Immunohistochemical and biochemical assessment of caspase-3 activation and DNA fragmentation following transient focal ischemia in the rat

PII: S 0 3 0 6 - 4 5 2 2 ( 0 2 ) 0 0 3 7 6 - 7

Neuroscience Vol. 115, No. 1, pp. 125^136, 2002 C 2002 IBRO. Published by Elsevier Science Ltd All rights reserved. Printed in Great Britain 0306-4522 / 02 $22.00+0.00

www.neuroscience-ibro.com

IMMUNOHISTOCHEMICAL AND BIOCHEMICAL ASSESSMENT OF CASPASE-3 ACTIVATION AND DNA FRAGMENTATION FOLLOWING TRANSIENT FOCAL ISCHEMIA IN THE RAT M. A. DAVOLI,1 J. FOURTOUNIS,1 J. TAM, S. XANTHOUDAKIS, D. NICHOLSON, G. S. ROBERTSON, G. Y. K. NG and D. XU Merck Frosst Center for Therapeutic Research, P.O. Box 1005, Pointe Claire-Dorval, QC, Canada H9R 4P8

Abstract6In the present study, we evaluated the time-course of caspase-3 activation, and the evolution of cell death following focal cerebral ischemia produced by transient middle cerebral artery occlusion in rats. Ischemia-induced active caspase-3 immunoreactivity in the striatum but not the cortex at 3 and 6 h time points post-reperfusion. Furthermore, using a novel approach to visualize enzymatic activity, vC-APP, a C-terminal cleavage product of APP generated by caspase-3, was found to immunolocalize to the same areas as active caspase-3. Double-labeling studies demonstrated colocalization of these two proteins at the cellular level. Further double-labeling experiments revealed that active caspase-3 was con¢ned to neuronal cells which were still viable and thus immunoreactive for NeuN. DNA fragmentation, assessed histologically by terminal dUTP nick-end labeling (TUNEL), was observed in a small number of cells in the striatum as early as 3 h, but only began to appear in the cortex by 6 h. DNA fragmentation was progressive, and by 24 h postreperfusion, large portions of both the striatum and cortex showed TUNEL positive cells. However, double-labeling of active caspase-3 with TUNEL showed only minimal co-localization at all time-points. Thus, caspase-3 activation is an event that appears to occur prior to DNA fragmentation. As a con¢rmation of the histological TUNEL data, 24 h ischemia also induced the generation of nucleosome fragments, evidenced by cell death enzyme-linked immunosorbent assay. Using a novel ischemia-induced substrate cleavage biochemical approach, spectrin P120 fragment, a caspasespeci¢c cleavage product of alpha II spectrin, a cytoskeletal protein, was shown to be elevated by western blotting. Brain concentrations of both nucleosomes and spectrin P120 correlate with the degree of injury previously assessed by triphenyltetrazolium chloride staining and infarct volume calculation. Together, our ¢ndings suggest a possible association between caspase-3 activation and ischemic cell death following middle cerebral artery occlusion brain injury. C 2002 IBRO. Published by Elsevier Science Ltd. All rights reserved. Key words: apoptosis, nucleosomes, spectrin, amyloid precursor protein, infarct.

and -7), and in£ammatory mediators (caspase-1, -4, -5, -11, -13). There is some amount of evidence in the literature linking ischemia-induced neuronal apoptosis to the caspase family of enzymes; however, their direct role is still under scrutiny (for review see Loetscher et al., 2001). The most notable association is seen in experimental models that elicit mild injury in the adult brain (Endres et al., 1998; Chen et al., 1998; Himi et al., 1998), and in neonate models of hypoxic ischemia (Pulera et al., 1998; Cheng et al., 1998; Liu et al., 1999; Nakajima et al., 2000). Caspase-3 is believed to be the main executioner protease of the apoptotic cascade, as evidenced in studies of caspase-3 knockout mice that die approximately 3 weeks after birth due to massive brain anomalies including ectopic growths and hyperplasias resulting in disorganized, overdeveloped structures (Kuida et al., 1996). Caspase-3 exists as a 32-kDa proenzyme that can be activated through catalytic cleavage by a number of proteases, thus yielding two smaller sub-units of 17 kDa (resulting from a subsequent autocatalytic cleavage of a 20-kDA pro-domain containing fragment) and 12 kDA. The active form of the enzyme exists as a tetrameric enzyme complex composed of two large (17-kDA) and two small (12-kDA) sub-units (Nicholson et al., 1995).

Caspases are homologous cysteine proteases that exhibit a primary speci¢city to cleave after aspartic acid residues in peptide substrates, allowing for highly selective proteolysis in animal cells (Stennicke and Salvesen, 1998). At least 14 members of the caspase family have thus far been identi¢ed (Eldadah and Faden, 2000). Based on their preferred substrate speci¢cities (tetrapeptide motifs), which have been de¢ned using a positional scanning combinatorial library (Rano et al., 1997; Thornberry et al., 1997), these enzymes can be divided into three general groups: apoptotic initiators (caspase-8 and caspase-9), apoptotic executioners (caspase-3, -6,

1 Both authors contributed equally to the work. *Corresponding author. Tel. : +1-514-428-3661; fax: +1-514-4283921. E-mail address: [email protected] (D. Xu). Abbreviations : vC-APP, C-terminal neo-epitope of APP generated by caspase cleavage ; APP, amyloid precursor protein; CCA, common carotid artery; CDE, cell death ELISA; EDTA, ethylenediaminetetra-acetate ; ELISA, enzyme-linked immunosorbent assay ; FITC, £uorescein isothiocyanate; MCA, middle cerebral artery; MCAo, MCA occlusion ; PBS, phosphate-bu¡ered saline ; TTC, triphenyltetrazolium chloride; TUNEL, terminal dUTP nick-end labeling.

