poisons (28,43), they are also referred to as “stress” proteins. Although HSPs are expressed by organisms ranging from bacteria to man, their physiological function is not completely understood. HSPs synthesized by cells exposed to sublethal levels of heat increase the resistance of these cells to subsequent lethal or harmful stimuli (10, 28, 43). This protective role of HSPs has been demonstrated in the mammalian nervous system. HSP induction in hyperthermic rats partially protects retinal cells from subsequent injurious light stimulation (1). Recently, Pelham (35) has hypothesized that HSPs may exert this protective function by to Dr. Frank R. Sharp,
Heat shock protein
Hippocampus
Rat
maintaining the tertiary structure of normal or partially denatured proteins. This hypothesis has been supported by studies showing that 70 kDa HSPs facilitate the movement of large proteins or peptides across mitochondrial and ribosomal membranes (13.14). In mammals, 70 kDa HSPs are among the most abundant and inducible (35). Some of these proteins are not present in resting cells. but are highly induced following heat shock or hypoglycemia. Others are expressed constitutively in resting cells and are induced to high levels following stress (35). The present study examined the induction of HSP72-like immunoreactivity (HSP72LI) in the rat’s hippocampal formation following global ischemia. The hippocampus is selectively vulnerable to this procedure (4. 20, 23-26. 36-39, 44, 49. 51, 52) and has populations of cells relatively susceptible or resistant to ischemia (37-39, 49). Our goal was to determine if HSP72 expression in the hippocampal formation is correlated with cellular vulnerability to ischemic injury.
EXPOSING cells to elevated temperatures induces the synthesis of a family of proteins, collectively known as the heat shock proteins (HSPs) (10, 28, 43). However, since these proteins are
‘Requests for reprints should be addressed cisco, CA 94121.
rrunsient
Neurology
241
Service (V127). VA Medical Center. 4150 Clement Street. San Fran-
FIG. I. Coronal sections of rat brains illustrating the patterns of HSP724ike immunoreactivity 18 hours after transient ischemia. Rats were subjected to global &hernia for 5 (A, C, D) or 10 (E, F) minutes by combining bilateral carotid occlusions with systemic hypotension. Panel B demonstrates control background staining in a hippocampus of a globaliy ischemic subject (10 min) in which primary antiserum was omitted during performance of the i~unoc~~hemis~. IX=dentate gyms; H = hip~campus: PO = posterior thalamus; H = hitus (in panels B-F). Calibration bars: Iz = 2 mm: BF= t mm.
METHOD Global ischemia was produced in 11 male Sprague-Dawley rats (Bantin and Kingman, 300-400 g) by a combination of carotid clamping and hypotension using the method of Smith et al.
(49). Briefly, subjects were anesthetized with 3.5% halothane using a nose cone and intubated. Anesthesia was maintained using a halothane, oxygen, nitrous oxide mixture so that no pain or cornea1 responses were elicitable. Both common carotid arteries
STRESS PROTEINS INDUCED BY ISCHEMIA
. /
/ .
*
. SLM
-
FIG. 2. The first pattern of HSP’IZ-like immunoreactivity observed following 5 minutes of global ischemia. Immunostained pyramidal CAla neurons were clustered in double (A) or single (B) columns. Single neurons occasionally occurred adjacent to these columns CC); their pyramidal morphology i< seen at higher magnification in (D). The asterisk in (D) indicates an axon of a hippocampal pyramidal cell. P= stratum pyramidale: SO = stratum miens; SR = stratum radiatum: SLM = stratum lacunosumimoleculare. Calibration bars: A-C = 100 pm; D = 25 pm.
