• Heme oxygenase isozymes;
  • Bile pigments;
  • Reactive oxygen species;
  • Cell death;
  • Glutamate neurotoxicity.


  1. Top of page
  2. Abstract
  5. Comparison of HO-1 expression and response to glutamate of Ntg and Tg mouse CGNs
  7. Acknowledgements

Abstract : This is the first report on the protective effect of heme oxygenase-1 (HO-1) overexpression against oxidative stress-mediated neuronal cell death and demonstration of a decreased production of oxygen free radicals when HO-1 levels are increased. HO-1 is the heat shock/stress cognate of the heat shock protein 32 family of proteins. A known function of these proteins is α-meso bridge-specific cleavage of the heme molecule. For the present study, we used cerebellar granular neurons (CGNs) isolated from homozygous transgenic (Tg) mice that overexpress HO-1 under neuron-specific enolase control and nontransgenic (Ntg) littermates. The Tg mouse CGNs were characterized by increased levels of HO-1 mRNA and protein, a lower resting intracellular calcium concentration, and a reduced HO-1 transcriptional response to glutamate-mediated oxidative stress. Compared with the Ntg neurons, when exposed to glutamate (30 μM or 3 mM), the magnitude of cell viability was increased and the number of cells exhibiting membrane permeability and chromatin condensation were significantly decreased in the Tg CGN cultures. The population of neurons surviving glutamate toxicity decreased when HO-1 activity was inhibited by a peptide inhibitor. The neuroprotective effect by HO-1 was extended to H2O2-induced cell death. The mechanism of protection may involve in part a reduced production of reactive oxygen species upon exposure to glutamate. We suggest that induction of HO-1 by pharmacological means may be a novel approach to amelioration of oxidative insults to neurons.

Heme oxygenase (HO) is the heat shock/stress-inducible cognate of the heat shock protein 32 family (EC The family comprises three isozymes : HO-1, HO-2, and the newly discovered HO-3 (Maines et al., 1986 ; McCoubrey et al., 1997). HO-1 and HO-2 have been fully characterized, and the isozymes represent different gene products (Shibahara et al., 1985 ; Cruse and Maines, 1988). Except for the heme-binding domain, known as “HO signature” (GenBank), which is a highly conserved domain, they share little similarity in primary structure, gene organization, or regulation (for review, see Maines, 1992). The HO isozymes cleave the heme molecule at the α-meso carbon bridge and produce the open tetrapyrrole biliverdin. The carbon bridge is converted to CO, the chelated iron is released, and biliverdin is subsequently reduced to bilirubin by biliverdin reductase.

The products of HO activity are biologically active molecules (for review, see Maines, 1997). Both biliverdin and bilirubin possess potent antioxidant properties (McDonagh, 1990 ; Dennery et al., 1995). CO, like NO, is suspected to be a signal molecule and a gaseous modulator of guanylyl cyclase activity (Marks et al., 1991 ; Maines, 1997 ; Wolin et al., 1998 ; Snyder et al., 1998 ; Denninger and Marletta, 1999), and iron is a gene regulator (Smith et al., 1998). Besides generating active molecules, HO activity decreases the levels of heme, which is well known as the most potent catalyst for lipid peroxidation and oxygen radical formation. In addition, HO-1, like other heat shock proteins, may function as a molecular chaperone.

HO-1 expression is exquisitely responsive to all types of stimuli that cause oxidative stress (Keyse and Tyrell, 1989 ; Maines, 1992 ; Abraham et al., 1995). HO-2, on the other hand, is not induced by oxidative stress (Maines et al., 1999). Oxidative stress has been postulated to be the underlying basis for neuronal cell death in neurodegenerative diseases such as Alzheimer's disease and Parkinson's disease (Simonian and Coyle, 1996 ; Keller and Mattson, 1998). Among cell types, neurons are particularly vulnerable to oxidative stress, which may in part reflect their having low levels of glutathione when compared with other cell populations in the brain (Raps et al., 1989). In this context, the oxidative stress mediated by the excitotoxic amino acid glutamate has been postulated to trigger neurotoxicity (Mattson, 1996). Glutamate-induced neuronal cell death in discrete brain areas has been suggested to be of significance to the etiology of various neurodegenerative diseases including stroke (Meldrum and Garthwaite, 1990 ; Mattson, 1996 ; Simonian and Coyle, 1996). Neurotoxicity by glutamate has been attributed to the excess activation of a subtype of NMDA receptor, subsequently leading to the generation of reactive oxygen species and ultimately to cell death (Lafon-Cazal et al., 1993 ; Nicotera et al., 1997). Recent reports have shown that HO-1 is highly inducible by the glutamate agonist kainic acid in rat brain and cultured glial cells, which suggests the activation of glutamate receptors is mediated partly by oxidative stress (Matsuoka et al., 1998, 1999). Neuronal cell death can be linked to changes in membrane permeability and nuclear morphology (Ankarcrona et al., 1995 ; Tan et al., 1998).

Because of the diverse functions that can be ascribed to HO-1, a good argument can be made in support of its involvement in cellular defense against, or exacerbation of, oxidative injury (Platt and Nath, 1998). This has been done in various studies using different systems and cell types. A previous study, which used HO-1 transgenic (Tg) mice that overexpressed the protein in neurons, showed HO-1 to be neuroprotective against ischemic stroke (Panahian et al., 1999).

In this study, we used primary cultures of cerebellar granule neurons (CGNs) obtained from these mice to investigate whether HO-1 overexpression protects against glutamate-mediated oxidative damage and, if so, to identify a cellular basis for protection. The observations made in this study led us to conclude that HO-1 is a vital component of neuronal defense mechanisms.


  1. Top of page
  2. Abstract
  5. Comparison of HO-1 expression and response to glutamate of Ntg and Tg mouse CGNs
  7. Acknowledgements

Neuronal cell cultures and treatment

Homozygous HO-1 Tg mice used in this study were from the same colony used for previous studies (Maines et al., 1998 ; Panahian et al., 1999). These mice were generated from the strain of DNX mice (Princeton, NJ, U.S.A.) by placing rat HO-1 cDNA (Shibahara et al., 1985) immediately after the neuron-specific enolase promoter, as described previously. The nontransgenic (Ntg) littermates were used as controls in this study. Seven-day-old Tg and Ntg mice born within 24 h of each other were used for CGN preparation (Miller et al., 1997).

