Neuropathology 1999; 19, 356–365

Review Article Ischemic delayed neuronal death: Role of the cysteine and

Anton B Tontchev and Tetsumori Yamashima Department of Neurosurgery, Kanazawa University School of Medicine, Kanazawa, Japan

A few days after a transient brain ischemia, the pyramidal (DND) and was first described in gerbils in 1982.1,2 Later, neurons in the cornu Ammonis (CA) 1 sector of the its clinical relevance was also confirmed.3,4 Both necrosis1–5 hippocampus undergo selective death, a process named and apoptosis have been reported as a type of cell death delayed neuronal death (DND). Cell death may occur as of DND, but its molecular mechanisms remain largely necrosis and/or apoptosis, and both have been reported to unclear.6,7 take place in DND. The cell’s decision between apoptosis The role of calcium (Ca2) in ischemic neuronal death is and necrosis may depend on the strength of the insult, the well recognized.8 An excessive postischemic Ca2 mobi- balance of downstream signal transduction systems, and lization occurs in CA1 neurons9,10 by both glutamate- the expression level of pro- and anti-apoptotic or necrotic induced Ca2 influx11,12 and/or inositol 1,4,5-trisphosphate 2 2 13,14 factors. Cytosolic calcium (Ca ) overload specifically (IP3)-induced Ca release from intracellular stores. occurs in the CA1 neurons after ischemia and thus is Elevated intracellular Ca2 concentration in turn activates considered a common triggering event of the death a series of including proteases, phospholipases, cascade. As Ca2 activates a wide array of intracellular phosphatases and protein kinases. It may also trigger the enzymes, many Ca2 -targeted enzymes have been im- synthesis of the free radical nitric oxide (NO), involved in plicated in DND. Among these, the present review will neuronal apoptosis.15 focus on the cysteine proteases calpain and cathepsins (B Among the Ca2-activated enzymes, the authors will and L). In addition, their possible interactions with another focus on the cysteine proteases calpain (EC 3.4.22.17) and family of cysteine proteases, , will be discussed in cathepsins (particularly cathepsins B (EC 3.4.22.1) and relation to the cellular fate toward apoptosis or necrosis. L (EC 3.4.22.15)), because of the recent progress in understanding their role in the pathogenesis of DND.16–18 Key words: brain ischemia, calpain, , delayed Calpain is a cytosolic cysteine ubiquitously and neuronal death. constitutively expressed in mammalian cells.19 Two forms of the are known: µ-calpain (calpain I) and m-calpain INTRODUCTION (calpain II). They differ in their sensitivity to Ca2 , µ- calpain being activated at a low (micromolar) Ca2 Cerebral ischemia has profound effects on the brain, concentration, while m-calpain is activated at a higher inducing both immediate and delayed injuries. Even a (millimolar) Ca2 concentration. are considered transient ischemic insult is sufficient to cause selective to participate in various intracellular signaling pathways damage to certain sensitive brain regions. The cornu mediated by Ca2. They are heterodimers, consisting of Ammonis (CA) 1 sector of the hippocampus is considered 30 and 80 kDa subunits. The small subunits are identical such a region, since a transient global ischemia is able to among one and the same tissue and/or animal species, while cause neuronal death a few days after cerebral ischemia. the large subunits are similar but not identical. A very This phenomenon is designated as delayed neuronal death specific endogenous inhibitor, , which interacts with the Ca2-binding domains of both large and small subunits, inhibits calpain. The coexistence of calpain and calpastatin in cells suggests a pivotal role of this inhibitor in Correspondence: Tetsumori Yamashima, MD, PhD, Department the regulation of calpain activity.20 of Neurosurgery, Kanazawa University School of Medicine, 13-1 Like calpains, cathepsins also belong to the Takaramachi, Kanazawa, 920-8641, Japan. Email: yamashim@med. 21 kanazawa-u.ac.jp superfamily of cysteine proteases. They are synthesized as Received 30 August 1999; accepted 9 September 1999. inactive precursors and require proteolytic activation that Cysteine proteases in neuronal death 357 generally requires an acidic pH. Thus, they are localized Table 1 A list of selected calpain substrates predominantly to the lysosomes (cathepsins B, H, L, S, C Substrate Reference and K), but also to the nucleus (cathepsins B and L) and 22 Cytoskeletal proteins cytosol (cathepsins B and E). Most cathepsins have Tropomyosin 20 molecular weight (MW) values of 20–35 kDa (with the Actin 23 exception of cathepsin C, which