Journal of Neurochemistry
Raven Press,Ltd., New York 0 1992 International Society for Neurochemistry
Regional and Temporal Alterations in Ca2+/CalmodulinDependent Protein Kinase I1 and Calcineurin in the Hippocampus of Rat Brain After Transient Forebrain Ischemia *tMotohiro Morioka, tKohji Fukunaga, TSetsuko Yasugawa, *Shinji Nagahiro, *Yukitaka Ushio, and tEishichi Miyamoto Departments of *Neurosurgery and ?Pharmacology, Kumarnoto University Medical School, Kumamoto, Japan
Abstract: We have investigated regional and temporal alterations in Ca2+/calmodulindependent protein kinase I1 (CaM kinase 11) and calcineunn (Ca2+/calmodulin-dependentprotein phosphatase) after transient forebrain ischemia. Immunoreactivity and enzyme activity of CaM kinase I1 decreased in regions CAI and CA3, and in the dentate gyrus, of the hippocampus early (6-12 h) after ischemia, but the decrease in immunoreactivity gradually recovered over time, except in the CA 1 region. Furthermore, the increase in Ca2+/ calmodulin-independent activity was detected up to 3 days after ischemia in all regions tested, suggesting that the concentration of intracellular Ca2+increased. In contrast to CaM kinase 11, as immunohistochemistry and regional immunoblot analysis revealed, calcineurin was preserved in the CA 1
region until 1.5 days and then lost with the increase in morphological degeneration of neurons. Immunoblot analysis confirmed the findings of the immunohistochemistry. These results suggest that there is a difference between CaM kinase I1 and calcineurin in regional and temporal loss after ischemia and that imbalance of Ca2+/calmodulin-dependentprotein phosphorylation-dephosphorylation may occur. Key Words: Transient cerebral ischemia-Hippocampus-Immunohistochemistry-Ca*+/calmodulin-dependent protein kinase II-Calcineurin. Morioka M. et al. Regional and temporal alterations in Ca2+/calmodulin-dependentprotein kinase I1 and calcineunn in the hippocampus of rat brain after transient forebrain ischemia. J. Neurochem. 58, 1798-1809 (1992).
Transient cerebral ischemia induces selective degeneration of certain populations of neurons, including pyramidal cells in the CA 1 region of the hippocampal formation, striatal medium-sized neurons, neocortical neurons, and cerebellar Purkinje cells (Brierley, 1976; Pulsinelli et al., 1982). Although the pathogenesis of hypoxic neuronal injury remains unclear, several lines of evidence have demonstrated that the concentration of intracellular Ca2+increases after cerebral ischemia and that the increase of CaZf is a pivotal event leading to irreversible cellular damage during the reperfusion phase and to delayed neuronal death (Kirino, 1982; Cheung et al., 1986; Siesjo and Bengtsson, 1989; Choi, 1990). Ca2+/calmodulin-dependentprotein kinase I1 (CaM
kinase 11), which is highly enriched in neural tissues, is a multifunctional protein kinase and is therefore involved in regulation of many cellular processes (for reviews, see Nairn et a]., 1985; Miyamoto, 1986; Schulman, 1988; Colbran et al., 1989; Soderling, 1990). In the hippocampal formation, the kinase is expressed at an unusually high level (Ouimet et al., 1984; Erondu and Kennedy, 1985; Fukunaga et al., 1988) and possibly is involved in induction of long-term potentiation (Malenka et al., 1989; Malinow et al., 1989). This enzyme is autophosphorylated in the presence of Ca2+/ calmodulin (CaM), and autophosphorylation of CaM kinase I1 converts the kinase from the Ca2+/CaM-dependent form to the Ca2+/CaM-independent form (Miyamoto, 1986; Schulman, 1988; Colbran et al.,
Received June 18, 1991 ; accepted October 2 1, 1991. Address correspondence and reprint requests to Dr. M. Morioka at Department of Neurosurgery, Kumamoto University Medical School, 1-1 Honjo I, Kumamoto 860, Japan. Abbreviations used: BSA, bovine serum albumin; CaM, calmod-
ulin; CaM kinase 11, Ca’+/calmodulin-dependent protein kinase 11; DAB, 3,Ydiaminobenzidine;NMDA, N-methyl-aaspartate;PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; SDS, sodium dodecyl sulfate.
I798
Ca”/CaM-DEPENDENT ENZYMES AFTER ISCHEMIA
1989). Furthermore, it has been reported that a transient elevation of intracellular Ca2+in cultured cells results in autophosphorylation of CaM kinase I1 to form the Ca2+/CaM-independent active form of the kinase (Fukunaga et al., 1989;Fukunaga and Soderling, 1990; MacNicol et al., 1990; Jefferson et al., 1991). Calcineurin is a neuron-specific Ca2+/CaM-dependent protein phosphatase (Wang and Desai, 1977;Klee et al., 1979; Stewart et al., 1982; Goto et al., 1985)and is abundant in the stnatum, hippocampus, substantia nigra, and cerebellum (Wallace et al., 1980; Goto et al., 1986a,b, 1987; Kincaid et al., 1987). The enzyme is a heterodimer containing calcineurin A of 61 kDa and calcineurin B of 19 kDa. The enzyme may be related to several neuronal actions by utilizing many functionally important substrates, such as microtubuleassociated protein 2 , factor, ~ tubulin (Goto et al., 1985; Halpain and Greengard, 1990), synapsin I and G substrate (King et al., 1984), type I1 regulatory subunit of cyclic AMP-dependent protein kinase (Klee et al., 1983), and DARRP-32 (Halpain et al., 1990). Several proteins are common to CaM kinase I1 and calcineurin as substrates for phosphorylation and dephosphorylation. This suggests that both enzymes may be involved complementady in brain functions by forming the cascade of phosphorylation and dephosphorylation of common substrates. Recently, decreases in Ca2+-dependent protein phosphorylation and CaM kinase I1 in the brain after ischemia have been reported (Taft et al., 1988; Churn et al., 1990).We have reported that forebrain ischemia causes the loss of immunoreactivity of CaM kinase I1 in pyramidal cells of CAI and CA3, and in granule cells of the dentate gyrus (Onodera et al., 1990). The present study describes regional and temporal alterations in immunoreactive staining produced by antibodies to CaM kinase I1 and calcineurin in the hippocampus after transient ischemia. The difference observed between both enzymes and the alteration in the activity of CaM kinase I1 after ischemia suggests the involvement of protein phosphorylation and dephosphorylation in postischemic Ca2+-dependentreactions. MATERIALS AND METHODS Materials Affi-gel blue was obtained from Bio-Rad Laboratories; cyanogen bromide-activated Sepharose 4B from Pharmacia Fine Chemicals; [y3’P]ATP from ICN Radiochemicals; Pcellulose paper from Whatman; CaM kinase I1 (28 1-309), syntide 2, cyclic AMP-dependent protein kinase inhibitor, and protein kinase C (19-36) from Bachem, Inc.; and Vectastain ABC kit from Vector Laboratories, Inc. The affinitypurified polyclonal antibody against rat brain CaM kinase I1 was prepared as previously described (Fukunaga et al., 1988; Ohta et al., 1988). CaM was purified from bovine brain (Gopalakrishna and Anderson, 1982). CaM kinase I1 was purified from rat brain by the method previously described (Fukunaga et al., 1982). CaM-dependent cyclic nucleotide phosphodi-
I799
esterase was purified from rat brain by the method of Klee et al. (1983).