were isolated via neck incisions and catheters were inserted into the right femoral artery and vein. EEG was monitored with needle electrodes inserted in muscles lateral to the skull, blood gases were monitored from samples taken from the arterial catheters. and body temperature was maintained at 37°C. Once these preparations were completed, halothane administration was discontinued and rats were maintained in a steady state for 30 min. Then. rats were administered heparin (50 U, IV) and ischemia was induced by clamping both carotid arteries with vascular clamps, infusing Arfonad (5 mgiml IV), and withdrawing venous blood to rapidly lower systemic blood pressure to approximately 50 mmHg. Six rats were rendered ischemic for 5 min, and 5 for IO min. Cessation of EEG activity was confirmed in all subjects. Following the desired period of ischemia the carotid clamps were removed, and blood pressure was restored by rapidly infusing the subject’s blood and stopping the Arfonad. Sodium bicarbonate solution (0.5 ml of 0.6M IV) was injected to
reverse systemic acidosis. When blood pressure and EEG activity were restored. rats were extubated, all catheters were removed. and incisions were sutured. Once fully recovered, subjects were returned to their cages and given free access to food and water. Two control rats were subjected to the same surgical procedures. but not rendered ischemic. Eighteen to 24 h after carotid occlusions or sham surgeries, all subjects were deeply anesthetized with ketamine (80 mg/kg) and xylazine (20 mg/kg IP). given heparin (100 U/kg IP) and perfused through the aorta with 100 ml of 0.9% saline, followed by 500 ml of 4% paraformaldehyde dissolved in 0.1 M pH 7.4 phosphate buffer (PB). Their brains were removed, and postfixed in the same fixative for 2L4 h. One hundredqm thick coronal sections were cut on a vibratome and collected in PB. The sections were then immersed in the following solutions at room temperature: 1% avidin in PB for 5 min: 1% biotin in PB for 5 min; 3 ml of PB with 2 drops of 0.001% hydrogen peroxide for 5 min: PB for
!;OiX\;%AL,E/
244
El
41
PIG. 3. The second pattern of HSP72-like immunoreactivity observed following 5 minutes of global ischemia. Immunostained pyramtdal neurons were evident throughout sector CA1 (A, B). HSP72-like immunoreactivity was also present in pyramidal neurons of sector CA4 (hilust (C). and in granule neurons of the dentate gyrus (D, upper border). Asterisks in A and C point to a pyramidal cell located in the border zone between CA4 and CA?c. DC = dentate gyms; G = granule layer: H = hilus; ibl= inner blade of hippocampal hilus; obl= outer blade of hippocampal hilua; F =~\traturn pyramidale: SO = stratum oriens; SR = stratum radiatum. Calibration bars: A = I mm: B-D = 200 km.
and PB containing 2% horse serum, 0.2% Triton X100. and 0.1% bovine serum albumin (BSA) for 2 h. This was followed by a 48 h incubation at 5°C in monoclonal antibody to HSP72 (Amersham) diluted 1:50-2ooO in the above horse serumtriton X/BSA PB mixture. Lastly, the sections were washed in PB and reacted at room temperature by the avidin-biotin-peroxidase (ABC) method using a Vectastain Kit (Vector Laboratories, Burlingame. CA). Sections were incubated in a biotinylated horse anti-mouse second antibody, sometimes preabsorbed with 2% rat serum, for 2-3 h. Sections were then placed in a avidin-horseradish peroxidase solution for 3 h, and reacted in 15 mg of DAB (Sigma) dissolved in 100 ml of PBS to which 0.001% hydrogen peroxide was added a drop at a time. Reacted sections were washed, mounted, dehydrated, and coverslipped. Representative control sections were processed as described above except that first antibody was deleted. The regions of the hippocampal formation were named according to the terminology of Lorente de N6 (29). The primary antibody used in this study is a mouse monoclonal IgG produced against purified human HeLa cell HSP72 5 min;
protein that appears to specifically recognize stress inducible HSP72 protein (56,58). On Western blots this antibody recognizes one major protein, and either a degradation product or a second protein (56). RESULTS A robust induction of HSP72LI occurred in individual brain cells following global ischemia (Figs. l-4). Golgi-like HSP72 immunostained neurons were found in the cortex, striatum, septum, hippocampal formation and thalamus of ischemic brains. However, the distribution of the immunostained cells in these regions varied between subjects (Figs. l-4), and between the left and right sides of individual brains (Fig. 1A). Because of these side-to-side differences, results on the hippocampal formation will be described in terms of individual hippocampi (12 subjected to 5 min ischemia, and 10 to 10 min ischemia). Sections from brains subjected to global ischemia not incubated in primary antiserum did not show any cellular immuno-
STRESS PROTEINS INDUCED BY ISCHEMIA
FIG. 4. The third pattern of HSP72-like immunoreactivity observed following 5 or IO minutes carotid occlusions. Intense HSP72like nnmunostaining was evident in most pyramidal neurons of CA3 (A, B, D), and in many granule cells of the dentate gyrus (C). Scattered @ial cells m the mnrr blade 01 the dentate gyms (A) and pyramidal cells throughout CA1 (B) also exhibited HSP72 immunostaining. There uas little tmmunostaining m the CA2 wctor (A. B). G = granule layer; H = hilus: HF= hippocampal fissure; ibl = inner blade of hippocampal hilus; obl= outer blade of hippocampal hilus: f’= stratum pyramidale; SO = stratum oriens: SR= stratum radiatum. Calibration bars: A, B = 500 pm; C. D= SO pm.
