Cell Death After Exposure to Subarachnoid Hemolysate Correlates Inversely With Expression of CuZn–Superoxide Dismutase
Background and Purpose—Subarachnoid hemolysate (SAH) has been associated with oxidative brain injury, cell death, and apoptosis. We hypothesized that over-expression of CuZn–superoxide dismutase (CuZn-SOD) would protect against injury after SAH, whereas reduction of its expression would exacerbate injury.
Methods—Saline (n=16) or hemolysate (n=50) was injected into transgenic mice overexpressing CuZn-SOD (SOD1-Tg), CuZn-SOD heterozygous knockout mutants (SOD1+/−), and wild-type littermates (Wt). Mice were killed at 24 hours. Stress gene induction was evaluated by immunocytochemistry and Western blotting for hemeoxygenase-1 and heat shock protein 70. Apoptosis was evaluated by 3′-OH nick end-labeling and DNA gel electrophoresis. Cell death was quantified through histological assessment after cresyl violet staining.
Results—Histological assessment demonstrated neocortical cell death in regions adjacent to the blood injection. Overall cell death was reduced 43% in SOD1-Tg mutants (n=6) compared with Wt littermates (n=6; P<0.02). In contrast, cell death was increased >40% in SOD1+/− mutants (n=6; P<0.05). Both hemeoxygenase-1 and heat shock protein 70 were induced after SAH. Apoptosis was also present after SAH, as evidenced by 3′-OH end-labeling and DNA laddering. However, the degree of stress gene induction and apoptosis did not vary between Wt, SOD1-Tg, and SOD1+/− mice.
Conclusions—The extent of CuZn-SOD expression in the cytosol correlates with cell death after exposure to SAH in a manner separate from apoptosis. Overexpression of CuZn-SOD may potentially be an avenue for therapeutic intervention.
Blood may be released into the subarachnoid space after a variety of pathological conditions.1 Lysis of extravasated red blood cells permits the release of blood constituents, including free hemoglobin, into the subarachnoid space.2 Subarachnoid blood products have been linked to the development of vasospasm, cell injury, stress gene induction, and apoptosis.2 3 4 5 6 Sublethal cell injury after subarachnoid hemolysate (SAH) has been ameliorated by pretreatment with exogenous antioxidants, a result that implicated oxidative stress as a pathway to injury after SAH.6
Hemoglobin is an iron-containing protein that is rapidly sequestered and metabolized by neurons and microglia after SAH.7 Hemoglobin has been associated with oxidative injury in different ways. The iron in hemoglobin has been linked to development of lipid peroxidation and oxidative cell membrane injury.8 9 Through the Fenton reaction, which develops in the presence of ferrous iron and H2O2, oxidative DNA damage may occur.10 Independent of iron, hemoglobin has been shown to initiate lipid peroxidation with subsequent release of catalytically active iron and propagation of lipid peroxidation.11 Exposure of neurons to hemoglobin has been associated with cell death in a dose-dependent manner.11 12
Superoxide is an oxygen-derived free radical that has been implicated in oxidative brain injury.13 Depending on cell environment, superoxide dismutase (SOD), an endogenous free radical scavenger, has been shown to both ameliorate and exacerbate such oxidative injury.14 15 16 In particular the cytosolic isoform, CuZn-SOD, has been shown to reduce cell injury after cold-induced trauma and ischemia.17 18 CuZn-SOD has also been associated with prolonged induction of stress genes after focal and global ischemia.19 20 In the setting of chronic oxidative stress due to excitotoxicity or prolonged superoxide production, overexpression of CuZn-SOD has been shown to exacerbate neuronal death.15 16 Because oxidative stress has been implicated in the pathogenesis of cell injury after exposure to subarachnoid blood products, we investigated stress gene induction and cell injury after SAH in transgenic mice that overexpress CuZn-SOD and in heterozygous mutants deficient in CuZn-SOD. We hypothesized that overexpression of this endogenous antioxidant would ameliorate cell injury after exposure to subarachnoid blood products, while a reduction in its expression would exacerbate cell injury.
Materials and Methods
The population consisted of 3-month-old male nontransgenic normal littermates, heterozygous CuZn-SOD transgenic mice of the SOD1 TgHS/SF-218-3 strain carrying human CuZn-SOD genes (SOD1-Tg), and heterozygous CuZn-SOD knockout mice (SOD1+/−). Transgenic mice were derived from founder stock previously described.21 They were bred on a CD-1 mouse background. The SOD1-Tg and SOD1+/− mice were identified by qualitative demonstration of SOD1 with the use of nondenaturing gel electrophoresis followed by nitro blue tetrazolium staining.21 There were no phenotypic differences between SOD1-Tg mice, SOD1+/− mice, and wild-type (Wt) normal littermates.
All procedures were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by Stanford University’s Administrative Panel on Laboratory Animal Care. SAH was induced in mutant mice (n=33) and Wt littermates (n=33) by a method previously described (weight, 35 to 40 g; age, 3 months).5 Experimental animals were anesthetized with chloral hydrate (350 mg/kg IP) and xylazine (4 mg/kg IP). The rectal temperature was controlled at 37±0.5°C with a homeothermic blanket. The left femoral artery was cannulated with a PE-10 catheter, and 50 μL of autologous blood was withdrawn. The blood was then placed into a sterile container and lysed by freezing and thawing in dry ice.22 An equivalent volume of normal saline was then replaced intraperitoneally.
