Address correspondence and reprint requests to Prof. D. A. Butterfield at Department of Chemistry and Center of Membrane Sciences, University of Kentucky, Lexington, KY 40506-0055, U.S.A.
Abstract : Peroxynitrite (ONOO-) is a highly reactive, oxidizing anion with a half-life of <1 s that is formed by reaction of superoxide radical anion with nitric oxide. Several reports of ONOO- -induced oxidation of lipids, proteins, DNA, sulfhydryls, and inactivation of key enzymes have appeared. ONOO- has also been implicated as playing a role in the pathology of several neurodegenerative disorders, such as Alzheimer's disease (AD) and amyotrophic lateral sclerosis, among others. Continuing our laboratory's interest in free radical oxidative stress in brain cells in AD, the present study was designed to investigate the damage to brain neocortical synaptosomal membrane proteins and the oxidation-sensitive enzyme glutamine synthetase (GS) caused by exposure to ONOO-. These synaptosomal proteins and GS have previously been shown by us and others to have been oxidatively damaged in AD brain and also following treatment of synaptosomes with amyloid β-peptide. The results of the current study showed that exposure to physiological levels of ONOO- induced significant protein conformational changes, demonstrated using electron paramagnetic resonance in conjunction with a protein-specific spin label, and caused oxidation of proteins, measured by the increase in protein carbonyls. ONOO- also caused inactivation of GS and led to neuronal cell death examined in a hippocampal cell culture system. All these detrimental effects of ONOO- were successfully attenuated by the thiol-containing antioxidant tripeptide glutathione. This research shows that ONOO- can oxidatively modify both membranous and cytosolic proteins, affecting both their physical and chemical nature. These findings are discussed with reference to the potential involvement of ONOO- in AD neurodegeneration.
Nitric oxide (NO) is synthesized from l-arginine (Palmer et al., 1987) by the action of NO synthase (NOS), which is present as two different isoforms—the calcium-independent inducible (iNOS) isoform and the calcium-dependent constitutively expressed isoform. The latter is synthesized in either the endothelium or the neurons (nNOS). NO is quite stable and benign for a free radical, with a lifetime of several seconds, but when produced in an oxidizing environment in the vicinity of superoxide radical anion, NO can react with superoxide, at diffusion controlled rates, to produce an extremely strong oxidant, peroxynitrite (ONOO-) (Beckman, 1996). ONOO- is extremely unstable with a half-life of <1 s and in its protonated form, peroxynitrous acid (ONOOH), is extremely reactive (Pryor and Squadrito, 1995). ONOOH has been regarded as the active intermediate, with hydroxyl radical-like reactivity (Koppenol et al., 1992), responsible for ONOO- -mediated oxidative events. There are also reports claiming that ONOOH undergoes homolytic cleavage to form hydroxyl radicals and that the deleterious effects of ONOO- are actually due to the formation of ·OH (Beckman, 1996) : NO + O2·-→ ONOO- + H+→ ONOOH → HO· + ·NO2.
ONOO- has been known to cause oxidative damage to lipids, DNA, carbohydrates, and proteins, particularly the amino acids cysteine, methionine, phenylalanine, and tyrosine (reviewed by Beckman, 1996, and references cited therein). ONOO- also can affect cellular energy status by inactivating key mitochondrial enzymes (Radi et al., 1994) and can trigger intracellular calcium release from the mitochondria (Packer and Murphy, 1994). The signature of ONOO- involvement is the presence of nitrated tyrosines in proteins, demonstrated in atherosclerosis (White et al., 1994), amyotrophic lateral sclerosis (ALS) (Beckman, 1996), ischemia/reperfusion injury, and other neurodegenerative disorders (reviewed by Halliwell, 1997). Only ONOO- and possibly a few other reactive nitrogen species (RNS), but not NO, can nitrate tyrosines (Halliwell, 1997) at the orthoposition, thereby preventing its phosphorylation, a step that can render proteins dysfunctional (Ischiropoulos et al., 1992) and can kill neurons in cell culture (Lafon-Cazal et al., 1993).