is approximately 200 kDa). Spectrin 24, 25 Unlike calpains, they lack Ca2-binding domains. NF-68 26 Tubulin 20 Numerous cathepsin inhibitors exist, the most abundant of Tau 27 which are the cystatins. The latter interact with active Ca2-binding proteins proteases to bind them tightly and essentially irreversibly. Caldesmon 34 Their function is to inhibit cathepsins that escape Calponin 34 Enzymes compartmentalization and protect cells, tissues and the Calpastatin 35 circulation from unwarranted activity.21 Nitric oxide synthase 36 PKC 20 Calcineurin 20 SUBSTRATES OF CALPAIN AND Growth factor/cytokine receptors CATHEPSIN EGF receptor 28 PDGF receptor 20 The ubiquitous and constitutive expression of µ- and Transcription and translation factors m-calpains makes it difficult to precisely define their AP-1 (c-Fos/c-Jun) 29 c-Myc 29 physiological functions. A large number of calpain sub- CREB 29 strates have been demonstrated, mainly in vitro, including eIF-4E 30 cytoskeletal proteins,20,23–27 growth factor receptors,20,28 and Cell cycle/tumor suppressor proteins transcription- and cell cycle-related proteins29–33 (Table 1). Cyclin D1 31 p53 32 In relation to the latter, selective nuclear transport of NF2 (merlin, schwannomin) 33 µ-calpain has been demonstrated.38 NF-68, neurofilament 68 kDa protein; PKC, protein kinase C; EGF, It is worth noting that calpastatin itself is also a target epidermal growth factor; PDGF, platelet-derived growth factor;AP-1, of calpain,35 and µ-calpain is capable of activating m- activator protein-1; CREB, cyclic AMP response element-binding calpain.37 Thus, a calpain cascade may coordinate the protein; eIF-4E, eukaryotic initiation factor-4E; c-Myc, c-myelocytoma oncogene product. functions of calpains in cells. A variety of calpain functions have been proposed including intracellular protein deg- radation, platelet aggregation, erythrocyte aging, neu- trophil activation as well as long-term potentiation in the hippocampus. Detailed data on the physiological functions of calpains are available in other reviews.19–21 Cathepsins were shown to participate in the extracel- lular matrix remodeling, and is considered the most potent elastolytic enzyme known.21 Collagen cleavage by cathepsin B39 and fibronectin degradation by cathepsin D40 have been demonstrated. Several cytoskeletal proteins are substrates of cathepsin D including myelin basic protein,41 neurofilament proteins,42 microtubule-associated protein 2,43 tubulin43 and -amyloid protein.44 Cathepsins are essential for the generation of antigenic peptides in the major histocompatibility complex class II antigen presenta- tion ()21 as well as for the processing of prohor- mone molecules, such as pro-endothelin-1 (cathepsin D)45 and prorenin ().46 Cathepsin B is also capable of cleaving the myristoylated alanine-rich C kinase sub- 2 Fig. 1 Time course of [Ca ]i elevation in the monkey strate (MARCKS),47 a major cellular substrate of protein hippocampal slice during hypoxia–hypoglycemia (adapted 2 kinase C. from Fig. 7a of Reference 16). The largest [Ca ]i elevation demonstrated by relative fluorescence intensity was observed has been implicated in tumor progression 2 21,22 preferentially in the CA1 sector ( ), in contrast to the [Ca ]i and bone resorption. Interestingly, members of the elevation in the CA3, CA4 and the dentate gyrus (a summing (the other major cysteine protease) superfamily, graph shown by ()). 358 AB Tontchev and T Yamashima

2 16 namely caspase-11 and to a lesser extent caspase-1, have ([Ca ]i) during hypoxia–hypoglycemia. The area of 48 2 recently been shown to be substrates of cathepsin B, while the maximum [Ca ]i elevation in the slice (Fig. 1) was caspase-3 has been shown to be a substrate of cathepsin L.49 selectively localized to the CA1 sector, the area of DND in These data, together with the observation of caspase- vivo. In previous studies, indirect evidence of Ca2-induced mediated calpastatin degradation,50 suggest a potential calpain involvement in DND has been demonstrated interaction between the caspase and papain superfamilies by the finding of calpain-mediated fodrin degradation, in integrating their cellular effects. associated with Ca2-induced calpain activation in CA1 neurons after brain ischemia.24 Using an antibody that specifically reacts with an activated form of µ-calpain51 and CALPAIN AND primate model of 20-min whole-brain complete ischemia, INVOLVEMENT IN ISCHEMIC DELAYED Yamashima et al. provided direct evidence of calpain NEURONAL DEATH activation prior to DND.16 At immunohistochemistry, the Analyzing postischemic neuronal Ca2 mobilization by perikarya of CA1 neurons, immediately after the ischemic an in vitro ischemia experiment using acute hippocampal insult, showed a specific staining of activated µ-calpain slices, Yamashima et al. measured intracellular free Ca2 (Fig. 2a,b) as compared with both the controls and the