Experimental animals and induction of ischemia Adult male Wistar rats weighing 200-260 g were kept under constant environmental conditions (temperature 22 k 2°C; humidity 55 k 5%)in the Laboratory Animal Research Center, Kumamoto University Medical School. The rats were kept for at least 7 days prior to the study under a 12-h light/ dark cycle. Transient forebrain ischemia was induced by the method of hlsinelli and Brierley (1979) with slight modifications. After animals were anesthetized with pentobarbital (50 mg/kg, i.p.), the vertebral arteries were electrocauterized within the alar foramina of the first cervical vertebra. The common carotid arteries were isolated, and then surrounded by 5-0 silk sutures loosely looped. The animals were fasted overnight and on the following day the bilateral carotid arteries were quickly occluded for 20 min by miniature aneurysmal clips. During the period of ischemia, rectal temperature and EEG were monitored and normothermia was maintained. Animals were excluded from the study if they failed to remain completely unresponsive or failed to show normothermia, bilateral loss of righting reflex, or flat EEG during the 20-min period of ischemia, or if they developed clinically evident seizures during the postischemic period. After release of the carotid occlusion, the animals were maintained to survive for 6 h, 12 h, 1 day, 1.5 days, 2 days, 3 days, or 7 days (n = 5).
Preparation of the polyclonal antibody against calcineurin A Calcineurin was prepared from fresh rat brain, as described by Klee et al. (1983), with slight modifications. The Affi-gel blue column chromatography was conducted after CaMSepharose affinity column. Purified calcineurin was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in 10% acrylamide by the method of Laemmli (1970). After protein staining and destaining, the gel containing only calcineurin A was cut. The protein was electrophoretically eluted from the gel and extensively dialyzed against distilled water. Female New Zealand rabbits were injected with 100 j t g of purified calcineurin A emulsified in complete Freund’s adjuvant at multiple intradermal sites. Then the rabbits were further immunized with 100 j t g of calcineurin A emulsified with incomplete Freund‘s adjuvant at 2-week intervals. The rabbits were bled (50 ml per bleed) from marginal ear veins 3 weeks after the fourth immunization. The blood was allowed to clot and the whole sera were stored at -70°C. Another female New Zealand rabbit was bled without immunization (nonimmune serum). The antibody was purified by an affinity chromatography column packed with cyanogen bromide-activated Sepharose 4B coupled with purified rat calcineurin, as described by Tallant and Cheung (1 983). The immunoblot analysis showed that the affinity-purified antibody reacted with calcineurin A of 6 1 kDa, but not with calcineurin B of 19 kDa of purified calcineurin, and recognized only calcineunn A of the brain homogenate (Fig. 1). No other protein in the homogenate cross-reacted with the antibody. Furthermore, the antibody did not recognize CaM kinase I1 P-subunit (60 kDa) or phosphodiesterase (60 kDa), which were easily contaminated in the purification steps and on the SDS-PAGE gel. The nonimmune serum did not recognize calcineurin A. The immunoreactive distribution of calcineurin in rat brain with the antibody prepared in the present study was essentially identical to those previously
J. Neurochem.. Vol. 58, No. 5 , I992
1800
M. MORIOKA ET AL. 12% acrylamide by SDS-PAGE, according to the method of Laemmli ( 1970), and then transferred electrophoretically at 80 V for 5 h to a Durapolla membrane by Trans Blot (BioRad) by the method of Towbin et al. (1979). After the membrane was incubated overnight at 4°C with 2.5% (wt/vol) bovine serum albumin (BSA) in Tris-bufferedsaline (pH 7.4) to block nonspecific binding sites, it was incubated for 1 h with the antibodies diluted in Tris-buffered saline containing 2.5% BSA. Immunostaining procedures were camed out according to the instruction of Vectastain ABC kit using DAB as a chromogen.
Assay for CaM kinase I1 activity FIG. 1. lrnrnunospecificity of the affinity-purified antibody against calcineurin A. Duplicate samples of purified calcineurin (2.0 pg, lanes 1 and 5), whole rat brain extract (30.0 pg, lanes 2 and 6), purified rat brain CaM kinase II (2.0 pg, lanes 3 and 7), and purified rat brain phosphodiesterase (2.0 pg. lanes 4 and 8) were electrophoresed in 12% acrylarnide by SDS-PAGE and transferred to a Durapolla membrane. Half of the membranewas stained with amido black (lanes 1,2, 3, and 4), and the other half was irnrnunostained with the affinity-purified antibody against calcineurin A (lanes 5, 6, 7, and 8). Molecular weight markers include phosphorylase b (94K). BSA (68K), ovalbumin (42K), carbonic anhydrase (30K). trypsin inhibitor (21K), and lysozyrne (14K).
reported (Wallace et al., 1980; Goto et al., 1986b, 1987; Kincaid et d., 1987).
Immunohistochemical procedures After ischemia, the rats were perfused and fixed at 120 mm Hg through the ascending aorta with cold 0.1 M phosphate-buffered saline (PBS; pH 7.4) for 5 min, and with 200 ml of 4% formaldehyde in 0.1 M PBS. The brains were removed and postfixed in the same fixative solution overnight and embedded in paraffin. The control animals were subjected to the same surgical procedures except for occlusion of bilateral carotid arteries (sham-operatedanimal). The perfusion, fixation, and decapitation of the control animals were performed on the day of ischemia induction. The sections (3-pm thick) were prepared and sequentially incubated with the affinity-purified polyclonal antibodies against CaM k i n a I1 and calcineurin A of rat brain at dilutions of 1:30 for anti-CaM kinase I1 antibody and 1:40 for anti-calcineurin A antibody at 4°C overnight. The immunostaining procedures were carried out according to the instruction of Vectastain ABC kit, using 3,Y-diaminobenzidine (DAB) as a chromogen. Staining specificity was assessed by incubation with the antibodies absorbed with the purified antigens and with nonimmune rabbit immunoglobulin G. Immunoreactivity was completely inhibited by preincubation with purified CaM kinase I1 or calcineurin. Adjacent sections were stained with Nissl and hematoxylin-eosin for comparison.
