Tourniquet-Induced Cerebral Ischemia in the Rat. Dasa Cizkova,1,4 Ivo Vanicky,1 Toshizo Ishikawa,2 and Martin Marsala3. Received April 16, 1999; accepted ...
Cellular and Molecular Neurobiology, Vol. 20, No. 3, 2000
Time Course of Brain Neuronal Degeneration and Heat Shock Protein (72) Expression Following Neck Tourniquet-Induced Cerebral Ischemia in the Rat Dasa Cizkova,1,4 Ivo Vanicky,1 Toshizo Ishikawa,2 and Martin Marsala3 Received April 16, 1999; accepted June 24, 1999 SUMMARY 1. The present study was designed to examine the regional expression of HSP72/73 protein after a 7.5-min period of cerebral ischemia and to compare the distribution of HSP neurons with the localization of irreversible neuronal degeneration as analyzed by silver impregnation technique. 2. During 6–24 hr after cerebral ischemia clear-cut neuronal argyrophilia developed in several brain regions including the hippocampal hilus, nucleus reticularis thalami, and colliculi inferiores. With the exception of the hippocampal hilus, the structures which showed silver impregnability were HSP72 negative at 6–24 hr. 3. Despite the clear HSP72 expression seen in hippocampal CA1 neurons, a significant loss of these neurons was seen at 7 days after ischemia. 4. These data show that in some structures the presence of HSP72 is indicative of higher resistance of these neurons to ischemia-induced degeneration, however, the process of delayed neuronal degeneration appears to be independent of the accelerated synthesis of HSP72 seen during the early period of reflow. KEY WORDS: cerebral ischemia; silver impregnation; HSP72 expression; neuronal degeneration.
INTRODUCTION During the past several years considerable experimental data have accumulated which show that, depending on the completeness and the duration of cerebral ischemia, there is a significant upregulation of a variety of proteins encoding genes including FOS, JUN, and heat shock protein (HSP) in specific brain regions during 1
Institute of Neurobiology, SAS, Soltesovej 6, 040 01 Kosice, Slovak Republic. The School of Allied Health Sciences, Yamaguchi University, 1-1-1 Minami-Kogushi, Ube, Yamaguchi 755-8554, Japan. 3 Anesthesiology Research Laboratory-0818, University of California, San Diego, 9500 Gilman Drive, San Diego, California 92093. 4 To whom correspondence should be addressed at Anesthesiology Research Laboratory-0818, University of California, San Diego, 9500 Gilman Drive, San Diego, California 92093. e-mail: cizkova@linux1. saske.sk 2
367 0272-4340/00/0600-0367$18.00/0 2000 Plenum Publishing Corporation
368
Cizkova, Vanicky, Ishikawa, and Marsala
reflow (Sharp et al., 1993; Kogure and Kato, 1993; Ikeda et al., 1994). While the precise role of HSP in the modulation of neuronal survivability after transient ischemic insult is not clear, it has been shown that under control conditions HSP plays an important role in the stabilization of unfolded proteins, mediating their proper folding, inhibiting aggregation, and facilitting transport to and across the membranes (Mestril et al., 1994). More importantly, it has also been shown in vivo that exposure of gerbil or rat brain to sublethal periods of ischemia will induce significant tolerance against an otherwise injurious interval of cerebral ischemia (Kato et al., 1991; Nishi et al., 1993) and that this state of ischemic tolerance covaries with the presence of HSP at the moment of exposure to lethal ischemic insult. In addition, it has been demonstrated that block of HSP synthesis (as achieved by treatment with protein synthesis inhibitors or intraventricular infusion of HSP72 antibodies) after application of preconditioning stressor will block the development of this ischemia-tolerant state (Nakata et al., 1993). The above characteristic on the functional properties of HSP would predict that if HSP represents a primary ‘‘rescue’’ mechanism, then the presence of HSP should be indicative of higher resistance of HSP expressing neurons against ischemia-induced degeneration. However, the temporal and the distributional profile of the appearance of HSP and the corresponding distribution of irreversible neuronal degeneration after single exposure to injurious intervals of cerebral ischemia are not consistent and differ among animal species (Kogure and Kato, 1993). For example, in the rat cerebral ischemia model it has been demonstrated that the expression of HSP72 (and HSP72 mRNA) is present not only in the areas which are resistant to global ischemia but also in the areas which show delayed degeneration (CA1 pyramidal cells) (Simon et al., 1991; Kawagoe et al., 1993; Tomioka et al., 1993). In contrast, in the gerbil cerebral ischemia model much lower immunoreactivity for HSP70 was seen in CA1 neurons following 10 min of ischemia (Vass et al., 1988). More recent studies using the same cerebral ischemia model in gerbils showed a persistent HSP70 mRNA upregulation during extended period of reflow, indicating that the translation block can account for the loss of protein immunoreactivity in this model (Abe et al., 1991; Nowak et al., 1990). While the mechanism accounting for such differences is not clear, there is indication that interspecies differences and the variability in the severity of ischemia in several brain regions among different models can, in part, account for these discrepancies. To address these questions, in the present study, we sought to employ a welldefined neck-tourniquet model of global cerebral ischemia in combination with systemic hypotension and to characterize the time course of HSP72 expression and corresponding presence of irreversible neuronal degeneration as assessed by the silver impregnation technique. In our previous study by using this model we have shown that a 7.5-min ischemic period causes (i) a reproducible pattern of neuronal injury restricted to a few extremely sensitive neuronal populations including the hippocampal hilar region, nucleus reticularis thalami, and colliculi inferiores and developing as early as 6 hr after reflow; and (ii) behavioral changes expressed as the presence of audiogenic seizures and persisting for several weeks to months after ischemia (Vanicky et al., 1997). Ischemic intervals shorter than 7 min had a minimal or no effect. Importantly, in previous studies using this ischemia model coupled