Animals were then placed into a stereotaxic frame (Kopf Instruments). The posterior scalp was incised on the midline, and the skull was exposed at the junction of the parietal and occipital bones. A small burr hole was made 2 mm to the right of midline. The lysed blood was then placed in a sterile 1-mL syringe with a 30-gauge needle. The needle was introduced through the burr hole, and 50 μL was placed into the subarachnoid space over the dorsal aspect of the cortex beneath the coronal suture. Hemolysate (n=50) or sterile saline for controls (n=16) was then slowly injected over 1 minute. The needle was removed, and the burr hole was filled with sterile bone wax. The incision was then closed.
Hemeoxygenase-1 and Heat Shock Protein 70 Immunocytochemistry
At 24 hours after SAH or saline, subjects (n=6 SOD1-Tg-SAH, Wt-SAH; n=4 SOD1-Tg-saline, Wt-saline) were anesthetized and killed by transcardiac perfusion with the use of heparinized 0.1 mol/L PBS followed by 3.7% formaldehyde. Afterward, 25-μm sections were cut on a vibratome. Endogenous peroxidase was blocked with 0.1 mol/L sodium azide and 0.2% H2O2 in PBS with 0.3% Triton-X. After blocking with 20% goat serum, sections were incubated with either a rabbit polyclonal anti–hemeoxygenase-1 (HO-1) antibody (1/10 000, StressGen Biotechnologies) or a mouse monoclonal anti–heat shock protein 70 (HSP70) antibody (C92 1/4000, StressGen) for 48 hours at 4°C.
HO-1 sections were washed and reacted with biotinylated goat anti-rabbit IgG antibody (1/200, Vector Laboratories) for 60 minutes at 25°C. HO-1 and HSP70 staining was visualized with the use of avidin-biotin-horseradish peroxidase (ABC kit, Vector Laboratories) followed by 0.02% diaminobenzidine and 0.06% H2O2 in PBS. To avoid background, HSP70 staining was visualized with the use of the Dako ARK peroxidase system (Dako). As a negative control, some sections were incubated without primary antibody.
Western Blot Analysis of HO-1 and HSP70
Whole-cell protein extraction was performed. Samples from neocortex were cut into pieces 24 hours after SAH or saline injection (n=6 SOD1-Tg-SAH, Wt-SAH, SOD1+/−-SAH; n=4 SOD1-Tg-saline, Wt-saline). Tissue was gently homogenized by douncing ×20 in a Teflon homogenizer (Wheaton) in ×7 cold suspension buffer (20 mmol/L HEPES-KOH [pH 7.5], 250 mmol/L sucrose, 10 mmol/L KCl, 1.5 mmol/L MgCl2, 1 mmol/L EDTA, 1 mmol/L EGTA, 1 mmol/L dithiothreitol, 0.1 mmol/L phenylmethylsulfonyl fluoride, 2 mg/L aprotinin, 10 mg/L leupeptin, 5 mg/L pepstatin, 12.5 mg/L N-acetyl-leu-leu-norleucinal). Protein concentrations were determined by the Bradford method (Bio-Rad). Extracts were then placed in an equal volume of Tris-glycine-SDS sample buffer (Novex) and heated to 85°C for 10 minutes.
Proteins (25 μg) were separated by SDS-PAGE on a 10% to 20% Tris-glycine gel (Novex) and transferred to a polyvinylidene difluoride membrane (Novex). Western blots were performed with either a rabbit polyclonal antibody against mouse HO-1 at a dilution of 1:2000 or a mouse monoclonal antibody to HSP70 at a dilution of 1:1000 (StressGen). Membranes were washed, incubated with horseradish peroxidase–conjugated secondary antibody (Boehringer Mannheim), and visualized with the use of enhanced chemiluminescence reagents (Amersham). A densitometric analysis was made on immunopositive bands from both hemorrhage and control hemispheres. The films were scanned by GS-700 imaging densitometer (Bio-Rad), and the results were quantified with the use of the Multi-analyst software program (Bio-Rad). Western blot analysis of β-actin was performed with a mouse monoclonal antibody with horseradish peroxidase–conjugated anti-mouse IgG reagents (Amersham).