125

NSC 5800 14-10-02

Cyaan Magenta Geel Zwart

126

M. A. Davoli et al.

Caspase-3 is known to exert its detrimental e¡ects by cleaving and disabling DNA repair proteins and other substrates involved in maintaining cellular structural integrity. For instance, caspase-3 is known to target the inhibitor of caspase-activated deoxyribonuclease, causing its dissociation from the inhibitor of caspase-activated deoxyribonuclease/caspase-activated deoxyribonuclease complex, thus allowing the endonuclease to cleave chromosomal DNA, which is considered to be a hallmark of caspase activation (Nagata, 2000). Furthermore, some structural cytoskeletal proteins, such as spectrin, for example (Wang et al., 1998), are known to be cleaved by caspase-3 thus leading to unfavorable e¡ects on cellular integrity. Several models of experimental cerebral ischemia have been reported to induce the activation of caspase-3 (Li et al., 1995a,b; Namura et al., 1998; Chen et al., 1998; Velier et al., 1999). Several areas of the brain have been reported to exhibit caspase-3 activation in animal models of global ischemia, namely the CA1 region of the hippocampus (Kihara et al., 1994; Ni et al., 1998) and the striatum (Chen et al., 1998). Much of the documented data is histologically based and there is a great lack of biochemical data in the literature (Loetscher et al., 2001). Thus, in this study, we set out to investigate the activation of caspase-3 in a rat model of focal ischemia by both immunohistochemical detection of the active enzyme and a cleaved fragment of its amyloid precursor protein (APP) substrate, as well as by using novel biochemical markers, such as the generation of caspase-3speci¢c spectrin P120 band which was detected in our studies by western blot. The relationship between caspase-3 activation and ischemic cell death was assessed by monitoring the progression of DNA fragmentation both by terminal dUTP nick-end labeling (TUNEL) and cell death ELISA (CDE; ELISA: enzyme-linked immunosorbent assay) in this rat model of middle cerebral artery occlusion (MCAo).

EXPERIMENTAL PROCEDURES

MCAo surgical procedure Transient focal ischemia was produced in male Wistar rats weighing between 250 and 300 g by unilateral occlusion of the middle cerebral artery (MCA), and bilateral occlusion of the common carotid artery (CCA) as previously described by Buchan et al. (1992). In brief, rats were anaesthetized with 2% iso£urane and were placed on a heating pad to maintain body temperature to within normal limits. Furthermore, core body temperature was monitored through a rectal probe and maintained at a constant 37‡C throughout the surgery by external heating if necessary. The rats were then placed on their left side and the head was immobilized in a stereotaxic frame. Following this, a small 1 cm incision was made between the right eye and the ear. Following craniotomy using a surgical drill (2 mm burr), the dura was removed to reveal the underlying MCA. After this, the MCA was occluded for 90 min using a Codman^Sundt micro aneurysm clip (1 mm). The common carotids were also occluded during this time, with the right (ipsilateral) CCA being permanently ligated using a suture and the left artery transiently occluded using a micro clip during the MCAo procedure. Following a 90 min occlusion, the rats were re-anaesthetized to remove the clips on the MCA and the left

NSC 5800 14-10-02

CCA for reperfusion. There were a total of 18 animals in the histological time-course study (n = 6 for each reperfusion timepoint: 3, 6, and 24 h). Tissue preparation for histological studies Following 3-, 6- and 24-h reperfusion intervals, the MCAo rats were killed using pentobarbital overdose (six animals/timepoint). The brains were removed from the skull and £ash frozen in cold isopentane over dry ice, and stored at 380‡C until sectioning. Sections were collected in a cryostat at a thickness of 12 Wm from four di¡erent levels, two encompassing the striatum (bregma+1.60 mm and bregma+0.70 mm) and two more from the region spanning part of the hippocampus (bregma31.80 mm and bregma32.30 mm; Paxinos and Watson, 1998). Single-labeling immunohistochemistry Histological assessment for active caspase-3 following focal ischemia was performed on fresh frozen brain slices. In brief, sections were incubated with A92790K neo-epitope polyclonal rabbit active caspase-3 antibody (Simpson et al., 2001) (1:32000; Merck Frosst, Kirkland, QC, Canada) in phosphate-bu¡ered saline (PBS) containing 0.3% Triton X-100 for 24 h at 4‡C in a humidi¢ed chamber. After washing, the tissue was incubated with Cy3-conjugated donkey anti-rabbit IgG (1:1000; Jackson Immunoresearch, West Grove, PA, USA) for 3 h at room temperature, and then mounted with Vectashield mounting medium for £uorescence (Vector Laboratories, Burlingame, CA, USA). To con¢rm the speci¢city of A92790K, a monoclonal antibody that recognizes the active form of caspase3 (1:1000; Pharmingen Canada, Mississauga, ON, Canada) was used on serial brain sections in the same manner. A rabbit polyclonal antibody E82684K or vC-APP (a C-terminal cleavage product of APP generated by caspase-3) (1:5000; Merck Frosst, Kirkland, QC, Canada), was used to further solidify the activity of caspase-3 in tissue sections (Gervais et al., 1999). For all experiments, the contralateral side of the brain was used as an intrinsic negative control since it is una¡ected by ischemia during the MCAo surgery. Immunohistochemistry was also performed without incubation with the primary antibody as an experimental negative control. This allowed us to determine the experimental signal to noise ratio. Double-labeling immunohistochemistry with vC-APP and active caspase-3/NeuN and active caspase-3 Since both active caspase-3 and vC-APP in-house antibodies are raised in the same species, posing a challenge for recognition by secondary anti-rabbit antibodies, as a primary antibody, a commercially available biotinylated form of the active caspase-3 antibody was used. In previous single-labeling experiments, this antibody has been shown to have similar localization to our inhouse antibody (unpublished observations). Double-labeling was thus carried out by ¢rst incubating the tissue sections with vC-APP at 1:5000, as mentioned above, followed by incubation with Cy3-conjugated donkey anti-rabbit IgG (1:1000; Jackson Immunoresearch, West Grove, PA, USA) for 3 h at room temperature. Sections were then washed stringently to remove all unbound secondary antibody, ¢xed in 4% paraformaldehyde, and then incubated with a biotinylated active caspase-3 polyclonal rabbit antibody (1:1000; Pharmingen Canada, Mississauga, ON, Canada) for 24 h at 4‡C in a humidi¢ed chamber. Following washing, sections were incubated with streptavidin £uorescein isothiocyanate (FITC) (1:200; Amersham Pharmacia Biotech, Baie d’Urfe¤, QC, Canada) for 3 h at room temperature. To determine the presence of active caspase-3 in neurons following MCAo, double-labeling immunohistochemistry was performed using the A92790K antibody with the A60 mouse monoclonal antibody that recognizes the neuron-speci¢c protein NeuN (Mullen et al., 1992). In brief, sections were incubated with an equal mixture of A92790K (1:16000) and A60 (1:100)