In brief, the dissected cerebellum was treated with 1 mg/ml papain and 12.5 μg/ml DNase I at 37°C for 20 min and then plated on poly-L-lysine-coated culture plates. Cells were seeded at a density of 3 × 105 cells/cm2 in basal medium Eagle (GibcoBRL) containing 10% fetal calf serum and 25 mM KCl and maintained at 37°C at an atmosphere of 5% CO2/95% air in a humidified incubator. Cytosine arabinofuranoside (10 μM) was added to the cultures ~24 h after plating to arrest the growth of nonneuronal cells. Cultures were maintained for 7 days before treatment. The purity of neuronal cultures was assessed using neurofilament 200 monoclonal antibody (Sigma, St. Louis, MO, U.S.A.) immunostaining. At this time, neurons accounted for >98% of the cultured cells. Cells were treated with 30 μM, 100 μM, or 3 mM glutamate (Ankarcrona et al., 1995) for 30 min in Locke's solution (134 mM NaCl, 25 mM KCl, 4 mM NaHCO3, 5 mM HEPES, 2.3 mM CaCl2, 5 mM glucose, pH 7.4).

Cells were also treated for 30 min with 1 and 10 μM H2O2 followed by incubation in original medium for 3 h. Neuronal viability was measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay. As addition of fresh serum is toxic to CGNs (Schramm et al., 1990), neurons were reincubated in the original culture medium after exposure to glutamate. The D2 peptide inhibitor of HO-1 and the noninhibitor peptide D2RP (Iyer et al., 1997) were dissolved in dimethyl sulfoxide and added to the culture at a final concentration of 50 μM.

HO assay

PC12 rat pheochromocytoma cells (ATCC, Manassas, VA, U.S.A.) were cultured on rat tail collagen-coated dishes in RPMI 1640 medium containing 5% fetal bovine serum and 10% horse serum. D2RP peptide or D2 peptide (50 μM) was added to the medium, and HO activity was measured 4 h later. For measurement of HO activity, cells were homogenized in a phosphate buffer containing 0.5% Nonidet P-40 and centrifuged at 14,000 rpm for 20 min ; the supernatant fraction was then collected. HO activity in the supernatant fraction was measured in the presence of purified preparations of biliverdin reductase and NADPH-cytochrome P450 reductase (Iyer et al., 1997).

Assessment of neuronal viability

The MTT assay was used to determine cell survival in a quantitative colorimetric assay (Mosmann, 1983). This assay is based on the capacity of mitochondrial enzymes in viable neurons to reduce MTT to form the insoluble product formazan. The MTT reduction, measured by change in absorbency at 570 nm, has been used as a measure of energy production and mitochondrial reducing potential (Musser and Oseroff, 1994). In brief, MTT was dissolved in serum-free culture medium at a concentration of 0.25 mg/ml and then added to the CGNs for 30 min at 37°C. The reaction was terminated by adding a solubilization solution consisting of 50% N,N′-dimethylformamide and 20% sodium dodecyl sulfate (pH 4.8).

Assessment of neuronal membrane damage and chromatin condensation

A combination of two fluorescent dyes, fluorescein diacetate (FDA ; Sigma) and propidium iodide (PI ; Sigma), was used for assessment of membrane damage. FDA, a membrane-permeant dye, yields green fluorescence after deesterification by the living cell. PI is membrane impermeant ; hence, only cells with membrane damage display orange/red fluorescence, which is produced upon binding to DNA (Du et al., 1997). In brief, CGNs, after treatment with glutamate for 3 h, were incubated with 10 μg/ml each FDA and PI for 5 min ; thereafter, micrographs were obtained by fluorescence microscopy. The percentage of surviving neurons in 10 fields (>400 cells) of each monolayer was estimated by assessing the FDA/PI staining (Favaron et al., 1988).

Cells displaying pyknotic nuclei with condensed chromatin were detected by PI fluorescence staining (Ankarcrona et al., 1995). Neurons were fixed and permeabilized in methanol/water (4 : 1) for 15 min, washed in phosphate-buffered saline (PBS), and subsequently stained with PI (10 μg/ml) for 5 min. The percentage of cells displaying a condensed nucleus was calculated. Counts were made without knowledge of treatment history.

Northern blot analysis

Total RNA was extracted from CGNs by using Trizol (GibcoBRL) according to the manufacturer's instructions, separated by electrophoresis on denaturing formaldehyde gels, and transferred onto a Nytran membrane. The HO-1 probe and an actin probe were labeled using [α-32P]dCTP with the Rediprime random primer labeling kit (Amersham, Arlington Heights, IL, U.S.A.). Prehybridization and hybridization were performed as described previously (Ewing and Maines, 1991). Laser densitometry (LKB Ultroscan XL, Rockville, MD, U.S.A.) was used for quantitation of signal intensity.

Western blot analysis

Cells were rinsed with PBS and lysed by addition of lysis buffer consisting of 1% Nonidet P-40, 0.1% sodium dodecyl sulfate, 50 mM Tris (pH 8.0), 50 mM NaCl, 0.05% deoxycholate, and protease inhibitors (Boehringer Mannheim, Indianapolis, IN, U.S.A.). The supernatant fraction was obtained by centrifugation at 12,000 rpm for 15 min, and 20-μg aliquots were used for analysis. Nonspecific binding was blocked by incubation in PBS containing 3% bovine serum albumin, 3% nonfat milk, and 0.1% Tween 20 for 3 h at room temperature. Blots were probed with rabbit anti-HO-1 polyclonal antibody (Maines et al., 1986). The secondary antibody was a 1 : 2,000 diluted horseradish peroxidase-conjugated goat anti-rabbit antibody (Bio-Rad). Detection was made by the enhanced chemiluminescence method using ECL western blotting reagents (NEN, Boston, MA, U.S.A.).

Measurement of reactive oxygen species

Oxygen radical production was detected using the dye dichlorofluorescein diacetate (DCF). The cleavage product of DCF, 2′,7′-dichlorofluorescein, fluoresces upon oxidation by reactive oxygen species (Rosenkranz et al., 1992 ; Kane et al., 1993). In brief, the cultures, in a 48-well plate, were washed with a modified Krebs-Ringer solution (20 mM HEPES, 10 mM glucose, 127 mM NaCl, 5.5 mM KCl, 1 mM CaCl2, and 2 mM MgSO4, pH 7.4). Subsequently, Krebs-Ringer buffer, containing DCF (1 μg/ml) plus 30 μM or 3 mM glutamate, in a total volume of 200 μl was added. Fluorescence was measured using a Wallac 1420 Counter (Gaithersburg, MD, U.S.A.) with an excitation wavelength of 485 nm and an emission wavelength of 535 nm (Rosenkranz et al., 1992).