Fig. 2 Calpain in ischemic hippocampus. (a,b) Immunohistochemistry using anti-activated µ-calpain antibody. A significant increase in immunoreactivity is observed in ischemic CA1 neurons (b) compared with controls (a) showing no staining (700; bar 10 µm. (c) Western blot analysis of control (C) and postischemic (I) hippocampal sectors using antibodies against inactive (pre-µ, upper half) or activated (post-µ, lower half) µ-calpain. Proteins extracted from the CA1, CA2/3 and CA4/DG (dentate gyrus), respectively, are presented from left to right. The 80 kDa (inactive) µ-calpain (pre-µ) is significantly decreased in the ischemic CA1 sector compared with the control. In contrast, 76 kDa (activated) µ-calpain (post-µ) showed a significant increase, specifically in the postischemic CA1. (d) Immunoelectron micrographs of CA1 neurons immediately after 20 min hypoxia–hypoglycemia, showing that activated µ-calpain is localized at the vacuolated membrane of lysosomes (arrows) (11 500; bar 1 µm). Cysteine proteases in neuronal death 359 remaining ischemia-resistant hippocampal neurons (data had returned to the normal level. Immunoelectron micro- not shown). This observation was confirmed by western scopy of the hippocampal slices using anti-activated blot analysis of the respective hippocampal sectors µ-calpain antibody revealed that activated µ-calpain was 2 (Fig. 2c). Most importantly,the [Ca ]i levels increased only localized at the vacuolated or disrupted membrane of transiently during hypoxia–hypoglycemia in vitro (Fig. 1), lysosomes (Fig. 2d). Taken together, these data suggest 2 while calpain activation in vivo persisted long after [Ca ]i that excessive calpain activation in postischemic CA1

Fig. 3 Immunohistochemistry for cathepsin B in CA1 (a,b,e) and CA3 (c,d) neurons. CA1 neurons 1 h after the ischemic insult (b) show a presence of extralysosomal cathepsin B immunoreactivity in contrast to the ischemia-resistant CA3 neurons (d).After CA- 074 treatment, cathepsin B immunoreactivity in the CA1 neurons (e) is weak and comparable with the control (a). C, control CA3 neurons (a–d, 690; e, 660, bar 20 µm). 360 AB Tontchev and T Yamashima neurons may cause lysosomal membrane disruption with the resultant release of lysosomal enzymes triggering the cascade of DND. Neuronal lysosomes are known to contain both cathepsins B and L,52,53 and the enzyme activity of both is increased in the CA1 sector 3–5 days after ischemia.17,18,54 Therefore, Yamashima et al. and Tsuchiya et al. aimed to identify the contribution of cathepsins B and L to ischemic CA1 DND.17,18 Immediately after ischemia, the cathepsin B immunoreactivity of CA1 neurons was significantly increased, and was localized not only to the lysosomes but also to the perikarya or the neuropil (Fig. 3a,b). In contrast, the cathepsin B immunoreactivity in the ischemia-resistant hippocampal neurons also increased after ischemia but did not show extralysosomal distribution (Fig. 3c,d). Such a difference in the immunolocalization of cathepsin B was also evident at the ultrastructural level.When the cathepsin B-selective inhibitor CA-074 was applied to the monkeys immediately after the ischemic insult, the immunoreac- tivity of cathepsin B in the lysosomal granules of CA1 neurons became much weaker, and no extralysosomal immunoreactivity was observed (Fig. 3e). Quantification of cathepsin B enzyme activity in each hippocampal sector displayed a similar pattern showing an increase in all sectors 3–5 days after the ischemic insult, the application of CA-074 diminishing this effect (Fig. 4a).18 Cathepsin L activity showed a similar pattern (Fig. 4b). Therefore, Ca2-induced calpain overactivation after transient brain ischemia may lead to disruption of lyso- somes with the resultant release of cathepsins. The latter may gradually degrade cell constituents, thus contri- buting to the development of DND (calpain–cathepsin Fig. 4 Enzymatic activities of cathepsins B and L in monkey hypothesis).16–18 The extralysosomal release of cathepsin hippocampal sectors as measured by fluorescence spectro- enzymes occurring specifically in the CA1 sector may photometry of 7-amino-4-methylcoumarin (AMC) release (see Reference 54 for more details). () CA1; ( ) CA2; ( ) explain the restricted localization of DND to this hip- CA3; () DG CA4. The residual cathepsin activity in the pocampal region. CA1 sector at postischemic day 5 after CA-074 (a selective cathepsin B inhibitor) treatment was 36% for cathepsin B (a) and 80% for cathepsin L (b), compared with the respective POTENTIAL INTERACTIONS OF CALPAIN, untreated day 5 groups. However, after E-64c (a non-selective CATHEPSIN AND CASPASES IN DELAYED inhibitor) treatment, the residual cathepsin activity in the CA1 NEURONAL DEATH sector was 14% for cathepsin B (a) and 43% for cathepsin L (b), compared with the respective untreated day 5 groups. 