Cyaan Magenta Geel Zwart

Caspase mediated neuronal death following transient MCAo

for 24 h at 4‡C (resulting in a ¢nal concentration of 1:32000 for A92790K and 1:200 for A60). The sections were washed in PBS and incubated with equal mixtures of Cy3-conjugated donkey anti-rabbit IgG (1:500; Jackson Immunoresearch, West Grove, PA, USA) and FITC-conjugated donkey anti-mouse IgG (1:100; Jackson Immunoresearch, West Grove, PA, USA) for 3 h at room temperature (resulting in a ¢nal concentration of 1:1000 for the Cy3 and 1:200 for the FITC). In situ labeling of DNA fragmentation The Apoptag0 £uorescein in situ apoptosis detection kit (Intergen Company, Purchase, NY, USA) was used to determine DNA damage following transient MCAo. Tissue sections on slides were ¢xed in 1% paraformaldehyde in 0.01 M PBS, pH 7.4 in a coplin jar for 10 min at room temperature. Sections were then washed in two changes of 0.01 M PBS for 5 min each. Sections were post-¢xed in pre-cooled ethanol:acetic acid (2:1 v/v) for 5 min at 320‡C. Sections were then washed in 2 changes of 0.01 M PBS for 5 min each. The TUNEL procedure was conducted according to the manufacturer’s protocol as speci¢ed in the Apoptag0 manual. Double-labeling immunohistochemistry with TUNEL/caspase-3 and TUNEL/NeuN To establish whether caspase-3 immunoreactive cells co-localized with TUNEL positive cells, double-labeling was performed using the A92790K active caspase-3 antibody with TUNEL. The TUNEL procedure was performed ¢rst followed by caspase-3 immunohistochemistry. A double-labeling immunohistochemistry with NeuN and TUNEL was carried out following the same protocol as above to see whether the cells showing signs of DNA damage were neurons. Infarct assessment by Cresyl Violet staining Cresyl Violet, a dye that stains the Nissl bodies in the stellate somas of viable neurons, was used to assess infarct size. Sections were added to a solution containing 1% Cresyl Violet in 0.25% acetic acid for 3 min (Sigma-Aldrich Canada, Oakville, ON, Canada). The sections were then rinsed in tap water followed by serial dehydrations in 50%, 75%, and 95% ethanol, and two changes of xylenes for 5 min each. The sections were mounted with D.P.X. neutral mounting medium (Sigma-Aldrich Canada, Oakville, ON, Canada) and were visualized both with the naked eye and under light microscopy for infarct development as a result of the MCAo surgery. Measurement of nucleosomes by CDE MCAo brains from 34 animals were harvested and cut into 2 mm slices and stained with 1.5% triphenyltetrazolium chloride (TTC), 24 h following MCAo. The slices were digitally scanned and the infarct volume measured. Cortices from TTC-stained brain slices were then dissected and frozen in liquid nitrogen. Brain extracts were prepared from the slices by homogenization in a lysis bu¡er (1% NP40, 2 mM EDTA, 50 mM Tris pH 7.5) containing Complete1 protease inhibitor cocktail (Roche Diagnostics, Laval, QC, Canada), a Merck caspase-speci¢c inhibitor, M920, and a calpain inhibitor, MDL28, 170, at a ¢nal concentration of 1 WM each. Protein concentration in the supernatant was quanti¢ed using the Pierce BCA assay (Rockford, IL, USA) according to manufacturer’s protocol. Ten microlitres of supernatant equivalent to 50 Wg of protein was used for CDE on a 96well plate using a CDE kit from Roche Diagnostics (Laval, QC, Canada). Arbitrary nucleosome units of the lysates, relative to the positive control curve, were calculated using the Softmax software, based on the absorbance at 650 nm on a spectromax (Molecular Devices, Sunnyvale, CA, USA) plate reader.

NSC 5800 14-10-02

127

Western blotting analysis of spectrin P120 Western blotting analysis for spectrin cleavage by active caspase-3 was performed on a subgroup of TTC-stained animals (n = 14 of 34) that were used for the CDE analysis above. Cellular lysates (generated as indicated above) containing 50 Wg protein were loaded onto a 4^20% poly-acrylamide gradient gel for electrophoresis. Next, the separated proteins were transferred onto a nitrocellulose membrane which was then blotted by sequential incubations with Mab1622 anti-spectrin mouse monoclonal antibody (1:1000; Chemicon International, Temecula, CA, USA), and anti-mouse-HRP (1:5000; Amersham Pharmacia Biotech, Baie d’Urfe¤, QC, Canada), for 1 h each. Labeling was developed in the dark using the SuperSignal Femto Substrate (Pierce, Rockford, IL, USA).

RESULTS

Immunolabeling Using A92790K, scattered caspase-3 immunoreactive neurons were observed in a narrow zone near the edge of the ischemic infarct in the ipsilateral striatum for four out of six rats at 3 h, and four out of six rats at 6 h postreperfusion (Fig. 1B, C, G). There was no caspase-3 staining seen at 24 h for any of the animals studied (data not shown). Other antibodies that recognize the active form of the caspase-3 enzyme were used on serial sections and yielded immunolabeling patterns within the striatum that were consistent with that observed with A92790K (data not shown). The contralateral hemisphere (uninjured side, Fig. 1A) yielded no comparable staining, and omission of the primary antibody also produced no staining (data not shown), con¢rming the speci¢city of the A92790K antibody for active caspase-3. We did not observe any immunoreactivity for active caspase-3 in the cortex at any of the three time-points investigated. Single-labeling immunohistochemistry for vC-APP showed a staining pattern that was localized to the same areas of the striatum where active caspase-3 staining was observed in the 3 and 6 h positive animals (Fig. 1E, F, G). Only very weak staining for vC-APP was observed in the cortex which does not correlate with the active caspase-3 signal (data not shown). Doublelabeling studies showed an almost perfect co-localization of caspase-3 positive cells with vC-APP positive cells at both 3 and 6 h post-reperfusion (Fig. 2A^C). Doublelabeling experiments for active caspase-3 and NeuN showed that the active caspase-3 immunoreactive cells were also stained with the neuron-speci¢c marker NeuN (Fig. 2D^F) at both time-points. TUNEL labeling TUNEL revealed scattered DNA fragmentation by 3 h in the ipsilateral striatum for ¢ve out of six rats (Fig. 3B). At 6 h, an increase in TUNEL reactivity was observed for ¢ve out of six rats in the striatum with some minor labeling in the cortex (Fig. 3C, G). By 24 h, ¢ve out of the six rats investigated showed massive TUNEL labeling in the striatum and all of the rats exhibited TUNEL throughout the cortex delineating the core of the infarct