Measurement of intracellular free calcium

The [Ca2+]i was measured as described before (Gunter et al., 1990) using fluorescence microphotometry and the Ca2+-sensitive indicator fura-2. For this purpose, CGNs were plated onto 25-mm round glass coverslips and were loaded with 5 μM cell-permeable fura-2 acetoxymethyl ester (Molecular Probes, Eugene, OR, U.S.A.) dissolved in dimethyl sulfoxide at 1 mM in Locke's solution (128 mM NaCl, 5 mM KCl, 1.2 mM Na2HPO4, 2.7 mM CaCl2, 10 mM glucose, 20 mM HEPES, pH 7.4) at 37°C for 30 min. Neurons were then washed with Locke's solution and left to equilibrate for 20 min at room temperature. Fluorescence from a group of neurons was detected using a photon-counting photomultiplier system (PTI, D104 microscope photometer, South Brunswick, NJ, U.S.A.) equipped with a Nikon Diaphot-TMD microscope and a Fluor 40×, 1.3 numerical aperture oil immersion objective. Fura-2 was excited at 340 and 375 nm with its emission monitored at 510 nm ; the increase in 340/375-nm excitation ratio, as a function of [Ca2+]i, was calculated. For each coverslip, fluorescence from 20-30 neurons in a given microscopic field was obtained.

Immunocytochemical staining

CGNs grown on glass coverslips for 7 days were fixed by treatment with 4% paraformaldehyde for 10 min. Cells were washed with PBS and were subsequently incubated (overnight at 4°C) with rabbit anti-rat HO-1 polyclonal antibodies (Stress-Gen, Vancouver, Canada). The secondary antibody, horseradish peroxidase-conjugated goat anti-rabbit antibody (Vector, Burlingame, CA, U.S.A.), was used according to the manufacturer's instruction. A peroxidase substrate kit (VIP ; Vector) was used as the chromogen.


Data were analyzed by Student's t test with statistical significance established at p < 0.05.

Comparison of HO-1 expression and response to glutamate of Ntg and Tg mouse CGNs

  1. Top of page
  2. Abstract
  5. Comparison of HO-1 expression and response to glutamate of Ntg and Tg mouse CGNs
  7. Acknowledgements

The Tg mice used in the present study were previously characterized for overexpression of HO-1 in neuronal populations by immunohistochemical staining of brain and analyses of heme oxidation activity and western blotting of the hippocampal tissue (Maines et al., 1998). These Tg mice express HO-1 under the control of neuronal enolase. This promoter is not very robust, but it is specific to neurons. To confirm that isolated neurons in the culture continue to express increased levels of HO-1, CGNs from Ntg and Tg mice were subjected to northern blotting for analysis of HO-1 message levels (Fig. 1a), western blotting (Fig. 1b), and immunocytochemistry for analysis of protein expression (Fig. 1c and d). Quantitation of the ratio of HO-1 mRNA relative to that for actin mRNA by laser densitometry showed nearly twofold higher levels of HO-1 message in CGNs isolated from the Tg mouse brain than from Ntg mouse CGNs. Similarly, western blot analysis showed a twofold increase in intensity of the HO-1-immunoreactive band. HO-1 immunostaining of CGNs isolated from the Tg mice was also more robust than that of the Ntg mouse neurons. These results confirmed elevated expression of the HO-1 gene in Tg neurons and indicated their suitability for the following experiments.


Figure 1. Comparative expression of HO-1 mRNA and protein in CGNs from Ntg and Tg mice. CGNs from 7-day-old mouse brains were cultured for 7 days and subjected to northern and western blot analyses and immunocytochemical staining for HO-1 expression. a : Northern blot analysis of HO-1 mRNA expression in the Ntg and the HO-1 Tg mouse CGNs. Total RNA (15 μg/lane) was blotted onto a Nytran membrane and consecutively probed with HO-1 and actin cDNA probes. Actin was used as the loading control. b : Western blot analysis of HO-1 protein expression in the Ntg and the Tg CGNs. Each lane contained 20 μg of supernatant protein. Purified Escherichia coli-expressed HO-1 lacZ fusion protein (His-tagged) was used as the standard. The protein migrates at ~34 kDa due to the additional sequences encoded by the vector and histidine tag. c and d : Demonstrations of increased HO-1 immunoreactivity in Tg CGNs. HO-1 immunostaining is shown of CGNs isolated from the Ntg (c) and the Tg (d) mice. Objective ×50. e : Induction of HO-1 mRNA by glutamate in the Ntg and the Tg mouse CGNs. The panel shows the time course of change in HO-1 mRNA ; the value of 1.0 is arbitrarily assigned to Ntg neurons at the zero time and is a representative of three independent experiments. Experimental details are provided in Materials and Methods.

Download figure to PowerPoint

Although glutamate toxicity to neurons is suspected to be associated with oxidative stress (Lafon-Cazal et al., 1993), its effect on HO-1 expression is not known. Therefore, it was relevant first to examine the effect of glutamate on HO-1 expression. For this, CGNs were obtained from the Ntg and Tg mice and exposed to glutamate for 0, 1, 3, 6, and 12 h at a concentration of 100 μM. Total RNA was isolated and probed sequentially for the expression of HO-1 and actin. As shown in Fig. 1e, a time-dependent induction of HO-1 mRNA was observed, with the maximum level being reached at 6 h in both the Ntg and the Tg mouse neurons. Relative to the zero time point of Ntg neurons, however, the magnitude of HO-1 mRNA induction was lower in the Tg neurons when compared with the response of the Ntg neurons.

HO-1 overexpression is neuroprotective

After establishment of a decreased responsiveness of the Tg mouse CGNs to glutamate-mediated oxidative stress, analyses of viability and morphological parameters were carried out to compare their response to glutamate exposures. CGNs were exposed to 30 μM and 3 mM concentrations of glutamate, which have been established to be toxic to CGNs (Ankarcrona et al., 1995 ; Nicotera et al., 1997), for 30 min. Cell viability was assessed at 3 and 24 h after removal of glutamate using the MTT assay and trypan blue exclusion assay (Fig. 2a and b, respectively). A higher proportion of the Tg mouse CGNs was found to be viable under all experimental conditions. As shown, <20% of the Tg mouse neurons lost the ability to metabolize MTT when exposed to 30 μM glutamate, whereas 40-50% of the Ntg mouse neurons, under similar conditions, lost MTT-metabolizing activity. When exposed to 3 mM glutamate, 70% of the Tg mouse CGNs retained their resistance to the killing effect of glutamate at the 3-h time point ; under similar conditions, ~40% of the Ntg mouse neurons were viable. Similarly, at 24 h, a significantly higher percentage of the Tg mouse neurons survived exposure to an extremely high concentration of glutamate when compared with the Ntg neurons (~50 vs. 30%). The findings of the trypan blue exclusion test were consistent with those obtained by MTT assay ; that is, regardless of the different assessing methods, the neuroprotective effect of HO-1 overexpression was evident. The possibility that HO-1 expression interferes with the MTT assay was excluded in preliminary experiments that did not detect any differences between Tg and Ntg CGNs in MTT reduction.


