SEARCH

SEARCH BY CITATION

Keywords:

  • apoptosis;
  • apoptosis-inducing factor;
  • caspase;
  • hypoxia;
  • ischemia;
  • neonatal

Abstract

  1. Top of page
  2. Abstract
  3. References

Apoptosis-inducing factor (AIF) triggers apoptosis in a caspase-independent manner. Here we report for the first time involvement of AIF in neuronal death induced by cerebral ischemia. Unilateral cerebral hypoxia-ischemia (HI) was induced in 7-day-old rats by ligation of the left carotid artery and hypoxia (7.7% O2) for 55 min. AIF release from mitochondria and AIF translocation to nuclei was detected immediately after HI, and only in damaged areas, as judged by the concurrent loss of MAP-2. AIF release was detected earlier than that of cytochrome c. Cells with AIF-positive nuclei displayed nuclear condensation and signs of DNA damage. The number of AIF-positive nuclei showed a positive correlation with the infarct volume 72 h post-HI, and this was not changed by treating the animals with boc-Asp-fmk (BAF), a multicaspase inhibitor. BAF treatment reduced the activity of caspase-3, -2 and -9 (78, 73 and 33%, respectively), and prevented caspase-dependent fodrin cleavage in vivo, but did not affect AIF release from mitochondria or the frequency of positive nuclear AIF or DNA damage 72 h post-HI, indicating that these processes occurred in a caspase-independent fashion. In summary, AIF-mediated cell death may be an important mechanism of HI-induced neuronal loss in the immature brain.

Abbreviations used
AIF

apoptosis-inducing factor

BAF

boc-aspartyl-(OMe)-fluoromethyl ketone

BSA

bovine serum albumin

CAD

caspase-activated DNase

COX

cytochrome c oxidase

DMSO

dimethyl sulfoxide

DTT

dithiothreitol

HI

hypoxia-ischemia

HPP

hairpin probe

ICAD

inhibitor of caspase-activated DNase

MNNG

N-methyl-N′-nitro-N-nitrosoguanidine

SDS–PAGE

sodium dodecyl sulfate

TBS

Tris-buffered saline

During normal brain development more than half of the neurons are lost in certain brain regions (Raff et al. 1993). This process, whereby nerve cells are discretely removed without interfering with the further development of the remaining cells, is known as apoptosis. The immature brain has been suggested to retain part of this developmental cell death program, which may be activated following hypoxia-ischemia (HI), irradiation or excitotoxicity (Sidhu et al. 1997; Yue et al. 1997). Many of the key elements of apoptosis have been demonstrated to be strongly up-regulated in the immature brain, such as caspase-3 (Blomgren et al. 2001), caspase-12 (unpublished data), APAF-1 (Ota et al. 2002), Bcl-2 (Merry et al. 1994) and Bax (Vekrellis et al. 1997). Mitochondria are key regulators in the process of cell death through their capacity to release a number of pro-apoptotic proteins from their intermembrane space, such as cytochrome c (Kroemer et al. 1998), SMAC/Diablo (Chai et al. 2000; Du et al. 2000), caspase-2 and -9, and apoptosis-inducing factor (AIF) (Susin et al. 1999). Unlike cytochrome c, AIF acts in a caspase-independent manner (Susin et al. 1999). AIF translocates to the nucleus and induces chromatin condensation and large-scale (∼50 kb) DNA fragmentation (Daugas et al. 2000).

Hypoxic-ischemic (HI) brain frequently causes long-term neurological sequelae in term and pre-term infants (Vannucci 1989; Hagberg et al. 1996; Volpe 2000). Some data suggest that mitochondria are critical in the pathophysiology of HI injury in the immature brain. Mitochondrial respiration is depressed after HI and, after a partial transient recovery, a secondary decrease occurs in parallel with the loss of MAP-2 immunostaining, decrease of tissue glucose utilization, activation of caspase-3 and DNA fragmentation (Gilland et al. 1998a; Puka-Sundvall et al. 2000c). Treatment that improves mitochondrial respiration also prevents secondary energy failure and brain injury (Gilland and Hagberg 1996; Gilland et al. 1998b; Puka-Sundvall et al. 2000b). After HI, mitochondria exhibit a swollen ultrastructure, as well as an increased permeability to deoxyglucose and glutathione, which may implicate a phase of membrane permeability transition (Puka-Sundvall et al. 2000a; Wallin et al. 2000; Puka-Sundvall et al. 2001). The relative contributions of necrosis and apoptosis to the injury that develops after cerebral HI has been a matter of much debate (Lee et al. 1999). Depending on the severity of the insult and the anatomical region, neurons may exhibit signs of necrosis or apoptosis, or both (Puka-Sundvall et al. 2000a; Northington et al. 2001a; Northington et al. 2001b), and attempts have been made to classify these hybrid modes of cell death (Leist and Jäättelä 2001).

The purpose of this study was to characterize the time course and extent of AIF release from mitochondria and its correlation to tissue damage in a model of neonatal hypoxia ischemia.

Experimental procedures

  1. Top of page
  2. Abstract
  3. References

Induction of HI

  1. Top of page
  2. Abstract
  3. References

Unilateral HI was induced in 7-day-old Wistar rat pups of either sex. Rats were anesthetized with halothane (3.0% for induction and 1.0–1.5% for maintenance) in a mixture of nitrous oxide and oxygen (1 : 1), and the duration of anesthesia was < 10 min. The left common carotid artery was cut between double ligatures of prolene sutures (6–0). After the surgical procedure the wounds were infiltrated with a local anesthetic, and the pups were allowed to recover for 1–2 h (Rice et al. 1981; Bona et al. 1998). The litters were placed in a chamber perfused with a humidified gas mixture (7.7% oxygen in nitrogen) for 55 min. The temperature in the incubator, and the temperature of the water used to humidify the gas mixture, was kept at 36°C. After hypoxic exposure the pups were returned to their biological dams and were allowed to recover for 0 min, 30 min, 1 h, 3 h, 8 h, 14 h, 24 h or 72 h. Control pups, subjected to neither ligation nor hypoxia, were killed at postnatal day 7 or 10 (P7 or P10). All animal experimentation was approved by the Ethical committee of Göteborg (183–99).

In vivo caspase inhibition

  1. Top of page
  2. Abstract
  3. References

The multicaspase inhibitor, boc-aspartyl-(OMe)-fluoromethyl ketone (BAF) (Enzyme Systems Products, Livermore, CA, USA) was prepared as a 100-mm stock solution in dimethyl sulfoxide (DMSO; Sigma, St Louis, MO, USA). Intraparenchymal BAF injections (n = 33) or vehicle (n = 33), were performed according to Cheng et al. (1998), with some modifications. P7 pups were anesthetized and received 5 µL vehicle (2 µL DMSO, 3 µL PBS, pH 7.4) or BAF solution (2 µL of stock solution, 3 µL PBS) at 2 h and 12 h post-HI, using a syringe attached to a microinjection pump (CMA, Stockholm, Sweden) at a speed of 1 µL/min. In preliminary experiments, injections of blue dye were performed to find the appropriate location. The injection site in relation to lambda was 5.5 mm rostral, 2.0 mm lateral, 3.0 mm deep from the skin surface. For morphological evaluation, pups were killed at 72 h post-HI. For immunoblotting and activity assays, pups were killed at 24 h post-HI.