Global Cerebral Ischemia and HSP Expression
369
with the labeled butanol technique to measure regional cerebral blood flow (CBF), it has also been shown that at 5 min after ischemia there is a consistent reduction of blood flow (below 1%) in several brain regions including the cerebellum and pons (Siemkowicz and Gjedde, 1980). Similarly, using 14C-iodoantipyrine autoradiography no flow was detected in the forebrain and only up to 0.5 ml/100 g/min was measured in the pons and cerebellar structures (Siemkowicz, 1985; Singh et al., 1992). This tight grouping of a reproducible pattern of neuronal degeneration across the ischemic intervals suggests the reliability of this model in inducing critical reduction in blood flow in specific brain areas and indicates that this model represents an appropriate preparation to define the modulatory role (if any) of HSP in the development of irreversible neuronal degeneration after an injurious interval of global cerebral ischemia in the rat.
MATERIALS AND METHODS This study was performed with approval from the Institutional Animal Care and Use Committee of the University of California, San Diego. Induction of Cerebral Ischemia Transient global ischemia was induced in adult male Wistar rats (n ⫽ 11) (300–400 g). The rats were anaesthetized with 4% halothane in 100% O2, endotracheally intubated, and mechanically ventilated with 1–2% halothane in a 70% N2O/ 30% O2 mixture. To monitor arterial blood pressure (MABP) the tail artery was catheterized. For blood withdrawal and drug administration a polyethylene catheter (PE-50) was inserted into the femoral vein. Rectal and tympanic temperatures were maintained at 37 and 36⬚C, respectively, by using an underbody heating pad. After surgical preparation, animals were immobilized with pancuronium bromide (0.5 mg/kg) and the halothane concentration was decreased to 0.5%. MABP was then decreased to 40 mm Hg by i.v. administration of trimethaphan camsylate (Arfonad; 2 mg). At this moment the cuff, previously wrapped around the neck, was inflated to 1.1 atm and halothane was discontinued. During the ischemic period (7.5 min), the blood volume was manipulated to maintain the MABP below 40 mm Hg. Shortly before the end of ischemia, the venous blood was partially reinjected to increase the MABP (75 mm Hg), and afterward the tourniquet was deflated. Postischemia MABP was maintained at the preischemic level by reinfusion of the rest of the venous blood together with norepinephrine (0.1 ml, diluted 1 : 100,000). Arterial blood gasses and pH were checked between 10 and 30 min after reflow. During this period the rats were reanesthetized with 1% halothane, the catheters removed, and the skin incisions sutured. After animals regained spontaneous breathing, which was sufficient to keep the acid base in the physiological range, the animals were extubated. After extubation all animals were thermoregulated on an automatic heating pad until they recovered righting reflex. The animals were then allowed to survive for 6 hr (n ⫽ 3), 24 hr (n ⫽ 5), and 7 days (n ⫽ 3). In control, sham-operated animals
370
Cizkova, Vanicky, Ishikawa, and Marsala
(n ⫽ 2 for each survival period) all surgical preparations were performed as described, however, the cuff was not inflated.
Histological Preparation and Immunohistochemistry After appropriate survival periods all animals were deeply anesthetized with ketamine (80 mg/kg) and xylazine (20 mg/kg, i.p.) and transcardially perfused with saline for 1–2 min, followed by 4% paraformaldehyde in 0.1 M phosphate buffer (PB) (pH 7.4). Four hours after perfusion fixation, the brains were removed from the skull and immersed in fixative overnight at 4⬚C. The whole brains were then embedded in egg yolk (Vanicky, 1996) and frozen coronal sections (40 애m thick) were cut. Sections were processed for basic histological hematoxylin–eosin staining and additional selected sections were reacted for the inducible HSP72/73 using a monoclonal mouse antibody (Oncogene Science, U.S.A.). At the same time bulk parallel sections were also prepared and stored in the fixative for 7 days. These sections were impregnated using the silver impregnation method (Gallyas et al., 1980; Nadler and Evenson, 1983). Immunohistochemistry was performed using the avidin–biotin/horseradish peroxidase technique (Unitect Mouse ABC kit; Oncogene Science). Briefly, after washing with PBS, free-floating sections were incubated in 0.3% H2O2 for 30 min to block endogenous peroxidase activity. To block background staining, sections were placed in 0.1 M PBS (pH 7.4) containing 3% horse serum (HS), 0.2% Triton X-100, and 0.2% bovine serum albumin for 4 hr. This was followed by overnight incubation at 4⬚C with the HSP72/73 monoclonal antibody, diluted 1 : 400. The next day, sections were washed in PBS and incubated in biotinylated horse anti-mouse secondary antibody (1 : 200) for 2 hr. Bound specific antibodies reacted with the avidin–biotin peroxidase solution for 1 hr and were visualized using 0.05% 3,3,4,4diaminobenzidine hydrochloride (DAB) in 0.05 M Tris buffer containing 0.001% H2O2. Finally, all sections were dehydrated and coverslipped with Depex. To test the specificity of the immunolabeling procedure, in some sections, the primary antibody was replaced by normal horse serum.