Figure 2. Protection by HO-1 against glutamate-induced neurotoxicity. CGNs were isolated from the Ntg and the Tg mice, were cultured in 48-well plates for 7 days, and thereafter were exposed to 30 μM or 3 mM final concentration of glutamate for 30 min. Neuronal viability was determined 3 and 24 h after removal of glutamate using the MTT colorimetric assay (a) and trypan blue exclusion assay (b). The results are expressed as percent of viable neurons. Open columns represent the Ntg mouse CGN values and filled columns those of the Tg mouse CGNs. Data are the means ± SD of measurements from three cultures. *p < 0.05 compared with the Ntg neurons. Neuronal morphology was assessed by phase-contrast microscopy 3 h after exposure (30 min) to 3 mM glutamate in control Ntg mouse CGNs (c), control Tg mouse CGNs (d), Ntg mouse CGNs 3 h after glutamate exposure (e), and Tg mouse CGNs 3 h after glutamate exposure (f). Objective ×22. Experimental details are provided in Materials and Methods.

Download figure to PowerPoint

Morphologically, under normal conditions, both the Ntg and the Tg mouse CGNs were differentiated into round cell bodies with a complex network of neurites (Fig. 2c and d). Exposure of the Ntg mouse CGNs to 30 μM glutamate after 3 h resulted in the typical appearance of compromised cells, which were small and translucent with the dissolution of neuritic processes (Fig. 2e). In comparison, the Tg mouse neurons appeared notably more resistant to the adverse effects of glutamate (Fig. 2f).

To examine whether the increased cellular levels of HO-1 also protect neurons against other stimuli that mediate oxidative stress, additional experiments, using hydrogen peroxide (30 min) as the oxidant, were performed. The MTT assay indicated that, as with glutamate, hydrogen peroxide was less cytotoxic to the Tg mouse CGNs than to the Ntg mouse neurons. The difference was observed with both the low (1 μM) and the high (10 μM) concentrations of hydrogen peroxide. The percent of viable CGNs for the Tg versus the Ntg mice was 83 ± 10 vs. 62 ± 7% (p < 0.05), respectively, when exposed to 1 μM H2O2 and 80 ± 5 vs. 50 ± 4%, respectively, using 10 μM H2O2.

The CGNs isolated from the Tg and the Ntg mouse brains were further compared for response to glutamate toxicity by assessing membrane permeability and nuclear morphology. In one set of experiments (Fig. 3a—d), neurons were exposed to 30 μM glutamate for 30 min and thereafter were incubated for 3 h in the absence of glutamate. Cultures were then loaded with the membrane-permeant FDA dye (green fluorescent) and the membrane-impermeant chromatin dye PI (orange/red fluorescent). As shown in Fig. 3, for the most part, both the untreated Ntg (a) and the Tg (b) mouse neurons displayed green fluorescence and excluded PI. After exposure to glutamate, neurons displaying orange fluorescence were detected in both cell lines ; however, a notably greater number of such nuclei were present among the Ntg neuronal population (c vs. d).


Figure 3. HO-1 overexpression attenuates the morphological features of cell death. CGNs were isolated from the HO-1 Tg and the Ntg mice as described in Materials and Methods. Neurons were exposed to 30 μM glutamate for 30 min. a-d : The effect of glutamate treatment on membrane integrity was assessed after 3 h by measuring the percentage of cells that displayed orange fluorescence of the nucleus subsequent to FDA and PI treatment. e-h : The effect of glutamate treatment on nuclear morphology was assessed after 24 h by measuring the percentage of neurons displaying condensed nuclei, as visualized by PI staining after methanol fixation. a and e : Control Ntg CGNs. b and f : Control Tg CGNs. c and g : Glutamate-treated Ntg CGNs. d and h : Glutamate-treated Tg CGNs.

Download figure to PowerPoint

Next, the effect of glutamate on chromatin structure was examined by assessing nuclear condensation. CGNs were again exposed to 30 μM glutamate for 30 min but were maintained in the culture for 24 h. The cultures were subsequently fixed and permeabilized with methanol followed by staining with PI. As noted in Fig. 3e and f, in the absence of glutamate, both the Ntg (e) and the Tg (f) mouse neurons displayed a similar appearance. The number of condensed nuclei, however, was notably higher in the culture of the Ntg neurons when compared with that in the Tg preparation (g vs. h). This experiment was extended to include 3 mM glutamate exposure. The findings were tabulated and the data are shown in Fig. 4. As suggested by the tabulated data, a number of cells display both membrane damage and nuclear condensation. In the case of both the Ntg and the Tg neurons, the sum of the cells that displayed the membranes and the nuclear effects of glutamate toxicity exceed the percentage of viable cells (Fig. 2a).


Figure 4. HO-1 protects neuron from glutamate-induced death related to membrane damage and chromatin distortion. After 7 days in the culture, CGNs isolated from the Tg and the Ntg mouse cerebella were treated with 30 μM or 3 mM glutamate for 30 min. a : Membrane permeability was measured after 3-h incubation, at which time cells were double-stained with Pl and FDA. The percentage of cells that displayed PI nuclear fluorescence staining was tabulated. b : Nuclear condensation was measured 24 h after glutamate exposure, at which time cells were fixed, permeabilized, and stained with PI. The results are those obtained from 10 fields scored (>400 cells). Experimental details are described in Materials and Methods. Open columns represent the Ntg mouse CGN values and filled columns the Tg mouse CGN values. Data are the means ± SD. *p < 0.05 compared with the Ntg CGNs.