staining (Fig. I B ). This was true when immunocytochemistry was performed using horse anti-mouse second antisera that had been preabsorbed with rat serum. As previously noted by Vass et al. (57). some injured neurons appear to take up immunoglobulins (and other proteins), and probably because of cross-reactivity of rodent immunoglobulins, incubation with anti-mouse second antisera which have not been preabsorbed can produce nonspecific cellular staining in injured neurons and other cells (57). A diffuse brown staining was evident in some regions where there was frank cell necrosis. The light staining in the hippocampus in Fig. IB from a section not exposed to primary antibody (particularly CA I and CA3) is an example of this nonspecific artifact. This diffuse reaction product is likely produced by endogenous peroxidase activity present in injured tissue (18), which can be decreased by pretreating sections with dilute solutions of hydrogen peroxide prior to immunocytochemistry. Finally, sections of brains of control subjects not subjected to ischemia did not exhibit either background staining or specific HSP72 cellular staining. Following global ischemia, HSP72-like immunostaining ranged from virtually no staining (left hippocampus, Fig. IA) to intense staining of all hippocampal regions (right hippocampus. Fig. IA; E. F). Aside from the absence of staining observed in 4 hippo-
campi following 5 min bilateral carotid occlusions, HSP72-like immunostaining occurred in 4 patterns that roughly correlated with the duration of ischemia. The first pattern. with the least HSP72 induction, was observed in 4 hippocampi following 5 min occlusions. HSP72LI was induced only in CAla pyramidal neurons (Figs. IC: IA-D). These neurons were easily identifiable because of their stained dendrites, axons. and perikarya (Fig. ?A-D). They were clustered into columns approximately 0.2 mm wide separated by columns of similar size in which HSP72 was not induced (Figs. IC; 2A). A single column of CAla pyramidal cells was induced in one subject (Fig. 2B). Isolated immunostained pyramidal neurons were occasionally found adjacent to these columns of HSP72 induction (Fig. 2C. D). HSP72 was not induced in pyramidal neurons of regions CA1 b and CAlc or in hippocampal plia in these subjects. The second pattern was observed in two hippocampi of rats subjected to a S-min carotid occlusion. HSP72LI was induced in most, if not all. pyramidal neurons of CA1 (Figs. ID; 3A. B). Marked induction was also observed in some cells within CA4 (Fig. ID: 3C. D). This hippocampal region, which is also referred to as the hilus of the dentate gyrus. contains neurons of different morphologies (2). HSP72 was induced in the modified
FIG. 5. The fourth pattern of HSP724ike immunor~activity crbserved foliowing 10 minutes carotid occlusions. Immunostaining was evident in many blood vessels of both the CAI sector (upper portion) of the hippocampuv and the dentnre gyros (lower portion). The open arrow points to one of these vessels crossing the stratum radiatum (SR), and the filled array point to necrotic pyramidal cells. DG = dentate gyrus; ibl = inner blade of hilus; P = stratum pyramidale: SLM = stratum lacunosurn!moleculare; SO = stratum oriens. Calibration bar= 200 km.
STRESS PROTEINS INDUCED BY ISCHEMIA
pyramidal (Fig. 3C) and in hilar neurons. A few immunostained pyramidal neurons extended into sector CA3c (Fig. 3A. C, indicated by asterisks). Some HSP72 induction also occurred in granule cells in the inner blade of the dentate gyrus (Fig. 3A, D. upper border). The third pattern occurred in 10 hippocampi. 2 after 5 min occlusions. and 8 after IO min occlusions. Intense HSP72 immunostaining was present in most pyramidal neurons of sector CA3 (Figs. IE. F: 4). and in granule cells of the dentate gyrus (Figs. IE, F: 3A. C). particularly those of its inner blade (Fig. 4A). HSP72 immunostaining was evident in individual astrocytes of the inner blade of the dentate gyrus. and in scattered pyramidal neurons of all regions of CAI as well (Fig. 4B, C). However. these pyramidal neurons were not organized in columns, and their numbers were significantly less than in the first pattern. The CAI region also exhibited moderate to intense fragmentation of the neuropil (unstained holes. Figs. IE: JA), suggesting necrosis of cells in this region (18). Finally, between areas CA1 and CA3 there was a region about 0.5 mm wide, corresponding to sector pyramidal neurons (Fig. CA2. that had few immunostained 4A. B). The fourth pattern was observed in 2 hippocampi following 10 minutes of ischemia. No HSP72-positive neurons or astrocytes were observed but many immunostained blood vessels were present through all the layers of the hippocampus and dentate gyrus (Fig. 5). The spacing and radial orientation of most stained hippocampal vessels was reminiscent of the columns of pyramidal cells observed in pattern I. There were also necrotic pyramidal cells with apparent cystic lesions in all sectors of the hippocampus proper (solid arrows. Fig. 5). The distribution of HSP77 immunostaining in neocortex, striatum and other forebrain regions will be the subject of a separate report. DISCUSSION
This study shows that ischemia induces HSP72LI in pyramidal neurons of the CA1 . CA3 and CA4 sectors of the hippocampus. as well as in hilar intemeurons. dentate gyrus granule cells, and occasionally in glia. These data confirm previous results showing that HSP synthesis in the brain is markedly increased following global ischemia (16, 32, 33. 56). This increase takes place even though global ischemia causes a decrease in total protein synthesis in the brain (5. 15. 22. 55). Vass et al. (56) first examined cellular HSP72 induction in the brain following global ischemia with the same antibody used in this study. They reported that following 10 min bilateral carotid occlusions in the gerbil, HSP72LI was first detected in some dentate granule cells 4 h after ischemia. At 24 h postischemia, HSP72 induction peaked in these cells and a few scattered immunostained pyramidal cells were evident in CA1 and CA3. At 48 h. while a few stained cells were still present in the dentate gyms and CA I, immunoreactivity in CA3 was at its peak, where it continued to be evident for up to 96 h after ischemia. A comparison of their results and ours is difficult because we used a different species and model of global ischemia. did not study the time course of HSP72 induction, and examined the effects of carotid occlusions of different duration. However, while differences between the two studies exist. both essentially agree on the hippocampal regions and the types of cells that express HSP72LI 18-24 h after ischemia. Localtarion pul
oj’ HSP72-Like
Immunoreactivity
in the Hippocan-
Formation CA1
.vector.