Sample preparation

  1. Top of page
  2. Abstract
  3. References

Animals were killed by decapitation 0 min, 30 min, 1 h, 3 h, 8 h, 14 h, 24 h or 72 h after HI. Control animals were killed on postnatal day 0, 3, 7, 8, 14, 21, 42 or adult. The brains were rapidly dissected out on a bed of ice. The parietal cortex and diencephalon were dissected out from each hemisphere and ice-cold isolation buffer was added [15 mm Tris-HCl, pH 7.6, 320 mm sucrose, 1 mm dithiothreitol (DTT), 3 mm EDTA-K, 0.5% protease inhibitor cocktail (Sigma) and 2.5 µm cyclosporin A (Sigma)]. Homogenization was performed either by sonication or gently by hand (to preserve mitochondrial integrity) in a 2-mL glass/glass homogenizer (Merck Eurolab, Göteborg, Sweden) using, sequentially, two different pestles with a total clearance of 0.12 mm and 0.05 mm, respectively (five strokes each). The homogenates were centrifuged at 800 g at 4°C for 10 min. The supernatants were further centrifuged at 9200 g for 15 min at 4°C, producing mitochondrial and synaptosomal fractions in the pellets (P2) and crude cytosolic fractions in the supernatants (S2). The P2 fractions were washed and recentrifuged. All fractions were stored at − 80°C.

Immunohistochemistry

  1. Top of page
  2. Abstract
  3. References

Pups were deeply anesthetized and perfusion-fixed with 5% formaldehyde in 0.1 m PBS (Histofix, from Histolab, Sweden). The brains were rapidly removed and immersion-fixed in 5% formaldehyde at 4°C for 24 h. After dehydration with graded ethanol and xylene, the brains were paraffin-embedded and cut into 5 or 10 µm sections. Sections were deparaffinized in xylene and rehydrated in graded ethanol before staining.

AIF

  1. Top of page
  2. Abstract
  3. References

Antigen retrieval was performed by boiling the sections in 10 mm sodium citrate buffer (pH 6.0) for 10 min. Nonspecific binding was blocked for 30 min with 4% goat serum in PBS. Anti-AIF (Susin et al. 1999) was applied diluted 1 : 150 in Tris-buffered saline (TBS) containing 1% bovine serum albumin (BSA) and 0.1% Triton X-100, incubated for 60 min at room temperature, followed by a biotinylated goat anti-rabbit antibody (6 µg/mL in PBS) for 60 min. Endogenous peroxidase activity was blocked with 3% H2O2 in PBS for 5 min. Visualization was performed using Vectastain ABC Elite with 0.5 mg/mL 3,3′-diaminobenzidine (DAB) enhanced with 15 mg/mL ammonium nickel sulfate, 2 mg/mL β-d-glucose, 0.4 mg/mL ammonium chloride and 0.01 mg/mL β-glucose oxidase (Sigma). Negative controls, where the primary antibody was omitted, were completely blank. Also, the staining obtained with this AIF antibody and the D-20 antibody (2 µg/mL, sc-9416, from Santa Cruz Biotechnology, Santa Cruz, CA, USA) were virtually indistinguishable, and preabsorbtion with the peptide provided for the latter antibody abolished the staining.

MAP-2 and cytochrome c

  1. Top of page
  2. Abstract
  3. References

MAP-2 was used as a marker of intact neuronal cell bodies. Antigen retrieval was performed as above. Non-specific binding was blocked with 4% horse serum in PBS for 30 min Anti-MAP-2 (clone HM-2, Sigma), diluted 1: 2000 (4 µg/mL) in PBS, or anti-cytochrome c (clone 7H8.2C12, Pharmingen, San Diego, CA, USA), diluted 1 : 500 (2 µg/mL) in PBS, was incubated for 60 min at room temperature, followed by another 60 min with a biotinylated horse anti-mouse antibody (2 µg/mL) diluted with PBS. Peroxidase blocking and visualization as above.

HPP staining

  1. Top of page
  2. Abstract
  3. References

An oligonucleotide hairpin probe with one base overhang in the 3′ end was used as a marker of DNA damage, presumably apoptotic (Didenko and Hornsby 1996; Zhu et al. 2000). Staining was performed essentially according to Didenko et al. (Didenko et al. 1998). Following deparaffinization and rehydration, sections were incubated for 90 min at 65°C in 10 mm sodium citrate (pH 6.0), washed and treated with proteinase K (Roche), 25 µg/mL in PBS for 10 min at room temperature. A mixture of 50 mm Tris-HCl (pH 7.8), 10 mm MgCl2, 10 mm DTT, 1 mm ATP, 15% polyethylene glycol 6000, with 35 µg/mL hairpin oligonucleotide and 250 U/mL T4 DNA ligase (Roche) was added, and the sections were incubated for 2 h at room temperature. After washing, sections were incubated with 3% H2O2 in PBS for 5 min, followed by washing and 0.1% BSA in PBS for 30 min. Visualization was performed using Vectastain ABC. Nonspecific labeling was checked by omitting the T4 DNA ligase.

Double labeling immunofluorescence

  1. Top of page
  2. Abstract
  3. References

AIF–HPP–Hoechst 33342

  1. Top of page
  2. Abstract
  3. References

HPP staining was performed as described above. After incubation with the HPP and T4 DNA ligase and washing, sections were incubated with Texas red-conjugated avidin D (20 µg/mL in 0.1 m sodium bicarbonate (pH 8.5) and 0.15 m sodium chloride), for 60 min. After washing, sections were blocked with 4% goat serum and incubated with anti-AIF as above, followed by fluorescein-conjugated goat anti-rabbit IgG diluted 1 : 250 (6 µg/mL) in PBS for 60 min. After washing, the sections were placed in Hoechst 33342 (Molecular Probes, Eugene, OR, USA) 1 µg/mL in PBS for 10 min at room temperature with gentle agitation, washed and mounted using Vectashield mounting medium.

AIF–TUNEL–Hoechst 33342

  1. Top of page
  2. Abstract
  3. References

After incubation with TUNEL enzyme and fluorescein-dUTP (Roche) as described previously (Zhu et al. 2000), the sections were stained for AIF and with Hoechst 33342 as described above.

AIF–cytochrome c–Hoechst 33342

  1. Top of page
  2. Abstract
  3. References

Incubation with anti-AIF as described above. After washing, cytochrome c staining was performed as above, using fluorescein-conjugated horse anti-mouse IgG diluted 1 : 100 (5 µg/mL) in PBS. Hoechst 33342 staining as above.

AIF staining was performed as described above. Cytochrome c oxidase subunit IV (COX) staining was started by blocking with 4% normal horse serum for 30 min followed by a mouse monoclonal antibody against COX diluted 1 : 100 (12C4-F12, 2.5 µg/mL, Molecular Probes, Eugene, OR, USA) in 1% BSA-TBS over night at 4°C. After washing, the sections were incubated with fluorescein-conjugated horse anti-mouse 1 : 100 (5 µg/mL) for 60 min at room temperature.

AIF and MAP-2 stainings were performed as above, and the latter was followed by fluorescein-conjugated horse anti-mouse as above.