Measurement of Cerebral Blood Flow Cerebral blood flow was measured in a separate group of animals (n ⫽ 5) by laser Doppler flowmetry (Periflux Master 4001; Perimed, Inc.). In each animal two craniotomies were carefully carried out under a stereomicroscope and two laser probes were placed, one over the parietal cortex and one over the cerebellar vermis, respectively. As this technique allows only assessment of relative flow differences, our experimental design was aimed to study the density of ischemia by comparing the intraischemic signal with that recorded after cardiac arrest. After 10 min of cerebral ischemia, cardiac arrest was induced by intracardial KCl injection, and the average values from three consecutive periods (preischemic, intraischemic, and circulation stop) were compared. The recordings were analyzed using the producer’s software (Perisoft).
Global Cerebral Ischemia and HSP Expression
371
Fig. 1. Polygraph tracing of the output of the laser Doppler probes positioned on the surface of the parietal cortex (A) or cerebellar vermis (B). Note the significant, near-complete loss of blood flow immediately after the induction of cerebral ischemia. The magnitude of the reduction is similar to that seen after induction of cardiac arrest.
Statistics Statistical analysis of physiological data was carried out by one-way ANOVA for multiple comparisons followed by the Dunnett post hoc test; P ⬍ 0.05 was considered significant.
RESULTS Blood Flow Changes (Figs. 1A and B) The average no-flow value (cardiac arrest period) was considered to be 0% related to the average of preischemic value representing 100% flow. In five animals, the average residual flow throughout a 7.5-min ischemic period represented 1.4 ⫾ 0.9 and 3.6 ⫾ 1.3% in the parietal cortex and cerebellar vermis, respectively. Consistently, in both regions the residual flow ceased within the first minutes of ischemia so that no demarcation could be seen between ischemia and complete loss of flow after cardiac arrest.
372
Cizkova, Vanicky, Ishikawa, and Marsala
Fig. 2. Distribution of the HSP72 immunoreactivity in coronal brain sections in a sham-operated rat (A) and a rat subjected to 7.5 min of global cerebral ischemia followed by 24 hr of recirculation (B–N). (A) In a sham-operated rat, no HSP72 expression was seen in any brain structure analyzed. Original magnification, ⫻2. (B) After ischemia and 24 hr of reflow, intense HSP72 immunoreac-
Global Cerebral Ischemia and HSP Expression
373
Fig. 2. (Continued) tivity in several brain structures [cortex, hippocampus, thalamus (thalamic nuclei), amygdala (amygdalar nuclei)] was seen. Original magnification, ⫻10. (C) Higher-power photomicrograph of the parietal cortex after ischemia and 24 hr of reperfusion. Note the presence of Golgi-like HSP72 immunopositivity in the pyramidal neurons in layers II and IV. Original magnification, ⫻40. (D) Higher-power photomicrograph of the cingular cortex after ischemia and 24 hr of reperfusion. Numerous HSP72-positive neurons in layer II were seen. Original magnification, ⫻40. (E) The basal lateral amygdaloid nucleus contained heavily HSP72-positive neurons of a variety of shapes including bipolar and star shape-like neurons. Original magnification, ⫻40. (F, G) The pyramidal layer of the hippocampal CA1 (F) and CA3 (G) sector revealed comparable HSP72 positivity of neuronal bodies and their processes. Original magnification, ⫻40. (H) Dentate hilus: numerous intrahilar neurons with a mossy or bipolar shape revealed dense HSP72 positivity, but with relatively modest HSP72 positivity seen in dentate granulae cells. Original magnification, ⫻30. (I) Ventrolateral posterior thalamic nucleus: HSP72-positive neurons were seen predominantly in the dorsomedial part. Original magnification, ⫻12. (J) Dorsolateral posterior thalamic nucleus: comparable HSP72 positivity as seen in the posterior thalamic nucleus, but with a more diffuse pattern, was seen. Original magnification, ⫻30. (K) Inferior colliculus: clear-cut HSP 72 immunoreactivity in the dorsomedial part can be seen. Note also the loss of HSP72 positivity in the ventral region. Original magnification, ⫻10. (L) Cerebellar cortex: a homogeneous HSP72 staining pattern in the granular cell layer, with a few densely stained Purkinje cell bodies (M), can be seen. Original magnification, ⫻3(L); ⫻40 (M). (N) Cochlear nucleus: clear-cut HSP72 positivity was seen. Original magnification, ⫻18.