5d, Recently, apoptosis (programmed cell death) has been untreated day 5 group; 5d(CA-074), CA-074-treated group; 5d(E-64c), E-64c-treated group. (a) *P < 0.01 versus control; proposed as a mechanism of neuronal damage in DND, **P 0.05 versus control and day 5; ***P 0.01 versus day 5. 6,7 particularly in rodents. Dying ischemic neurons in (b) *P 0.05 versus control; **not significant; ***P 0.01 different brain regions, including CA1, showed morpho- versus day 5. logical features of apoptosis and a positive labeling of nuclear double-strand DNA breaks in situ (terminal deoxynucleotidl dUTP nick end labeling conditions,58 their involvement in ischemic DND is (TUNEL) assay) together with DNA fragmentation.55,56 currently a major subject of investigation. Based on their Since another superfamily of cysteine proteases, the regulation and sequence of activation, caspases are aspartyl-specific cysteine proteases (caspases),57 is well classified as initiator (e.g. caspase-8, - 9, -10) and effector known to have an important role in neuronal apoptosis (e.g. caspase-3, -6, -7) classes.57 In rodents, caspase-3 during development and also in various pathological activation in transient global ischemia has been reported as Cysteine proteases in neuronal death 361 showing a selective and prolonged caspase-3 mRNA and disruption of mitochondrial membrane integrity, is protein induction in the CA1 neurons prior to considered a key step in the execution of apoptosis.65 apoptosis.59–61 Furthermore, a specific inhibitor of caspase-3 Mitochondrial permeability transition is thought to be significantly reduced CA1 neuronal death.59,61 Therefore, caused by factors such as mitochondrial calcium overload up-regulation of caspase-3 in dying CA1 neurons may be and results in release of cytochrome c and caspase-9 from an important step in the development of the postischemic the intermembrane space of mitochondria, which triggers DND of rodents. the caspase cascade prior to apoptosis.64 In postischemic As mentioned, caspase-3 and calpain have common hippocampal slices, cytochrome c was released into the substrates in neurons, including α-spectrin25 and neu- cytosol66 and the activation of MPT among various brain rofilament proteins.26 A temporal relationship between regions correlated with their susceptibility to ischemic calpain/caspase-3 activation and α-spectrin degradation damage.67 In addition, the MPT inhibitor cyclosporin A prior to DNA fragmentation and apoptosis was estab- ameliorated hippocampal CA1 neuronal damage after lished.62 Calpain activation in apoptosis upstream of transient global ischemia in rodents.68 Thus, it is tempting to caspases,63 and caspase-3-mediated calpastatin degrada- speculate that in the primate postischemic DND, both tion35 inducing calpain activation in non-neuronal cells, calpain–cathepsin- and MPT-triggered caspase activation were shown. Furthermore, cathepsin B activates caspase-11 might induce neuronal degeneration and death (Fig. 5). upstream of caspase-3-mediated liver cell apoptosis,48 and cathepsin L activates caspase-3-like protease, also in liver cells.49 These data suggest that a mutual communication THERAPEUTIC STRATEGIES AND may exist between members of the two large cysteine FUTURE PESPECTIVES protease superfamilies in executing their roles in DND Assuming that cysteine proteases are involved in the (Fig. 5). Thus, a calpain–cathepsin–caspase cascade might development of ischemic brain injury, it is reasonable be involved in the primate postischemic CA1 DND. to predict that their inhibition, even if applied after the The role of mitochondria in necrosis has been ischemic insult, could have a protective effect on the de- recognized for decades, while their implication in apoptosis generating postischemic neurons.69 In particular, calpain70 has recently also become clear.64 The induction of mito- and caspase71 inhibition has successfully blocked ischemic chondrial permeability transition (MPT), a process of CA1 DND in rodents. As described, selective cathepsin B