Gel electrophoresis and immunoblotting The brains were quickly removed after decapitation and dissected into each region of the hippocampus in cold 0.1 M PBS (pH 7.4) under a microscope. Each brain region was homogenized and sonicated with a Sonifier-250(Branson) at 0°C in 0.2 ml of 50 mMHEPES buffer (pH 7.4), 0.1%Triton X-100,4 mMEGTA, 10 mMEDTA, 0.1 Mieupeptin, 75 pM pepstatin A, and 0.1 mg/ml of aprotinin. Insoluble materials were removed by centrifugation at 15,000g for 5 min. Proteins in the supernatant were electrophoresed in 10 or J. Neurochem., Vol. 58, No. 5. 1992
The brains were quickly removed and frozen in liquid nitrogen after decapitation and stored at -70°C until used. Coronal sections (200-300-pm thick) were cut from frozen brain using a cryostat at -2O"C, and four regions, including CAI, CA3, the dentate gyrus of the hippocampus, and the parietal cortex, were dissected microscopically at -20°C. Each sample was quickly homogenized by a hand homogenizer at 0°C in 0.3 ml of 50 mM HEPES buffer (pH 7.4), 0.1% Triton X-100, 4 mMEGTA, 10 mMEDTA, 15 mM NQ207, 100 mMP-glycerophosphate, 25 mM NaF, 0.1 mM leupeptin, 75 pM pepstatin A, and 0.1 mg/ml of aprotinin. After sonication with a Sonifier-250(Branson), the insoluble materials were removed by centrifugation at 15,000 g for 5 min, and the supernatant was assayed for CaM kinase I1 by the method of Fukunaga et al. (1989). The standard kinase assay contained 50 mM HEPES buffer (pH 7.5), 10 mM magnesium acetate, 0.1 mM [y-32P]ATP,1 mg/ml of BSA, 40 pM syntide 2, and 5.0 p M cyclic AMP-dependent protein kinase inhibitor with or without (1 mMEGTA) 1 mMCaClz and 3 p M CaM in a volume of 25 pl. The reaction was initiated by the addition of 2 pl of the supernatant fractions. After 2 min, 15-pl aliquots were spotted on P-cellulose paper squares and processed as described by Roskoski (1983). The activity was considered to be the real activity of CaM kinase 11, although syntide 2 served as a good substrate for both CaM kinase I1 and protein kinase C, and the activity was assayed for the extract of the brain tissue. The activity was dependent on Ca2+/CaM and almost completely inhibited by CaM-dependent protein kinase I1 (281-309), which inhibited the purified CaM kinase 11. Addition of 2.0 pMprotein kinase C (19-36) (a pseudosubstrate for the enzyme) which inhibited more than 90% of the activity of purified brain protein kinase C to the assay, had no effect on the activity determined in the present study. The data were consistent with those obtained with extract of cultured cerebellar granule cells (Fukunaga et al., 1989). The activity of CaM kinase I1 was determined in the samples from five sham-operated animals, and from groups of five animals at 12 h, 1.5 days, 3 days, and 7 days after ischemia. Data for CaM kinase I1 activity was assessed by oneway analysis of variance. When a significant F value was obtained, Fisher's modified t test was used to determine statistically significant differences in means for each variable.
Other procedures Protein concentration was determined by the method of Bradford ( 1 976) with BSA as standard.
RESULTS Histological alteration after transient ischemia In the hippocampus, CA 1 pyramidal cells exhibited delayed neuronal death in all animals after ischemia.
Ca2+/CaM-DEPENDENT ENZYMES AFTER ISCHEMIA The cells appeared intact by light microscopy until 1 day after ischemia and most of cells were necrotic 23 days after ischemia. Other regions including CA3 and the dentate gyms revealed no remarkable alteration of morphology 7 days after ischemia. In the striatum, two of five animals sacrificed 7 days after ischemia revealed neuronal damage. One of the
1801
two animals had only a dorsolateral crescent lesion, and the other had a whole striatum lesion (data not shown). The other three animals showed normal striatum. In the parietal cortex, two of five animals killed 7 days after ischemia revealed moderate neuronal damage in layers 3, 5 , and 6 (data not shown). The other
FIG. 2. Time course of the imrnunohistochernical reaction of CaM kinase I1 in the hippocampus of rat brain after forebrain ischemia. Photomicrographs of the immunohistochernicalreactionfor CaM kinase I1 are shown for a sham-operated animal (A), and for animals killed 6 h (B), 12 h (C), 1 day (D),1.5 days (E), 2 days (F), 3 days (G), and 7 days (H) after transient ischemia. Magnification X20. The stratum pyramidale of CAI (arrows) and CA3 (arrow heads), and the granule cells of the dentate gyrus (double arrows) lost imrnunoreactivity 6 h after ischemia (B). DG,dentate gyrus.
J. Neuroehem., Vol. 58, No. 5, 1992
1802
M. MORIOKA ET AL.
FIG. 3. Time course of the immunohistochemicalreaction of CaM kinase II in the CAI region of a sham-operated animal (A), and of animals killed 6 h (B), 12 h (C). 1 day (D). 1.5 days (E), 2 days (F), 3 days (G), and 7 days (H) after transient ischemia. SO, stratum oriens; SP, stratum pyramidale; SR, stratum radiatum. Magnification X248.