374
Cizkova, Vanicky, Ishikawa, and Marsala
HSP Immunohistochemistry (Figs. 2A–N) In control sham-operated animals, no stainability for the HSP72/73 protein in any brain areas was observed (Fig. 2A). Analysis of the time course of HSP expression after 7.5 min of ischemia showed that the 24-hr survival period was the only time point when increased HSP expression in several brain regions was observed (Fig. 2B). At 6 hr or at 7 days of survival, comparable negativity for HSP as in the control animals was seen. In the following paragraphs a topographical description of HSP positivity in several brain regions (cerebral cortex, hippocampus, thalamus, inferior colliculus, and cerebellum) observed at 24 hr of reflow is provided. Cerebral Cortex (Figs. 2C and D). Induction of HSP72/73 was observed in different regions of the cerebral cortex. In the frontoparietal, parietal, and occipital cortex the HSP was induced in neurons in layers II–IV, with typical localization in the somatosensory area. In these regions the HSP-positive pyramidal neurons displayed a typical Golgi-like staining pattern, with clear outlining of neuronal bodies and their apical dendrites, however, HSP72/73 expression in neuronal cytoplasm was more intense than in neuronal nuclei. In contrast, layers II–IV in the motor cortex did not show induction of HSP. Expression of HSP in the cingulate gyrus was even more intense, particularly in the granular cells of layer II (Fig. 2D). The pattern of laminar expression of HSP in the piriform cortex was also present, however, the close area of the lateral amygdaloid nucleus contained scattered, densely stained neurons (Fig. 2E). Hippocampus (Figs. 2F–H). The HSP was expressed in the pyramidal neurons involving their dendrites throughout the entire CA1, CA2, and CA3 region. In the intrahilar area of the dentate gyrus, hilar cells along the superior and inferior margins of the dentate granule cells including the CA3 pyramidal neurons expressed the highest density of HSP staining, with completely stained cell bodies and their processes (Fig. 2H). However, in the dentate granule cells, only faint HSP immunoreactivity was seen. Thalamus (Figs. 2I and J). The HSP-positive neurons were disseminated predominantly in two thalamic nuclei; part of the ventrolateral thalamic nucleus (adjacent to the reticular thalamic nucleus) and part of the dorsolateral thalamic nucleus (Fig. 2J). Other parts of the thalamus were virtually unstained. Inferior Colliculus (Fig. 2K). There was a large number of HSP-positive neurons accumulated in the whole region, with particular localization in the dorsomedial part of the central and external nucleus. The only exception represented the ventral part of the central nucleus, where only a few positive neurons were observed, mostly on the periphery; however, the central region was unstained. Cerebellum (Figs. 2L and M). In the cerebellar area pronounced homogeneous immunoreactivity was seen in the granular layer of the cerebellar cortex, with occasional positive Purkinje cells. Regularly, strong immunopositivity was also observed in the ventral cochlear nucleus (Fig. 2N). The deep cerebellar nuclei were unstained. Silver Impregnation (Figs. 3A–C and 4A and B) In contrast to the above data on HSP expression, the time course and the regional distribution of neuronal argyrophilia or neuronal loss during reflow showed a different
Global Cerebral Ischemia and HSP Expression
375
pattern. In the following sections a description of these differences is provided with respect to the time course and the regional distribution of these changes. Six-Hour Survival. At this time point a few scattered argyrophilic neurons were seen in the hilar region of the hippocampus, mostly in the cells along the superior margin of the inner surface of the dentate granule cell layer (Fig. 3A). The argyrophilic pattern was also present in most neurons of the nucleus reticularis thalami (NRT), particularly in its ventral part (Fig. 3B). Thus, the argyrophilia which developed in this region strongly demarcated the nucleus from the surrounding structures. In the colliculi inferiores numerous impregnated neurons localized predominantly in the ventral part of the central nucleus were seen (Fig. 3C). The cortex revealed an unstained pattern. Twenty-Four-Hour Survival. In comparison with 6-hr survival the distributional pattern of neuronal argyrohilia in all brain regions analyzed was essentially unchanged. However, the number of necrotic cells appeared to be slightly increased. Seven-Day Survival. At this time point a dense argyrophilia in hippocampal CA1 sector was seen (Fig. 4A), which contrasted with the essentially complete lack of silver impregnability in control sham-operated animals (Fig. 4B). Similarly occasional argyrophilic Purkinje cells in cerebellum were detected. DISCUSSION In several early studies it was speculated that the expression of HSP in specific neuronal pools after ischemia either will be indicative of higher resistance of these
Fig. 3. Development of selective degenerative changes in the rat brain following 7.5 min of ischemia and 6 h of reperfusion visualized by the Nadler silver impregnation technique. Representative 40-애m coronal sections were taken from the anterior hippocampus (A), nucleus reticularis thalamus (B), and inferior colliculus (C). (A) Dentate hilus: heavily stained argyrophilic neurons in the hilar tip and along the superior margin of the dentate granule cells can be seen. Original magnification, ⫻30. (B) Nucleus reticularis thalami: high impregnability in the entire nucleus outlines this structure from the surrounding unstained area. Original magnification, ⫻30. (C) Inferior colliculus: dense silver impregnability in the ventral part of the central nucleus can be seen. Compare with the near-complete loss of HSP72 immunoreactivity at 24 hr of reflow (Fig. 1K). Original magnification, ⫻18.