Fig. 5 Potential interactions between cysteine proteases in ischemic CA1 neurons prior to delayed neuronal death 2 (DND). Excessive [Ca ]i may lead to calpain–cathepsin-mediated activation of caspase-11 (cathepsin B) or caspase-3 (cathepsin L) and, on the other hand, to mitochondrial calcium overload resulting in release of proteins such as cytochrome c and caspase-9 through the mitochondrial permeability transition (PT) pore. Both pathways converge on caspase-3-induced DNase activation leading to apoptosis. Concurrently, de- gradation of still unknown cellular con- stituents by extralysosomal cathepsins B and L in addition to mitochondrial factors may lead to necrosis. In addition to its interaction with cathepsins and caspases, calpain may also mediate cleavage of cytoplasmic (such as cyto- skeleton) and nuclear (such as c-Fos/ c-Jun, p53, cyclin D1) proteins. Caspases (via calpastatin) may provide a positive feedback link to calpain, perpetuating its activity. PL, plasmalemma; NL, nucleolemma; IM, inner mitochondrial membrane; OM, outer mitochondrial membrane. 362 AB Tontchev and T Yamashima inhibition was able to prevent ischemic DND in monkeys.17 Also, the non-selective cysteine protease inhibitor E-64c that inhibits not only cathepsins B, H and L, but also calpain had a greater potency in reversing DND compared with the selective cathepsin B blocker (Fig. 4b).18 This is further evidence implicating calpain and cathepsin enzymes in the development of DND. In addition to inhibitors of cysteine proteases and cyclo- sporin A, several drugs, such as the gamma-aminobutyric acid (GABA) agonist diazepam,72 the microtubule- disrupting agent colchicine73 and the immunosuppressant FK-506,74 displayed a neuroprotective effect on CA1 neu- rons after transient global ischemia. A similar effect was confirmed also for acidic fibroblast growth factor,75 hepat- ocyte growth factor,76 and interleukin-3.77 The effect of diazepam is probably mediated via the mitochondrial ben- zodiazepine receptor,64 in a similar way to cyclosporin A. Both cyclosporin A and FK-506 inhibit the cytoplasmic protein phosphatase calcineurin; however, FK-506 lacks the ability of inhibiting MPT. Therefore, calcineurin, a calpain substrate,20 may be involved in postischemic DND in a MPT-independent fashion. It is now recognized that cell death may take place by necrosis or apoptosis. The prevalence of either type of death in various neuropathologies including ischemia is debatable at present.78,79 In particular, both necrotic and apoptotic features have been reported in CA1 DND. In Fig. 6 Two hypothetical models of relation of necrosis and apoptosis to delayed neuronal death (DND). (a) Separate various models of total brain ischemia, including the death cascades execute necrosis and apoptosis: severe insults 16–18 primate model described above, the features of CA1 activate calpain–cathepsin cascade leading to necrosis, while neuronal death were essentially necrotic. However, evi- milder insults activate mitochondrial factors–caspase cascade dence of apoptosis has also been reported in gerbils.6 leading to apoptosis. The two systems may interact and thus regulate their activity. (b) A common calpain–cathepsin– In the context of the involvement of calpain-cathepsin caspase cascade (with the involvement of mitochondrial cascade in CA1 neuronal necrosis and of mitochondrial factors) executes cell death. In this case, apoptosis is con- factors–caspase cascade in apoptosis, we would like to sidered a step prior to development of necrosis. propose two possible models of the occurrence and pre- valence of necrosis and apoptosis in CA1 DND (Fig. 6). In both models an interaction between the two systems, of apoptosis and necrosis to the development type of calpain–cathepsin and mitochondrial factors–caspase, neuronal damage. takes place prior to cell death. In one model, the two systems may separately lead to necrosis or apoptosis depending on the severity of the insult (Fig. 6a), and the cell ACKNOWLEDGMENT fate is determined by the relative prevalence of their This work was supported by a grant-in-aid for the Strategic components. In the other model, a common calpain– Promotion System for Brain Science from the Science and cathepsin–caspase cascade (with the participation of Technology Agency of Japan. mitochondrial factors) may execute cell death (Fig. 6b). 48,49 That is, a cathepsin-mediated apoptosis and a caspase- REFERENCES mediated necrosis80–82 may simultaneously take place, and it is the severity of insult that determines whether cell death 1. Kirino T. Delayed neuronal death in the gerbil will be restricted to apoptosis only or whether necrosis will hippocampus following ischemia. Brain Res 1982; 239: eventually follow. Future studies on the temporal and 57–69. quantitative pattern of the expression of various pro- 2. Pulsinelli WA, Brierly JB, Plum F. Temporal profile and anti-apoptotic and necrotic factors will shed more light of neuronal damage in a model of transient brain on their precise role in DND as well as the contribution ischemia. Ann Neurol 1982; 11: 491–498. Cysteine proteases in neuronal death 363