Download figure to PowerPoint

Inhibition of HO exacerbates glutamate-induced neurotoxicity

Recently, a peptide inhibitor of HO-1 was described (Iyer et al., 1997). This peptide, referred to as D2, was shown to inhibit HO-1 activity in a reconstituted system as well as in the cell culture system. This D2 peptide was used to assess the effect of HO-1 inhibition on glutamate-induced toxicity. First, however, the D2 peptide and D2RP peptide (Iyer et al., 1997), a modified form of D2, were tested in PC12 cells, which display a neuronal phenotype, to confirm the inhibitory activity of D2. (The number of 7-day-old mouse pups that are required to be killed for isolating CGNs for the measurement of HO activity is prohibitive.) The inhibitory activity of the D2 peptide was confirmed, and as shown in Fig. 5a, when used at equimolar concentrations (50 μM), the D2 peptide inhibited HO activity to the same extent as was reported before with other cell culture systems (~45%), whereas the same concentration of D2RP peptide did not inhibit the activity. HO-1 is a highly conserved gene product, and its primary structure is >90% identical among mammalian species. Cultured Ntg CGNs were treated with D2 or D2RP peptides for 1 h before the addition of 30 μM glutamate. Neuronal survival was assessed 3 h after glutamate treatment by the MTT assay. The results are shown in Fig. 5b. In the presence of the D2 peptide, CGNs showed significantly less resistance to the glutamate challenge, and nearly 60% of the neurons lost viability. The same concentration of D2RP peptide did not cause an exacerbation of glutamate-induced neuronal death.


Figure 5. HO peptide inhibitor increases the sensitivity of CGNs to glutamate-induced neurotoxicity. a : PC12 cells were cultured in RPMI 1640 medium in the presence of 50 μM D2 or D2RP peptide (inhibitor and noninhibitor, respectively) for 4 h. Thereafter, the supernatant fraction was prepared and used for determination of HO activity as described in Materials and Methods. b : CGNs from the Ntg mice were cultured for 7 days and exposed to glutamate (30 μM) in the presence or absence of pretreatment with 50 μM D2 or D2RP peptides. Cell survival was measured using the MTT assay 3 h after glutamate treatment. *p < 0.05 compared with the control (C) group ; ¶p < 0.05 compared with the glutamate-treated alone or the glutamate plus D2RP-treated groups.

Download figure to PowerPoint

Cellular basis for protection offered by HO-1 overexpression

As noted above, [Ca2+]i levels are suspected to play a significant role in cell death (Ankarcrona et al., 1995 ; Juin et al., 1998) ; therefore, [Ca2+]i in the Tg and the Ntg mouse CGNs under normal conditions and subsequent to glutamate exposure were assessed. The resting [Ca2+]i in both groups of CGNs showed steady levels without spontaneous fluctuations ; however, as shown in Fig. 6, the resting [Ca2+]i was significantly lower in the Tg than in the Ntg mouse CGNs. Addition of glutamate (100 μM) caused a rapid rise in [Ca2+]i in both neuronal cell lines, and 1 min after glutamate treatment, [Ca2+]i levels in both lines of CGNs were comparable. A stable [Ca2+]i plateau was similarly maintained in both cell lines during the 10-min exposure to glutamate.


Figure 6. Resting [Ca2+]i is lower in neurons that overexpress HO-1. CGNs isolated from the Ntg and the HO-1 Tg 7-day-old mouse brain were plated onto 25-mm glass coverslips and loaded with 5 μM cell-permeable fura-2 acetoxymethyl ester at 37°C for 30 min. Neurons were then washed and left to equilibrate for 20 min at room temperature. The 340/375 fluorescence ratio from 20-30 neurons was collected in a given microscopic field. Experimental details are provided in the text. *p < 0.05.

Download figure to PowerPoint

Next, to determine whether the protective effect conferred by HO-1 overexpression to the Tg CGNs involved a decreased generation of reactive oxygen species, the time course of oxygen radical generation by CGNs after glutamate exposure was measured (Fig. 7). For this analysis, the DCF fluorescent probe was used. This probe measures cellular levels of hydrogen peroxide and hydroxyl radicals and likely that of nitric oxide oxygen derivatives (Kane et al., 1993 ; Wolin et al., 1998). The DCF fluorescence was measured at 30, 60, or 90 min after exposure to 30 μM or mM glutamate. As shown, the values obtained for the Ntg mouse CGNs, were significantly greater than those obtained with the Tg neurons at the indicated time points with both concentrations of glutamate. As noted, the precipitous time-dependent increase in fluorescence detected in the Ntg mouse CGNs was replaced by a shallow incline in the Tg mouse neurons. The findings suggest that increased cellular levels of HO-1 are associated with a decrease in accumulation and/or formation of oxygen radicals.


Figure 7. The accumulation of reactive oxygen species in CGNs exposed to glutamate is attenuated by HO-1 overexpression. CGNs obtained from the Ntg and the Tg mice were exposed to 30 μM or 3 mM glutamate in the presence of the fluorescence probe DCF (1 μg/ml). The intracellular reactive oxygen levels were assessed by measuring the fluorescence intensity at the indicated time points. Experimental details are described in Materials and Methods. Ntg (○) and Tg (•) CGNs exposed to 30 μM glutamate ; Ntg (•) and Tg (▪) CGNs exposed to 3 mM glutamate. Data are the means ± SD of determinations made in three cultures. *p <0.05 compared with corresponding time point value for the Tg mouse CGNs.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  5. Comparison of HO-1 expression and response to glutamate of Ntg and Tg mouse CGNs
  7. Acknowledgements

This report, for the first time, describes the protective effects offered by HO-1 overexpression to neurons against glutamate-mediated oxidative stress. The HO system is a highly conserved physiological catalyst for heme degradation. Its evolutionary conservation is undoubtedly related to the catalytic activity of the HO isozymes and the stress responsiveness of the HO-1 gene expression. The biological significance of HO-1 induction remains a matter of debate. Reports have appeared identifying the induction as a prelude to cell death and others show protection (da Silva et al., 1996 ; Dwyer et al., 1998 ; Ye and Laychock, 1998 ; Panahian et al., 1999). For example, administration of HO-1 cDNA via viral vectors has been reported to protect against heme/hemoglobin toxicity in endothelial cells and hyperoxia-induced lung injury (Abraham et al., 1995 ; Otterbein et al., 1999). Conversely, with use of metalloporphyrin inhibitors of HO isozymes, neuroprotection has been reported against ischemic injury (Dwyer et al., 1998).