Cyaan Magenta Geel Zwart

128

M. A. Davoli et al.

Fig. 1. Active caspase-3 (A^C) and vC-APP (D^F) immunostaining of 12-Wm cryosections from adult Wistar rats subjected to MCAo and subsequently reperfused for various lengths of time. Intense caspase-3 immunoreactivity is seen scattered throughout the ipsilateral striatum at 3 (B) and 6 h (C), but not at 24 h post-reperfusion (not shown). The same pattern is seen in the ipsilateral striatum stained with an antibody to the caspase-3 cleaved APP protein (E, F; 24 h, not shown). The corresponding contralateral controls (A, D) show no immunoreactivity, as do experimental negative controls with the omission of the primary antibodies (not shown). In G, a schematic representation of pooled data (n = 6 animals per time-point) shows the scattered staining pattern observed in the striatum with active caspase-3 and vC-APP at the 3- and 6-h postreperfusion time-points. When comparing individual animals, serial sections stained with the two antibodies show strikingly similar patterns of staining.

(Fig. 3D, H). TUNEL labeling as seen with £uorescence microscopy for striatal and cortical tissue at the various reperfusion times can be seen in Fig. 3A^H. A diagrammatic summary of the TUNEL-labeling pattern for all the rats examined in this study is reported in Fig. 3I. Double-labeling immunohistochemistry of TUNEL

NSC 5800 14-10-02

with A92790K anti-active caspase-3 revealed that only a very small number of cells co-localized in the striatum (Fig. 4, arrows). Double-labeling with TUNEL and NeuN showed that all TUNEL positive cells at the various reperfusion times were con¢ned to neurons (Fig. 5). Thus, as neuronal viability is a¡ected by ischemia, NeuN

Cyaan Magenta Geel Zwart

Caspase mediated neuronal death following transient MCAo

129

Fig. 2. Double-labeling of active caspase-3 and either vC-APP (A^C) or NeuN (D^F). The strikingly similar pattern of staining between caspase-3 (A) and vC-APP (B) reported in the previous ¢gure is con¢rmed by this double-labeling study. The results indicate that there is almost perfect co-localization between these two markers for active caspase-3 in the ipsilateral striatum (C). Co-localization of active caspase-3 in the striatum (D) is seen to occur with cells expressing NeuN (E) as visualized in the merged image (F). This suggests that most of the caspase-3 positive cells in the striatum are neurons. Only background staining is seen in sections in which the primary antibodies have been omitted (not shown).

expression is decreased and TUNEL labeling is increased (Fig. 5C, F, respectively). Infarct assessment by Cresyl Violet staining The Cresyl Violet-stained tissue allowed for infarct detection (Fig. 3J). At 24 h, four out of six animals displayed striatal and cortical damage typical of transient MCAo injury. One animal showed only striatal damage and another showed no detectable infarct (although some cortical TUNEL positivity was observed in this animal). Although ischemic damage as determined by Cresyl Violet is most convincing at 24 h, there was some indication of damage in the striatum in three out of six animals at 6 h, but at 3 h, no macroscopic infarct was noticeable by this method (data not shown).

NSC 5800 14-10-02

Ischemia-induced nucleosome generation Nucleosomes in ischemic brain tissue harvested 24 h after reperfusion were measured by CDE. The CDE assay was conducted using brain slices previously stained with TTC for infarct volume measurement. Upon examination of 34 MCAo animals, it was observed that cortical nucleosome concentration correlated strongly with infarct volume measured by TTC staining (r = 0.89, Fig. 6A). However, it is noteworthy that nucleosomes were sometimes seen in animals without detectable infarcts by TTC staining (points along the graphical y-axis). This may be accounted for by a relatively small number of dead or dying cells within the ischemic area that are not yet detectable by the less sensitive TTC staining methodology.

Cyaan Magenta Geel Zwart

130

M. A. Davoli et al.

Fig. 3. TUNEL staining, as an indicator of ischemic cell death, in the striatum (A^D) and cortex (E^F) of MCAo animals. Scattered TUNEL positive cells can be seen in the striatum as early as 3 h post-reperfusion (B), become more numerous by 6 h (C), and encompass almost the entire striatum by 24 h (D). In the cortex, TUNEL positive cells are only observed after 6 h post-reperfusion (G), and become maximal within the infarcted area by 24 h (H). No staining is seen in either brain region in the contralateral side (A, E). Macroscopically at 24 h post-reperfusion, the infarcted regions can be visualized by Cresyl Violet staining as lighter stained areas in the right hemisphere (J). TUNEL staining is observed within these lighter stained areas of the cortex and striatum. In I, the diagrammatic representation of pooled data (n = 6 animals per time-point) shows scattered TUNEL positive cells in the ventrolateral striatum of animals as early as 3 h post-reperfusion but not in the cortex at the same time-point. Some scattered cortical labeling is observed by 6 h post-reperfusion as the number of TUNEL positive cells increases in the striatum, but the signal in both brain regions becomes most apparent by 24 h.

Caspase-3-mediated spectrin cleavage To further determine evidence for caspase-3 activation in the brain and apoptosis in ischemic brain tissues, we examined the appearance of spectrin P120 by western

NSC 5800 14-10-02

blot in a subgroup of the animals used for the CDE analysis above (Fig. 6B). A band, at 120 kDa, corresponding to spectrin P120 was detected in the ipsilateral cortex, whereas very little of this fragment was seen in the contralateral, non-injured side. In addition to P120,

Cyaan Magenta Geel Zwart

Caspase mediated neuronal death following transient MCAo

Fig. 4. Double-labeling of active caspase-3 (A^C) and TUNEL (D^F) in the ipsilateral striatum of MCAo rats following 3 and 6 h post-reperfusion. Virtually no co-localization is seen at 3 h post-reperfusion (H). However, by 6 h we observe a few caspase-3 positive cells co-localizing with TUNEL in the striatum (arrows). At 24 h, no active caspase-3 can be detected ; however, TUNEL labeling is abundant in the striatum (not shown). This suggests that the expression of active caspase-3 appears to be an event that precedes DNA fragmentation and subsequent cell death.