three animals showed no remarkable alteration of morphology. Thus, only CA1 pyramidal cells of the hippocampus consistently revealed delayed neuronal death without degeneration of other hippocampal regions, whereas the striatum and the parietal cortex revealed neuronal degeneration in two of five animals. Therefore, mainly the hippocampus was examined in the following study. Immunohistochemical alteration of CaM kinase I1 after transient ischemia As shown in Fig. 2A, the immunoreactivity of the hippocampus was relatively strong. In coronal sections of the dorsal hippocampus, pyramidal cells in CA 1, granule cells in the dentate gyrus, and the stratum lucidum of the CA3 region were strongly immunostained with anti-CaM kinase I1 antibody. Pyramidal cells in CA3 were moderately immunostained. In the parietal cortex, neuronal somata were moderately immunostained (data not shown). Immunohistochemical alteration of CaM kinase I1 in the hippocampus was examined after transient ischemia (Fig. 2B-H). At 6 h after ischemia, pyramidal cells in CA1 and CA3 and granule cells in the dentate gyrus lost the immunoreactivity (Fig. 2B). On the other J. Neurochem.. Vol. 58, No. 5 , 1992
hand, the immunoreactivity in the strata oriens and radiatum of CA 1 and CA3 and the stratum moleculare of the dentate gyrus was relatively preserved (Fig. 2B). The immunoreactivity of pyramidal cells in CA1 for CaM kinase I1 was completely lost and the immunoreactivity of the strata oriens and radiatum of CA 1 was moderately lost 7 days after ischemia and not recovered (Fig. 2H). In contrast, the immunoreactivity of pyramidal cells in CA3 and granule cells in the dentate gyrus was gradually recovered to the level of the shamoperated animal 3 days after ischemia (Fig. 2F). Higher magnification of the figures more clearly revealed the alteration of the immunoreactivity of CAI pyramidal cells (Fig. 3) and CA3 pyramidal cells (Fig. 4). The loss of immunoreactivity of neuronal cytoplasm for CaM kinase I1 was prominently observed 6 h after ischemia (Fig. 3B) and did not recover from 12 h (Fig. 3C) to 7 days (Fig. 3H). In contrast, the immunoreactivity of CA 1 apical dendrites recovered 12 h after ischemia (Fig. 3C). In CA3 pyramidal cells, the immunoreactivity for CaM kinase I1 was lost 6 h after ischemia (Fig. 4B), and gradually recovered during 2 days (Fig. 4F) to 3 days (Fig. 4G) after ischemia. The immunoreactivity of neuronal somata in the
Ca”/CaM-DEPENDENT ENZYMES AFTER ISCHEMIA
1803
FIG. 4. Time course of the immunohistochemical reaction of CaM kinase II in the CA3 region of a sham-operated animal (A), and of animals killed 6 h (B), 12 h (C), 1 day (D), 1.5 days (E), 2 days (F), 3 days (G), and 7 days (ti) after transient ischemia.SL, stratum lucidum; SP,stratum pyramidale. Magnification X248.
parietal cortex was lost 6 h after ischemia and recovered 3 days after ischemia (data not shown). Immunoblot analysis of CaM kinase I1 after ischemia Immunoblot analysis was performed to confirm the immunohistochemical alteration of CaM kinase 11. The immunostaining of CaM kinase I1 slightly decreased in the CA1 and CA3 regions 12 h after ischemia (Fig. 5 , lane 2 of CA1 and CA3) and more decreased in the CAI region 7 days after ischemia (Fig. 5, lane 5 of CAI). In contrast to the CA1 region, the immunostaining of CaM kinase I1 in the CA3 region was recovered to the level of the sham-operated animal 7 days after ischemia (Fig. 5, lane 5 of CA3). There was
no remarkable fragmentation of CaM kinase I1 after ischemia in the CA1 and CA3 regions. However, it is possible that the epitope recognized by the anti-CaM kinase I1 antibody was not included in the fragment separated from CaM kinase 11. Alteration in activity of CaM kinase I1 after ischemia Figure 6 shows the activity of CaM kinase I1 in the extracts of regions of CAI, CA3, the dentate gyrus, and the parietal cortex obtained from sham-operated animals. The total activity of the enzyme determined in the presence of Ca*+/CaMwas highest in the CA1 region and lowest in the parietal cortex. The differences in total activity were statistically significant in each reJ. Neuroehem., Vol. 58, No. 5 , 1992
1804
M. MORIOIU ET AL. CA1
CA3
1 2 3 4 5
1 2 3 4 5
r - - - - - l n
Ongin+ ZOOK116K+ 94K+
30K + Front-
FIG. 5. Western blot analysis of CaM kinase II in the CAI and CA3 regions of the hippocampus of rat brain after transient forebrain ischemia. The samples were prepared from the CAI and CA3 regions in the hippocampus of a sham-operated animal (lane l), and of animals killed 12 h (lane Z ) , 1.5 days (lane 3). 3 days (lane 4), and 7 days (lane 5) after transient forebrain ischemia. Each sample of 30 Ng/lane was dectrophoresedin 10%acrylamide by SDSPAGE and transferred to a Durapolla membrane.Molecular weight markers include myosin (200K). p-galactosidase (116K). phosphorylaseb (94K), BSA (68K), ovalbumin (42K), and carbonic anhydrase (30K). (Y and 0 represent positions of +subunit (49K) and @-subunit(60K) of CaM kinase II, respectively.
gion, except for the difference between CA 1 and CA3. However, the Ca2+/CaM-independent activity of the enzyme determined in the presence of EGTA was not significantly different among the four regions (Fig. 6). The total activity of CaM kinase I1 decreased to 3436% in regions CAI, CA3, and the dentate gyrus 12 h after ischemia (Fig. 7A). Then the total kinase activity increased slightly except for the CAI region, which showed neuronal death 7 days after ischemia (Fig. 6A). The total kinase activity in regions CA3 and the dentate gyrus was recovered to 7040% of the control activity 7 days after ischemia (Fig. 7A), confirming the results observed by immunohistochemistry and immunoblot analysis (Figs. 2-5). On the other hand, about 46% of the total kinase activity in the CA1 region was maintained 7 days after ischemia, when the neuronal death was observed. This may be due to the fact that the intact tissues were contaminated with neurons of the CAI region during collection of the sample. To test the possibility that the decrease in total activity of CaM kinase I1 determined with Ca2+/CaM was due to the production of inhibitory compounds in the extract of the brain tissues, we determined the kinase activity of purified CaM kinase I1 from rat brain with an aliquot of the extract from the tissue at each time after ischemia. As the extract of the brain tissue had no effect on the kinase activity of the purified enzyme, we concluded that the inhibitory compounds were not produced in the ischemic tissue (data not shown). The percentage of the Ca"/CaM-independent activity for the total activity increased until 3 days after ischemia (Fig. 7B). Thus, the highest Ca2'/CaM-in-
J. Neurochem.. Vol. 58, No. 5 , 1992
dependent activity was about 400% of the control activity in the CAI region 3 days after ischemia. The results suggest that the concentration of intracellular Ca2' may increase in association with degeneration by ischemia. The independent activity, rather, decreased 7 days after ischemia (Fig. 7B). Immunohistochemical alteration of calcineurin after transient ischemia The coronal section of unaffected dorsal hippocampus from the sham-operated rat showed that the strata pyramidale, radiatum, and oriens, and apical dendrites of the CA 1 region were strongly immunostained (Fig. 8A). The dentate gyrus was moderately immunostained, but the neuronal somata were strongly immunostained (Fig. 8A). In the CA3 region, the pyramidal cells were not strongly immunostained, but the stratum lucidum showed strong immunoreactivity (Fig. 8A). After ischemia, no remarkable alteration of the immunostaining was observed in regions CA 1, CA3, and the dentate gyrus until 24 h (Fig. 8B-D). Then, pyramidal cells of the CAI region lost immunoreactivity 1.5 days after ischemia (Fig. 8E). The immunoreactivity of dendrites in the stratum radiatum was also lost. These findings affected by ischemia were more potentiated during the time course of the neuronal death