Immunoblotting

  1. Top of page
  2. Abstract
  3. References

The protein concentration was determined according to Whitaker and Granum (Whitaker and Granum 1980), adapted for microplates, using a Spectramax Plus fluorometer (Molecular Devices, Sunnyvale, CA, USA). Samples were mixed with an equal volume of concentrated (3 ×) sodium dodecyl sulfate–polyacryamide gel electrophoresis (SDS–PAGE) buffer and heated (96°C) for 5 min. Pooled samples were run on 8–16% Tris-glycine gels (Novex, San Diego, CA, USA) and transferred to reinforced nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany). All samples were run 3–5 times to verify the performance. The membranes were blocked in 30 mm Tris-HCl (pH 7.5), 100 mm NaCl and 0.1% Tween 20 (TBS-T) containing 5% fat-free milk powder for 1 h at room temperature. After washing in TBS-T, the membranes were incubated with anti-AIF, either the same antibody as above or the D-20 (sc-9416, 1 : 1000, 0.2 µg/mL, goat polyclonal antibody; Santa Cruz), anti-alpha fodrin (FG 6090, 1 : 500, Affiniti Research Products Ltd, Mamhead, UK), anti-cytochrome oxidase (COX) subunit IV (12C4-F12, 1 : 500, Molecular Probes), anti-caspase-3 (H-277, 1 : 1000, Santa Cruz), anti-caspase-9 (H-83, 1 : 300, Santa Cruz), anti-α-tubulin (TU-01, 1 : 1000, Monosan, Uden, Netherlands) at 4°C over night. After washing, the membranes were incubated with the appropriate peroxidase-labeled, secondary antibody diluted in blocking solution, for 30 min at room temperature (goat anti-rabbit, 1: 2000, horse anti-goat, 1: 2000, or horse anti-mouse 1 : 4000). All secondary antibodies were from Vector (Burlingame, CA, USA). Immunoreactive species were visualized using the Super Signal Western Dura substrate (Pierce, Rockford, IL, USA) and a LAS 1000 cooled CCD camera (Fujifilm, Tokyo, Japan). Immunoreative bands were quantified using the Image Gauge software (Fujifilm, Tokyo, Japan).

SYPRO ruby protein blot stain

  1. Top of page
  2. Abstract
  3. References

The SYPRO ruby protein blot stain (S-11791, Molecular Probes) was used to assure equal protein loading on the immunoblotting membranes. Membranes were immersed in 7% acetic acid, 10% methanol and incubated at room temperature for 15 min with gentle agitation, followed by 4 × 5-min washes in deionized water. Finally, membranes were immersed in SYPRO ruby blot stain for 15 min, and washed 6 × 1 min in deionized water. The staining was visualized using epifluorescence and a LAS 1000 cooled CCD camera (Fujifilm, Tokyo, Japan).

Caspase activity assays

  1. Top of page
  2. Abstract
  3. References

The protein concentrations were determined as above. Samples of crude cytosolic fractions (S2) 25 µL were mixed with 75 µL of extraction buffer as described earlier (Wang et al. 2001). Cleavage of Ac-DEVD-AMC (for caspase 3, from Peptide Institute, Osaka, Japan) or Ac-YVAD-AMC (for caspase 1, from Alexis Corp., San Diego, CA, USA) was measured at 37°C using a Spectramax Gemini microplate fluorometer (Molecular Devices) with an excitation wavelength of 380 nm and an emission wavelength of 460 nm, and expressed as pmol AMC released per mg protein and minute. Cleavage of Ac-VDVAD-AFC (for caspase-2), Ac-IETD-AFC (for caspase-8), Ac-LEHD-AFC (for caspase-9) (Enzyme Systems Products, Livermore, CA, USA) were measured as for the AMC conjugates, but with an excitation wavelength of 400 nm and an emission wavelength of 505 nm, and expressed as pmol AFC released per mg protein and minute. Every sample was analyzed 3–4 times and the average value was used as n = 1.

Cell counts, infarct volumes and brain injury scoring

  1. Top of page
  2. Abstract
  3. References

Cell counting was performed in the cortex, hippocampus, striatum and thalamus (nucleus habenularis). The CA1, CA3 and dentate gyrus subfields of the hippocampus were counted separately. Positive cells were counted at 400 × magnification (one visual field = 0.196 mm2). In the hippocampal subfields and the nucleus habenularis counting was performed throughout the entire region. In the cortex and striatum, three visual fields within an area displaying loss of MAP-2 (if any) were counted and expressed as average number per visual field. This average was used as n = 1 when comparing different brains. Parallel sections were used for all different stainings.

Infarct volumes were measured 3 days post-HI by sectioning the entire brains into 10 µm sections and staining every 50th section for MAP-2. The areas in the cortex, striatum, thalamus and hypothalamus displaying loss of MAP-2 staining were measured using Micro Image (Olympus, Japan) and the volumes calculated according to the Cavalieri principle using the following formula

  • image

where V = total volume, ΣA is the sum of the areas measured, P = the inverse of the sampling fraction, and T is the section thickness. The investigator measuring the MAP-2-negative areas and calculating the volumes was blinded to the treatment of the animals.

Brain injury in different regions was also evaluated using a semiquantitative neuropathological scoring system modified from Bona et al. (Bona et al. 1998). Briefly, sections were stained with thionin/acid fuchsin (Mallard et al. 1993) and scored by an observer blinded to the treatment of the animals. The cortical injury was graded from 0 to 4, 0 being no observable injury and 4 confluent infarction encompassing most of the cerebral cortex. The damage in hippocampus, striatum and thalamus was assessed both with respect to hypotrophy (shrinkage) (0–3) and observable cell injury/infarction (0–3) resulting in a neurpathological score for each brain region (0–6). The total score (0–22) was the sum of the scores for all 4 regions.

Statistics

  1. Top of page
  2. Abstract
  3. References

The Mann–Whitney U-test with Bonferroni correction, preceded by a Kruskal–Wallis test when comparing more than two groups, was used when comparing cell counts, infarct volumes, injury scores and the results from activity assays. anova with Bonferroni correction was used when comparing cell counts from different time points of recovery (AIF and cytochrome c).

Developmental changes

  1. Top of page
  2. Abstract
  3. References

The total levels of AIF were virtually unchanged during normal brain development from postnatal day 0 (P0) to adult (Fig. 1). The mitochondrial marker cytochrome c oxidase (COX), a mitochondrial, membrane-bound protein, on the other hand, displayed an increase, and so did cytochrome c and caspase-9 (Fig. 1). All these mitochondrial proteins, except AIF, were low at P0 and increased several-fold during the brain growth spurt (P7-21), remaining on a higher level in the adult brain (Fig. 1). Caspase-3, the most abundant effector caspase in the immature brain, decreased as the brain growth spurt leveled out (Fig. 1), as demonstrated earlier (Blomgren et al. 2001), displaying a virtually inverse correlation with COX, cytochrome c and caspase-9.

image

Figure 1. Postnatal expression of some apoptosis-related proteins in the brain. Immunoblots of pooled (n = 6 from each age) homogenate samples from postnatal day (P) 0, 3, 7, 14, 21, 42 and adult (Ad) rats, demonstrating the developmental patterns of AIF, cytochrome c oxidase, subunit IV (COX), caspase-9, cytochrome c and caspase-3. A tubulin immunoblot and a SYPRO Ruby total protein staining of the same membrane are included to verify equal loading.