The aim of this study was to examine the consequences of exposing neocortical synaptosomal membrane proteins, known to be oxidized in senile plaquerich regions of AD brain (Hensley et al., 1995), to various levels of ONOO-. The membrane proteins were examined for conformational changes after ONOO- treatment, using electron paramagnetic resonance (EPR) in association with a protein-specific spin label, and for oxidatively mediated structural modifications, by measuring the amount of carbonyls on the proteins. The effects of ONOO- on the activity of the enzyme glutamine synthetase (GS), which is extremely susceptible to oxidation and known to have decreased activity in AD brain (Hensley et al., 1995), and on hippocampal neuronal survival also were studied. Thiol groups have been suggested as efficient scavengers of ONOO- (Pryor and Squadrito, 1995) ; consequently, we reasoned that if this were so, then protection by glutathione, a thiol-containing tripeptide, against the damage caused by ONOO- to the membrane and cytosolic proteins should be observed.
MATERIALS AND METHODS
Sucrose for synaptosome isolation was obtained from Sigma Chemical Co. Protease inhibitors and chelating agents used were from ICN Biochemicals. All other chemicals were of the highest quality from either Sigma or Aldrich Chemical Co.
Synthesis of ONOO-
ONOO- was synthesized using sodium azide and ozone as described earlier (Pryor et al., 1995). In brief, ozone from a Welsbach Ozonator was bubbled through 100 ml of a 0.1 M solution of sodium azide. Sodium azide was earlier dissolved in water, which was adjusted to pH 12 using 1 M NaOH, and remained cooled on ice throughout the reaction. The progress of the reaction and the formation of ONOO- were monitored by drawing aliquots of the solution at 5-min intervals and reading the UV absorbance of ONOO- at 302 nm (ε = 1,670 M-1 cm-1). The peak concentration of ONOO- obtained was 42 mM after 100 min of reaction time. The solution was then divided into 5-ml aliquots and stored in glass vials at -80°C. Before each experiment the solution was thawed on ice, and the absorbance was measured at 302 nm to determine the concentration of ONOO-· ONOO- was found to be quite stable, and its concentration decreased from 40 to 30 mM only after 4 months of storage under these conditions. To confirm further the existence of ONOO- in the solution, we replicated the EPR spectrum of the nitroxide, previously obtained by Denicola et al. (1995), after addition of 1 mM ONOO- to 0.5 mM desferrioxamine in 0.1 M phosphate-buffered saline (data not shown).
All protocols followed were approved by the University of Kentucky Animal Care and Use Committee. For this study male Mongolian gerbils, 3-5 months old, obtained from Tumblebrook farms (West Brookfield, MA, U.S.A.), kept under 12-h light/dark conditions in the University of Kentucky Central Animal Facility, and fed standard Purina rodent laboratory chow, were used. The animals were decapitated during the light phase, and the brain was quickly dissected on ice according to the method previously described (Hensley et al., 1994). The neocortex was isolated and immediately suspended in 20 ml of isolation buffer (0.32 M sucrose with the protease inhibitors, 4 μg/ml leupeptin, 4 μg/ml pepstatin, 5 μg/ml aprotinin, and 20 μg/ml trypsin inhibitor, 0.2 mM phenylmethylsulfonylfluoride, 2 mM EDTA, 2 mM EGTA, and 20 mM HEPES and homogenized by 12 passes of a motor-driven Teflon pestle.
The procedure for synaptosome isolation has been described earlier (Hensley et al., 1994). In brief, the homogenized neocortices were centrifuged at 3,500 rpm for 10 min at 4°C in a DuPont Sorvall RC5C refrigerated centrifuge. The pellet obtained was discarded, and the supernatant was spun at 13,500 rpm in a similar fashion. The resulting pellet was then resuspended in a sucrose isolation buffer and layered onto discontinuous sucrose density gradients. These were spun at 22,000 rpm for 2 h at 4°C in an SW28 rotor in a Beckman L7-55 refrigerated ultracentrifuge. Purified synaptosomes were obtained at the 1.18/1.10 M sucrose gradient interface.