NSC 5800 14-10-02

Cyaan Magenta Geel Zwart

131

132

M. A. Davoli et al.

Fig. 5. Double-labeling of NeuN, a neuronal marker, and TUNEL, an indicator of cell death, in the ipsilateral striatum of MCAo animals. Healthy neurons appear to be expressing an intense signal for NeuN at 3 (A) and 6 h post-reperfusion (B). Most of the TUNEL positive cells observed at the same time-points (D, E) appear to co-localize with the NeuN signal suggesting death of the neuronal population (arrows G, H). However, by 24 h NeuN immunoreactivity is dramatically reduced (C). This is correlated with an extensive increase in TUNEL labeling (F) indicating a loss of neuronal viability at this time-point.

two bands of molecular weight 145 kDa and 150 kDa were also detected, corresponding to larger cleavage products of caspase-3 (P150) and calpain (P150 and P145) (Pike et al., 1998). The amount of the spectrin

NSC 5800 14-10-02

P120 fragment generated by the ischemia was obtained by densitometry of the 120-kDA band. Cortical spectrin P120 level correlated well with infarct volume assessed by prior TTC staining (r = 0.84; -Fig. 6C).

Cyaan Magenta Geel Zwart

Caspase mediated neuronal death following transient MCAo

Fig. 6. In A, the concentration of nucleosomes was assayed by CDE and compared with infarct volumes measured by TTC staining in animals subjected to transient MCAo (n = 34). The correlation between the two parameters was seen to be signi¢cant (r = 0.89). In B, caspase-mediated cleavage of spectrin was assayed by western blot analysis. Low levels of P150 and P145 were detected in the contralateral cortical tissues, consistent with basal calpain activity. In the ipsilateral or ischemic cortex, a new band of molecular weight 120 kDa was detected, indicative of a caspase-3-generated P120 fragment. Levels of P150 and P145 were also signi¢cantly elevated. Densitometry in a subgroup of animals reported above (n = 14) revealed that the spectrin P120 band correlates well with infarct size measured by TTC staining (C; r = 0.84).

DISCUSSION

Transient focal ischemia produced by distal MCAo resulted in ischemic infarcts in the ipsilateral hemisphere, a¡ecting large portions of both the cortex and striatum, as visualized by both TTC staining and TUNEL. Ischemic injury is associated with progressive DNA fragmentation in both striatal and cortical neurons within the

NSC 5800 14-10-02

133

infarct area. Ischemia was also seen to elevate active caspase-3 immunoreactivity in striatal, but not cortical neurons. In contrast to the lack of active caspase-3 immunoreactivity in the cortex, caspase-3-mediated spectrin cleavage and nucleosome generation was induced in both cortex and striatum. Moreover, nucleosome and spectrin P120, markers for apoptosis and caspase-3 activation, correlated quite well with infarct size determined by TTC staining. In the striatum, the staining for active caspase-3 was localized to the nucleus, con¢rming the role of active caspase-3 as an inactivator of nuclear substrates such as poly ADP^ribose polymerase, commonly known as PARP (Martins et al., 1997; Velier et al., 1999). In addition to this, immunoreactivity for vC-APP was observed in the same neurons as caspase-3, consistent with our previous observation that caspase-3 is implicated with APP cleavage (Gervais et al., 1999). Active caspase-3 labeling was present in the border zone between ischemic and non-ischemic striatal tissues. Furthermore, doublelabeling experiments for caspase-3 and TUNEL showed co-localization of active caspase-3 in a small number of cells in the striatum with TUNEL positive cells, which was observed in the transition zone between areas harboring active caspase-3 and TUNEL. Thus, it would appear as though the event is a temporal one in which caspase-3 activation appears to precede cell death. This was not an unexpected ¢nding given the known role of the caspase family of enzymes in DNA fragmentation. This ¢nding has been con¢rmed and documented by others (Sasaki et al., 2000). The results were also con¢rmed by co-localization studies using Fluoro-Jade (Schmued et al., 1997; Schmued and Hopkins, 2000), a £uorochrome capable of e¡ectively distinguishing degenerating neurons from other cell populations (data not shown). The observation that the appearance of caspase-3 immunoreactivity preceded DNA fragmentation and thus subsequent cell death would suggest that caspase-3 serves as the executioner of cell death in this brain structure following ischemia. The lack of active caspase-3 immunoreactivity in the cortex in this MCAo rat model is evident, and several reasons may underlie these results. Our earliest assessment of active caspase-3 following reperfusion was at 3 h. It is possible that caspase-3 activation occurred earlier in cortical neurons and that a rapid enzymatic turnover rate may have impeded its detection at the 3-h time-point investigated. However, if this were the case, then it would be expected that DNA fragmentation and thus TUNEL labeling would also appear earlier in the cortex than in the striatum. In contrast, TUNEL labeling was ¢rst observed in the striatum (few scattered cells at 3 h) and only appeared in the cortex at 6 h postreperfusion. However, this may re£ect a temporary resistance of cortical neurons to cell death induced by ischemia as compared to striatal neurons. In the cortex, where neuronal death was more delayed as evidenced by TUNEL, less amount of caspase-3 enzyme may be required for cell death. Thus perhaps these low levels of active caspase-3 may not be detectable by the antibodies used in the present study. However, consistent