We observed
that HSP72 can be induced in col-
241
umns of CAla pyramidal neurons when induction is absent in other hippocampal regions (Fig. 2). In addition. HSP72 can be induced in most pyramidal neurons throughout CA1 with sparing of most other regions (Fig. 3). Vass et al. (56) did not observe either phenomenon following global ischemia in the gerbil. This could be due to differences in the duration of ischemia used in the two studies or to species differences. Vass et al. (56) did observe scattered immunostained pyramidal neurons throughout CA I similar to those seen in the third pattern of this study. Silver impregnation studies of degenerating neurons following 5 min carotid occlusions in the gerbil have shown differences between region CAla and regions CAIb and c. While argyrophilic neurons appear sooner in region a, regions b and c exhibit the most intense neuronal degeneration in the hippocampal formation ( I I ). The columnar induction of HSP72 in neurons of CAla (Figs. IC, and 2) cannot be explained on a neuroanatomicai or functional basis (53.54). However. traditional staining techniques have demonstrated selective survival of islands of CAI pyramidal neurons following global ischemia using a four vessel occlusion model (5) and following cardiac arrest (4). It is possible that during episodes of moderate ischemia blood flow differences in the radially oriented individual blood vessels of CA1 a (9) cause HSP72 induction in columns of pyramidal cells. The fact that CA1 sustains a selective and marked decrease of blood flow compared to other hippocampal sectors during diffuse ischemia supports this po\sibility (49). CA3 sec’tor. HSP72 was induced in the CA3 sector in most subjects. In agreement with results of Vass et al. (56) CA3 neurons exhibited the greatest degree of HSP72 induction in the hippocampal formation (Fig. I). In the gerbil CA3 neurons sure ive IO min ischemia. (23) and in the rat are “resistant’* to global ischemia following cardiac arrest (4) or 20-30 nun four vessel occlusion (26. 36. 37, 39). Because CA3 cells are resistant and exhibit maximal expression of HSP72. Vass et al. (56) hypothesized that transient ischemic insults induce HSP72 primarily in neurons destined to survive. However, Smith et al. (49) using the same model used in this study showed that though CAI and CA4 neurons are more vulnerable. many CA3 pyramidal neurons are destroyed following IO min of systemic hypotension and bilateral carotid occlusion. Therefore. it is possible that some immunostained CA3 neurons of this study were destined to die. Several hypotheses may explain why HSP72 is markedly induced in CA3 pyramidal neurons. (a) CA3 pyramidal neurons survive the ischemic insult better than other cells (25.37): therefore they may be able to synthesize proteins for a longer period than other cells (5. 22. 55). (b) Since it has been shown in vitro that cells differ in their capability to synthesize HSP72 (10,28), CA3 cells may have a greater intrinsic ability to synthesize HSP72. (c) Finally. if intracellular Ca’ ’ concentration is one of the signals that initiates HSP72 synthesis, the fact that CA3 pyramidal cells may have a greater influx of calcium or less binding of intracellular calcium than other neurons (27.47 1may enhance HSP72 synthesis. This preferential CA3 HSP72 induction is not due to greater vulnerability. since CA I and CA4 are more vulnerable to ischemic injury than CA3 (11. 37. 38. 44, 39). or to regional distribution of excitatory amino acid receptors. since these are distributed uniformly among hippocampal pyramidal neurons ( 19, 30. 31). CA4 sector (hilus) and dentate gvus. HSP77 was also induced in sector CA4 neurons in most subjects. The rat hilus contains several types of cells, including modified pyramidal. basket and mossy neurons (2). Some of the immunostained neurons show these morphologies (Fig. 3C) and could be GABA (I 7.62) or cholecystokinin (20) containing cells. However. it is unlikely that they are xomatostatin containing neurons since hilar somatostatin cells are particularly vulnerable to global ischemia and may be
24x
TABLE 1 SUMMARY OF HSP72 IMMUNOSTAINING IN THE HIPPOCAMPI OF ADULT RATS SUBJECTED TO GLOBAL ISCHEMIA PRODUCED BY A COMBINATION OF TEMPORARY HYPOTENSION COMBINED WITH TEMPORARY CAROTID OCCLUSION ___. _~I_.__._.._.._. --