I
""I
'
p
Raven Press,Ltd., New York 0 1992 International Society for Neurochemistry
Regional and Temporal Alterations in Ca2+/CalmodulinDependent Protein Kinase I1 and Calcineurin in the Hippocampus of Rat Brain After Transient Forebrain Ischemia *tMotohiro Morioka, tKohji Fukunaga, TSetsuko Yasugawa, *Shinji Nagahiro, *Yukitaka Ushio, and tEishichi Miyamoto Departments of *Neurosurgery and ?Pharmacology, Kumarnoto University Medical School, Kumamoto, Japan
Abstract: We have investigated regional and temporal alterations in Ca2+/calmodulindependent protein kinase I1 (CaM kinase 11) and calcineunn (Ca2+/calmodulin-dependentprotein phosphatase) after transient forebrain ischemia. Immunoreactivity and enzyme activity of CaM kinase I1 decreased in regions CAI and CA3, and in the dentate gyrus, of the hippocampus early (6-12 h) after ischemia, but the decrease in immunoreactivity gradually recovered over time, except in the CA 1 region. Furthermore, the increase in Ca2+/ calmodulin-independent activity was detected up to 3 days after ischemia in all regions tested, suggesting that the concentration of intracellular Ca2+increased. In contrast to CaM kinase 11, as immunohistochemistry and regional immunoblot analysis revealed, calcineurin was preserved in the CA 1
region until 1.5 days and then lost with the increase in morphological degeneration of neurons. Immunoblot analysis confirmed the findings of the immunohistochemistry. These results suggest that there is a difference between CaM kinase I1 and calcineurin in regional and temporal loss after ischemia and that imbalance of Ca2+/calmodulin-dependentprotein phosphorylation-dephosphorylation may occur. Key Words: Transient cerebral ischemia-Hippocampus-Immunohistochemistry-Ca*+/calmodulin-dependent protein kinase II-Calcineurin. Morioka M. et al. Regional and temporal alterations in Ca2+/calmodulin-dependentprotein kinase I1 and calcineunn in the hippocampus of rat brain after transient forebrain ischemia. J. Neurochem. 58, 1798-1809 (1992).
Transient cerebral ischemia induces selective degeneration of certain populations of neurons, including pyramidal cells in the CA 1 region of the hippocampal formation, striatal medium-sized neurons, neocortical neurons, and cerebellar Purkinje cells (Brierley, 1976; Pulsinelli et al., 1982). Although the pathogenesis of hypoxic neuronal injury remains unclear, several lines of evidence have demonstrated that the concentration of intracellular Ca2+increases after cerebral ischemia and that the increase of CaZf is a pivotal event leading to irreversible cellular damage during the reperfusion phase and to delayed neuronal death (Kirino, 1982; Cheung et al., 1986; Siesjo and Bengtsson, 1989; Choi, 1990). Ca2+/calmodulin-dependentprotein kinase I1 (CaM
kinase 11), which is highly enriched in neural tissues, is a multifunctional protein kinase and is therefore involved in regulation of many cellular processes (for reviews, see Nairn et a]., 1985; Miyamoto, 1986; Schulman, 1988; Colbran et al., 1989; Soderling, 1990). In the hippocampal formation, the kinase is expressed at an unusually high level (Ouimet et al., 1984; Erondu and Kennedy, 1985; Fukunaga et al., 1988) and possibly is involved in induction of long-term potentiation (Malenka et al., 1989; Malinow et al., 1989). This enzyme is autophosphorylated in the presence of Ca2+/ calmodulin (CaM), and autophosphorylation of CaM kinase I1 converts the kinase from the Ca2+/CaM-dependent form to the Ca2+/CaM-independent form (Miyamoto, 1986; Schulman, 1988; Colbran et al.,
Received June 18, 1991 ; accepted October 2 1, 1991. Address correspondence and reprint requests to Dr. M. Morioka at Department of Neurosurgery, Kumamoto University Medical School, 1-1 Honjo I, Kumamoto 860, Japan. Abbreviations used: BSA, bovine serum albumin; CaM, calmod-
ulin; CaM kinase 11, Ca’+/calmodulin-dependent protein kinase 11; DAB, 3,Ydiaminobenzidine;NMDA, N-methyl-aaspartate;PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; SDS, sodium dodecyl sulfate.
I798
Ca”/CaM-DEPENDENT ENZYMES AFTER ISCHEMIA
1989). Furthermore, it has been reported that a transient elevation of intracellular Ca2+in cultured cells results in autophosphorylation of CaM kinase I1 to form the Ca2+/CaM-independent active form of the kinase (Fukunaga et al., 1989;Fukunaga and Soderling, 1990; MacNicol et al., 1990; Jefferson et al., 1991). Calcineurin is a neuron-specific Ca2+/CaM-dependent protein phosphatase (Wang and Desai, 1977;Klee et al., 1979; Stewart et al., 1982; Goto et al., 1985)and is abundant in the stnatum, hippocampus, substantia nigra, and cerebellum (Wallace et al., 1980; Goto et al., 1986a,b, 1987; Kincaid et al., 1987). The enzyme is a heterodimer containing calcineurin A of 61 kDa and calcineurin B of 19 kDa. The enzyme may be related to several neuronal actions by utilizing many functionally important substrates, such as microtubuleassociated protein 2 , factor, ~ tubulin (Goto et al., 1985; Halpain and Greengard, 1990), synapsin I and G substrate (King et al., 1984), type I1 regulatory subunit of cyclic AMP-dependent protein kinase (Klee et al., 1983), and DARRP-32 (Halpain et al., 1990). Several proteins are common to CaM kinase I1 and calcineurin as substrates for phosphorylation and dephosphorylation. This suggests that both enzymes may be involved complementady in brain functions by forming the cascade of phosphorylation and dephosphorylation of common substrates. Recently, decreases in Ca2+-dependent protein phosphorylation and CaM kinase I1 in the brain after ischemia have been reported (Taft et al., 1988; Churn et al., 1990).We have reported that forebrain ischemia causes the loss of immunoreactivity of CaM kinase I1 in pyramidal cells of CAI and CA3, and in granule cells of the dentate gyrus (Onodera et al., 1990). The present study describes regional and temporal alterations in immunoreactive staining produced by antibodies to CaM kinase I1 and calcineurin in the hippocampus after transient ischemia. The difference observed between both enzymes and the alteration in the activity of CaM kinase I1 after ischemia suggests the involvement of protein phosphorylation and dephosphorylation in postischemic Ca2+-dependentreactions. MATERIALS AND METHODS Materials Affi-gel blue was obtained from Bio-Rad Laboratories; cyanogen bromide-activated Sepharose 4B from Pharmacia Fine Chemicals; [y3’P]ATP from ICN Radiochemicals; Pcellulose paper from Whatman; CaM kinase I1 (28 1-309), syntide 2, cyclic AMP-dependent protein kinase inhibitor, and protein kinase C (19-36) from Bachem, Inc.; and Vectastain ABC kit from Vector Laboratories, Inc. The affinitypurified polyclonal antibody against rat brain CaM kinase I1 was prepared as previously described (Fukunaga et al., 1988; Ohta et al., 1988). CaM was purified from bovine brain (Gopalakrishna and Anderson, 1982). CaM kinase I1 was purified from rat brain by the method previously described (Fukunaga et al., 1982). CaM-dependent cyclic nucleotide phosphodi-
I799
esterase was purified from rat brain by the method of Klee et al. (1983).