376
Cizkova, Vanicky, Ishikawa, and Marsala
Fig. 4. (A) Development of neuronal degeneration in the hippocampal region after 7.5 min of ischemia and 7 days of survival: heavy somatodendritic argyrophilia in the pyramidal CA1 sector can be seen. Original magnification, ⫻10. (B) control nonischemic animal: a lack of silver impregnability in the hippocampal CA1 region can be seen. Original magnification, ⫻10.
neurons against ischemia-induced degeneration or can simply be used as a marker of neuronal injury resulting from injurious ischemic stress (Gonzalez et al., 1989). However, the principal question, whether or not the neurons which express HSP after ischemia will survive for an extended period of reflow, is still not clear. It has been shown, in the four-vessel occlusion model in the rat, that there is a region-specific expression of HSP72 in several brain structures at 24 hr after ischemia and that the magnitude (density) of this expression depends on the duration of the ischemic interval. Thus, shorter ischemic intervals (4–10 min) evoked clear HSP72 expression in the hilar interneurons, lateral thalamic nuclei, and CA1 sector of the hippocampus. After intermediate ischemic intervals (10–20 min) expression was observed in the CA3 hippocampal sector and in the cortical mantle, while after longer ischemia (30 min) the most dense expression was seen in dentate granule cells (Simon et al., 1991). Although this study provided a detailed definition on the ischemic threshold which is effective to induce HSP72 in brain, the question of whether or not these HSP expressing neurons are more resistant to degeneration after ischemia was not defined. In a more recent study, using the two-vessel occlusion model combined with systemic hypotension, it has similarly been demonstrated that after 10 min of global cerebral ischemia in the rat there was clear HSP70 immunostaining in neurons and glial cells in the cortex (most dense in layers 2 and 3), hippocampus, striatum, and basal ganglia at 24 hr after ischemia. Importantly, staining of adjacent sections with acid fuchsin, to detect irreversible degenerative changes, revealed that the majority (more then 90%) of neurons in layer 3 in the cortex which were HSP70 positive were acid fuchsin negative. These data suggest that neurons which are able to synthesize HSP72 after ischemia will likely survive (Gaspary et al., 1995). Correlative Analysis of HSP Expression and Irreversible Neuronal Degeneration Data from the present study show that at 6 hr of reflow following 7.5 min of global cerebral ischemia, there is a clear-cut development of heavy somatodendritic
Global Cerebral Ischemia and HSP Expression
377
argyrophilia (suggestive of irreversible degeneration) in several brain regions including the hilar region of the hippocampus, nucleus reticularis thalami (ventral part), and colliculi inferiores (ventral part of the central nucleus). No neuronal argyrophilia in CA1 pyramidal neurons at 6 or 24 hr after reflow was observed. Similarly, staining for an inducible form of heat shock protein (HSP72) at 24 hr after ischemia (but not at 6 hr) showed nonselective positivity in several brain regions including the sensory cortex, hippocampus (CA1, CA2, CA3, and hilus), cerebellum, ventrolateral and dorsolateral posterior thalamic nucleus, and colliculus inferior (dorsomedial part of the central and external nucleus) (see Table I for details). One significant observation obtained in the present study is that in some structures the development of neuronal argyrophilia was clearly present at 6–24 hr after reflow, while HSP was exclusively detected at 24 hr after ischemia. Although neuronal argyrophilia was still present at 24 hr, we were not able to do double staining in the same section and therefore it was impossible to confirm whether or not heavy neuronal argyrophilia corresponds with the loss of HSP stainability in the same neurons. In addition, no obvious correlation between the localization of argyrophilic neurons seen at 6–24 hr and the loss/presence of HSP positivity at 24 hr was observed (see Table I). First, in some structures (hippocampal hilus neurons) neuronal argyrophilia developed at 6 hr and these neuronal pools also showed HSP72 expression at 24 hr. Second, in other structures (nucleus reticularis thalami, colliculus inferior–ventral part of the central nucleus), neuronal argyrophilia was seen at 6–24 hr, however, no HSP72 was expressed at 24 hr or 7 days. Third, in the cerebral cortex and hippocampal CA1 sector, no neuronal argyrophilia was observed at 6–24 hr, however, clear expression of HSP72 was seen at 24 hr. Before any conclusion can be drawn based on the data from the present study, one obvious question which needs to be answered is the nature and significance of neuronal argyrophilia seen during the early period of reflow. Are the argyrophilic neurons going to survive