In Situ Labeling of DNA Fragmentation
At 24 hours after SAH or saline injection, subjects (n=6 SOD1-Tg-SAH, Wt-SAH; n=4 Tg-saline, Wt-saline) were anesthetized and killed by transcardiac perfusion. Sections were cut to 25 μm, mounted on slides, and stained with the use of an in situ technique to detect the DNA-free 3′-OH ends by the terminal deoxynucleotidyl transferase–mediated uridine 5′-triphosphate-biotin nick end-labeling (TUNEL) reaction.23
Sections were first dehydrated, placed in chloroform to remove lipid, and rehydrated. After they were washed, sections were reacted with proteinase K (20 mg/L, Boehringer Mannheim) in 0.01 mol/L Tris-HCl (pH 8.0) at 25°C for 60 minutes, and endogenous peroxidase was blocked with the use of sodium azide and H2O2 in PBS with 0.3% Triton-X for 60 minutes. Sections were equilibrated with 1× terminal deoxynucleotidyl transferase (TdT) buffer (Life Technologies) for 15 minutes and reacted with TdT enzyme (25 mL/L buffer, Life Technologies) and biotinylated 16-dUTP (60 mL/L buffer, Boehringer Mannheim) in 1× TdT buffer at 37°C for 60 minutes. The slides were washed in 2× SSC (150 mmol/L sodium chloride, 15 mmol/L sodium citrate, pH 7.4) for 2× 15 minutes and blocked with 2% bovine serum albumin in PBS 2× 15 minutes. Staining was visualized with the use of avidin-biotin horseradish peroxidase solution (ABC kit, Vector Laboratories) followed by 0.025% diaminobenzidine, 0.04 mol/L nickel sulfate, and 0.075% H2O2 in 0.175 mol/L sodium acetate.
At 24 hours after SAH or saline injection, subjects (n=6 SOD1-Tg-SAH, Wt-SAH; n=4 SOD1-Tg-saline, Wt-saline) were anesthetized and killed by decapitation. Brains were removed, and 40 to 50 mg wet weight of neocortex was removed from the region underlying the area of injection. Samples were incubated in 0.6 mL of lysis buffer (0.5% SDS, 0.01 mol/L Tris-HCL, and 0.1 mol/L EDTA) with 0.6 mg proteinase K at 55°C for 20 hours. The DNA was extracted with equal volumes of phenol and phenol-chloroform-isoamyl alcohol (25:24:1) and precipitated in 0.2 mol/L sodium chloride in 100% ethanol at −80°C for 24 hours. The DNA was washed twice with 75% ethanol, air dried, and resuspended in DNase-free H2O (Sigma Chemical). The DNA concentration was measured with the use of To-Pro-1 dye (Molecular Probes).
Gel electrophoresis for detecting DNA laddering was performed according to the instructions of the manufacturer (Trevigen) and as described.24 Before electrophoresis, 1 μg of DNA was incubated with 50 mg/L DNase-free RNase (Boehringer Mannheim) for 30 minutes at 37°C. The samples were then reacted with Klenow enzyme and dNTP in 1× Klenow buffer (Trevigen) for 10 minutes at 25°C. After they were mixed with loading buffer, samples were subjected to electrophoresis on a 1.5% agarose gel. After sequential washes with 0.25 mol/L HCl for 20 minutes, 0.4 mol/L NaOH/0.8 mol/L NaCl for 20 minutes, and 0.5 mol/L Tris buffer, pH 7.5, for 20 minutes, DNA was transferred to a nylon membrane overnight in 10× SSC. The membrane was incubated in 5% powdered milk (Bio-Rad) and then reacted with Strept-horseradish peroxidase conjugate (Trevigen) for 30 minutes. Bands were visualized by the chemiluminescence method with the use of PeroxyGlow (Trevigen) followed by exposure to chemiluminescence film. Bands were compared by densitometry as described in the Western blotting section.
Histological Assessment of Cell Death
At 24 hours after SAH or saline injection (n=6 SOD1-Tg-SAH, Wt-SAH, SOD1+/−-SAH; n=4 Tg-saline, Wt-saline), subjects were anesthetized and killed. After transcardiac perfusion fixation, 25-μm sections were mounted on slides and dried overnight. Sections were then dehydrated, delipidized, rehydrated, and stained with cresyl violet for 5 minutes. Sections were dehydrated, coverslipped, and evaluated for cell death.
Counts of immunopositive, TUNEL-positive, and dead cells were quantified by light microscopy. Beginning at the level of the coronal suture, serial sections (n=12) were cut at intervals of 300 μm to encompass an area of 3.6 mm of neocortex beneath the region of hemorrhage. Sections were stained and evaluated by light microscopy equipped with a 250×250-μm grid. An average cell count was obtained for each animal after review of 12 sections from that animal. The ear of each animal was tagged with a number, and the genotype of the particular animal was referenced to its number in a central database. The reviewer (P.G.M.) was blinded to the genotype of the animal during all histological evaluation. When counts were completed, an average was obtained for each group (Wt, SOD1-Tg, and SOD1+/−) with respect to HO-1 immunopositive cells, HSP70 immunopositive cells, TUNEL-positive cells, and dead cells. Counts of cells were compared by an unpaired Student’s t test (2-tailed). Significance was determined as P<0.05.
For Western blots, densitometry was performed on HO-1, HSP70, and β-actin bands (n=6 for SOD1-Tg, Wt, SOD1+/−). The optical density of the bands was corrected for the intensity of the β-actin band. For DNA electrophoresis, densitometry was performed on both SOD1-Tg and Wt bands (n=6). For both Western blots and gel electrophoresis, the optical densities were averaged, and means were compared by an unpaired Student’s t test (2-tailed). Significance was determined as P<0.05.