Cyaan Magenta Geel Zwart

134

M. A. Davoli et al.

with this ¢nding is the lack of DEVD-AMC cleavage by the ischemic extracts (not shown). In contrast to the cortical labeling reported by Krupinski et al. (2000) using an antibody generated against the inactive pro-enzyme, we were not able to observe similar staining using an antibody that recognizes the active form of the enzyme. Interestingly, however, other reports have shown the active caspase-3 p20 cleavage product to appear in the ischemic cortex in a similar model of transient ischemia as early as 5 min, and remain distinguishable even at 24 h post-reperfusion (Namura et al., 1998). Strong labeling has also been reported in the small-to-medium-sized pyramidal neurons in lamina II/III of the cortex 24 h following permanent MCAo (Velier et al., 1999). The lack of caspase-3 activation in the cortex in our hands might also be related to the severity of the MCAo procedure. It has been previously reported that apoptosis is best triggered after milder forms of prolonged ischemia (Endres et al., 1998), suggesting that the model used in our experiments elicited too severe an insult, predominantly in the cerebral cortex, that brought forth cell death that was not primarily caspase-3 dependent in nature. In support of this, in a permanent MCAo model, van Lookeren Campagne and Gill (1996) using electron microscopy techniques report cell death that is di¡erent than the classical apoptotic phenotype. However, in the current study, ultrastructural investigations were not conducted to decipher a possible necrotic phenotype; however, such studies could provide insight. It is noteworthy, that in an attempt to induce milder ischemic injury restricted to the cortex by altering the position of the microaneurysm clip on the MCA, or by occluding the MCA for shorter periods of time (30 and 60 min), no caspase-3 activation could be seen in the cortex (data not shown). Lastly, the discrepancy between these previous observations and our failure to detect active caspase-3 by immunohistochemistry may be related to technical reasons such as di¡erence in antibodies used and/or the processing of tissue which may a¡ect accessibility of speci¢c epitopes by the antibodies. The time-course of DNA fragmentation as measured by TUNEL revealed that cell death occurred as early as 3 and 6 h in the striatum and cortex, respectively. By 24 h, massive portions of striatum and cortex had succumbed to ischemic cell death as visualized by TUNEL staining and con¢rmed by Cresyl Violet staining (Fig. 3). Also noteworthy, was that TUNEL labeling appeared to be most intense, and most abundant by 24 h post-reperfusion, pointing to the fact that with increasing time post-surgery, more DNA fragmentation was occurring in these damaged cells. This was also seen in both the neurons of the cortex and the striatum. Ischemia-induced DNA fragmentation contains not only internucleosomal fragmentation but also random breakdown into both larger (30^50 kbp) and smaller fragments represented by smears (Heron et al., 1993; Tominaga et al., 1993; Charriaut-Marlangue et al., 1995a). Random DNA breakdown products are generally associated with necrosis. The TUNEL procedure labels all types of DNA fragments and cannot discriminate accurately between apoptotic and necrotic cells (de Torres et al.,

NSC 5800 14-10-02

1997; Ferrand-Drake and Wieloch, 1999). Therefore, TUNEL positivity should not be considered as a marker for apoptosis (Charriaut-Marlangue and Ben-Ari, 1995b). We also observed that NeuN staining was reduced dramatically by 24 h post-reperfusion (the opposite of what was observed with TUNEL) in both striatal and cortical neurons (Fig. 5C), con¢rming the decreasing viability of the ischemic cells over time as a result of the ischemic injury. Despite the lack of active caspase-3 immunoreactivity in the cortex, spectrin P120 was detected in cortical extracts of ischemic brain tissues. Alpha-II spectrin is a cytoskeletal protein that contains a DXXD sequence and thus is a substrate for caspase-3 (Nath et al., 2000). It is noteworthy that spectrin is cleaved by both calpain and caspases in apoptotic neuronal cells. Caspases cleave spectrin at two di¡erent sites thus generating two fragments, with molecular weights at 120 and 150 kDa, respectively (Nath et al., 2000). Calpain cleaves spectrin to generate two bands at 150 and 145 kDa (Pike et al., 1998; Wang et al., 1998). Unlike P150, which was observed in cases of both necrotic and apoptotic cell death, spectrin P120 was detected exclusively during apoptosis in a wide variety of cell types (Zhao et al., 1999). Inhibition of caspase-3 by selective inhibitors dose dependently attenuated the generation of P120 but not P145, whereas application of a calpain inhibitor selectively blocked the production of P145 (Pike et al., 1998; Tam et al., unpublished results), supporting the view that the caspase-speci¢c p120 fragment is a speci¢c marker for caspase-3 activity. Compared to P150 and P145, the level of P120 in the cortex is quite low, which is consistent with the proposition that levels of active caspase-3 in the cortex were too low to be detected by immunohistochemistry. Nonetheless, the presence of P120 is indicative of caspase-3 activity following ischemia in the cortex. The accumulation and persistence of the spectrin P120 fragment may have helped its detection by western blot. In support of this hypothesis is the observation that the calpain-mediated spectrin cleavage product has been reported to be very stable (Vanderklish and Bahr, 2000). Thus, given the increased stability of the substrate cleavage product compared to the activated enzyme, immunohistochemical localization using antibodies speci¢c for the P120 fragment would be meaningful. Furthermore, given the abundant signal of P145 in the ischemic cortex, the possible involvement of calpains cannot be discounted at this time. Future studies are warranted to determine whether the P145 fragment could account for a more necrotic phenotype in the ischemic cortex. Consistent with the appearance of P120, nucleosomal DNA fragments were generated in ischemic brain tissues as observed by both TUNEL (described previously) and CDE. Internucleosomal cleavage of chromatin by endonucleases during apoptosis is known to result in the generation of nucleosomes (Earnshaw, 1995). Nucleosomes are basic structural units of chromatin composed of DNA double helix surrounding a histone octamer. Nucleosomal fragments are released into the cytoplasm of cells undergoing apoptosis. Cytoplasmic nucleosomal

Cyaan Magenta Geel Zwart

Caspase mediated neuronal death following transient MCAo

fragments can thus be isolated from nuclear chromatin or DNA, and detected by ELISA using antibodies for both DNA and histones (Salgame et al., 1997). Unlike TUNEL, which non-discriminately labels DNA fragments of all sizes, CDE detects cytoplasmic nucleosomes (Salgame et al., 1997). Therefore, it is felt to be a more speci¢c measure of apoptosis. Together with the appearance of spectrin P120 in ischemic cortex, the presence of nucleosomes in the tissue supports the view that caspase-

135

3-mediated apoptosis is involved in ischemic brain injury following MCAo. Moreover, the strong correlation between these apoptotic events and the infarcts produced by the MCAo surgery suggests that apoptosis may be an important component of ischemic cell death. In light of the current results, studies to examine whether selective caspase-3 inhibitors might be bene¢cial in protecting brain tissues from ischemic injury induced by MCAo are warranted.