or does this simply represent the first morphological sign of irreversible degeneration? In this regard, in our previous study, using the cardiac arrest model in the dog and the silver impregnation technique, we showed that at 8 hr after 15 min of cardiac arrest there is a development of Golgi-like neuronal argyrophilia, with corkscrew-like impregnated dendrites typically seen in the second and third neocortical layer. This initial somatodendritic argyrophilia is later followed by a gradual disintegration of neurons in the same areas, and at 7 days after ischemia, only drop-like argyrophilic fragments are seen (Vanicky et al., 1992). Similarly, after transient injurious spinal ischemia there is a rapid development (30 min–6 hr after ischemia) of neuronal argyrophilia in small and medium-sized interneurons, followed by complete loss of these neurons between 24 and 72 hr after reflow. This histopathological picture corresponds with fully developed spastic paraplegia (Marsala et al., 1989, 1994). Based on these data we believe that the neuronal argyrophilia seen in the present study at 6 hr after reflow represents irreversible degenerative changes which are followed by a gradual loss of these neurons. However, despite the fact that the majority of argyrophilic neurons is likely irreversibly damaged, both silver impregnability (at 6–24 hr) and HSP positivity (at 24 hr) were seen in the hippocampal hilus. It is possible that not all neurons in
b
a
⫺ ⫺ ⫹ ⫺ ⫹ ⫺
⫺ ⫺ ⫺ ⫺ ⫹ ⫺
CA1
⫺ ⫹ ⫺
⫺ ⫺ ⫺
CA2
⫺ ⫹ ⫺
⫺ ⫺ ⫺
CA3
Hippocampus
⫺ ⫹ ⫺
⫹ ⫹ ⫹
Hilus
⫺ ⫹ ⫺
⫺ ⫺ ⫺
VTN
⫺ ⫹ ⫺
⫺ ⫺ ⫺
DTN
Thalamusa
⫺ ⫺ ⫺
⫹ ⫹ ⫹
NRT
⫺ ⫹ ⫺
⫺ ⫺ ⫺
DM-CN
⫺ ⫹ ⫺
⫺ ⫺ ⫺
DM-EN
⫺ ⫺ ⫺
⫹ ⫹ ⫹
VNT-CN
Inferior colliculusb
⫺ ⫾ ⫺
⫺ ⫺ ⫾
Cerebellum (Purkinje cells)
VTN, ventrolateral posterior nucleus; DTN, dorsolateral posterior nucleus; NRT, nucleus reticularis thalami. DM-CN, dorsomedial part of the central nucleus; DM-EN, dorsomedial part of the external nucleus; VNT-CN, ventral part of the central nucleus.
Silver impregnation 6 hr 24 hr 7 days HSP72 positivity 6 hr 24 hr 7 days
Somatosensory cortex
Table I.
Global Cerebral Ischemia and HSP Expression
379
this region undergo acute degeneration and that a fraction of these neurons is capable of HSP synthesis. This can be associated with a higher resistance of this subpopulation of neurons against ischemic injury. In contrast, a fully developed neuronal argyrophilia was seen in the nucleus reticularis thalami and colliculus inferior (ventral part of the central nucleus) at 6 hr, however, no HSP72 was expressed in these regions at 24 hr (compare Fig. 3C vs. Fig. 2K and Fig. 3B vs. Fig. 2I). Based on the above commentary we believe that the majority of argyrophilic neurons seen in these areas at 6–24 hr of reflow is dead or in the process of irreversible degeneration and with significantly altered protein synthesis. Interestingly, in the NRT and the IC region extremely high HSP72 positivity in areas adjacent to irreversibly damaged neurons was observed. Comparable selective HSP72 expression in surviving neurons around the areas of necrosis after insulininduced hypoglycemia or after transient occlusion of the middle cerebral artery in the rat was reported (Li et al., 1993; Bergstedt et al., 1993). Although the role of selective regional vulnerability in the induction of HSP72 in these ‘‘penumbral’’ regions is not clear, it was hypothesized that an increased release of several ‘‘active factors’’ in the site of developing neuronal degeneration can play an active role (Bergstedt et al., 1993). We hypothesize that one of these ‘‘active factors’’ can be represented by an increased release of excitatory amino acids (such as glutamate) resulting from loss of local inhibition and/or leakage of these neurotransmitters from degenerating neurons. This would be in accord with the previous observation on the selective loss of small GABAergic neurons in IC after global cerebral ischemia and the presence of corresponding audiogenic seizures resulting from the loss of local inhibition (Kawai et al., 1995; Ribak and Morin, 1995). In addition, it has been demonstrated that intraventricular injections of kainic acid (a non-NMDA receptor agonist) is a potent stimulus for induction of HSP72 (Planas et al., 1994). We have observed similar HSP72 induction in spinal interneurons after intrathecal delivery of 1–3 애g of NMDA (unpublished observation). Delayed Neuronal Death in CA1 Hippocampal Neurons and HSP Expression Similarly as demonstrated in several previous studies, no degenerative changes in the hippocampal CA1 sector were seen during the early (6- and 24-hr) period of reflow and these neurons displayed a clear HSP72 immunoreactivity at 24 hr, suggesting sufficient recovery of protein synthesis at this time point. However, at 7 days, near-complete loss of these neurons was seen. In numerous previous studies it has clearly been established that