Download figure to PowerPoint

AIF and cytochrome c translocation after HI

  1. Top of page
  2. Abstract
  3. References

In tissue sections, diffuse and relatively weak AIF immunoreactivity was located in the cytoplasm of normal neurons (Figs 2, 3 and 4). In damaged areas, as judged by the loss of MAP-2, AIF was translocated to nuclei immediately following HI, producing a distinct nuclear staining with little or no background (Fig. 2). This staining grew increasingly stronger and more condensed during reperfusion, eventually outlining only pyknotic nuclei (Figs 2, 3 and 4). Double labeling revealed colocalization of AIF and COX in mitochondria of normal neurons, and AIF nuclear staining after HI in damaged areas (Fig. 3), indicating AIF translocation from mitochondria to the nucleus. This translocation was not complete, though, as a considerable amount of AIF remained colocalized with COX after HI (Fig. 3). The correlation with DNA damage was investigated using Hoechst staining, TUNEL and an oligonucleotide hairpin probe (HPP) with a one base (A) overhang in the 3′ end (Didenko et al. 1998). Double strand DNA breaks with one base (T) overhang in the 3′ end, detected using this HPP, have been demonstrated to be selective for apoptosis (Didenko and Hornsby 1996), and to occur early after neonatal cerebral HI, earlier than TUNEL (Zhu et al. 2000). During early recovery (0–1 h), there was a very close correlation between AIF- and HPP-positive cells, but a larger number of TUNEL-positive cells (not shown). Furthermore, AIF translocation appeared to precede DNA damage, because larger cells displaying strong AIF immunoreactivity did not always display strong HPP staining, and small, pyknotic cells displaying strong HPP staining showed less intense AIF immunoreactivity, particularly at later time points of recovery (Fig. 5). Strong AIF and HPP stainings were also colocalized with chromatin condensation, as judged by Hoechst staining (Fig. 5).

image

Figure 2. Nuclear AIF immunostaining in the cerebral cortex after HI. Microphotographs of cortical layer V demonstrating AIF staining in control animals and after 0 min, 30 min, 1 h, 3 h and 8 h of recovery. Normal cells (control) display a weak, granular cytoplasmic staining. During the first hour of reperfusion the nuclei exhibited increasing AIF immunostaining, and during the following hours after HI the nuclei also turned more pyknotic. Bar = 10 µm.

Download figure to PowerPoint

image

Figure 3. Translocation of AIF from mitochondria to nuclei. Representative stainings of AIF (red), COX (green), and overlays from normal cortex (control) and ipsilateral, injured cortex 1 h post-HI (1 h). The control panels (upper panel) show undamaged cortex with a virtually complete overlap between AIF and COX in neurons. The lower panel shows an injured cortical area after 1 h of reperfusion, where extensive nuclear translocation of AIF can be seen. Bar = 20 µm.

Download figure to PowerPoint

image

Figure 4. Differential release of AIF and cytochrome c in injured cortex. Typical stainings of cytochrome c (Cyt c, green), AIF (red) and overlays (Cyt c/AIF) plus Hoechst 33342 staining (chromatin staining, blue) in normal cortex from control animals (Ctrl) and ipsilateral, injured cortex from 0 and 3 h post-HI to demonstrate the gradual appearance of cyt c- and AIF-positive cells during reperfusion (arrows). During early reperfusion (0–3 h) there were more AIF- than cyt c-positive cells. The chromatin was more condensed in cyt c- and AIF-positive cells and the nuclei grew more pyknotic during late reperfusion. Bar = 20 µm.

Download figure to PowerPoint

image

Figure 5. AIF and HPP immunostaining of ipsilateral cortex after 24 h of recovery. Immunostaining of AIF (green) and detection of DNA damage using an oligonucleotide hairpin probe (HPP) (red) as described in materials and methods. After 24 h recovery large cells staining strongly for AIF (arrows) usually displayed less intense HPP staining, whereas smaller cells with pyknotic nuclei, displaying strong HPP staining, usually appeared only weakly AIF-positive, indicating that AIF redistribution precedes HPP staining, and that the HPP-positive cells accumulate in the damaged areas. Hoechst 33342 staining (blue) demonstrated a close correlation between AIF immunoreactivity, HPP staining and chromatin condensation. Bar = 20 µm.

Download figure to PowerPoint

Cytochrome c staining was found to be more intense and distinct in damaged areas (Fig. 4), presumably indicating release from mitochondria. Release of AIF apparently occurred earlier than that of cytochrome c. Immediately following HI (0 min recovery) numerous cells in injured areas were positive for both AIF and cytochrome c (Fig. 4). Cells considered AIF-positive displayed a strongly stained nucleus whereas cytochrome c-positive cells displayed a strongly stained cytoplasm (Fig. 6a). The number of cells positive for AIF was several fold higher than the number of cytochrome c-positive cells in the cortex and nucleus habenularis immediately after HI (Figs 6b and c), but not clearly so in the dentate gyrus (Fig. 6d). The numbers of AIF-positive cells were significantly increased already at 0 min of recovery in the cortex and nucleus habenularis, but the numbers of cytochrome c-positive cells were not significantly increased until 1 h and 3 h of recovery, respectively, in these regions (Figs 6b and c). All cytochrome c-positive cells were also positive for AIF. Inversely, not all AIF-positive cells were cytochrome c-positive (Fig. 4). During early recovery (0 min to 3 h), AIF-positive cells without cytochrome c staining could be found in the damaged areas (Fig. 4). Later, from about 3 h onwards, the numbers of cells positive for AIF and cytochrome c were approximately equal and overlapping (not shown). In summary, these data indicate that injured neurons released first AIF and then cytochrome c. The nuclear condensation, as judged by Hoechst staining, was more pronounced in cells displaying redistribution of AIF and cytochrome c (Figs 4 and 5). Western blots of subcellular fractions revealed the presence of both the newly synthesized 67 kDa and the mature 57 kDa forms of AIF in homogenates (Fig. 7a). In crude cytosolic (S2) fractions, only the 67 kDa band was found, and in the mitochondrial (P2) fractions only the 57 kDa band was detected, as expected (Fig. 7a). Approximately 25% less AIF was present in P2 fractions from cortical samples on the ipsilateral, damaged side than the contralateral, undamaged side 1 h post-HI (Fig. 7b), indicating release from mitochondria.

image

Figure 6. Counts of AIF- and cytochrome c-positive cells in the cortex, nucleus habenularis and dentate gyrus after HI. AIF and cytochrome c immunostaining in control animals (P7 and P10) and at different time points of recovery (from 0 min to 72 h post-HI). Typical immunopositive cells, as indicated by arrows in (a), were counted in the cortex (b), in the nucleus habenularis of the thalamus (c) and in the dentate gyrus of the hippocampus (d). Data are presented as mean ± SD (n = 6 animals per time point). *p <  0.05, **p <  0.01, ***p <  0.001 compared with normal control animals.