3. Zola-Morgan S, Squire LR, Amaral DG. Human 17. Yamashima T, Kohda Y, Tsuchiya K et al. Inhibition amnesia and the medical temporal region: Enduring of ischemic hippocampal neuronal death in primates memory impairment following a bilateral lesion limited with cathepsin B inhibitor CA-074: a novel strategy to field CA1 of the hippocampus. J Neurosci 1986; 6: for neuroprotection based on ‘calpain-cathepsin hypo- 2950–2967. thesis’. Eur J Neurosci 1998; 10: 1723–1733. 4. Petito CK, Feldmann E, Pulsinelli WA, Plum F.Delayed 18. Tsuchiya K, Kohda Y, Yoshida M et al. Postictal hippocampal damage in humans following cardiores- blockade of ischemic hippocampal neuronal death piratory arrest. Neurology 1987; 37: 1281–1286. in primates using selective cathepsin inhibitors. Exp 5. Deshpande JK, Siesjö BK, Wieloch T. Calcium ac- Neurol 1999; 155: 187–194. cumulation and neuronal damage in the rat hippo- 19. Sorimachi H, Ishiura S, Suzuki K. Structure and campus following cerebral ischemia. J Cereb Blood physiological function of calpains. Biochem J 1997; 328: Flow Metab 1987; 7: 89–95. 721–732. 6. Nitatori T, Sato N, Waguri S et al. Delayed neuronal 20. Croall DE, Demartino GN. Calcium-activated neutral death in the CA1 pyramidal cell layer of the gerbil protease (calpain) system: structure, function, and hippocampus following cerebral ischemia is apoptosis. regulation. Physiol Rev 1991; 81: 813–847. J Neurosci 1995; 15: 1001–1011. 21. Chapman HA, Riese RJ, Shi GP. Emerging roles for 7. Li Y, Chopp M, Powers C. Granule cell apoptosis and cysteine proteases in human biology. Annu Rev Physiol protein expression in hippocampal dentate gyrus after 1997; 59: 63–88. forebrain ischemia in the rat. J Neurol Sci 1997; 150: 22. Keppler D, Sameni M, Moin K et al. Tumor progression 93–102. and angiogenesis: cathepsin & Co. Biochem Cell Biol 8. Kristián T, Siesjö BK. Calcium in ischemic cell death. 1996; 74: 799–810. Stroke 1998; 29: 705–718. 23. Villa PG, Henzel WJ, Sensenbrenner M et al. Calpain 9. Martins E, Yanase H, Sakai K et al. Accumulation of inhibitors, but not caspase inhibitors, prevent actin calcium and loss of potassium in the hippocampus and DNA fragmentation during apoptosis. following transient cerebral ischemia: a proton micro- J Cell Sci 1998; 111: 713–722. probe study. J Cereb Blood Flow Metab 1988; 8: 24. Roberts-Lewis JM, Savage MJ, Marcy VR et al. 531–538. Immunolocalization of calpain I-mediated spectrin 10. Mitani A, Takeyasu S, Yanase H et al. Changes in degradation to vulnerable neurons of the ischemic intracellular Ca2 and energy levels during in vitro gerbil brain. J Neurosci 1994; 14: 3934–3944. ischemia in the gerbil hippocampal slice. J Neurochem 25. Pike BR, Zhao X, Newcomb JK et al. Regional calpain 1994; 62: 626–634. and caspase-3 proteolysis of α-spectrin after traumatic 11. Nakamura K, Hatakeyama T, Furuta S et al. The role brain injury. Neuroreport 1998; 9: 2437–2442. of early Ca2 influx in the pathogenesis of delayed 26. Posmantur RM, Zhao X, Kampfl A et al. Immunoblot neuronal death after brief forebrain ischemia in gerbils. analyses of the relative contributions of cysteine and Brain Res 1993; 613: 181–192. aspartic proteases to neurofilament breakdown pro- 12. Colbourne F, Li H, Buchan AM et al. Continuing ducts following experimental brain injury in rats. postischemic neuronal death in CA1: influence of Neurochem Res 1998; 23: 1265–1276. ischemia duration and cytoprotective doses of NBQX 27. Canu N, Dus L, Barbato C et al. Tau cleavage and and SNX-111 in rats. Stroke 1999; 30: 662–668. dephosphorylation in cerebellar granule neurons 13. Ishida A, Shimazaki K, Kawai N. Ischemia-induced undergoing apoptosis. J Neurosci 1998; 18: 7061–