delayed neuronal death typically occurs between 24 and 48 hr after global cerebral ischemia (Pulsinelli et al., 1982; Kirino et al., 1984). These data indicate that the presence of HSP72 has no protective effect and that other mechanisms are involved in the process of progressive degeneration in this neuronal population. One hypothetical possibility is that, in addition to the synthesis of ‘‘rescue’’ proteins, there is parallel upregulation of ‘‘killer’’ proteins, which can then mediate degeneration by initiating a process of programmed cell death. In this line, it has recently been shown that after global cerebral ischemia in the gerbil, there is positivity for single-stranded DNA protein (a marker of apoptotic cell death) in CA1 neurons at 24 hr after ischemia. In the same study it
380
Cizkova, Vanicky, Ishikawa, and Marsala
was also suggested that this process precedes the appearance of double-strand breaks (Maeda et al., 1998). However, in another study, using the in situ end-labeling technique to detect fragmented DNA, it has been shown that DNA fragmentation occurs after the development of delayed neuronal death after 10 min of global cerebral ischemia in the rat (Petito et al., 1997). Taking these data together suggests that mechanisms other then apoptosis are involved in the process of delayed neuronal death in CA1 neurons. In conclusion, the data from the present study show that after a 7.5-min period of global cerebral ischemia there is an early development of neuronal degeneration in specific brain regions including the hippocampal hilus, nucleus reticularis thalami (ventral part), and colliculi inferiores 6–24 hr after ischemia. With the exception of the hippocampal hilus, the structures which showed early degenerative changes showed no HSP72 positivity at any time examined. Although clear HSP72 expression was seen in hippocampal CA1 neurons, a significant loss of these neurons was seen at 7 days after ischemia. These data showed that in some structures the presence of HSP72 could (might) play a protective role against ischemia-induced neurodegeneration, while the process of delayed neuronal death seen in the CA1 sector appeared to be independent of the presence of HSP72 seen during the early period of postischemic reflow. REFERENCES Abe, K., Tanzi, R. E., and Kogure, K. (1991). Induction of HSP70 mRNA after transient ischemia in gerbil brain. Neurosci. Lett. 125:166–168. Bergstedt, K., Hu, B. R., and Wieloch, T. (1993). Initiation of protein synthesis and heat-shock protein72 expression in the rat brain following severe insulin-induced hypoglycemia. Acta Neuropathol. (Berl.) 86:145–153. Gallyas, F., Wolff, J. R., Bottcher, H., and Zaborszky, L. (1980). A reliable and sensitive method to localize terminal degeneration and lysosomes in the central nervous system. Stain. Technol. 55:299–306. Gaspary, H., Graham, S. H., Sagar, S. M., and Sharp, F. R. (1995). HSP70 heat shock protein induction following global ischemia in the rat. Brain Res. Mol. Brain Res. 34:327–332. Gonzalez, M. F., Shiraishi, K., Hisanaga, K., Sagar, S. M., Mandabach, M., and Sharp, F. R. (1989). Heat shock proteins as markers of neural injury. Brain Res. Mol. Brain Res. 6:93–100. Ikeda, J., Nakajima, T., Osborne, O. C., Mies, G., and Nowak, T. S. J. (1994). Coexpression of c-fos and hsp70 mRNAs in gerbil brain after ischemia: Induction threshold, distribution and time course evaluated by in situ hybridization. Brain Res. Mol. Brain Res. 26:249–258. Kato, H., Liu, Y., Araki, T., and Kogure, K. (1991). Temporal profile of the effects of pretreatment with brief cerebral ischemia on the neuronal damage following secondary ischemic insult in the gerbil: Cumulative damage and protective effects. Brain Res. 553:238–242. Kawagoe, J., Abe, K., and Kogure, K. (1993). Regional difference of HSP70 and HSC70 heat shock mRNA inductions in rat hippocampus after transient global ischemia. Neurosci. Lett. 153:165–168. Kawai, K., Penix, L. P., Kawahara, N., Ruetzler, C. A., and Klatzo, I. (1995). Development of susceptibility to audiogenic seizures following cardiac arrest cerebral ischemia in rats. J. Cereb. Blood Flow. Metab. 15:248–258. Kirino, T., Tamura, A., and Sano, K. (1984). Delayed neuronal death in the rat hippocampus following transient forebrain ischemia. Acta Neuropathol. (Berl.) 64:139–147. Kogure, K., and Kato, H. (1993). Altered gene expression in cerebral ischemia. Stroke 24:2121–2127. Li, Y., Chopp, M., Zhang, Z. G., Zhang, R. L., and Garcia, J. H. (1993). Neuronal survival is associated with 72-kDa heat shock protein expression after transient middle cerebral artery occlusion in the rat. J. Neurol. Sci. 120:187–194. Maeda, M., Sugiyama, T., Akai, F., Jikihara, I., Hayashi, Y., and Takagi, H. (1998). Single stranded DNA as an immunocytochemical marker for apoptotic change of ischemia in the gerbil hippocampus. Neurosci. Lett. 240:69–72.