Pattern of HSP72 Staining No Staining CA 1a neurons CAla, b, c; CA4 neurons CA3, CAla-c, CA4, dentate granule neurons CA3, CAla-c, CA4, dentate granule neurons Hippocampai and dentate vessels
lost prior to CA1 rteurons (20). HSP72 induction in this region is not described by Vass et al. (56). This discrepancy may be due to a species difference since many, but not all, hilar basket cells in the gerbil die rapidly following carotid occlusion (11) and may not express HSP72. HSP72 was induced in granule neurons of the dentate gyrus in subjects exhibiting patterns 2 and 3 in this study (Figs. 1, 2, 3) and in all gerbils of Vass et al.‘s study (56). The fact that HSP72 i~unos~ning in granule cells occurred only after the protein was expressed in virtually all other hippocampal regions and was absent after moderate &hernia suggests that these cells require severe degrees of ischemia to induce the HSP72 protein. Dentate granule neurons are among the hippocampal neurons most resistant to global ischemia (11, 37, 38, 44, 49, 60, 61). Hippocampal blood vessels. HSW2 was induced in the blood vessels but not in neurons or glia of two hippocampi and dentate gyri following 10 min occlusions (Fig. 5). This induction is compatible with previous results showing ind~tion of HSPs in cerebral vessels in vivo and in vitro ( 1259). We hypothesize that the neurons and glia of these hippocampi were lethally injured shortly after global ischemia and rendered unable to synthesize HSP72. The fact that CA1 neurons appeared cystic (solid arrows, Fig 5) supports this assumption. We also hypothesize that endothelial cells were more resistant to ischemia and were able to synthesize HSP72 after neurons and glia were unable to do so.
Most studies of global ischemia have shown that CA1 pyramidal neurons and perhaps certain subpopulations of neurons in CA4 are the most vulnerable, CA3 neurons have intermediate resistance, and dentate granule cells are the most resistant to global ischemia (4, 20, 23, 26, 37-39, 43, 44, 51, 52). Significantly, Smith et al. (49), who developed the model used in this study, agree with this ranking of the vuInerabili~ of hippocampa1 cells. Our results suggest that induction of HSP72 in hip~ampus occurs in this order. However, the relationship is imperfect possibly because of the variability of the model used, and because severely injured cells may not be able to synthesize stress proteins. HSP72 as a Marker of Cell Injury Our suggestion that HSP72 immunostaining labels stressed cells is not supported by the fact that more CA1 neurons immunostained following 5 min compared to 10 min occlusions, or by the two instances that CA1 cells did not stain at dl foilowing 10 min occlusions. Nonetheless, the fact that cells exhibiting HSWZLI occur in areas selectively vulnerable to ischemia, as discussed above, supports the idea that HSP72 expression occurs in stressed
No. of ~ippocampi
Duration of Ischemia
4 4 2 2 8 2
5 min 5 min 5 min 5 min 10 min 10 min
or injured cells. The exception to this may occur when cells are so severely injured that transcription or translation of heat shock genes is impaired (15, 22, 55). Since HSPs enhance the resistance of cells to lethal stimuli (1,28), it could be maintained that cells expressing the protein have a greater probability of surviving. This study and previous work of Vass et al. (56) show that HSP72 expression was greatest in CA3 pyramidal neurons destined to survive. On the other hand, we have shown that neurons within brain areas infarcted by complete middle cerebral artery occlusions clearly destined to die may or may not express HSP72 ( 18). The fact that HSP72 can be induced both in cells destined to die or survive has also been documented in cultured cells (28.57). Therefore, it is not possible to answer the question of whether cells that express HSP72Ll are destined to survive or to die. Selective Vu~nerabil~~ in Relation to the Excitutox~c Hypothesis of ~sc~~i~ Damage