Download figure to PowerPoint

image

Figure 7. AIF immunoblots of pooled cortical homogenates and subcellular fractions. Samples of cerebral cortex were fractionated (n = 6 animals per time point) and subsequently pooled. (a) shows an AIF immunoblot from a control animal, demonstrating the presence of two bands in homogenates (67 and 57 kDa), but only the heavier of the two (the newly synthesized proform) was present in the crude cytosolic fraction (S2) and only the lighter one (mature form) in the mitochondrial fraction (P2), as expected. The purity of the fractions, with respect to mitochondria, was demonstrated by reprobing the same membrane with an antibody against the mitochondrial marker COX, showing no detectable amounts of COX in the S2 fraction (lower panel). (b) shows a representative AIF immunoblot of pooled P2 fractions from control animals (C), and from ipsilateral (IL) and contralateral (CL) cortex after 1 h of recovery (upper panel). The two middle panels show an α-tubulin blot and a SYPRO ruby staining of the same membrane as in the top panel, demonstrating equal loading. The lower panel shows quantification of AIF immunoblots from individual samples (n = 6), demonstrating an approximately 23% decrease in the ipsilateral P2 fraction 1 h post-HI. **p < 0.01.

Download figure to PowerPoint

Caspase-independence

  1. Top of page
  2. Abstract
  3. References

Pups treated with the caspase inhibitor boc-Asp-fmk (BAF) (0.2 + 0.2 µmol at 2 h and 12 h post-HI, respectively) displayed significantly decreased levels of caspase-3 (78%, p = 0.007), caspase-2 (73%, p = 0.012) and caspase-9 (33%, p = 0.042) activity at 24 h post-HI (n = 9 in each group) (Fig. 8), the time point when caspase-3 (Wang et al. 2001), -2 and -9 (data not shown) activities are maximal in this model. Caspase-8 and -1 activities at 24 h post-HI were not significantly altered by BAF treatment (Fig. 8). Other dose regimens (0.1 µmol at 0 h; 0.2 + 0.2 µmol at 0 and 6 h, respectively; 0.2 µmol at 12 h; 0.2 + 0.2 µmol at 0 and 12 h, respectively) were all less effective with respect to caspase inhibition at 24 h post-HI (data not shown). The caspase-specific 120 kDa cleavage product of fodrin was evident in vehicle-treated animals, but was almost undetectable after BAF treatment, whereas the calpain-specific 150/145 kDa fragments were much less affected by caspase inhibition (Fig. 9a). This further confirmed that caspases were inhibited by the treatment. The amount of AIF released from mitochondria 24 h post-HI, as judged by the loss of AIF from the P2 fractions, was not changed by BAF treatment (Fig. 9b). Neither the number of AIF-positive nuclei, nor the number of HPP-positive cells, in the cortex, hippocampal subfields, the nucleus habenularis or the striatum, at 72 h post-HI were significantly changed by caspase inhibition (n = 8 animals per group, data not shown). The number of AIF-positive nuclei in the damaged areas of the cortex showed a linear correlation with the infarct volume at 72 h post-HI and this was not changed by BAF treatment (Fig. 10). The BAF treatment did not confer tissue protection, as judged by the infarct volumes or injury scoring (Table 1). The vehicle (5 µL of 40% DMSO) did not produce any damage per se, rather, there was an 18% significant reduction in total infarct volume in the vehicle-treated animals compared with untreated animals (n = 33 animals per group, data not shown).

image

Figure 8. Caspase-3, -2, -9, -8 and -1 activity in cortical samples 24 h post-HI. Crude cytosolic fractions from naïve P8 control animals (Control), from animals subjected to HI only (HI), animals subjected to HI and BAF treatment (BAF) and animals subjected to HI and vehicle injections (Vehicle) were assayed for their ability to cleave fluorogenic peptide substrates (DEVD, VDVAD, LEHD, IETD and YVAD, selective for caspase-3, -2, -9, -8 and -1, respectively). All animals subjected to HI were killed 24 h post-HI. The data presented represent the average from nine animals in each group ± SD. **p < 0.01 and *p < 0.05.

Download figure to PowerPoint

image

Figure 9. Effects of BAF on fodrin degradation and AIF release from mitochondria. (a) A typical fodrin immunoblot of cortex homogenates from naïve P8 control animals (C), from animals subjected to HI only (HI) and 24 h recovery, animals subjected to HI and BAF treatment (BAF) and animals subjected to HI and vehicle injections (Veh), demonstrating that the appearance of the caspase-dependent 120 kDa degradation product could be virtually abolished by BAF treatment. The calpain-dependent 150/145 fragments were much less affected. (b) An AIF immunoblot of the P2 fractions from the same animals as in Fig. 10a, demonstrating that the AIF contents decreased after HI (upper panel). The two middle panels show an α-tubulin immunoblot and SYPRO Ruby staining of the same membrane as in the upper panel, demonstrating equal loading. The bottom panel reflects the density of the AIF bands (average of 4–7 assays) showing that caspase inhibition had no appreciable effect on the extent of AIF loss from the mitochondrial fraction.

Download figure to PowerPoint

image

Figure 10. Correlation between the numbers of AIF-positive nuclei and cortical infarct volumes. The average numbers of AIF-positive nuclei per visual field in the cortex, plotted against the infarct volumes in the cortex for each of 32 animals. The correlation coefficients were very similar (r2 = 0.656 and 0.630 for vehicle- and BAF-treated animals, respectively), and no apparent changes were induced by caspase inhibition. Vehicle-treated (●) and BAF-treated (□).

Download figure to PowerPoint

Table 1.  Brain damage was evaluated by infarct volumes and brain injury scoring in different brain regions 72 h after HI
 Infarct volumeBrain injury scoring
BAFVehiclepBAFVehiclep
  1. The infarct volumes (in mm3), as well as the brain injury scores, for 4 different brain regions, plus the total volume or total score, are listed for the vehicle-treated (n = 33) and BAF-treated (n = 33) animals. The Mann–Whitney U-test was used to check for statistical differences, but all comparisons were non-significant (ns).

Cortex29.1 ± 14.128.9 ± 11.7ns2.76 ± 1.122.45 ± 0.94ns
Striatum5.2 ± 2.05.8 ± 1.7ns3.91 ± 1.313.33 ± 1.14ns
Thalamus5.2 ± 2.44.6 ± 2.1ns2.79 ± 1.792.45 ± 0.97ns
Hippocampus3.1 ± 1.53.0 ± 1.3ns3.18 ± 1.512.73 ± 1.13ns
Total42.5 ± 18.842.3 ± 15.4ns12.64 ± 4.6410.97 ± 3.62ns

AIF expression during normal development

  1. Top of page
  2. Abstract
  3. References

The total levels of AIF in homogenates did not change during normal brain development from P0 to adulthood. This stands in contrast to the increase of the mitochondrial marker COX, and the apoptosis-related, mitochondrial intermembrane proteins cytochrome c and caspase-9 (Fig. 1). It appears that the apoptosis-related proteins cytochrome c and caspase-9 were up-regulated in parallel with the general mitochondrial marker COX. This up-regulation with age stands in sharp contrast to the extensive down-regulation of the downstream effectors caspase-3 (Fig. 1) (Blomgren et al. 2001) and caspase-12 (unpublished data). Other investigators have not found an up-regulation of caspase-9 with age. There are two reports demonstrating unchanged levels of the proform of caspase-9 in the rat brain during development (Shimohama et al. 2001; Yakovlev et al. 2001) and one report even a decrease (Ota et al. 2002). The constant levels of AIF and increasing levels of COX demonstrate a relative down-regulation of AIF with age, when normalized to COX, indicating that AIF may be relatively more important in the early developmental neuronal apoptosis.