REFERENCES

Buchan, A.M., Xue, D., Slivka, A., 1992. A new model of temporary focal neocortical ischemia in the rat. Stroke 23, 273^279. Charriaut-Marlangue, C., Ben-Ari, Y., 1995b. A cautionary note on the use of the TUNEL stain to determine apoptosis. NeuroReport 29, 61^64. Charriaut-Marlangue, C., Margaill, I., Plotkine, M., Ben-Ari, Y., 1995a. Early endonuclease activation following reversible focal ischemia in the rat brain. J. Cereb. Blood Flow Metab. 15, 385^388. Chen, J., Nagayama, T., Jin, K., Stetler, R.A., Zhu, R.L., Graham, S.H., Simon, R.P., 1998. Induction of caspase-3-like protease may mediate delayed neuronal death in the hippocampus after transient cerebral ischemia. J. Neurosci. 18, 4914^4928. Cheng, Y., Deshmukh, M., D’Costa, A., Demaro, J.A., Gidday, J.M., Shah, A., Sun, Y., Jacquin, M.F., Johnson, E.M., Holtzman, D.M., 1998. Caspase inhibitor a¡ords neuroprotection with delayed administration in a rat model of neonatal hypoxic-ischemic brain injury. J. Clin. Invest. 101, 1809^1810. de Torres, C., Munell, F., Ferrer, I., Reventos, J., Macaya, A., 1997. Identi¢cation of necrotic cell death by the TUNEL assay in the hypoxicischemic neonatal rat brain. Neurosci. Lett. 230, 1^4. Earnshaw, W.C., 1995. Nuclear changes in apoptosis. Curr. Opin. Cell Biol. 7, 337^343. Eldadah, B.A., Faden, A.I., 2000. Caspase pathways, neuronal apoptosis, and CNS injury. J. Neurotrauma 17, 811^829. Endres, M., Namura, S., Shimizu-Sasamata, M., Waeber, C., Zhang, Z.L., Gomez-Isla, T., Hyman, B.T., Moskowitz, M.A., 1998. Attenuation of delayed neuronal death after mild focal ischemia in mice by inhibition of the caspase family. J. Cereb. Blood Flow Metab. 18, 238^247. Ferrand-Drake, M., Wieloch, T., 1999. The time-course of DNA fragmentation in the choroid plexus and the CA1 region following transient global ischemia in the rat brain. The e¡ect of intra-ischemic hypothermia. Neuroscience 93, 537^549. Gervais, F.G., Xu, D., Robertson, G.S., Vaillancourt, J.P., Zhu, Y., Huang, J., Leblanc, A., Smith, D., Rigby, M., Shearman, M.S., Clarke, E.E., Zheng, H., Van Der Ploeg, L.H., Ru¡olo, S.C., Thornberry, N.A., Xanthoudakis, S., Zamboni, R.J., Roy, S., Nicholson, D.W., 1999. Involvement of caspases in proteolytic cleavage of Alzheimer’s amyloid-beta precursor protein and amyloidogenic A beta peptide formation. Cell 97, 395^406. Heron, A., Pollard, H., Dessi, F., Moreau, J., Lasbennes, F., Ben-Ari, Y., Charriaut-Marlange, C., 1993. Region variability in DNA fragmentation after global ischemia evidenced by combined histological and gel electrophoresis observations in the rat brain. J. Neurochem. 61, 1973^1976. Himi, T., Ishizaki, Y., Murota, S., 1998. A caspase inhibitor blocks ischemia-induced delayed neuronal death in the gerbil. Eur. J. Neurosci. 10, 777^781. Kihara, S., Shiraishi, T., Nakagawa, S., Toda, K., Tabuchi, K., 1994. Visualization of DNA double strand breaks in the gerbil hippocampal CA1 following transient ischemia. Neurosci. Lett. 175, 133^136. Krupinski, J., Lopez, E., Marti, E., Ferrer, I., 2000. Expression of caspases and their substrates in the rat model of focal cerebral ischemia. Neurobiol. Dis. 7, 332^342. Kuida, K., Zheng, T.S., Na, S., Kuan, C., Yang, D., Karasuyama, H., Rakic, P., Flavell, R.A., 1996. Decreased apoptosis in the brain and premature lethality in CPP32-de¢cient mice. Nature 384, 368^372. Li, Y., Chopp, M., Jiang, N., Zaloga, C., 1995b. In situ detection of DNA fragmentation after focal cerebral ischemia in mice. Brain Res. Mol. Brain Res. 28, 164^168. Li, Y., Sharov, V.G., Jiang, N., Zaloga, C., Sabbah, H.N., Chopp, M., 1995a. Ultrastructural and light microscopic evidence of apoptosis after middle cerebral artery occlusion in the rat. Am. J. Pathol. 146, 1045^1051. Liu, X-H., Kwon, D., Schielke, G.P., Yang, G-Y., Silverstein, F.S., Barks, J.D.E., 1999. Mice de¢cient in interleukin-1 converting enzyme are resistant to neonatal hypoxic-ischemic brain damage. J. Cereb. Blood Flow Metab. 19, 1099^1108. Loetscher, H., Niederhauser, O., Kemp, J., Gill, R., 2001. Is caspase-3 inhibition a valid therapeutic strategy in cerebral ischemia? Drug Discov. Today 6, 671^680. Martins, L.M., Mesner, P.W., Kottke, T.J., Basi, G.S., Sinha, S., Tung, J.S., Svingen, P.A., Madden, B.J., Takahashi, A., McCormick, D.J., Earnshaw, W.C., Kaufmann, S.H., 1997. Comparison of caspase activation and subcellular localization in HL-60 and K562 cells undergoing etoposide-induced apoptosis. Blood 90, 4283^4296. Mullen, R.J., Buck, C.R., Smith, A.M., 1992. NeuN, a neuronal speci¢c nuclear protein in vertebrates. Development 116, 201^211. Nagata, S., 2000. Apoptotic DNA fragmentation. Exp. Cell Res. 256, 12^18. Nakajima, W., Ishida, A., Lange, M.S., Gabrielson, K.L., Wilson, M.A., Martin, L.J., Blue, M.E., Johnston, M.V., 2000. Apoptosis has a prolonged role in the neurodegeneration after hypoxic ischemia in the newborn rat. J. Neurosci. 20, 7994^8004. Namura, S., Zhu, J., Fink, K., Endres, M., Srinivasan, A., Tomaselli, K.J., Yuan, J., Moskowitz, M.A., 1998. Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia. J. Neurosci. 18, 3659^3668. Nath, R., Huggins, M., Glantz, S.B., Morrow, J.S., McGinnis, K., Nadimpalli, R., Wang, K.K., 2000. Development and characterization of antibodies speci¢c to caspase-3-produced alpha II-spectrin 120 kDa breakdown product: marker for neuronal apoptosis. Neurochem. Int. 37, 351^361. Ni, B., Wu, X., Su, Y., Stephenson, D., Smalstig, E.B., Clemens, J., Paul, S.M., 1998. Transient global forebrain ischemia induces a prolonged expression of the caspase-3 mRNA in rat hippocampal CA1 pyramidal neurons. J. Cereb. Blood Flow Metab. 18, 248^256. Nicholson, D.W., Ali, A., Thornberry, N.A., Vaillancourt, J.P., Ding, C.K., Gallant, M., Gareau, Y., Gri⁄n, P.R., Labelle, M., Lazebnik, Y.A., Munday, N.A., Raju, S.M., Smulson, M.E., Yamin, T-T., Yu, V.L., Miller, D.K., 1995. Identi¢cation and inhibition of the ICE/CED-3 protease necessary for mammalian apoptosis. Nature 376, 37^43. Paxinos, G., Watson, C., 1998. The Rat Brain in Stereotaxic Coordinates (4th ed.). Academic Press, San Diego, CA.