It has been hypothesized that excitatory ammo acids (EAAs) may mediate cell death in the hippocampus (7, 21, 34, 45, 60, 61) and other regions and may be the cause of selective cellular vulnerability to ischemia (3, 8, 2 1, 34, 40, 41, 45, 60). Glutamate and aspartate (1950) may mediate cell death via intracellular accumulation of calcium and/or chloride (6, 40-42). While the EAA hypothesis may explain the vulnerability of the whole hip~ampal formation to global isehemia, it does not explain why ischemia induces HSP72 selectively in CA1 p~~dal neurons (Table I), why CA1 pyramidal neurons are more vulnerable than CA3 pyramidal neurons to ischemia (4, 20, 23-26, 37, 39, 43, 44, 51, 52). or why CA3 pyramidal neurons are more vulnerable than CA1 pyramidal neurons to damage caused by stimulation of the perforant path (46,48). All pyramidal neurons receive excitatory amino acid inputs (46, 60-62). Additional factors may determine the vulnerability of particular subpopulations of cells. For example, the induction of HSP72 during ischemia and sus~ptibility of CA1 cells to global ischemia may be related to regional differences of blood flow in the hip~c~pus (9); and the susceptibility of CA4 and CA3 neurons to damage caused by perforant path stimulation may be due to their greater electrical excitability produced by differences in calcium entry or binding (46-48) or to other unknown factors. ACKNOWLEDGEMENTS
We wish to thank Dr. Stephen M. Sagar for his helpfu1 ccunmettts on this manuscript, and Pat Jasper. Theresa Marsh, and Matt Morton fur excellent technical assistance. This work was supported by NIH grants NS24666 and NS28167 tu F.R.S., NS01424 to D.L., and NS24728 to R.S., and by the VA Medical Research Service (F.R.S.) and the WCSF Academic Senate (D.L.).
STRESS PROTEINS
INDUCED
BY ISCHEMIA
249
REFERENCES I. Barbe. M. F.: Tytell, M.: Cower, D. J.; Welch, W. J. Hyperthermia protects against light damage in the rat retina. Science 241: 18171820; 1988. 2. Bayer. S. A. Hippocampal region. In: The rat nervous system, vol. I. In: Paxinos, G., ed. Sydney: Academic Press; 1985:335-352. 3. Benveniste, H.; Drejer, J.; Schousboe, A.; 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: 1369-1374; 1984. 4. Blomqvist, P.; Wieloch, T. Ischemic brain damage in rats following cardiac arrest using a long-term recovery model. J. Cereb. Blood Flow Metab. 5:42043 I; 1985. 5. Bodsch. W.; Takahashi, K.: Barbier. A.; Grosse Ophoff. B.: Hossmann. K.-A. Cerebral protein synthesis and ischemia. Prog. Brain Res. 63~237-259: 1985. 6. Choi. D. W. Ionic dependence of glutamate neurotoxicity. J. Neurosci. 7:369-379; 1987. 7. Clark, G. D.; Rothman. S. M. Blockade of excitatory amino acid receptors protects anoxic hippocampal slices. Neuroscience 21:665671; 1987. 8. Collins, R. C. Selective vulnerability of brain: new insights from the excitatory synapse. Metab. Brain Dis. I:23 l-240; 1986. 9. Coyle, P. Vascular patterns of the rat hippocampal formation. Exp. Neurol. 52:447i158: 1976. IO. Craig, E. A. The heat shock response. CRC Crit. Rev. Biochem. 18:239-280; 1985. I I. Crain. B. J.: Westerkam. W. D.; Harrison, A. H.; Nadler, J. V. Selective neuronal death after transient forebrain ischemia in the Mongolian gerbil: a silver impregnation study. Neuroscience 27:3871102: 1988. 12. Currie, R. W .; White, F. P. Trauma-induced protein in rat tissues: a physiological role for a “heat shock” protein. Science 214:72-73: 1981. 13. Deshaies. R. J.; Koch, B. D.; Schekman. R. The role of stress proteins in membrane biogenesis. Trends Biol. Sci. 13:384-388; 1988. 14. Deshaies, R. J.: Koch. B. D.: Werner-Washbume, M.: Craig, E. A.; Schekman. R. A subfamily of stress proteins facilitates translocation of secretory and mitochondrial precursor polypeptides. Nature 332: 800-805; 19X8. 15. Dienel. G. A.; Cruz. N. F.; Rosenfeld, S. J. Temporal profiles of proteins responsive to transient ischemia. J. Neurochem. 44600-610; 1985. 16. Dienel. G. A.; Kiessling, M.; Jacewicz, M.; Pulsinelli, W. A. Synthesis of heat shock proteins in rat brain cortex after transient ischemia. J. Cereb. Blood Flow Metab. 6:505-510; 1986. 17. Gamrani, H.; Onteniente. B.; Seguela, P.: Geffard, M.; Calas. A. Gamma-aminobutyric acid immunoreactivity in the rat hippocampus. A light and electron microscopic study with antiGABA antibodies. Brain Res. 364:30-38: 1986. 18. Gonzalez, M. F.; Shiraishi, K.; Kisanaga, K.; Sagar, S. M.; Mandabach. M.; Sharp, F. R. Heat shock proteins as markers of neural injury. Mol. Brain Res. 6:93-100; 1989. 19. Greenamayre. J. T.: Olson, J. M.; Penney, J. B.; Young, A. B. Autoradiographic characterization of N-methyl-D-aspartate-, quisqualateand kainate-sensitive glutamate binding sites. J. Pharmacol. Exp. Ther. 233:254-263: 1985. 20. Johansen, F. F.; Zimmer, J.; Diemer, N. H. Early loss of somatostatin neurons in dentate hilus after cerebra1 ischemia in the rat precedes CA-I pyramidal cell loss. Acta Neuropathol. (Berl.) 73:110-l 14; 1987. 21. Jorgensen. M. B.; Johansen, F. F.; Diemer, N. H. Removal of the entorhinal cortex protects hippocampal CA-1 neurons from ischemic damage. Acta Neuropathol. (Berl.) 73:189-194; 1987. 22. Kiessling, M.; Dienel, G. A.; Jacewicz, M.; Pulsinelli, W. A. Protein synthesi\ in postischemic rat brain: a two-dimensional electrophoretic analysis. J. Cereb. Blood Flow Metab. 6:642649; 1986. 23. Kirino, T. Delayed neuronal death in the gerbil hippocampus following ischemia. Brain Res. 239:5749; 1982. 24. Kirino. T.; Sano, K. Selective vulnerability in the gerbil hippocam-