Experimental animals and induction of ischemia Adult male Wistar rats weighing 200-260 g were kept under constant environmental conditions (temperature 22 k 2°C; humidity 55 k 5%)in the Laboratory Animal Research Center, Kumamoto University Medical School. The rats were kept for at least 7 days prior to the study under a 12-h light/ dark cycle. Transient forebrain ischemia was induced by the method of hlsinelli and Brierley (1979) with slight modifications. After animals were anesthetized with pentobarbital (50 mg/kg, i.p.), the vertebral arteries were electrocauterized within the alar foramina of the first cervical vertebra. The common carotid arteries were isolated, and then surrounded by 5-0 silk sutures loosely looped. The animals were fasted overnight and on the following day the bilateral carotid arteries were quickly occluded for 20 min by miniature aneurysmal clips. During the period of ischemia, rectal temperature and EEG were monitored and normothermia was maintained. Animals were excluded from the study if they failed to remain completely unresponsive or failed to show normothermia, bilateral loss of righting reflex, or flat EEG during the 20-min period of ischemia, or if they developed clinically evident seizures during the postischemic period. After release of the carotid occlusion, the animals were maintained to survive for 6 h, 12 h, 1 day, 1.5 days, 2 days, 3 days, or 7 days (n = 5).
Preparation of the polyclonal antibody against calcineurin A Calcineurin was prepared from fresh rat brain, as described by Klee et al. (1983), with slight modifications. The Affi-gel blue column chromatography was conducted after CaMSepharose affinity column. Purified calcineurin was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in 10% acrylamide by the method of Laemmli (1970). After protein staining and destaining, the gel containing only calcineurin A was cut. The protein was electrophoretically eluted from the gel and extensively dialyzed against distilled water. Female New Zealand rabbits were injected with 100 j t g of purified calcineurin A emulsified in complete Freund’s adjuvant at multiple intradermal sites. Then the rabbits were further immunized with 100 j t g of calcineurin A emulsified with incomplete Freund‘s adjuvant at 2-week intervals. The rabbits were bled (50 ml per bleed) from marginal ear veins 3 weeks after the fourth immunization. The blood was allowed to clot and the whole sera were stored at -70°C. Another female New Zealand rabbit was bled without immunization (nonimmune serum). The antibody was purified by an affinity chromatography column packed with cyanogen bromide-activated Sepharose 4B coupled with purified rat calcineurin, as described by Tallant and Cheung (1 983). The immunoblot analysis showed that the affinity-purified antibody reacted with calcineurin A of 6 1 kDa, but not with calcineurin B of 19 kDa of purified calcineurin, and recognized only calcineunn A of the brain homogenate (Fig. 1). No other protein in the homogenate cross-reacted with the antibody. Furthermore, the antibody did not recognize CaM kinase I1 P-subunit (60 kDa) or phosphodiesterase (60 kDa), which were easily contaminated in the purification steps and on the SDS-PAGE gel. The nonimmune serum did not recognize calcineurin A. The immunoreactive distribution of calcineurin in rat brain with the antibody prepared in the present study was essentially identical to those previously
J. Neurochem.. Vol. 58, No. 5 , I992
1800
M. MORIOKA ET AL. 12% acrylamide by SDS-PAGE, according to the method of Laemmli ( 1970), and then transferred electrophoretically at 80 V for 5 h to a Durapolla membrane by Trans Blot (BioRad) by the method of Towbin et al. (1979). After the membrane was incubated overnight at 4°C with 2.5% (wt/vol) bovine serum albumin (BSA) in Tris-bufferedsaline (pH 7.4) to block nonspecific binding sites, it was incubated for 1 h with the antibodies diluted in Tris-buffered saline containing 2.5% BSA. Immunostaining procedures were camed out according to the instruction of Vectastain ABC kit using DAB as a chromogen.
Assay for CaM kinase I1 activity FIG. 1. lrnrnunospecificity of the affinity-purified antibody against calcineurin A. Duplicate samples of purified calcineurin (2.0 pg, lanes 1 and 5), whole rat brain extract (30.0 pg, lanes 2 and 6), purified rat brain CaM kinase II (2.0 pg, lanes 3 and 7), and purified rat brain phosphodiesterase (2.0 pg. lanes 4 and 8) were electrophoresed in 12% acrylarnide by SDS-PAGE and transferred to a Durapolla membrane. Half of the membranewas stained with amido black (lanes 1,2, 3, and 4), and the other half was irnrnunostained with the affinity-purified antibody against calcineurin A (lanes 5, 6, 7, and 8). Molecular weight markers include phosphorylase b (94K). BSA (68K), ovalbumin (42K), carbonic anhydrase (30K). trypsin inhibitor (21K), and lysozyrne (14K).
reported (Wallace et al., 1980; Goto et al., 1986b, 1987; Kincaid et d., 1987).
Immunohistochemical procedures After ischemia, the rats were perfused and fixed at 120 mm Hg through the ascending aorta with cold 0.1 M phosphate-buffered saline (PBS; pH 7.4) for 5 min, and with 200 ml of 4% formaldehyde in 0.1 M PBS. The brains were removed and postfixed in the same fixative solution overnight and embedded in paraffin. The control animals were subjected to the same surgical procedures except for occlusion of bilateral carotid arteries (sham-operatedanimal). The perfusion, fixation, and decapitation of the control animals were performed on the day of ischemia induction. The sections (3-pm thick) were prepared and sequentially incubated with the affinity-purified polyclonal antibodies against CaM k i n a I1 and calcineurin A of rat brain at dilutions of 1:30 for anti-CaM kinase I1 antibody and 1:40 for anti-calcineurin A antibody at 4°C overnight. The immunostaining procedures were carried out according to the instruction of Vectastain ABC kit, using 3,Y-diaminobenzidine (DAB) as a chromogen. Staining specificity was assessed by incubation with the antibodies absorbed with the purified antigens and with nonimmune rabbit immunoglobulin G. Immunoreactivity was completely inhibited by preincubation with purified CaM kinase I1 or calcineurin. Adjacent sections were stained with Nissl and hematoxylin-eosin for comparison.