AIF expression and translocation after HI

  1. Top of page
  2. Abstract
  3. References

Damaged neurons, as judged by the loss of MAP-2, demonstrated a conspicuous, predominantly nuclear AIF staining (Figs 2 and 3), similar to that reported after traumatic brain injury in rats (Zhang et al. 2002). Translocation from mitochondria to nuclei was evident early, immediately after HI, even before MAP-2 was lost, and gradually became more pronounced during the first few hours of recovery (Figs 2 and 4). The redistribution of AIF from mitochondria to nuclei was extensive, but not complete, as judged by tissue stainings (Fig. 3) and immunoblots (Fig. 7b), indicating either that only part of the AIF pool was released from each mitochondrion, or that only some of the mitochondria released their AIF. The latter alternative is more likely, supported by the finding that there is great variation in the extent of mitochondrial swelling at 30 min and 3 h post-HI (Puka-Sundvall et al. 2000a). The AIF and cytochrome c stainings indicate different modes of release, because redistribution of AIF could be detected earlier than that of cytochrome c. Changing the detection limits, using tyramide signal amplification or different dilutions of the primary antibodies, did not alter this relative order of appearance (data not shown), indicating that this was not just a matter of different sensitivities of the AIF and cytochrome c protocols. Furthermore, all cytochrome c-positive cells also contained AIF-positive nuclei, but no cells displayed cytochrome c redistribution only. This indicates that there is a transient stage where AIF translocation, but not cytochrome c release, has occurred, similar to what was reported earlier in Rat-1 cells treated with staurosporine (Daugas et al. 2000) and in fibroblasts after N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) treatment (Yu et al. 2002). The patterns of AIF and cytochrome c immunostaining appeared to follow the time-course for development of injury in the three selected brain regions (Fig. 6). Cortex displays tissue damage which usually develops 3–24 h after HI in this model, and the number of AIF-positive cells peaked at 8 h post-HI (Fig. 6b). The nucleus habenularis typically displays signs of damage (e.g. loss of MAP-2) earlier than other regions. Consequently, the AIF staining reached a maximum level already 1–3 h post-HI (Fig. 6c). The dentate gyrus reacts more slowly, and in this region AIF staining did not peak until around 24 h post-HI (Fig. 6d).

Correlation with DNA damage

  1. Top of page
  2. Abstract
  3. References

There are at least two parallel pathways leading to chromatin processing during apoptosis. One involves the intrinsic pathway with apoptosome formation, leading to activation of caspase-3, cleavage of inhibitor of caspase-activated DNase (ICAD) and subsequent activation of caspase-activated DNase (CAD). CAD, in turn, produces oligonucleosomal DNA fragmentation and advanced chromatin condensation (Enari et al. 1998). The other pathway, which is caspase-independent, involves AIF and leads to large-scale (∼50 kb) DNA fragmentation and peripheral chromatin condensation (Susin et al. 2000). The oligonucleotide hairpin probe (HPP) with one A overhang in the 3′ end can be specifically ligated to double-strand DNA breaks with one T overhang in the 3′ end, and has been demonstrated to be useful in the detection of cells undergoing apoptosis (Didenko and Hornsby 1996; Didenko et al. 1998). We have earlier compared three different markers of DNA damage in this model of neonatal HI, and found that the HPP was more sensitive (appeared earlier) than both TUNEL and an antibody against single-stranded DNA, and showed a closer correlation with caspase-3 activation (Zhu et al. 2000). In this study we found that the HPP also showed a close correlation with nuclear translocation of AIF, more so than TUNEL. During early recovery (1 h) there were more TUNEL- than AIF-positive cells, whereas during late recovery (24 h) AIF and TUNEL staining were highly colocalized in small, presumably pyknotic cells (not shown), but larger AIF-positive cells at this time point were usually HPP-negative (Fig. 5) and TUNEL-negative (not shown). Strong AIF immunoreactivity was colocalized with HPP at least as early as 1 h post-HI, concurrent with nuclear condensation (not shown). This indicates either that the type of DNA breaks detected by the HPP can be induced by AIF, or that CAD-induced such DNA breaks occur in parallel with AIF translocation in the same cells, or both. The finding that TUNEL stained both AIF-positive and AIF-negative cells at 1 h post-HI, whereas the HPP largely colocalized with AIF staining, indicates that the HPP is selective for AIF-induced DNA damage. The finding that caspase inhibition did not change the number of HPP-positive cells 3 days post-HI might also indicate that the HPP preferentially labels AIF-induced DNA breaks. It may also be that HPP, like TUNEL and the antibody against single-stranded DNA, is less selective at this late time point of reperfusion, when cellular (and nuclear) disintegration is advanced and presumably many different types of DNA breaks are present (Zhu et al. 2000). This is supported by the finding that the number of HPP-positive cells is much higher (approximately sevenfold) than the number of cells with AIF-positive nuclei at 3 days post-HI, indicating that the AIF-positive cells appear transiently, whereas HPP-positive cells accumulate.

Caspase-independence

  1. Top of page
  2. Abstract
  3. References

Caspase inhibition, using an inhibitor selective for caspase-3/7 immediately after HI (Han et al. 2002), or the multicaspase inhibitor BAF prior to (Adachi et al. 2001) or up to 3 h after HI (Cheng et al. 1998), has earlier been demonstrated to provide tissue protection in this model of neonatal HI. In our hands, the dose used in the latter two studies (0.1 µmol) did not produce significant caspase inhibition at 24 h post-HI (data not shown), the time point when caspase-3 activation is maximal in this model (Zhu et al. 2000; Wang et al. 2001). When we used 0.2 + 0.2 µmol at 2 and 12 h post-HI, respectively, we achieved significant inhibition of caspase-3, -2 and -9 (Fig. 8), and prevented the appearance of the caspase-dependent 120 kDa fodrin breakdown product (Fig. 9a), but we could see no tissue protection, as judged by the infarct volumes or damage scoring (Table 1). This discrepancy may be due to that a different vehicle composition was used (20% and 40% DMSO, respectively), and/or that the 12 h time point used for BAF administration in our case was too late to have a protective effect even though it was effective from the point of view of caspase inhibition. On the other hand, Cheng and coworkers also injected BAF intraperitoneally at 2.5 and 12 h post-HI and were able to demonstrate clear tissue protection (Cheng et al. 1998). A more severe insult may also provide less favorable circumstances for pharmacological intervention than a less severe one, and our insult was somewhat more severe than that used by Cheng et al. (e.g. 37.9% ± 15.3% average cortical loss in our study compared with approx. 25–35%). However, when using a similar, or even more severe, insult the NMDA receptor antagonist MK-801 affords protection in our hands (Hagberg et al. 1994; Puka-Sundvall et al. 2000b). There are studies demonstrating neuroprotection by caspase inhibition after ischemia in adult mice (Hara et al. 1997; Endres et al. 1998), but also conflicting reports (Gill et al. 2002). Nevertheless, the caspase inhibition demonstrated was useful when studying the caspase-dependence of AIF redistribution. It was clear that loss of AIF from the mitochondrial fraction was not affected by caspase inhibition. Neither the numbers of AIF-positive nuclei nor the numbers of HPP-positive cells at 72 h post-HI were affected by caspase inhibition in any of the six regions studied (not shown). Furthermore, the number of AIF-positive nuclei displayed a linear correlation with the infarct volume in the cortex, and BAF treatment did neither change the linear correlation nor the distribution along the correlation curve (Fig. 10). It is thus possible that the caspase-independent, AIF-mediated, cell death pathway plays a major role in the immature brain.