pus
25.
26.
27.
28. 29.
30.
31.
32. 33.
34.
35. 36.
37. 38.
39.
40. 41.
42. 43.
44.
45.
46.
47.
48.
following transient ischemia. Acta Neuropathol. (Berl.) 62:201208; 1984. Kirino, T.: Sane, K. Fine structural nature of delayed neuronal death following ischemia in the gerbil hippocampus. Acta Neuropathol. (Berl.) 62:209-218; 1984. Kirino. T.; Tamura. A.: Sano. K. Delayed neuronal death in the rat hippocampus following transient forebrain ischemia. Acta Neuropathol, (Berl.) 64:139-147; 1984. Kudo, Y.: Ogura, A. Glutamate-induced increase in intracellular Ca” concentration in isolated hippocampal neurones. Br. J. Pharmacol. 89:19l-198: 1986. Lindquist. S. The heat-shock response. Annu. Rev. Biochem 55: 1151-l 191; 1986. Lorente de No R. Studies on the structure of the cerebral cortex. II. Continuation of the study of the ammonic system. J. Psychol. Neuml. (Leipzig) 46:1 13-177: 1934. Monaghan. D. T.: Cotman, C. W. Distribution of ,Y-methyl-D-aspartate-sensitive L-[‘HIglutamate-binding \ites in rat main. J. Neut-osci. 5:2909-2919; 1985. Monaghan. D. T.; Holets. V. R.: Toy. D. W.; Cotman. C. W Anatomical distributions of four pharmacologically distinct ‘H-L.-glutamate binding sites. Nature 306: 176-l 79: 1983. Nowak, T. S. Synthesis of a stress protein following transient ischemia in the gerbil. J. Neurochem. 45:16351641: 19X5. Nowak. T. S.; Marini. A. M.; Kozuka, M. Differential localization of 70 kDa stress protein induction in brain after i\chemia and hyperthermia. J. Cereb. Blood Flow Metab. 9:S54: 1989. Onodera, H.; Sato. G.; Kogure, K. Le\iona to Schaffer collaterals prevent iachemic death of CAI pyramidal cells. Neurosci. Lett. 68: 169-17-l; 1986. Pelham. H. R. B. Speculations on the functions of the major heat shock and glucose-regulated proteins. Cell 46:959-96 I: 1986. Petito. C. K.: Pulsinelli. W. A. Delayed neuronal recovery and neuronal death in rat hippocampus following severe cerebral ischemia: possible relationship to abnormalities in neuronal processe\. J. Ccreb. Blood Flow Metab. 4:194-205: 1984. Pulsinelli. W. A. Selective neuronal vulnerability: morphological and molecular characteristics. Prog. Brain Res. 63:29-37; 1985. Pulsinelli. W. A.: Brierley, J. B.: Plum. F. Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann. Neurol. ll:491~98: 1982. Pulsinelli, W. A.; Levy, D. E. Duffy, T. E. Regional cerebral hlood flow and glucose metabolism following transient forehram ischemia. Ann. Neurol. Il:499-509: 1982. Rothman. S. M.: Olney, J. W. Glutamate and the pathophysiology of hypoxic-ischemic brain damage. Ann. Neurol. 19: 105-l II: 1986. Rothman. S. Synaptic release of excitatory amino acid neurotransmitter mediates anoxic neuronal death. J. Neurosci. 4:1884-1891: 1984. Rothman. S. M. The neurotoxicity of excitatory amino acids is produced by passive chloride influx. J. Neurosci. 5: 1483-1489; 1985. Schlesinger, M. J.; Ashbumer. M.: Tissierea. A. Heat shock from bacteria to man. Cold Spring Harbor. NY: Cold Spring Harbor Laboratory; 1982. Schmidt-Kastner, R.; Hossmann, K.-A. Distribution of ischemic neuronal damage in the dorsal hippocampus of rat. Acta Neuropathol. (Berl.) 76:4ll-t21; 1988. Simon, R. P.; Swan. J. H.; Griffiths. T.: Meldrum, B. S. Blockade of N-methyl-D-aspartate receptors may protect against ischemic damage in the brain. Science 226:850-852; 1984. Sloviter. R. S. “Epileptic” brain damage in rats induced by sustained electrical stimulation of the perforant path. I. Acute electrophysiological and light microscopic studies. Brain Res. Bull. IO: 675-697; 1983. Sloviter. R. S. Calcium-binding protein (Calbindin-DZXI) and parvalbumin immunocytochemistry: Localization in the rat hippocampus with specific reference to the selective vulnerability of hippocampal neurons to seizure activity. J. Comp. Neurol. 280:183-196: 1989. Sloviter. R. S.; Damiano. B. P. Sustained electrical stimulation of
49.