Synaptosomes obtained were then washed three times with ~30 ml of lysing buffer (10 mM HEPES, 2 mM EDTA, and 2 mM EGTA in distilled water, pH 7.4) containing 100 μM diethylenetriaminepentaacetic acid, a chelator for iron. After the three washes the synaptosome membranes were resuspended in ~1 ml of lysing buffer, and the protein concentration of each homogenate was measured by the method of Lowry et al. (1951).
Spin labeling and EPR spectrometry
Protein concentration in the purified synaptosomes was adjusted to 4 mg/ml. The protein aliquots were then centrifuged in a refrigerated Eppendorf table-top microcentrifuge at 14,000 rpm for 4 min. The protein pellets obtained were then either treated with ONOO- or preincubated with glutathione and then treated with ONOO-, in lysing buffer, depending on the studies described later. After the various treatments the homogenates were spin-labeled with the protein-specific spin label 2,2,6,6-tetramethyl-4-maleimidopiperidine-l-oxyl (MAL-6), and the EPR spectra were obtained by methods described previously (Hensley et al., 1994).
For this study ONOO- (200 μM) was added to the synaptosomal protein pellet (4 mg/ml) and vortex-mixed, and 1 ml of lysing buffer was added to it. The membranes with ONOO- were then incubated for 10 min, 30 min, and 1 h at room temperature. At the end of the respective incubation times the homogenate was spun at 14,000 rpm at 4°C for 4 min, and the spin label MAL-6 solution was added to the pellet obtained after the spin. The pellets were then kept in the refrigerator, and the EPR spectra of these samples were obtained after 18 h, after washing away the excess spin label.
For all initial studies performed, the synaptosomal control samples were also incubated with the same volume of NaOH solution, adjusted to the same pH as ONOO-. This was to rule out that the ONOO--induced changes were artifactual, resulting from the NaOH used to dissolve the ONOO-.
ONOO- solutions (50 μM-1 mM) were added to synaptosomal protein pellets, vortex-mixed, and incubated at room temperature for 10 min in 1 ml of lysing buffer. At the end of the incubations the samples were again spun as described above and labeled with MAL-6, and the EPR spectra were later obtained.
Glutathione protection studies
To determine the optimal time required for glutathione action, synaptosomal membranes were treated with glutathione (500 μM) 30 min before addition of ONOO-, immediately after addition of ONOO-, or 30 min after addition of ONOO-. The concentration of ONOO- used was 250 μM.
For ascertaining the amount of glutathione necessary for complete protection against ONOO--induced membrane alterations, 125, 250, and 500 μM glutathione was added to the synaptosomes 30 min before addition of ONOO- (250 μM). Glutathione used for this study was purchased from Sigma, and aliquots of a 50 mM stock solution in distilled water were used to obtain the required concentration.
Protein carbonyl assay
Protein concentration of synaptosomes, isolated and purified as described above, was adjusted to 4 mg/ml. The samples were then subjected to the various doses of ONOO- (100, 250, and 500 μM) and glutathione (500 μM, 30 min before ONOO- additions), as described above, and the protein carbonyl content was determined spectrophotometrically by measuring the absorbance of the resulting 2,4-dinitrophenylhydrazone at 390 nm (Oliver et al., 1987).
The GS assay was performed according to procedures described earlier (Rowe et al., 1970) and later modified (Miller et al., 1978). A dose-response study was carried out to determine the percentage of inactivation of GS with increasing concentrations of ONOO-. A 10-min incubation with ONOO-, as performed for the EPR spin labeling studies, did not give as reproducible and significant data as expected ; therefore, GS was incubated for 30 min with 100 μM, 250 μM, 500 μM, 1 mM, and 5 mM ONOO- in 1 ml of HEPES buffer. Again, GS was incubated with only NaOH to make sure that the decrease in activities observed was not due to the base. For studying the extent of protection offered by glutathione, GS was preincubated with glutathione (1 mM) for 30 min and then with ONOO- (500 μM) for another 30 min in HEPES buffer before performing the assay.
Cell survival assay
Hippocampal neuronal cultures were obtained from 18-day-old Sprague-Dawley rat fetuses. ONOO- (250 μM) and glutathione (500 μM) were added to the cells at the same time and incubated with the cells for 18 h. Neuronal toxicity was then evaluated by the trypan blue exclusion assay.