changes in PIP2 levels in gerbil hippocampus. Neurosci 7074. Res 1992; 15: 305–309. 28. Gregoriou M, Willis AC, Pearson MA et al. The calpain 14. Yamashima T, Takita M, Akaike S et al. Temperature- cleavage sites in the epidermal growth factor receptor dependent Ca2 mobilization induced by hypoxia- kinase domain. Eur J Biochem 1994; 223: 455–464. hypoglycemia in the monkey hippocampal slices. 29. Watt F, Molloy PL. Specific cleavage of transcription Biochem Biophys Res Commun 1994; 205: 1843–1849. factors by the thiol protease, m-calpain. Nucl Acid Res 15. Nomura Y. Involvement of glial NO synthase/NO in 1993; 21: 5092–5100. neuronal apoptosis in the brain. Neuropathology 1999; 30. Neumar RW, DeGracia DJ, White BC et al. Eukaryotic 19: 1–9. initiation factor 4E degradation during brain ischemia. 16. Yamashima T, Saido TC, Takita M et al. Transient brain J Neurochem 1995; 65: 1391–1394. 2 ischemia provokes Ca , PIP2 and calpain responses 31. Choi YH, Lee SJ, Nguyen P et al. Regulation of cyclin prior to delayed neuronal death in monkeys. Eur J D1 by calpain protease. J Biol Chem 1997; 272: Neurosci 1996; 8: 1932–1944. 28 479–28 484. 364 AB Tontchev and T Yamashima

32. Kubbutat MH, Vousden KH. Proteolytic cleavage of 47. Spizz G, Blackshear PJ. Identification and character- human p53 by calpain: a potential regulator of protein ization of cathepsin B as the cellular MARCKS stability. Mol Cell Biol 1997; 17: 460–468. cleaving enzyme. J Biol Chem 1997; 272: 23 833–23 842. 33. Kimura Y, Koga H, Araki N et al. The involvement of 48. Vancompernolle K,Van Herreweghe F, Pynaert G et al. calpain-dependent proteolysis of the tumor suppressor Atractyloside-induced release of cathepsin B, a pro- NF2 (merlin) in schwannomas and meningiomas. Nat tease with caspase-processing activity. FEBS Lett 1998; Med 1998; 4: 915–922. 438: 150–158. 34. Croall DE, Chacko S, Wang Z. Cleavage of caldesmon 49. Ishisaka R, Utsumi T, Yabuki M et al. Activation of and calponin by calpain: substrate recognition is not caspase-3-like protease by digitonin-treated lysosomes. dependent on calmodulin binding domains. Biochim FEBS Lett 1998; 435: 233–236. Biophys Acta 1996; 1298: 276–284. 50. Wang KK, Postmantur R, Nadimpalli R et al. Caspase- 35. Porn-Ares MI, Samali A, Orrenius S. Cleavage of the mediated fragmentation of calpain inhibitor protein calpain inhibitor, calpastatin, during apoptosis. Cell calpastatin during apoptosis. Arch Biochem Biophys Death Differ 1998; 5: 1028–1033. 1998; 356: 187–196. 36. Walker G, Pfeilschifter J, Kunz D. Mechanisms of 51. Saido TC, Nagao S, Shiramine M et al. Autolytic suppression of inducible nitric-oxide synthase (iNOS) transition of µ-calpain upon activation as resolved by expression in interferon (IFN)-gamma-stimulated antibodies distinguishing between the pre- and post- RAW 264.7 cells by dexamethasone. Evidence for autolysis forms. J Biochem 1992; 111: 81–86. glucocorticoid-induced degradation of iNOS protein by 52. Bernstein HG, Kirschke H, Wiederanders B et al. calpain as a key step in post-transcriptional regulation. Cathepsin B immunoreactive neurons in rat brain. J J Biol Chem 1997; 272: 16 679–16 687. Hirnforsch 1988; 29: 17–19. 37. Tompa P, Baki A, Schad E et al. The calpain cascade. 53. Bando Y, Kominami E, Katunuma N. Purification and m-calpain activates m-calpain. J Biol Chem 1996; 271: tissue distribution of cathepsin L. Biochem Tokyo 1986; 33 161–33 164. 100: 35–42. 38. Mellgren RL, Lu Q. Selective nuclear transport of 54. Kohda Y, Yamashima T, Sakuda K et al. Dynamic mu-calpain. Biochem Biophys Res Commun 1994; 204: changes of cathepsin B and L expression in the 544–550. monkey hippocampus after transient ischemia. 39. Creemers LB, Hoeben KA, Jansen DC et al. Partici- Biochem Biophys Res Commun 1996; 228: 616–622. pation of intracellular cysteine proteinases, in par- 55. Ferrer I, Tortoise A, Macaya A et al. Evidence of ticular cathepsin B, in degradation of collagen in nuclear DNA fragmentation following hypoxia- periosteal tissue explants. Matrix Biol 1998; 16: 575–584. ischemia in the infant rat brain, and transient forebrain 40. Lambert Vidmar S, Lottspeich F et al. Latent ischemia in the adult gerbil. Brain Pathol 1994; 4: fibronectin-degrading serine proteinase activity in N- 115–122. terminal heparin-binding domain of human plasma 56. Charriaut-Marlangue C, Aggoun-Zouaoui D, Represa fibronectin. Eur J Biochem 1991; 201: 71–77. A et al. Apoptotic features of selective neuronal death 41. Williams KR, Williams ND, Konigsberg W et al. Acidic in ischemia, epilepsy and gp120 toxicity. Trends lipids enhance cathepsin D cleavage of the myelin basic Neurosci 1996; 19: 109–114. protein. J Neurosci Res 1986; 15: 137–145. 57. Thornberry NA, Lazebnik Y.Caspases: enemies within. 42. Banay-Schwartz M, Dahl D, Hui KS, Lajtha A. The Science 1998; 281: 1312–1316. breakdown of the individual neurofilament proteins by 58. Cohen GM. Caspases: the executioners of apoptosis. cathepsin D. Neurochem Res 1987; 12: 361–367. Biochem J 1997; 326: 1–16. 