Gel electrophoresis and immunoblotting The brains were quickly removed after decapitation and dissected into each region of the hippocampus in cold 0.1 M PBS (pH 7.4) under a microscope. Each brain region was homogenized and sonicated with a Sonifier-250(Branson) at 0°C in 0.2 ml of 50 mMHEPES buffer (pH 7.4), 0.1%Triton X-100,4 mMEGTA, 10 mMEDTA, 0.1 Mieupeptin, 75 pM pepstatin A, and 0.1 mg/ml of aprotinin. Insoluble materials were removed by centrifugation at 15,000g for 5 min. Proteins in the supernatant were electrophoresed in 10 or J. Neurochem., Vol. 58, No. 5. 1992
The brains were quickly removed and frozen in liquid nitrogen after decapitation and stored at -70°C until used. Coronal sections (200-300-pm thick) were cut from frozen brain using a cryostat at -2O"C, and four regions, including CAI, CA3, the dentate gyrus of the hippocampus, and the parietal cortex, were dissected microscopically at -20°C. Each sample was quickly homogenized by a hand homogenizer at 0°C in 0.3 ml of 50 mM HEPES buffer (pH 7.4), 0.1% Triton X-100, 4 mMEGTA, 10 mMEDTA, 15 mM NQ207, 100 mMP-glycerophosphate, 25 mM NaF, 0.1 mM leupeptin, 75 pM pepstatin A, and 0.1 mg/ml of aprotinin. After sonication with a Sonifier-250(Branson), the insoluble materials were removed by centrifugation at 15,000 g for 5 min, and the supernatant was assayed for CaM kinase I1 by the method of Fukunaga et al. (1989). The standard kinase assay contained 50 mM HEPES buffer (pH 7.5), 10 mM magnesium acetate, 0.1 mM [y-32P]ATP,1 mg/ml of BSA, 40 pM syntide 2, and 5.0 p M cyclic AMP-dependent protein kinase inhibitor with or without (1 mMEGTA) 1 mMCaClz and 3 p M CaM in a volume of 25 pl. The reaction was initiated by the addition of 2 pl of the supernatant fractions. After 2 min, 15-pl aliquots were spotted on P-cellulose paper squares and processed as described by Roskoski (1983). The activity was considered to be the real activity of CaM kinase 11, although syntide 2 served as a good substrate for both CaM kinase I1 and protein kinase C, and the activity was assayed for the extract of the brain tissue. The activity was dependent on Ca2+/CaM and almost completely inhibited by CaM-dependent protein kinase I1 (281-309), which inhibited the purified CaM kinase 11. Addition of 2.0 pMprotein kinase C (19-36) (a pseudosubstrate for the enzyme) which inhibited more than 90% of the activity of purified brain protein kinase C to the assay, had no effect on the activity determined in the present study. The data were consistent with those obtained with extract of cultured cerebellar granule cells (Fukunaga et al., 1989). The activity of CaM kinase I1 was determined in the samples from five sham-operated animals, and from groups of five animals at 12 h, 1.5 days, 3 days, and 7 days after ischemia. Data for CaM kinase I1 activity was assessed by oneway analysis of variance. When a significant F value was obtained, Fisher's modified t test was used to determine statistically significant differences in means for each variable.
Other procedures Protein concentration was determined by the method of Bradford ( 1 976) with BSA as standard.
RESULTS Histological alteration after transient ischemia In the hippocampus, CA 1 pyramidal cells exhibited delayed neuronal death in all animals after ischemia.
Ca2+/CaM-DEPENDENT ENZYMES AFTER ISCHEMIA The cells appeared intact by light microscopy until 1 day after ischemia and most of cells were necrotic 23 days after ischemia. Other regions including CA3 and the dentate gyms revealed no remarkable alteration of morphology 7 days after ischemia. In the striatum, two of five animals sacrificed 7 days after ischemia revealed neuronal damage. One of the
1801
two animals had only a dorsolateral crescent lesion, and the other had a whole striatum lesion (data not shown). The other three animals showed normal striatum. In the parietal cortex, two of five animals killed 7 days after ischemia revealed moderate neuronal damage in layers 3, 5 , and 6 (data not shown). The other
FIG. 2. Time course of the imrnunohistochernical reaction of CaM kinase I1 in the hippocampus of rat brain after forebrain ischemia. Photomicrographs of the immunohistochernicalreactionfor CaM kinase I1 are shown for a sham-operated animal (A), and for animals killed 6 h (B), 12 h (C), 1 day (D),1.5 days (E), 2 days (F), 3 days (G), and 7 days (H) after transient ischemia. Magnification X20. The stratum pyramidale of CAI (arrows) and CA3 (arrow heads), and the granule cells of the dentate gyrus (double arrows) lost imrnunoreactivity 6 h after ischemia (B). DG,dentate gyrus.
J. Neuroehem., Vol. 58, No. 5, 1992
1802
M. MORIOKA ET AL.
FIG. 3. Time course of the immunohistochemicalreaction of CaM kinase II in the CAI region of a sham-operated animal (A), and of animals killed 6 h (B), 12 h (C). 1 day (D). 1.5 days (E), 2 days (F), 3 days (G), and 7 days (H) after transient ischemia. SO, stratum oriens; SP, stratum pyramidale; SR, stratum radiatum. Magnification X248.