The data were analyzed for statistical significance using Dunnett's test for one-way ANOVA followed by Student's t test. A value of p < 0.05 was considered to be statistically significant for comparison between data sets.
EPR is an extremely sensitive method for studying paramagnetic species, especially in opaque samples in which optical methods are subject to light scattering artifacts. Proteins are generally nonparamagnetic species, and hence the protein-specific spin label MAL-6 is used to study the protein microenvironment. MAL-6 covalently labels sulfhydryl groups on proteins located on the protein surface or in deep pockets. The former group of SH sites leads to spin label motion that is weakly immobilized, whereas reaction of MAL-6 with the latter group of SH sites leads to spin label motion that is strongly hindered (Butterfield, 1982). Accordingly, the EPR spectrum of MAL-6-labeled proteins in synaptosomal membranes has a weakly (W) and strongly (S) immobilized component. The ratio of the signal amplitudes of these two components of the MI = +1 low-field region of the EPR spectrum, the W/S ratio (Fig. 1), is extremely sensitive to alterations in the protein environment (Butterfield, 1982 ; Soszynski and Bartosz, 1997). A decrease in the W/S ratio results from increased protein-protein interactions, increased protein cross-linking, and changes in protein conformation (Butterfield, 1982). Previous models of oxidative stress, such as hydroxyl radical generation (Hensley et al., 1994), hyperoxia (Howard et al., 1996), ischemia/reperfusion (Hall et al., 1995), accelerated aging (Butterfield et al., 1997), and Aβ (Subramaniam et al., 1998), have all shown decreased W/S ratios of spinlabeled synaptosomes with increasing oxidative stress.
To determine the optimal time for incubation of ONOO- with synaptosomal membranes, a time-response study was performed. Using EPR and the protein-specific spin label MAL-6, a decrease in W/S ratios was observed after 10 min, 30 min, and 1 h of incubation with 200 μM ONOO-, and this decrease was the same for all the time points studied (data not shown). Hence, 10 min was chosen as the incubation interval for all the studies performed involving W/S ratio measurements.
The dose-response study showed an overall significant (p < 0.001) lowering in W/S ratios (71-41% of controls) with ONOO- concentrations ranging from 50 μM to 1 mM (Fig. 2A). Increasing concentrations of ONOO- caused greater lowering of W/S ratios for 0-250 μM, beyond which no further decrease was noted (Fig. 2A). Addition of the same volume of NaOH alone did not lead to any significant lowering of W/S ratios, when compared with the control. To demonstrate that the protein alteration caused by ONOO- is not unique to synaptosomal membranes, we exposed human erythrocyte membranes to the same treatments and observed similar results ; that is, a decrease in W/S ratios with increasing concentrations of ONOO- was found (data not shown). These latter data replicate the EPR results obtained earlier by Soszynski and Bartosz (1996), also in erythrocyte membranes.
Glutathione was shown to protect best against ONOO--mediated protein conformational changes when preincubated with the membrane for 30 min before ONOO- addition, compared with simultaneous or post-additions (data not shown). Increasing doses of glutathione protected membrane proteins to greater extents from the ONOO--induced protein alterations. Synaptosomes pretreated with 500 μM glutathione before treatment with 250 μM ONOO- had W/S ratio values similar to those of control samples (p < 0.5 ; Fig. 2B).
Protein carbonyls are a key marker of protein oxidation (Oliver et al., 1987 ; Butterfield et al., 1997). In this study, protein carbonyl level measurements provided evidence of oxidative stress, confirming the EPR finding for ONOO--treated samples. There was nearly a twofold increase in the amount of carbonyls measured on the proteins after treatment with 250 (188% of control, p < 0.003) and 500 μM (217% of control, p < 0.007) ONOO- (Fig. 3A). Even 100 μM ONOO- led to a > 50% increase in content of protein carbonyls (Fig. 3A). A 30-min preincubation of the membrane with glutathione (500 μM), before addition of 250 μM peroxynitrite, dramatically reduced the amount of carbonyls formed (116% of controls, p < 0.5 ; Fig. 3B), similar to the protective effect observed by EPR.