In summary, we found early redistribution of AIF from mitochondria to nuclei, starting immediately after HI. This translocation appeared earlier than that of cytochrome c, indicating differential release of these two mitochondrial apoptogenic proteins. AIF redistribution occurred in areas with neuronal damage and displayed a close correlation with a marker of DNA damage. The larger the infarct volume, the higher the number of AIF-positive nuclei, indicating involvement of AIF in the process of cell death. None of these parameters were affected using the broad spectrum caspase inhibitor BAF, confirming earlier findings that AIF acts in a caspase-independent manner. Taken together, these findings indicate that AIF may be an important player in the process of neuronal death after HI.

Acknowledgements

  1. Top of page
  2. Abstract
  3. References

This work was supported by the Swedish Research Council (to HH and KB), the National Natural Science Foundation of China (to CZ), the Åhlén Foundation, the Wilhelm and Martina Lundgren Foundation, the Sven Jerring Foundation, the Magnus Bergvall Foundation, the Frimurare Barnhus Foundation, the Excellent Youth Foundation of Henan Province, and a special grant from the Ligue Nationale contre le Cancer (to GK).

References

  1. Top of page
  2. Abstract
  3. References
  • Adachi M., Sohma O., Tsuneishi S., Takada S. and Nakamura H. (2001) Combination effect of systemic hypothermia and caspase inhibitor administration against hypoxic-ischemic brain damage in neonatal rats. Pediatr Res. 50, 590595.
  • Blomgren K., Zhu C., Wang X., Karlsson J. O., Leverin A. L., Bahr B. A., Mallard C. and Hagberg H. (2001) Synergistic activation of caspase-3 by m-calpain after neonatal hypoxia-ischemia: a mechanism of ‘pathological apoptosis’? J. Biol. Chem 276, 1019110198.
  • Bona E., Hagberg H., Loberg E. M., Bågenholm R. and Thoresen M. (1998) Protective effects of moderate hypothermia after neonatal hypoxia-ischemia: short- and long-term outcome. Pediatr Res. 43, 738745.
  • Chai J., Du C., Wu J. W., Kyin S., Wang X. and Shi Y. (2000) Structural and biochemical basis of apoptotic activation by Smac/DIABLO. Nature 406, 855862.
  • Cheng Y., Deshmukh M., D'Costa A., Demaro J. A., Gidday J. M., Shah A., Sun Y., Jacquin M. F., Johnson E. M. and Holtzman D. M. (1998) Caspase inhibitor affords neuroprotection with delayed administration in a rat model of neonatal hypoxic-ischemic brain injury. J. Clin. Invest. 101, 19921999.
  • Daugas E., Susin S. A., Zamzami N., Ferri K. F., Irinopoulou T., Larochette N., Prevost M. C., Leber B., Andrews D., Penninger J. and Kroemer G. (2000) Mitochondrio-nuclear translocation of AIF in apoptosis and necrosis. FASEB J. 14, 729739.
  • Didenko V. V. and Hornsby P. J. (1996) Presence of double-strand breaks with single-base 3′ overhangs in cells undergoing apoptosis but not necrosis. J. Cell Biol. 135, 13691376.
  • Didenko V. V., Tunstead J. R. and Hornsby P. J. (1998) Biotin-labeled hairpin oligonucleotides: probes to detect double-strand breaks in DNA in apoptotic cells. Am. J. Pathol. 152, 897902.
  • Du C., Fang M., Li Y., Li L. and Wang X. (2000) Smac, a mitochondrial protein that promotes cytochrome c-dependent caspase activation by eliminating IAP inhibition. Cell 102, 3342.
  • Enari M., Sakahira H., Yokoyama H., Okawa K., Iwamatsu A. and Nagata S. (1998) A caspase-activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD. Nature 391, 4350.
  • Endres M., Namura S., Shimizu-Sasamata M., Waeber C., Zhang L., Gomez-Isla T., Hyman B. T. and 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, 238247.
  • Gill R., Soriano M., Blomgren K., Hagberg H., Wybrecht R., Miss M. T., Hoefer S., Adam G., Niederhauser O., Kemp J. A. and Loetscher H. (2002) Role of caspase-3 activation in cerebral ischemia-induced neurodegeneration in adult and neonatal brain. J. Cereb. Blood Flow Metab. 22, 420430.
  • Gilland E. and Hagberg H. (1996) NMDA receptor-dependent increase of cerebral glucose utilization after hypoxia-ischemia in the immature rat. J. Cereb. Blood Flow Metab. 16, 10051013.
  • Gilland E., Bona E. and Hagberg H. (1998a) Temporal changes of regional glucose use, blood flow, and microtubule-associated protein 2 immunostaining after hypoxia-ischemia in the immature rat brain. J. Cereb. Blood Flow Metab. 18, 222228.
  • Gilland E., Puka-Sundvall M., Hillered L. and Hagberg H. (1998b) Mitochondrial function and energy metabolism after hypoxia-ischemia in the immature rat brain: involvement of NMDA-receptors. J. Cereb. Blood Flow Metab. 18, 297304.
  • Hagberg H., Gilland E., Diemer N. H. and Andine P. (1994) Hypoxia-ischemia in the neonatal rat brain: histopathology after post- treatment with NMDA and non-NMDA receptor antagonists. Biol. Neonate 66, 205213.
  • Hagberg B., Hagberg G., Olow I. and Van Wendt L. (1996) The changing panorama of cerebral palsy in Sweden. VII. Prevalence and origin in the birth year period 1987–90. Acta Paediatr. 85, 954960.
  • Han B. H., Xu D., Choi J., Han Y., Xanthoudakis S., Roy S., Tam J., Vaillancourt J., Colucci J., Siman R., Giroux A., Robertson G. S., Zamboni R., Nicholson D. W. and Holtzman D. M. (2002) Selective, reversible caspase-3 inhibitor is neuroprotective and reveals distinct pathways of cell death following neonatal hypoxic-ischemic brain injury. J. Biol. Chem. 277, 3012830136.
  • Hara H., Friedlander R. M., Gagliardini V., Ayata C., Fink K., Huang Z., Shimizu-Sasamata M., Yuan J. and Moskowitz M. A. (1997) Inhibition of interleukin 1beta converting enzyme family proteases reduces ischemic and excitotoxic neuronal damage. Proc. Natl Acad. Sci. USA 94, 20072012.
  • Kroemer G., Dallaporta B. and Resche-Rigon M. (1998) The mitochondrial death/life regulator in apoptosis and necrosis. Annu. Rev. Physiol. 60, 619642.
  • Lee J. M., Zipfel G. J. and Choi D. W. (1999) The changing landscape of ischaemic brain injury mechanisms. Nature 399, A7A14.