The present study has used neurons isolated from HO-1 Tg mice that overexpress the gene at a moderate level and a peptide that selectively inhibits HO-1, without damage to the protein, to examine the effect of HO-1 on cell survival/cell death when exposed to oxidative stress. CGNs have been extensively used for such characterization, and glutamate has been commonly used as the mediator of oxidative stress in this system (Ankarcrona et al., 1995 ; Du et al., 1997). The present study indicates that steady-state elevated levels of HO-1 dampen the stress response of the cell, as suggested by the attenuated increase in induction of this stress protein by glutamate ; this is reminiscent of the well established preconditioning protective effect offered by various stimuli. Moreover, Ankarcrona et al. (1995) reported that glutamate caused dose-dependent neuronal cell killing with a rapid onset. Neuronal cell death occurred within the first 3 h following exposure, and no further increase in cell death took place between 3 and 24 h, as assessed by measuring MTT and trypan blue exclusion (Ankarcrona et al., 1995). Consistent with the previous results, the higher level of HO-1 expression in Tg CGNs within 3 h following glutamate treatment may account for the protective effect against early neurotoxicity. It is reasonable to suggest that an increased HO-1 level in the cell, prior to exposure to stress, is crucial to protection offered against oxidative damage. Support for this suggestion is found in the observations that (a) the Tg CGNs were consistently more viable when exposed to different concentrations of glutamate at both early and late time points than the Ntg CGNs, (b) HO-1 inhibition exacerbated the toxicity of glutamate, and (c) the levels of reactive oxygen species were lower in the Tg CGNs, subsequent to glutamate exposure, than in Ntg neurons. The blunted induction of HO-1 mRNA in Tg CGNs is likely an indication of a lower level of stress caused by glutamatemediated generated reactive oxygen species, which in turn could reflect the enhanced antioxidant potential of Tg CGNs. The fact that the effect of HO-1 inhibition on the survival of cells exposed to glutamate was examined using a specific inhibitor of HO-1, the D2 peptide, gives further credence to the above suggestion. D2 peptide is derived from a certain region of the HLA class I heavy chain (Iyer et al., 1997). The mechanism by which the peptide inhibits HO-1 activity is not known at this time. However, because the heme-binding pocket of HO isozymes is conserved (Rotenberg and Maines, 1991), the specificity of the D2 peptide for HO-1 suggests its interaction with other regions of HO-1 protein. A possible mechanism by which the interaction inhibits HO-1 activity may involve change in the protein folding. Metalloprotoporphyrins were not used in this study because those compounds not only are nonspecific for HO isozymes but can also affect HO-1 mRNA levels along with the integrity of HO-2 (Maines and Trakshel, 1992 ; Mark and Maines, 1992) as well as interact with hemebinding proteins including NO synthase and guanylate cyclase (Maines, 1997). Moreover, metalloporphyrin inhibitors can activate molecular oxygen (Vreman et al., 1993).

The present study does not define the mechanism by which HO-1 overexpression provides neuroprotection against glutamate-induced oxidative cell death. The mechanisms, however, could involve any of the following functions of HO-1 : the chaperonin activity of the protein, which could be comprised by a change in folding of the protein ; an in-place and preexisting increased capacity to sequester iron released in the course of heme degradation ; the activities of the products of tetrapyrrole cleavage (CO and biliverdin) ; and/or the activity of the enzyme to degrade the heme moiety of cytochrome c (Kutty and Maines, 1982), hence inactivating the cytochrome/cytochrome c activity that is linked to cell death. The increased capacity to sequester iron relates to the reported induction of the iron storage protein, ferritin, by iron released by HO activity (Vile and Tyrrell, 1993). If not sequestered, free iron, liberated by HO activity, can act as a potent catalyst for lipid peroxidation and formation of oxygen radicals. Furthermore, additional functions in cellular defense mechanisms have been ascribed to the pyrrolic product of HO activity in connection with modulation of signal transduction pathways (Hansen et al., 1996). Bilirubin has inhibitory effects on protein phosphorylation and inhibits protein kinase C, protein kinase A, and cyclic AMP-dependent protein kinases (Kwak et al., 1991 ; Sano et al., 1985). Activation of protein kinases regulates the cascade of enzymes involved in programmed cell death (Herdegen et al., 1998). In this context, it is noteworthy that, as was shown before (Panahian et al., 1999), when this strain of HO-1 Tg mice is subjected to ischemic stroke, the nuclear localization of P53 is blocked. Those animals also show an increase in neuronal bcl2 expression. Nuclear localization of P53 is essential for its mediation of downstream effects leading to cell death (Wu and Momand, 1998), and bcl2 is generally considered to be cytoprotective. Regardless of which of the potential activities of HO-1 are involved in protection against oxidative stress-mediated cell death, data suggest that suppression of oxygen radical formation and the lowered resting levels of [Ca2+]i in the Tg mouse neurons are of relevance to the defense. The significance of the lower basal levels of [Ca2+]i in the Tg neurons than in the Ntg cells may lie in the possibility that the Tg neurons are less excitable at the “resting” condition. The mechanism(s) by which the HO-1 increase relates to a lowered calcium concentration in the cell is not evident at this time.

Based on the collective findings of the present study, we conclude that a moderate increase in HO-1, which is likely to be attainable by approaches such as gene therapy, is cytoprotective against oxidative stress-mediated cell death. The effect of the robust expression of HO-1 on neuronal survival from death remains to be elucidated.


  1. Top of page
  2. Abstract
  5. Comparison of HO-1 expression and response to glutamate of Ntg and Tg mouse CGNs
  7. Acknowledgements

This study was supported by NIEHS grant ES04391. We are grateful to Dr. Michael Fowler for assistance in providing transgenic newborn mice, Xiaojun Wang for assistance with northern blot analysis, and Suzanne Bono for manuscript preparation. We thank Drs. Robert S. Freeman and Nariman Panahian for helpful discussion and comments and Sangstat Medical Corporation for the generous gift of peptides.