43. Johnson GV, Litersky JM, Whitaker JN. Proteolysis 59. Chen J, Nagayama T, Jin K et al. Induction of caspase-3- of microtubule-associated protein 2 and tubulin by like protease may mediate delayed neuronal death in cathepsin D. J Neurochem 1991; 57: 1577–1583. the hippocampus after transient cerebral ischemia. J 44. McDermott JR, Gibson AM. Degradation of Neurosci 1998; 18: 4914–4928. Alzheimer’s b-amyloid protein by human cathepsin D. 60. Ni B, Wu X, Su X et al. Transient global forebrain Neuroreport 1996; 7: 2163–2166. ischemia induces a prolonged expression of the 45. Savage P, Shetty SS, Martin LL et al. Conversion of caspase-3 mRNA in rat hippocampal CA1 pyramidal proendothelin-1 into endothelin-1 by asparylproteases. neurons. J Cereb Blood Flow Metab 1998; 18: 248– Int J Pept Protein Res 1993; 42: 227–232. 256. 46. Jutras I, Reudelhuber TL. Prorenin processing by 61. Himi T, Ishizaki Y,Murota S.A caspase inhibitor blocks cathepsin B in vitro and in transfected cells. FEBS Lett ischemia-induced delayed neuronal death in the gerbil. 1999; 443: 48–52. Eur J Neurosci 1998; 10: 777–781. Cysteine proteases in neuronal death 365

62. Pike BR, Zhao X, Newcomb JK et al. Temporal 73. Kimura M, Saji M. Protective effect of a low dose of relationship between de novo protein synthesis, calpain colchicine on the delayed cell death of hippocampal and caspase-3-like protease activation, and DNA CA1 neurons following transient forebrain ischemia. fragmentation during apoptosis in septo-hippocampal Brain Res 1997; 774: 229–233. cultures. J Neurosci Res 1998; 52: 505–520. 74. Tokime T, Nozaki K, Kikuchi H. Neuroprotective 63. Waterhouse NJ, Finucane DM, Green DR et al. Calpain effect of FK506, an immunosuppressant, on transient activation is upstream of caspases in radiation- global ischemia in gerbil. Neurosci Lett 1996; 206: induced apoptosis. Cell Death Differ 1998; 12: 81–84. 1051–1061. 75. Sasaki K, Oomura Y, Suzuki K et al. Acidic fibroblast 64. Green DR, Reed JC. Mitochondria and apoptosis. growth factor prevents death of hippocampal CA1 Science 1998; 281: 1309–1312. pyramidal cells following ischemia. Neurochem Int 65. Hirsch T, Susin SA, Marzo I et al. Mitochondrial 1992; 21: 397–402. permeability transition in apoptosis and necrosis. Cell 76. Miyazata T, Matsumoto K, Ohmichi H et al. Protection Biol Toxicol 1998; 14: 141–145. of hippocampal neurons from ischemia-induced 66. Perez-Pinzon MA, Xu GP, Born J et al. Cytochrome c delayed neuronal death by hepatocyte growth factor: a is released from mitochondria into the cytosol after novel neurotrophic factor. J Cereb Blood Flow Metab cerebral anoxia or ischemia. J Cereb Blood Flow Metab 1998; 18: 345–348. 1999; 19: 39–43. 77. Wen TC, Takaka J, Peng H et al. Interleukin 3 prevents 67. Friberg H, Connern C, Halestrap AP et al. Differences delayed neuronal death in the hippocampal CA1 field.J in the activation of the mitochondrial permeability Exp Med 1998; 188: 635–649. transition among brain regions in the rat correlate with 78. Nicotera P, Leist M, Manzo L. Neuronal cell death: a selective vulnerability. J Neurochem 1999; 72: 2488– demise with different shapes. Trends Pharmacol Sci 2497. 1999; 20: 46–51. 68. Li PA, Uchino H, Elmer E et al. Amelioration by 79. Martin LJ, Al-Abdula NA, Brambrink AM et al. cyclosporin A of brain damage following 5–10 min of Neurodegeneration in excitotoxicity, global cerebral ischemia in rats subjected to preischemic hyper- ischemia, and target deprivation: a perspective on the glycemia. Brain Res 1997; 753: 133–140. contributions of apoptosis and necrosis. Brain Res Bull 69. Lee KS,Yanamoto H, Fergus A et al. Calcium-activated 1998; 46: 281–309. proteolysis as a therapeutic target in cerebrovascular 80. Shimizu S, Eguchi Y, Kamiike W et al. Retardation disease. Ann N Y Acad Sci 1997; 825: 95–103. of chemical-induced necrotic death by bcl-2 and 70. Lee KS, Frank S, Vanderklish P et al. Inhibition of pro- ICE inhibitors: possible involvement of common teolysis protects hippocampal neurons from ischemia. mediators in apoptotic and necrotic signal trans- Proc Natl Acad Sci USA 1991; 88: 7233–7237. ductions. Oncogene 1996; 12: 2045–2050. 71. Himi T, Ishizaki Y,Murota S.A caspase inhibitor blocks 81. Suzuki A. Amyloid b-protein induces necrotic cell ischemia-induced delayed neuronal death in the gerbil. death mediated by ICE cascade in PC12 cells. Exp Cell Eur J Neurosci 1998; 10: 777–781. Res 1997; 234: 507–511. 72. Schwartz RD, Yu X, Katzman MR et al. Diazepam, 82. Higuchi M, Honda T, Proske RJ et al. Regulation of given postischemia, protects selectively vulnerable reactive oxygen species-induced apoptosis and necrosis neurons in the rat hippocampus and striatum. J by caspase 3-like proteases. Oncogene 1998; 26: Neurosci 1995; 15: 529–539. 2753–2760.