  • Leist M. and Jäättelä M. (2001) Four deaths and a funeral: from caspases to alternative mechanisms. Nat. Rev. Mol. Cell Biol. 2, 589598.
  • Mallard E. C., Williams C. E., Gunn A. J., Gunning M. I. and Gluckman P. D. (1993) Frequent episodes of brief ischemia sensitize the fetal sheep brain to neuronal loss and induce striatal injury. Pediatr Res. 33, 6165.
  • Merry D. E., Veis D. J., Hickey W. F. and Korsmeyer S. J. (1994) bcl-2 protein expression is widespread in the developing nervous system and retained in the adult PNS. Development 120, 301311.
  • Northington F. J., Ferriero D. M. and Martin L. J. (2001a) Neurodegeneration in the thalamus following neonatal hypoxia-ischemia is programmed cell death. Dev. Neurosci. 23, 186191.
  • Northington F. J., Ferriero D. M., Flock D. L. and Martin L. J. (2001b) Delayed neurodegeneration in neonatal rat thalamus after hypoxia-ischemia is apoptosis. J. Neurosci. 21, 19311938.
  • Ota K., Yakovlev A. G., Itaya A., Kameoka M., Tanaka Y. and Yoshihara K. (2002) Alteration of apoptotic protease-activating factor-1 (APAF-1)-dependent apoptotic pathway during development of rat brain and liver. J. Biochem. (Tokyo) 131, 131135.
  • Puka-Sundvall M., Gajkowska B., Cholewinski M., Blomgren K., Lazarewicz J. W. and Hagberg H. (2000a) Subcellular distribution of calcium and ultrastructural changes after cerebral hypoxia-ischemia in immature rats. Brain Res. Dev Brain Res. 125, 3141.
  • Puka-Sundvall M., Hallin U., Zhu C., Wang X., Karlsson J. O., Blomgren K. and Hagberg H. (2000b) NMDA blockade attenuates caspase-3 activation and DNA fragmentation after neonatal hypoxia-ischemia. Neuroreport 11, 28332836.
  • Puka-Sundvall M., Wallin C., Gilland E., Hallin U., Wang X., Sandberg M., Karlsson J., Blomgren K. and Hagberg H. (2000c) Impairment of mitochondrial respiration after cerebral hypoxia-ischemia in immature rats: relationship to activation of caspase-3 and neuronal injury. Brain Res. Dev. Brain Res. 125, 4350.
  • Puka-Sundvall M., Gilland E. and Hagberg H. (2001) Cerebral hypoxia-ischemia in immature rats: involvement of mitochondrial permeability transition? Dev. Neurosci. 23, 192197.
  • Raff M. C., Barres B. A., Burne J. F., Coles H. S., Ishizaki Y. and Jacobson M. D. (1993) Programmed cell death and the control of cell survival: lessons from the nervous system. Science 262, 695700.
  • Rice J. E., Vannucci R. C. and Brierley J. B. (1981) The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann. Neurol. 9, 131141.
  • Shimohama S., Tanino H. and Fujimoto S. (2001) Differential expression of rat brain caspase family proteins during development and aging. Biochem. Biophys. Res. Commun. 289, 10631066.
  • Sidhu R. S., Tuor U. I. and Del Bigio M. R. (1997) Nuclear condensation and fragmentation following cerebral hypoxia- ischemia occurs more frequently in immature than older rats. Neurosci. Lett. 223, 129132.
  • Susin S. A., Lorenzo H. K., Zamzami N., Marzo I., Snow B. E., Brothers G. M., Mangion J., Jacotot E., Costantini P., Loeffler M., Larochette N., Goodlett D. R., Aebersold R., Siderovski D. P., Penninger J. M. and Kroemer G. (1999) Molecular characterization of mitochondrial apoptosis-inducing factor. Nature 397, 441446.
  • Susin S. A., Daugas E., Ravagnan L., Samejima K., Zamzami N., Loeffler M., Costantini P., Ferri K. F., Irinopoulou T., Prevost M. C., Brothers G., Mak T. W., Penninger J., Earnshaw W. C. and Kroemer G. (2000) Two distinct pathways leading to nuclear apoptosis. J. Exp. Med. 192, 571580.
  • Vannucci R. (1989) Acute perinatal brain injury: hypoxia-ischemia, in Management of Labor (CohenW., AckerD. and FriedmanE., eds), pp. 186244. Aspen Publishers, Rockville.
  • Vekrellis K., McCarthy M. J., Watson A., Whitfield J., Rubin L. L. and Ham J. (1997) Bax promotes neuronal cell death and is downregulated during the development of the nervous system. Development 124, 12391249.
  • Volpe J. J. (2000) Neurology of the Newborn, 4th edn, p. 995. Saunders, Philadelphia.
  • Wallin C., Puka-Sundvall M., Hagberg H., Weber S. G. and Sandberg M. (2000) Alterations in glutathione and amino acid concentrations after hypoxia-ischemia in the immature rat brain. Brain Res. Dev. Brain Res. 125, 5160.
  • Wang X., Karlsson J. O., Zhu C., Bahr B. A., Hagberg H. and Blomgren K. (2001) Caspase-3 activation after neonatal rat cerebral hypoxia-ischemia. Biol. Neonate 79, 172179.
  • Whitaker J. R. and Granum P. E. (1980) An absolute method for protein determination based on difference in absorbance at 235 and 280 nm. Anal. Biochem. 109, 156159.
  • Yakovlev A. G., Ota K., Wang G., Movsesyan V., Bao W. L., Yoshihara K. and Faden A. I. (2001) Differential expression of apoptotic protease-activating factor-1 and caspase-3 genes and susceptibility to apoptosis during brain development and after traumatic brain injury. J. Neurosci. 21, 74397446.
  • Yu S. W., Wang H., Poitras M. F., Coombs C., Bowers W. J., Federoff H. J., Poirier G. G., Dawson T. M. and Dawson V. L. (2002) Mediation of poly (ADP-ribose) polymerase-1-dependent cell death by apoptosis-inducing factor. Science 297, 259263.
  • Yue X., Mehmet H., Penrice J., Cooper C., Cady E., Wyatt J. S., Reynolds E. O., Edwards A. D. and Squier M. V. (1997) Apoptosis and necrosis in the newborn piglet brain following transient cerebral hypoxia-ischaemia. Neuropathol. Appl. Neurobiol. 231625.
  • Zhang X., Chen J., Du Graham S. H. L., Kochanek P. M., Draviam R., Guo F., Nathaniel P. D., Szabo C., Watkins S. C. and Clark R. S. (2002) Intranuclear localization of apoptosis-inducing factor (AIF) and large scale DNA fragmentation after traumatic brain injury in rats and in neuronal cultures exposed to peroxynitrite. J. Neurochem. 82, 181191.
  • Zhu C., Wang X., Hagberg H. and Blomgren K. (2000) Correlation between caspase-3 activation and three different markers of DNA damage in neonatal cerebral hypoxia-ischemia. J. Neurochem. 75, 819829.