  • 1
    Abraham N.G., Lavrovsky Y., Schwartzman M.L., Stoltz R.A., Levere R.D., Gerritsen M.E., Shibahara S., Kappas A. (1995) Transfection of the human heme oxygenase gene into rabbit coronary microvessel endothelial cells : protective effect against heme and hemoglobin toxicity.Proc. Natl. Acad. Sci. USA 92,67986802.
  • 2
    Ankarcrona M., Dypbukt J.M., Bonfoco E., Zhivotovsky B., Orrenius S., Lipton S.A., Nicotera P. (1995) Glutamate-induced neuronal death : a succession of necrosis or apoptosis depending on mitochondrial function.Neuron 15,961973.
    Balla J., Jacob H.S., Balla G., Nath K., and Vercellotti G.M. (1992) Endothelial cell heme oxygenase and ferritin induction by heme proteins : a possible mechanism limiting shock damage.Trans Assoc. Am. Phys. 105,16.
  • 4
    Cruse I. & Maines M.D. (1988) Evidence suggesting that the two forms of heme oxygenase are products of different genes.J. Biol. Chem. 263,33483353.
  • 5
    Da Silva J.L., Morishita T., Escalante B., Staudinger R., Drummond G., Goligarsky M.S., Lutton J.D., Abraham N.G. (1996) Dual role of heme oxygenase in epithelial cell injury : contrasting effects of short-term and long-term exposure to oxidant stress.J. Lab. Clin. Med. 128,290296.
  • 6
    Dennery P.A., McDonagh A.F., Spitz D.R., Rodgers P.A., Stevenson D.K. (1995) Hyperbilirubinemia results in reduced oxidative injury in neonatal Gunn rats exposed to hyperoxia.Free Rad. Biol. Med. 19,395404.
  • 7
    Denninger J.W. & Marletta M.A. (1999) Guanylate cyclase and the NO/cGMP signaling pathway.Biochim. Biophys. Acta 1411,334350.
  • 8
    Du Y., Bales K.R., Dodel R.C., Hamilton-Byrd E., Horn J.W., Czilli D.L., Simmons L.K., Ni B., Paul S.M. (1997) Activation of a caspase 3-related cysteine protease is required for glutamatemediated apoptosis of cultured cerebellar granule neurons.Proc. Natl. Acad. Sci. USA 94,1165711662.
  • 9
    Dwyer B.E., Lu S., Laitinen J.T., Nishimura R.N. (1998) Protective properties of tin- and manganese-centered porphyrins against hydrogen peroxide-mediated injury in rat astroglial cells.J. Neurochem. 71,24972504.
  • 10
    Ewing J.F. & Maines M.D. (1991) Rapid induction of heme oxygenase-1 mRNA and protein by hyperthermia in rat brain : heme oxygenase-2 is not a heat shock protein.Proc. Natl. Acad. Sci. USA 88,53645368.
  • 11
    Favaron M., Manev H., Alho H., Bertolino M., Ferret B., Guidotti A., Costa E. (1988) Gangliosides prevent glutamate and kainate neurotoxicity in primary neuronal cultures of neonatal rat cerebellum and cortex.Proc. Natl. Acad. Sci. USA 85,73517355.
  • 12
    Gunter T.E., Zuscik M.J., Puzas J.E., Gunter K.K., Rosier R.N. (1990) Cytosolic free calcium concentration in avian growth plate chondrocytes.Cell Calcium 11,445457.
  • 13
    Hansen T.W., Mathiesen S.B., Walaas S.I. (1996) Bilirubin has widespread inhibitory effects on protein phosphorylation.Pediatr. Res. 39,10721077.
  • 14
    Herdegen T., Claret F.X., Kallunki T., Martin-Villalba A., Winter C., Hunter T., Karin M. (1998) Lasting N-terminal phosphorylation of c-Jun and activation of c-Jun N-terminal kinases after neuronal injury.J. Neurosci. 18,51245135.
  • 15
    Iyer S., Woo J., Cornejo M.C., Gao L., McCoubrey W., Maines M., Buelow R. (1997) Characterization and biological significance of immunosuppressive peptide D2702.75-84(E [RIGHTWARDS ARROW] V) binding protein. Isolation of heme oxygenase-1.J. Biol. Chem. 273,26922697.
  • 16
    Juin P., Pelletier M., Oliver L., Tremblais K., Gregoire M., Meflah K., Vallette F.M. (1998) Induction of a caspase-3-like activity by calcium in normal cytosolic extracts triggers nuclear apoptosis in a cell-free system.J. Biol. Chem. 273,1755917564.
  • 17
    Kane D.J., Sarafian T.A., Anton R., Hahn H., Gralla E.B., Valentine J.S., Ord T., Bredesen D.E. (1993) Bcl-2 inhibition of neural death : decreased generation of reactive oxygen species.Science 262,12741277.
  • 18
    Keller J.N. & Mattson M.P. (1998) Roles of lipid peroxidation in modulation of cellular signaling pathways, cell dysfunction, and death in the nervous system.Rev. Neurosci. 9,105116.
  • 19
    Keyse S.M. & Tyrell R.M. (1989) 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. USA 86,99103.
  • 20
    Kutty R.K. & Maines M.D. (1982) Oxidation of heme c derivatives by purified heme oxygenase : evidence for the presence of one molecular species of heme oxygenase in rat liver.J. Biol. Chem. 257,99449952.
  • 21
    Kwak J.Y., Takeshiege K., Cheung B.S., Minakami S. (1991) Bilirubin inhibits the activation of superoxide-producing NADPH oxidase in a neutrophil cell-free system.Biochim. Biophys. Acta 1076,369373.
  • 22
    Lafon-Cazal M., Pietri S., Culcasi M., Bockaert J. (1993) NMDA-dependent superoxide production and neurotoxicity.Nature 364,535537.
  • 23
    Maines M.D. (1992) Heme Oxygenase : Clinical Applications and Functions. CRC Press, Boca Raton, Florida.
  • 24
    Maines M.D. (1997) The heme oxygenase system : a regulator of second messenger gases.Annu. Rev. Pharmacol. Toxicol. 37,517554.
  • 25
    Maines M.D. & Trakshel G.M. (1992) Differential regulation of heme oxygenase isozymes by Sn- and Zn-protoporphyrins : possible relevance to suppression of hyperbilirubinemia.Biochim. Biophys. Acta 1131,166174.
  • 26
    Maines M.D., Trakshel G.M., Kutty R.K. (1986) Characterization of two constitutive forms of rat liver microsomal heme oxygenase ; only one molecular form is inducible.J. Biol. Chem. 261,411419.
  • 27
    Maines M.D., Polevoda B., Coban T., Johnson K., Stoliar S., Huang T.J., Panahian N., Cory-Slechta D.A., McCoubrey W.K.J r. (1998) Neuronal overexpression of heme oxygenase-1 correlates with an attenuated exploratory behavior and causes an increase in neuronal NADPH diaphorase staining.J. Neurochem. 70,20572069.
  • 28
    Maines M.D., Raju V.S., Panahian M. (1999) Spin trap (N-t-butyl-α-phenylnitrone)-mediated suprainduction of heme oxygenase-1 in kidney ischemia/reperfusion model : the role of the oxygenase in protection against oxidative injury. J. Pharmacol. Exp. Ther. 291,911919.