After injection of blood or saline, 60 of 66 animals survived (9% mortality rate). Differences in survival were not statistically significant between SOD1-Tg and Wt mice (P>0.10, 2-tailed Fisher exact test). After saline injection, a small amount of blood was apparent near the injection site (Figure 1A⇓). With SAH, hemorrhage was evident in the region overlying the dorsal, right neocortex beneath the parietal bone (Figure 1B⇓). The appearance of the hemolysate was consistent with prior reports in which a similar method was used.5
Stress Gene Induction
At 24 hours after saline injection, sparse HO-1 immunoreactivity was detected in neocortex. HO-1 immunoreactivity was observed adjacent to the site of injection and was similar in SOD1-Tg (n=4) and Wt (n=4) mice (Figure 2A⇓ and 2B⇓). At 24 hours after hemolysate injection, significant HO-1 immunoreactivity was observed in the ipsilateral neocortex. HO-1 immunoreactivity was also observed in the contralateral neocortex and in the corpus callosum, caudate putamen, hippocampus, and basal forebrain. HO-1 immunopositive cells were most abundant near the site of hemolysate injection (Figure 2C⇓ and 2D⇓). Immunostaining decreased with distance from the injection site. After SAH, the abundance of HO-1 immunoreactivity (expressed as HO-1 cells per square millimeter ±SD) was similar in SOD1-Tg (213±58) and Wt (225±20; P>0.50) mice (Figure 2C⇓ and 2D⇓).
At 24 hours after saline injection, HSP70 immunoreactivity was not detected in neocortex in SOD1-Tg (n=4) or Wt (n=4) mice (Figure 3A⇓ and 3B⇓). At 24 hours after hemolysate injection, significant HSP70 immunoreactivity was observed in the ipsilateral neocortex adjacent to the site of injection (Figure 3C⇓ and 3D⇓). Occasional HSP70 immunoreactivity was observed in the ipsilateral hippocampus and corpus callosum. HSP70 immunopositive cells had a morphology similar to that of neurons (Figure 3C⇓ and 3D⇓), a result consistent with prior studies.22 After SAH, the abundance of HSP70 immunoreactivity (expressed as HSP70 cells per square millimeter ±SD) was similar in SOD1-Tg (320±69) and Wt (349±84; P>0.50) mice (Figure 3C⇓ and 3D⇓).
At 24 hours after saline injection in SOD1-Tg and Wt mice, Western blotting of neocortex for HO-1 demonstrated a faint 32-kDa immunopositive band (Figure 4⇓). The intensity of this band was similar for SOD1-Tg and Wt mice (Figure 4⇓). After injection of hemolysate into Wt, SOD1-Tg, and SOD1+/− mice, Western blotting of neocortex demonstrated a strong 32-kDa immunopositive band. The intensity of the HO-1 band after SAH was significantly greater than in saline-injected animals (Figure 4⇓). However, the intensity of this band was similar for SOD1-Tg (OD=0.79±23), SOD1+/− (OD=73±22), and Wt mice (OD=0.95±29; P>0.10) after SAH.
Western blotting of neocortex for HSP70 demonstrated a weak 70-kDa band 24 hours after saline injection (Figure 4⇑). The intensity of the HSP70 band was similar for SOD1-Tg (n=4) and Wt mice (n=4) after saline injection (Figure 4⇑). Western blotting for HSP70 after injection of hemolysate in SOD1-Tg, SOD1+/−, and Wt mice revealed a strongly immunopositive band at 24 hours (Figure 4⇑). However, the intensity of this band was similar for SOD1-Tg (OD=0.86±0.19), SOD1+/− (OD=0.83±0.19), and Wt mice (OD=0.84±0.18; P>0.50). The overall intensity of β-actin bands was relatively similar for saline- and hemolysate-injected mice and for Wt and mutant mice. Band densities of HO-1 and HSP70 were corrected for any small differences in β-actin density (Figure 4⇑).
DNA fragmentation was analyzed by the in situ labeling of DNA breaks (TUNEL). DNA fragmentation was not evident after injection of saline in either SOD1-Tg (n=4) or Wt mice (n=4; Figure 5A⇓ and 5B⇓). DNA fragmentation was observed in the ipsilateral neocortex in all subjects after injection of hemolysate (Figure 5C⇓ and 5D⇓), a result consistent with prior studies.5 TUNEL-positive cells were observed in neocortical regions closest to the hemolysate injection and had densely labeled nuclei surrounded by small TUNEL-positive particles consistent with apoptotic bodies (Figure 5C⇓ and 5D⇓).23 However, the abundance of TUNEL-positive cells (expressed as TUNEL cells per square millimeter ±SD) did not differ significantly between SOD1-Tg (872±174) and Wt (841±61; P>0.50) mice (Figure 5C⇓ through 5E).
To investigate further the presence of apoptosis as characterized by intranucleosomal DNA fragmentation, we analyzed DNA extracted from neocortex adjacent to the region of SAH by gel electrophoresis. DNA laddering was not evident 24 hours after injection of saline in either Wt mice (Figure 6⇓, lane 1) or SOD1-Tg mice (Figure 6⇓, lane 2). DNA laddering was present in neocortex 24 hours after hemolysate injection in Wt mice (Figure 6⇓, lane 3) and SOD1-Tg mice (Figure 6⇓, lane 4). The intensity of the DNA laddering, however, was not significantly different between SOD1-Tg (OD=0.70± 0.08) and Wt mice (OD=0.68±0.08; P>0.50).