three animals showed no remarkable alteration of morphology. Thus, only CA1 pyramidal cells of the hippocampus consistently revealed delayed neuronal death without degeneration of other hippocampal regions, whereas the striatum and the parietal cortex revealed neuronal degeneration in two of five animals. Therefore, mainly the hippocampus was examined in the following study. Immunohistochemical alteration of CaM kinase I1 after transient ischemia As shown in Fig. 2A, the immunoreactivity of the hippocampus was relatively strong. In coronal sections of the dorsal hippocampus, pyramidal cells in CA 1, granule cells in the dentate gyrus, and the stratum lucidum of the CA3 region were strongly immunostained with anti-CaM kinase I1 antibody. Pyramidal cells in CA3 were moderately immunostained. In the parietal cortex, neuronal somata were moderately immunostained (data not shown). Immunohistochemical alteration of CaM kinase I1 in the hippocampus was examined after transient ischemia (Fig. 2B-H). At 6 h after ischemia, pyramidal cells in CA1 and CA3 and granule cells in the dentate gyrus lost the immunoreactivity (Fig. 2B). On the other J. Neurochem.. Vol. 58, No. 5 , 1992
hand, the immunoreactivity in the strata oriens and radiatum of CA 1 and CA3 and the stratum moleculare of the dentate gyrus was relatively preserved (Fig. 2B). The immunoreactivity of pyramidal cells in CA1 for CaM kinase I1 was completely lost and the immunoreactivity of the strata oriens and radiatum of CA 1 was moderately lost 7 days after ischemia and not recovered (Fig. 2H). In contrast, the immunoreactivity of pyramidal cells in CA3 and granule cells in the dentate gyrus was gradually recovered to the level of the shamoperated animal 3 days after ischemia (Fig. 2F). Higher magnification of the figures more clearly revealed the alteration of the immunoreactivity of CAI pyramidal cells (Fig. 3) and CA3 pyramidal cells (Fig. 4). The loss of immunoreactivity of neuronal cytoplasm for CaM kinase I1 was prominently observed 6 h after ischemia (Fig. 3B) and did not recover from 12 h (Fig. 3C) to 7 days (Fig. 3H). In contrast, the immunoreactivity of CA 1 apical dendrites recovered 12 h after ischemia (Fig. 3C). In CA3 pyramidal cells, the immunoreactivity for CaM kinase I1 was lost 6 h after ischemia (Fig. 4B), and gradually recovered during 2 days (Fig. 4F) to 3 days (Fig. 4G) after ischemia. The immunoreactivity of neuronal somata in the
Ca”/CaM-DEPENDENT ENZYMES AFTER ISCHEMIA
1803
FIG. 4. Time course of the immunohistochemical reaction of CaM kinase II in the CA3 region of a sham-operated animal (A), and of animals killed 6 h (B), 12 h (C), 1 day (D), 1.5 days (E), 2 days (F), 3 days (G), and 7 days (ti) after transient ischemia.SL, stratum lucidum; SP,stratum pyramidale. Magnification X248.
parietal cortex was lost 6 h after ischemia and recovered 3 days after ischemia (data not shown). Immunoblot analysis of CaM kinase I1 after ischemia Immunoblot analysis was performed to confirm the immunohistochemical alteration of CaM kinase 11. The immunostaining of CaM kinase I1 slightly decreased in the CA1 and CA3 regions 12 h after ischemia (Fig. 5 , lane 2 of CA1 and CA3) and more decreased in the CAI region 7 days after ischemia (Fig. 5, lane 5 of CAI). In contrast to the CA1 region, the immunostaining of CaM kinase I1 in the CA3 region was recovered to the level of the sham-operated animal 7 days after ischemia (Fig. 5, lane 5 of CA3). There was
no remarkable fragmentation of CaM kinase I1 after ischemia in the CA1 and CA3 regions. However, it is possible that the epitope recognized by the anti-CaM kinase I1 antibody was not included in the fragment separated from CaM kinase 11. Alteration in activity of CaM kinase I1 after ischemia Figure 6 shows the activity of CaM kinase I1 in the extracts of regions of CAI, CA3, the dentate gyrus, and the parietal cortex obtained from sham-operated animals. The total activity of the enzyme determined in the presence of Ca*+/CaMwas highest in the CA1 region and lowest in the parietal cortex. The differences in total activity were statistically significant in each reJ. Neuroehem., Vol. 58, No. 5 , 1992
1804
M. MORIOIU ET AL. CA1
CA3
1 2 3 4 5
1 2 3 4 5
r - - - - - l n
Ongin+ ZOOK116K+ 94K+
30K + Front-
FIG. 5. Western blot analysis of CaM kinase II in the CAI and CA3 regions of the hippocampus of rat brain after transient forebrain ischemia. The samples were prepared from the CAI and CA3 regions in the hippocampus of a sham-operated animal (lane l), and of animals killed 12 h (lane Z ) , 1.5 days (lane 3). 3 days (lane 4), and 7 days (lane 5) after transient forebrain ischemia. Each sample of 30 Ng/lane was dectrophoresedin 10%acrylamide by SDSPAGE and transferred to a Durapolla membrane.Molecular weight markers include myosin (200K). p-galactosidase (116K). phosphorylaseb (94K), BSA (68K), ovalbumin (42K), and carbonic anhydrase (30K). (Y and 0 represent positions of +subunit (49K) and @-subunit(60K) of CaM kinase II, respectively.
gion, except for the difference between CA 1 and CA3. However, the Ca2+/CaM-independent activity of the enzyme determined in the presence of EGTA was not significantly different among the four regions (Fig. 6). The total activity of CaM kinase I1 decreased to 3436% in regions CAI, CA3, and the dentate gyrus 12 h after ischemia (Fig. 7A). Then the total kinase activity increased slightly except for the CAI region, which showed neuronal death 7 days after ischemia (Fig. 6A). The total kinase activity in regions CA3 and the dentate gyrus was recovered to 7040% of the control activity 7 days after ischemia (Fig. 7A), confirming the results observed by immunohistochemistry and immunoblot analysis (Figs. 2-5). On the other hand, about 46% of the total kinase activity in the CA1 region was maintained 7 days after ischemia, when the neuronal death was observed. This may be due to the fact that the intact tissues were contaminated with neurons of the CAI region during collection of the sample. To test the possibility that the decrease in total activity of CaM kinase I1 determined with Ca2+/CaM was due to the production of inhibitory compounds in the extract of the brain tissues, we determined the kinase activity of purified CaM kinase I1 from rat brain with an aliquot of the extract from the tissue at each time after ischemia. As the extract of the brain tissue had no effect on the kinase activity of the purified enzyme, we concluded that the inhibitory compounds were not produced in the ischemic tissue (data not shown). The percentage of the Ca"/CaM-independent activity for the total activity increased until 3 days after ischemia (Fig. 7B). Thus, the highest Ca2'/CaM-in-
J. Neurochem.. Vol. 58, No. 5 , 1992
dependent activity was about 400% of the control activity in the CAI region 3 days after ischemia. The results suggest that the concentration of intracellular Ca2' may increase in association with degeneration by ischemia. The independent activity, rather, decreased 7 days after ischemia (Fig. 7B). Immunohistochemical alteration of calcineurin after transient ischemia The coronal section of unaffected dorsal hippocampus from the sham-operated rat showed that the strata pyramidale, radiatum, and oriens, and apical dendrites of the CA 1 region were strongly immunostained (Fig. 8A). The dentate gyrus was moderately immunostained, but the neuronal somata were strongly immunostained (Fig. 8A). In the CA3 region, the pyramidal cells were not strongly immunostained, but the stratum lucidum showed strong immunoreactivity (Fig. 8A). After ischemia, no remarkable alteration of the immunostaining was observed in regions CA 1, CA3, and the dentate gyrus until 24 h (Fig. 8B-D). Then, pyramidal cells of the CAI region lost immunoreactivity 1.5 days after ischemia (Fig. 8E). The immunoreactivity of dendrites in the stratum radiatum was also lost. These findings affected by ischemia were more potentiated during the time course of the neuronal death
I
""I
'
p