  • 29
    Mark J.A. & Maines M.D. (1992) Tin-protoporphyrin-mediated disruption in vivo of heme oxygenase-2 protein integrity and activity in rat brain.Pediatr. Res. 32,324329.
  • 30
    Marks G.S., Brien J.F., Nakatsu K., McLaughlin B.E. (1991) Does carbon monoxide have a physiological function ?Trends Pharmacol. Sci. 12,185188.
  • 31
    Matsuoka Y., Kitamura Y., Okazaki M., Kakimura J., Tooyama I., Kimura H., Taniguchi T. (1998) Kainic acid induction of heme oxygenase in vivo and in vitro.Neuroscience 85,12231233.
  • 32
    Matsuoka Y., Kitamura Y., Kakimura J., Taniguchi T. (1999) Expression of heme oxygenase-1 mediated by non-NMDA and metabotropic receptors in glial cells : possible involvement of reactive oxygen species production and protein kinase C activation.Neuropharmacology 38,825834.
  • 33
    Mattson M.P. (1996) Calcium and free radicals : mediators of neurotrophic factor and excitatory transmitter-regulated developmental plasticity and cell death.Perspect. Dev. Neurobiol. 3,7991.
  • 34
    McCoubrey W.J., Huang T.J., Maines M.D. (1997) Isolation and characterization of a cDNA from the rat brain that encodes hemoprotein heme oxygenase-3.Eur. J. Biochem. 247,725732.
  • 35
    McDonagh A.F. (1990) Is bilirubin good for you ?Clin. Perinatol. 17,359369.
  • 36
    Meldrum B. & Garthwaite J. (1990) Excitatory amino acid neurotoxicity and neurodegenerative disease.Trends Pharmacol. Sci. 11,379387.
  • 37
    Miller T.M., Moulder K.I., Knudson C.M., Creedon D.J., Deshmukh M., Korsmeyer S.J., Johnson E.M.J r. (1997) Bax deletion further orders the cell death pathway in cerebellar granule cells and suggests a caspase-independent pathway to cell death.J. Cell Biol. 139,205217.
  • 38
    Mosmann T. (1983) Rapid colorimetric assay for cellular growth and survival : application to proliferation and cytotoxicity assays.J. Immunol. Methods 65,5563.
  • 39
    Musser D.A. & Oseroff A.R. (1994) The use of tetrazolium salts to determine sites of damage to the mitochondrial electron transport chain in intact cells following in vitro photodynamic therapy with Photofrin II.Photochem. Photobiol. 59,621626.
  • 40
    Nicotera P., Ankarcrona M., Bonfoco E., Orrenius S., Lipton S.A. (1997) Neuronal necrosis and apoptosis : two distinct events induced by exposure to glutamate or oxidative stress.Adv. Neurol. 72,95101.
  • 41
    Otterbein L.E., Kolls J.K., Mantell L.L., Cook J.L., Alam J., Choi A.M. (1999) Exogenous administration of heme oxygenase-1 by gene transfer provides protection against hyperoxiainduced lung injury.J. Clin. Invest. 103,10471054.
  • 42
    Panahian N., Yoshiura M., Maines M.D. (1999) Overexpression of heme oxygenase-1 is neuroprotective in a model of permanent middle cerebral artery occlusion in transgenic mice.J. Neurochem. 72,11871203.
  • 43
    Platt J.L. & Nath K.A. (1998) Heme oxygenase : protective gene or Trojan horse.Nat. Med. 4,13641366.
  • 44
    Raps S.P., Lai J.C., Hertz L., Cooper A.J. (1989) Glutathione is present in high concentrations in cultured astrocytes but not in cultured neurons.Brain Res. 493,398401.
  • 45
    Rosenkranz A.R., Schmaldienst S., Stuhlmeier K.M., Chen W., Knapp W., Zlabinger G.J. (1992) A microplate assay for the detection of oxidative products using 2′,7′-dichlorofluorescin-diacetate. J. Immunol. Methods 156,3945.
  • 46
    Rotenberg M.O. & Maines M.D. (1991) Characterization of a cDNA encoding rabbit brain heme oxygenase-2 and identification of a conserved domain among mammalian heme oxygenase isozymes : possible heme-binding site ?Arch. Biochem. Biophys. 290,336344.
  • 47
    Sano K., Nakamura H., Matsuo T. (1985) Mode of inhibitory action of bilirubin on protein kinase C.Pediatr. Res. 19,587590.
  • 48
    Schramm M., Eimerl S., Costa E. (1990) Serum and depolarizing agents cause acute neurotoxicity in cultured cerebellar granule cells : role of the glutamate receptor responsive to N-methyl-D-aspartate. Proc. Natl. Acad. Sci. USA 87,11931197.
  • 49
    Shibahara S., Muller R., Taguchi H., Yoshida T. (1985) Cloning and expression of cDNA for rat heme oxygenase.Proc. Natl. Acad. Sci. USA 82,78657869.
  • 50
    Simonian N.A. & Coyle J.T. (1996) Oxidative stress in neurodegenerative diseases.Annu. Rev. Pharmacol. Toxicol. 36,83106.
  • 51
    Smith M.A., Wehr K., Harris P.L.R., Siedlak S.L., Connor J.R., Perry G. (1998) Abnormal localization of iron regulatory protein in Alzheimer's disease.Brain Res. 788,232236.
  • 52
    Snyder S.H., Jaffrey S.R., Zakhary R. (1998) Nitric oxide and carbon monoxide : parallel roles as neural messengers.Brain Res. Brain Res. Rev. 26,167175.
  • 53
    Stout A.K., Raphael H.M., Kanterewicz B.I., Klann E., and Reynolds I.J. (1998) Glutamate-induced neuron death requires mitochondrial calcium uptake.Nat. Neurosci. 1,366373.
  • 54
    Tan S., Wood M., Maher P. (1998) Oxidative stress induces a form of programmed cell death with characteristics of both apoptosis and necrosis in neuronal cells.J. Neurochem. 71,95105.
  • 55
    Vile G.F. & Tyrrell R.M. (1993) Oxidative stress resulting from ultraviolet A irradiation of human skin fibroblasts leads to a heme oxygenase-dependent increase in ferritin.J. Chem. Biol. 268,1467814681.
  • 56
    Vreman H.J., Ekstrand B.C., Stevenson D.K. (1993) Selection of metalloporphyrin heme oxygenase inhibitors based on potency and photoreactivity.Pediatr. Res. 33,195200.
  • 57
    Wolin M.S., Davidson C.A., Kaminski P.M., Fayngersh R.P., Mohazzab-H K.M. (1998) Oxidant—nitric oxide signalling mechanisms in vascular tissue. Biochemistry 63,810816.
  • 58
    Wu H. & Momand J. (1998) Pyrrolidine dithiocarbamate prevents P53 activation and promotes P53 cysteine residue oxidation.J. Biol. Chem. 273,1889818905.
  • 59
    Ye J. & Laychock S.G. (1998) A protective role for heme oxygenase expression in pancreatic islets exposed to interleukin-1beta.Endocrinology 139,41554163.