The presence of cell death was evaluated by histological assessment after cresyl violet staining. At 24 hours after injection of saline, no cell death was evident in either Wt or SOD1-Tg mice (Figure 7A⇓ and 7B⇓). Cell death was evident 24 hours after injection of hemolysate in Wt, SOD1-Tg, and SOD1+/− mice. Nonviable cells were evident closest to the site of hemolysate injection (Figure 7C⇓). The abundance of dead cells varied with expression of CuZn-SOD (Figure 7D⇓ through 7F). The abundance of nonviable cells (expressed as dead cells per square millimeter ±SD) was significantly less in SOD1-Tg mice (1087±409) versus their Wt littermates (1887±493; P<0.02; Figure 7G⇓). In contrast, the abundance of nonviable cells was significantly greater in SOD1+/− mice (2710±570; P<0.05; Figure 7G⇓).
Stress Gene Expression and CuZn-SOD
Stress gene induction has been described after SAH and has been linked to oxidative injury.4 6 22 Pretreatment with exogenous antioxidants has been shown to attenuate stress induction after SAH.6 Previous studies of brain hsp70 gene expression in SOD1-Tg mutants has shown an increase in hsp70 expression over Wt mice after ischemia but a decrease in hsp70 expression over Wt mice after trauma.19 20 25 In all prior studies, hsp70 expression in both SOD1-Tg mutants and Wt mice was increased over control subjects that did not receive an oxidative insult.19 20 25 Expression of the ho-1/hsp32 gene has not been studied in SOD1-Tg mutants.
In our study stress protein expression was increased in SOD1-Tg mice, SOD1+/− knockout mutants, and their Wt littermates 24 hours after SAH. However, the degree of stress protein expression was largely unchanged between SOD1-Tg mice, SOD1+/− mutants, and Wt littermates. This result suggests that the degree of sublethal stress after SAH does not vary despite changes in the expression of CuZn-SOD. Hydrogen peroxide, a product of SOD, is a known inducer of both hsp70 and ho-1 gene expression.26 27 28 29 Consequently, the stress protein induction observed in the SOD1-Tg mice may have been a manifestation of increased hydrogen peroxide production. However, the increased expression of these same stress proteins in the SOD1+/− mutants suggested the role of hydrogen peroxide to have been a minor one.
Apoptosis and CuZn-SOD
Recent evidence has linked apoptosis to mitochondrial oxidative stress and translocation of cytochrome c.24 30 31 32 33 Although SAH has been linked to development of oxidative stress, overexpression of CuZn-SOD did not attenuate apoptosis in our study. Although the principal scavenger of mitochondrial superoxide is manganese-SOD, CuZn-SOD may play a role in apoptosis. Overexpression of CuZn-SOD in the setting of chronic excitotoxicity has been linked to an increase in apoptosis.15 This result suggests that chronic oxidative stress, presumably due to increased production of hydrogen peroxide, may predispose cells to delayed cell death. In contrast, reduction of CuZn-SOD expression has also been associated with simultaneous increases in apoptosis and brain edema after focal ischemia.34 Exacerbation of brain edema with cellular accumulation of extravasated proteins has been hypothesized to play a role in the development of apoptosis after oxidative injury.35 Although brain edema was not specifically evaluated in our study, none of the brains appeared to have had any significant edema on gross examination. Therefore, the absence of significant edema in our model may have been the reason CuZn-SOD expression was not associated with an alteration in apoptosis.35 It is possible that a different model of more severe SAH and edema, such as that observed after endovascular arterial perforation, may demonstrate CuZn-SOD to be protective against apoptosis.36
After SAH, blood is rapidly distributed throughout the brain, and hemoglobin is sequestered by neurons and microglia.6 Hemoglobin has been linked to cell injury and oxidative stress.11 12 Furthermore, apoptosis has been observed in areas of highest blood concentration.5 37 Another scenario that may explain our results is a combination of acute and chronic oxidative stress. Cells that overexpress CuZn-SOD may be protected against DNA fragmentation from the acute oxidative stress produced by hemolysate exposure.5 14 However, when chronic oxidative stress develops thereafter, some of the viable SOD1-Tg neocortical cells may be predisposed to delayed cell death from DNA fragmentation.15 It is conceivable that the sequestration of blood products by the brain could produce significant cytosolic oxidative stress with resultant secondary injury to mitochondria. Despite the overexpression of CuZn-SOD in the cytosol, the degree of apoptosis did not vary after SAH in our study. This result raises the possibility that SAH may be able to produce mitochondrial injury through a mechanism unrelated to cytosolic oxidative stress. Exploring apoptotic mechanisms after SAH in Mn- and CuZn-SOD heterozygous knockout mutants will form the basis for further studies.