NSC 5800 14-10-02

Cyaan Magenta Geel Zwart

136

M. A. Davoli et al.

Pike, B.R., Zhao, X., Newcomb, J.K., Wang, K.K., Posmantur, R.M., Hayes, R.L., 1998. Temporal relationships between de novo protein synthesis, calpain and caspase 3-like protease activation, and DNA fragmentation during apoptosis in septo-hippocampal cultures. J. Neurosci. Res. 52, 505^520. Pulera, M.R., Adams, L.M., Liu, H., Santos, D.G., Nishimura, R.N., Yang, F., Cole, G.M., Wasterlain, C.G., 1998. Apoptosis in a neonatal rat model of cerebral hypoxia-ischemia. Stroke 29, 2622^2630. Rano, T.A., Timkey, T., Peterson, E.P., Rotonda, J., Nicholson, D.W., Becker, J.W., Chapman, K.T., Thornberry, N.A., 1997. A combinatorial approach for determining protease speci¢cities : application to interleukin-1beta converting enzyme (ICE). Chem. Biol. 4, 149^155. Salgame, P., Varadhachary, A.S., Primiano, L.L., Fincke, J.E., Muller, S., Monestier, M., 1997. An ELISA for detection of apoptosis. Nucleic Acids Res. 25, 680^681. Sasaki, C., Kitagawa, H., Zhang, W.R., Warita, H., Sakai, K., Abe, K., 2000. Temporal pro¢le of cytochrome C and caspase-3 immunoreactivities and TUNEL staining after permanent middle cerebral artery occlusion in rats. Neurol. Res. 22, 223^228. Schmued, L.C., Albertson, C., Slikker, W., Jr., 1997. Fluoro-Jade : a novel £uorochrome for the sensitive and reliable histochemical localization of neuronal degeneration. Brain Res. 751, 37^46. Schmued, L.C., Hopkins, K.J., 2000. Fluoro-Jade : novel £uorochromes for detecting toxicant-induced neuronal degeneration. Toxicol. Pathol. 28, 91^99. Simpson, M.T.W., MacLaurin, J.G., Xu, D.G., Ferguson, K.L., Vanderluit, J.L., Davoli, M.A., Roy, S., Nicholson, D.W., Robertson, G.S., Park, D.S., Slack, R.S., 2001. Caspase 3 De¢ciency Rescues Peripheral Nervous System Defect in Retinoblastoma Nullizygous Mice. J. Neurosci. 21, 7089^7098. Stennicke, H.R., Salvesen, G.S., 1998. Properties of the caspases. Biochim. Biophys. Acta 1387, 17^31. Thornberry, N.A., Rano, T.A., Peterson, E.P., Rasper, D.M., Timkey, T., Garcia-Calvo, M., Houtzager, V.M., Nordstrom, P.A., Roy, S., Vaillancourt, J.P., Chapman, K.T., Nicholson, D.W., 1997. A combinatorial approach de¢nes speci¢cities of members of the caspase family and granzyme. B. Functional relationships established for key mediators of apoptosis. J. Biol. Chem. 272, 17907^17911. Tominaga, T., Kure, S., Narisawa, K., Yoshimoto, T., 1993. Endonuclease activation following focal ischemic injury in the rat brain. Brain Res. 608, 21^26. van Lookeren Campagne, M., Gill, R., 1996. Ultrastructural morphological changes are not characteristic of apoptotic cell death following focal cerebral ischemia in the rat. Neurosci. Lett. 213, 111^114. Vanderklish, P.W., Bahr, B.A., 2000. The pathogenic activation of calpain: a marker and mediator of cellular toxicity and disease states. Int. J. Exp. Pathol. 81, 323^339. Velier, J.J., Ellison, J.A., Kikly, K.K., Spera, P.A., Barone, F.C., Feuerstein, G.Z., 1999. Caspase-8 and caspase-3 are expressed by di¡erent populations of cortical neurons undergoing delayed cell death after focal stroke in the rat. J. Neurosci. 19, 5932^5941. Wang, K.K., Posmantur, R., Nath, R., McGinnis, K., Whitton, M., Talanian, R.V., Glantz, S.B., Morrow, J.S., 1998. Simultaneous degradation of alphaII- and betaII-spectrin by caspase 3 (CPP32) in apoptotic cells. J. Biol. Chem. 273, 22490^22497. Zhao, X., Pike, B.R., Newcomb, J.K., Wang, K.K., Posmantur, R.M., Hayes, R.L., 1999. Maitotoxin induces calpain but not caspase-3 activation and necrotic cell death in primary septo-hippocampal cultures. Neurochem. Res. 24, 371^382. (Accepted 18 June 2002)

NSC 5800 14-10-02

Cyaan Magenta Geel Zwart