50.
5 1.
52.
53.
54.
55.
the perforant path duplicates kainate-induced electrophysiological effects and hippocampal damage in rats. Neurosci. Lett. 24:279-284; 1981. Smith, M.-L.; Auer, R. N.; Siesjo, B. K. The denstty and distribution of ischemic brain injury in the rat following 2-10 min of forebrain ischemia. Acta Neuropathol. (Berl.) 64:319-332: 1984. Storm-Mathisen, J. Localization of transmitter candidates in the brain: the hippocampal formation as a model. Prog. Neurobiol. 8:119-l8 1; 1977. Suzuki, R.; Yamaguchi, T.; Kirino, T.; Orzi, F.; Klatzo, 1. The effects of 5-minute ischemia in Mongolian gerbils: I. Blood-brain barrier, cerebral blood flow, and local cerebral glucose utilization changes. Acta Neuropathol. (Berl.) 60:207-216; 1983. Suzuki, R.; Yamaguchi, T.; Li, C.-L.; Klatzo. I. The effects of 5minute ischemia in Mongolian gerbils: II. Changes of spontaneous neuronal activity in cerebral cortex and CA1 sector of hippocampus. Acta Neuropathol. (Berl.) 60:217-222; 1983. Swanson, L. W.; Cowan, W. M. An autoradiographic study of the organization of the efferent connections of the hippocampal formation in the rat. J. Comp. Neurol. 172:49-84; 1977. Swanson, L. W.; Wyss, J. M.; Cowan, W. M. An autoradiographic study of the organization of intrahippocampal association pathways in the rat. J. Comp. Neurol. 181:681-716; 1978. Thilmann, R.; Xie. Y.; Kleihues. P.: Kiessling. M. Persistent inhi-
56.
57.
58.
59.
60. 61.
62.
bition of protein synthesis precedes delayed neuronal death m posttschemic gerbil hippocampus. Acta Neuropathol. (Berl. I 71:889~. 1986. Vass, K.; Welch, W. J.; Nowak, 7’. S. Localization of 70-kDa strebs protein induction in gerbil brain after ischemia. Acta Neuropathol. (Berl.) 77:128-135; 1988. Vass, K.; Berger, M. L.; Nowak, T. S., Jr.; Welch, W. J.; Lassmann. H. Induction of stress protein HSP70 in nerve cells after status epilepticus in the rat. Neurosci. Lett. 100:259-264; 1989. Welch, W. J.; Suhan, J. P. Cellular and biochemical events in mammalian cells during and after recovery from physiological stress. J. Cell Biol. 103:2035-2052; 1986. White, F. P. Differences in protein synthesized in viva and in vitro by cells associated with the cerebral vasculature. A protein synthesized in response to trauma. Neuroscience 5:1793-1799; 1980. Wieloch, T. Neurochemical correlates to selective neuronal vulnem bility. Prog. Brain Res. 63:69-85; 1985. Wieloch, T.; Lindvall, 0.; Blomqvist, P.; Gage, F. H. Evidence for amelioration of ischaemic neuronal damage in the hippocampal formation by lesions of the perforant path. Neurol. Res. 7:24-26; 1985. Woodson. W.; Nitecka, L.; Ben-Ari. Y. Organization of the GABAergic system in the rat hippocampal formation: a quantitative immunocytochemical study. J. Comp. Neurol. 280:254-271: 1989.