Cell Death and CuZn-SOD
In our study cell death was significantly reduced in SOD1-Tg mutants compared with Wt littermates. Investigation of apoptosis did not reveal a difference between SOD1-Tg mutants and Wt mice. Our results suggest that CuZn-SOD protects cells after SAH by attenuating necrosis. Both necrosis and apoptosis may occur after an oxidative insult.38 Although apoptosis may be studied through TUNEL staining along with DNA laddering, investigation of necrosis is less precise.38 Use of basic histology may have permitted apoptotic cells to be counted along with necrotic cells. However, since the degree of apoptosis did not vary significantly between Wt mice and SOD1-Tg mutants, these cells were unlikely to account for the difference.
Reactive oxygen species and superoxide in particular have been associated with cell death and edema.16 18 The extent of cell death and brain edema has been significantly decreased in mutant mice that overexpress CuZn-SOD, especially under conditions that promote peroxynitrite formation.16 18 However, under conditions of chronic excitotoxicity or massive superoxide production, overexpression of CuZn-SOD may exacerbate apoptosis and cell death.15 16 In our study overexpression of CuZn-SOD significantly decreased cell death after SAH, while reduced expression exacerbated it. The exact mechanism by which CuZn-SOD attenuated damage was uncertain. Superoxide anions may form the toxic hydroxyl radical through the Haber-Weiss reaction in the presence of hydrogen peroxide and iron. Increased levels of CuZn-SOD coupled with the presence of glutathione peroxidase may have eliminated the precursors needed for the Haber-Weiss reaction.39 It is also possible that chronic oxidative stress, having developed in mice that overexpress CuZn-SOD, may have protected the cells paradoxically through preconditioning.15
Another possible route to cell death is through the formation of peroxynitrite from nitric oxide and superoxide.16 40 Peroxynitrite has the ability to attack cellular macromolecules directly. Peroxynitrite may also damage cells through formation of the hydroxyl radical and nitrogen dioxide, which may attack endothelial membranes.40 Elevated levels of CuZn-SOD may reduce the degree of peroxynitrite formation and attenuate cell death through this pathway.16 It is also possible that reduction of peroxynitrite formation in the endothelium may permit more nitric oxide to be available for endothelial relaxation; such an environment could attenuate vasospasm and secondary ischemia. Hemolysate has been shown to be a potent spasmogen,3 and overexpression of CuZn-SOD has been linked to a reduction in the degree of vasospasm after SAH.41
CuZn-SOD activity may also reduce brain injury after SAH by ameliorating glutamate neurotoxicity. SAH has been associated with vasospasm and secondary ischemia.3 In this setting, excitotoxicity may develop and lead to phospholipase activation and arachidonic acid metabolism. This cascade may result in superoxide formation and cell injury. Overexpression of CuZn-SOD has been shown to protect against glutamate neurotoxicity.39 The specific mechanism by which CuZn-SOD protects against hemolysate-induced cell damage will form the basis for further studies in these mutant mice.
In summary, this study has demonstrated that expression of CuZn-SOD correlates inversely with the extent of cell death after exposure to SAH. The degree of apoptosis and sublethal cellular stress does not significantly vary in the same setting.
This study was supported by a Bayer research fellowship (Dr Matz) and in part by grants NS14543, NS25372, NS36144, and NS 38653 (Dr Chan). Dr Chan is a recipient of a Javits neuroscience investigator award.
- Received March 31, 2000.
- Revision received June 15, 2000.
- Accepted July 18, 2000.
- Copyright © 2000 by American Heart Association
Bonita R, Tomson S. Subarachnoid hemorrhage: epidemiology, diagnosis, management and outcome. Stroke.. 1995;16:591–594.
MacDonald RL, Weir BK. A review of hemoglobin and the pathogenesis of cerebral vasospasm. Stroke.. 1991;22:971–982.
Delgado TJ, Brismar J, Svendgaard NA. Subarachnoid hemorrhage in the rat: angiography and fluorescent microscopy of the major cerebral arteries. Stroke.. 1985;16:595–602.
Balla J, Jacob HS, Balla G, Nath K, Eaton JW, Vercellotti GM. Endothelial-cell heme uptake from heme proteins: induction of sensitization and desensitization to oxidant damage. Proc Natl Acad Sci U S A.. 1993;90:9285–9289.
Braughler JM, Duncan LA, Chase RL. The involvement of iron in lipid peroxidation. J Biol Chem.. 1986;261:10282–10289.
Imlay JA, Chin SM, Linn S. Toxic DNA damage by hydrogen peroxide through the Fenton reaction in vivo and in vitro. Science.. 1988;240:640–642.
Bar-Peled O, Korkotian E, Segal M, Groner Y. Constitutive overexpression of Cu/Zn superoxide dismutase exacerbates kainic acid-induced apoptosis of transgenic-Cu/Zn superoxide dismutase neurons. Proc Natl Acad Sci U S A . 1996;93:8530–8535.
Ying W, Anderson CM, Chen Y, Stein BA, Fahlman CS, Copin JC, Chan PH, Swanson RA. Differing effects of copper, zinc superoxide dismutase overexpression on neurotoxicity elicited by nitric oxide, reactive oxygen species, and excitotoxins. J Cereb Blood Flow Metab.. 2000;20:359–368.