Global Cerebral Ischemia and HSP Expression
381
Marsala, J., Sulla, I., Santa, M., Marsala, M., Mechirova, E., and Jalc, P. (1989). Early neurohistopathological changes of canine lumbosacral spinal cord segments in ischemia-reperfusion-induced paraplegia. Neurosci. Lett. 106:83–88. Marsala, M., Sorkin, L. S., and Yaksh, T. L. (1994). Transient spinal ischemia in rat: characterization of spinal cord blood flow, extracellular amino acid release, and concurrent histopathological damage. J. Cereb. Blood Flow. Metab. 14:604–614. Mestril, R., Chi, S. H., Sayen, M. R., O’Reilly, K., and Dillmann, W. H. (1994). Expression of inducible stress protein 70 in rat heart myogenic cells confers protection against simulated ischemia-induced injury. J. Clin. Invest. 93:759–767. Nadler, J. V., and Evenson, D. A. (1983). Use of excitatory amino acids to make axon-sparing lesions of hypothalamus. Methods Enzymol. 103:393–400. Nakata, N., Kato, H., and Kogure, K. (1993). Inhibition of ischemic tolerance in the gerbil hippocampus by quercetin and anti-heat shock protein-70 antibody. NeuroReport 4:695–698. Nishi, S., Taki, W., Uemura, Y., Higashi, T., Kikuchi, H., Kudoh, H., Satoh, M., and Nagata, K. (1993). Ischemic tolerance due to the induction of HSP70 in a rat ischemic recirculation model. Brain Res. 615:281–288. Nowak, T. S. J., Ikeda, J., and Nakajima, T. (1990). 70-kDa heat shock protein and c-fos gene expression after transient ischemia. Stroke 21:III107–III111. Petito, C. K., Torres-Munoz, J., Roberts, B., Olarte, J. P., Nowak, T. S. J., and Pulsinelli, W. A. (1997). DNA fragmentation follows delayed neuronal death in CA1 neurons exposed to transient global ischemia in the rat. J. Cereb. Blood Flow. Metab. 17:967–976. Planas, A. M., Soriano, M. A., Ferrer, I., and Rodriguez Farre, E. (1994). Regional expression of inducible heat shock protein-70 mRNA in the rat brain following administration of convulsant drugs. Brain Res. Mol. Brain Res. 27:127–137. Pulsinelli, W. A., Brierley, J. B., and Plum, F. (1982). Temporal profile of neuronal damage in a model of transient forebrain ischemia. Ann. Neurol. 11:491–498. Ribak, C. E., and Morin, C. L. (1995). The role of the inferior colliculus in a genetic model of audiogenic seizures. Anat. Embryol. (Berl.) 191:279–295. Sharp, F. R., Kinouchi, H., Koistinaho, J., Chan, P. H., and Sagar, S. M. (1993). HSP70 heat shock gene regulation during ischemia. Stroke 24:I72–I75. Siemkowicz, E. (1985). The effect of glucose upon restitution after transient cerebral ischemia: A summary. Acta Neurol. Scand. 71:417–427. Siemkowicz, E., and Gjedde, A. (1980). Post-ischemic coma in rat: Effect of different preischemic blood glucose levels on cerebral metabolic recovery after ischemia. Acta Physiol. Scand. 110:225–232. Simon, R. P., Cho, H., Gwinn, R., and Lowenstein, D. H. (1991). The temporal profile of 72-kDa heatshock protein expression following global ischemia. J. Neurosci. 11:881–889. Singh, N. C., Kochanek, P. M., Schiding, J. K., Melick, J. A., and Nemoto, E. M. (1992). Uncoupled cerebral blood flow and metabolism after severe global ischemia in rats. J. Cereb. Blood Flow. Metab. 12:802–808. Tomioka, C., Nishioka, K., and Kogure, K. (1993). A comparison of induced heat-shock protein in neurons destined to survive and those destined to die after transient ischemia in rats. Brain Res. 612:216–220. Vanicky, I. (1996). Rapid egg yolk embedding of whole rat brain for frozen or vibratome sectioning. J. Histotechnol. 19:143–144. Vass, K., Welch, W. J., and Nowak, T. S. J. (1988). Localization of 70-kDa stress protein induction in gerbil brain after ischemia. Acta Neuropathol. (Berl.) 77:128–135. Vanicky, I., Marsala, M., Orendacova, J., and Marsala, J. (1992). Silver impregnability of ischemiasensitive neocortical neurons after 15 minutes of cardiac arrest in the dog. Anat. Embryol. (Berl.) 186:167–173. Vanicky, I., Cizkova, D., Prosbova, T., and Marsala, M. (1997). Audiogenic seizures after neck tourniquetinduced cerebral ischemia in the rat. Brain Res. 766:262–265.