Kinouchi H, Epstein CJ, Mizui T, Carlson E, Chen SF, Chan PH. Attenuation of focal cerebral ischemic injury in transgenic mice over-expressing CuZn superoxide dismutase. Proc Natl Acad Sci U S A.. 1991;88:11158–11162.
Epstein CJ, Avraham KB, Lovett M, Smith S, Elory-Stein O, Bry C, Groner Y. Transgenic mice with increased CuZn-superoxide dismutase activity: animal model of dosage effects in Down syndrome. Proc Natl Acad Sci U S A.. 1987;83:8044–8048.
Matz P, Weinstein P, States B, Honkaniemi J, Sharp FR. Subarachnoid injections of lysed blood induce the hsp70 stress gene and produce DNA fragmentation in focal areas of the rat brain. Stroke.. 1996;27:504–513.
Keyse SM, Tyrrell RM. Heme oxygenase is the major 32-kDa stress protein induced in human skin fibroblasts by UVA radiation, hydrogen peroxide, and sodium arsenite. Proc Natl Acad Sci U S A.. 1989;86:99–103.
Keyse SM, Tyrrell RM. Induction of the heme oxygenase gene in human skin fibroblasts by hydrogen peroxide and UVA (365 nm) radiation: evidence for the involvement of the hydroxyl radical. Carcinogenesis.. 1990;11:787–791.
Kukreja RC, Kontos MC, Loesser KE, Batra SK, Qian YZ, Gbur CJ, Naseem SA, Jesse RL, Hess ML. Oxidant stress increases heat shock protein 70 mRNA in isolated perfused rat heart. Am J Physiol.. 1994;267:H2213–H2219.
Morgan RW, Christman MF, Jacobson FS, Storz G, Ames BN. Hydrogen peroxide-inducible proteins in Salmonella typhimurium overlap with heat shock and other stress proteins. Proc Natl Acad Sci U S A.. 1986;83:8059–8063.
Liu X, Kim CN, Yang J, Jemmerson R, Wang X. Induction of apoptotic program in cell-free extracts: requirement for dATP and cytochrome c. Cell.. 1996;86:147–157.
Cai J, Jones DP. Superoxide in apoptosis: mitochondrial generation triggered by cytochrome c loss. J Biol Chem.. 1998;273:11401–11404.
Murakami K, Kondo T, Kawase M, Li Y, Sato S, Chen SF, Chan PH. Mitochondrial susceptibility to oxidative stress exacerbates cerebral infarction that follows permanent focal cerebral ischemia in mutant mice with manganese superoxide dismutase deficiency. J Neurosci.. 1998;18:3659–3668.
Fujimura M, Morita-Fujimura Y, Kawase M, Copin J-C, Calagui B, Epstein CJ, Chan PH. Manganese superoxide dismutase mediates the early release of mitochondrial cytochrome c and subsequent DNA fragmentation after permanent focal cerebral ischemia in mice. J Neurosci.. 1999;19:3414–3422.
Kondo T, Reaume AG, Huang TT, Carlson E, Murakami K, Chen SF, Hoffman EK, Scott RW, Epstein CJ, Chan PH. Reduction of CuZn-superoxide dismutase activity exacerbates neuronal cell injury and edema formation after transient focal cerebral ischemia. J Neurosci.. 1997;17:4180–4189.
Bederson JB, Germano IM, Guarino L. Cortical blood flow and cerebral perfusion pressure in a new noncraniotomy model of subarachnoid hemorrhage in the rat. Stroke.. 1995;26:1086–1092.
Chan PH, Chu L, Chen SF, Carlson EJ, Epstein CJ. Reduced neurotoxicity in transgenic mice overexpressing human copper-zinc superoxide dismutase. Stroke. 1990;21(suppl III):III-80–III-82.
Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci U S A.. 1990;87:1620–1624.
Kamii H, Kato I, Kinouchi H, Chan PH, Epstein CJ, Akabane A, Okamoto H, Yoshimoto T. Amelioration of vasospasm after subarachnoid hemorrhage in transgenic mice overexpressing CuZn-superoxide dismutase. Stroke.. 1999;30:867–872.
Matz and coworkers have developed a new model of subarachnoid hemorrhage (SAH) in mice. Hemolysed arterial whole blood (50 μl) was injected through a burr hole into the subarachnoid space over the convexity of the brain in wild-type mice, transgenic mice overexpressing CuZn–superoxide dismutase (CuZn-SOD), and heterozygous knockout mice with reduced CuZn-SOD. The effects on heme oxygenase-1 (HO-1), heat-shock protein 70 (HSP70), and cell death were quantified. Overexpression of CuZn-SOD was associated with decreased cell death, whereas underexpression was associated with increased cell death. There was no correlation of cell death or CuZn-SOD expression with induction of HO-1 or HSP70; both proteins were increased to a similar degree in all 3 groups.
These experiments raise several questions. Why was there a difference in cell death but no different in induction of HO-1 and HSP70? It has been suggested that both of these proteins are able to protect neurons from ischemia and, for HO-1, from oxidative stress.R1 R2 It may be that after SAH, neuronal toxicity is not mediated by reactions or processes that can be prevented by HO-1 and HSP70 or that their effect is maximal already, so that further induction does not prevent damage. This is a matter of speculation at present. There are complex reactions between SOD and free radicals that may increase or decrease free radical–mediated cell damage, depending on the type and concentration of compounds.R3 SOD may increase or decrease cellular injury. Clinical trials have been disappointingly negative, at least in head injury, and the ability of SOD to both reduce and increase neuronal injury suggests that the following questions be addressed before conducting more extensive clinical trials.R4
What component or components of hemolyzed blood cause neuronal death, and how do they do so? Investigations have shown that hemoglobin is toxic to neurons in vitro,R5 although direct hemoglobin toxicity has probably been overestimated in prior studies through use of contaminated, impure solutions.R6 Hemoglobin may mediate toxicity to brain after intracerebral hemorrhage.R7 Other components of blood are almost certainly involved in view of data from Matz et al, who have shown previouslyR8 R9 that induction of stress genes after SAH in rats is more marked in response to hemolysate that to hemoglobin. Furthermore, whole blood caused more cell death in an intracerebral hemorrhage model in rats than various blood fractions.R10 The mechanism remains unexplored, although this experiment and prior workR11 suggest that free radical mechanisms may be involved. More experiments are required to answer these questions.
This model examines changes in the brain and neurons after SAH. The hemolysate appears to exert most of its effects over the convexity and does not reach the larger cerebral arteries at the base of the brain, which is what occurs after human SAH. Therefore, it is unlikely to be a good model of vasospasm. Kamii et alR12 used an endovascular perforation model of SAH to demonstrate that transgenic mice overexpressing CuZn-SOD had less vasospasm than wild-type littermates. The magnitude of vasospasm was mild, on the order of 20% 3 days after SAH. The model of Matz et al, therefore, investigates potential effects of acute SAH on the brain. The most important cause of poor outcome after SAH is the brain damage from the initial hemorrhage. The etiology of this damage is probably multifactorial: factors suggested to be important include transient global ischemia from increased intracranial pressure or acute vasospasm occurring at the time of the hemorrhageR13 and direct toxic effects of the blood on the brain.R14 Acute vasospasm seems to occur after SAH in rats, but it is not clearly documented to occur in man.R13 R15 Matz et al did not report on the diameters of cortical vessels at the site of SAH, so whether acute vasospasm contributes in this model is not known. The skull is open, and it is unlikely that there is ischemia from increased intracranial pressure. This leaves some direct effect of the blood as a potential mechanism. It has been known for some time that SAH causes a variety of changes acutely, including increased cerebrospinal fluid lactate, depletion of brain high-energy phosphates, increased extracellular potassium, decreased extracellular calcium, inhibition of Na+,K+-ATPase, increased cerebrovascular resistance, and a decrease in cerebral blood flow independent of changes in intracranial pressure, all of which suggests that subarachnoid blood exerts deleterious effects on the brain.R14 R16 R17 What is responsible for this is unclear, which leads back to the questions of mechanism posed above. The development of a mouse model to investigate these questions is ideal because of the ability to use transgenic and knockout animals.
- Received March 31, 2000.
- Revision received June 15, 2000.
- Accepted July 18, 2000.
Ying W, Anderson CM, Chen Y, Stein BA, Fahlman CS, Copin JC, Chan PK, Swanson RA. Differing effects of copper, zinc superoxide dismutase overexpression on neurotoxicity elicited by nitric oxide, reactive oxygen species, and excitotoxins. J Cereb Blood Flow Metab. 2000;20:359–368.
Regan RF, Panter SS. Neurotoxicity of hemoglobin in cortical cell culture. Neurosci Lett. 1993;153:219–222.
Matz PG, Massa SM, Weinstein PR, Turner C, Panter SS, Sharp FR. Focal hyperexpression of hemeoxygenase-1 protein and messenger RNA in rat brain caused by cellular stress following subarachnoid injections of lysed blood. J Neurosurg. 1996;85:892–900.
Matz P, Weinstein P, States B, Honkaniemi J, Sharp FR. Subarachnoid injections of lysed blood induce the hsp70 stress gene and produce DNA fragmentation in focal areas of the rat brain. Stroke. 1996;27:504–512.
Turner CP, Panter SS, Sharp FR. Anti-oxidants prevent focal rat brain injury as assessed by induction of heat shock proteins (HSP70, HO-1/HSP32, HSP47) following subarachnoid injections of lysed blood. Brain Res. 1999;65(Mol Brain Res):87–102.
Kamii H, Kato I, Kinouchi H, Chan PH, Epstein CJ, Akabane A, Okamoto H, Yoshimoto T. Amelioration of vasospasm after subarachnoid hemorrhage in transgenic mice overexpressing CuZn-superoxide dismutase. Stroke. 1999;30:867–871.
Sehba FA, Ding WH, Chereshnev I, Bederson JB. Effects of S-nitrosoglutathione on acute vasoconstriction and glutamate release after subarachnoid hemorrhage. Stroke. 1999;30:1955–1961.