The activity of GS was monitored as a function of the concentration of ONOO- added. Increased concentrations of ONOO- showed greater inhibition of GS (overall significance, p < 0.01) : 250 μM was the minimal concentration required to inhibit significantly the activity of the enzyme (92% of control, p < 0.002), and a concentration of 5 mM ONOO- lowered the GS activity to 47% of the untreated enzyme activity (Fig. 4A). Preincubation of the GS with 1 mM glutathione for 30 min before the 30-min addition of 500 μM ONOO- prevented the significant loss of GS activity (102% of control, p < 0.4 ; Fig. 4B).
The obvious pathological change in neurodegenerative diseases is the death of neurons. To determine if ONOO- treatment would lead to neuronal death, cell viability was tested using the trypan blue exclusion assay. ONOO- (250 μM), with or without 500 μM glutathione, was added to hippocampal neuronal cell cultures and left overnight. After 18 h the cell viability had dropped to 37% of controls for cells treated with ONOO- alone (p < 0.007 ; Fig. 5). Neuronal cultures treated with both glutathione and ONOO- had ~70% cell survival (p < 0.03 ; Fig. 5), suggesting that glutathione partially protected cells against ONOO- -induced toxicity, consistent with the results above showing that glutathione protected against oxidative stress in brain membranes and GS. Treatment of cells with glutathione alone did not decrease cell survival significantly (90% cell survival, p < 0.1).
In this study ONOO- caused a decrease in the W/S ratios of MAL-6-labeled synaptosomes and increased levels of protein carbonyls. Based on previous studies with membranes subjected to oxidative stress (Hensley et al., 1994 ; Hall et al., 1995 ; Howard et al., 1996 ; Butterfield et al., 1997 ; Soszynski and Bartosz, 1997 ; Subramaniam et al., 1998), these results are consistent with ONOO- causing oxidative stress-induced protein conformational and structural changes in cortical synaptosomal membranes. We previously reported decreased W/S ratios of synaptosomal membranes, increased levels of protein carbonyls, and decreased GS activity from rapid autopsy of AD brain samples compared with those of controls (Hensley et al., 1995) and also in rodent cortical synaptosomes treated with Aβ (25-35) (Subramaniam et al., 1998). Because treatment with ONOO- led to similar observations, the possibility exists that ONOO- may be another factor responsible for the elevated oxidative stress evidenced in AD. Calculations performed considering the half-life of 1 s for ONOO- show that a bolus of 250 μM ONOO-, the concentration used in much of this study, is equivalent to a 7-min exposure to a steady-state concentration of 1 μM, which is physiologically relevant under inflammatory conditions (Beckman, 1996).
The inactivation of GS by ONOO- observed here and in other systems (Berlett et al., 1996 ; Hensley et al., 1997) may be the result of oxidation or nitration of the enzyme because GS activity, such as in Escherichia coli, is mediated by the cyclic attachment and detachment of the AMP moiety of ATP to the hydroxyl group of a unique tyrosine in each of the 12 subunits of the enzyme (Berlett et al., 1996). Treatment of GS with ONOO- has been shown to lead to nitration of tyrosine and oxidation of methionine residues (Berlett et al., 1996). This tyrosine nitration is irreversible, mimics the adenylation of the enzyme, and prevents tyrosine phosphorylation by protein kinases. Thus, these results are consonant with the notion that ONOO- can alter signal transduction and regulation of key enzymes.
Glutathione, known to protect cells against apoptosis (Walker et al., 1995 ; Jurma et al., 1997), protected hippocampal neurons studied in the current investigation, a finding consistent with the suggested involvement of ONOO- -mediated cellular events in neuronal apoptosis (Estevez et al., 1995 ; Nicotera et al., 1995 ; Keller et al., 1998). Glutathione is a tripeptide containing a SH group and, along with glutathione peroxidase, is involved in the primary defense mechanism for the removal of peroxides from the brain. Some reports suggest that the reaction of ONOO- with thiols generate thiyl radicals, shown by EPR (Berlett et al., 1996), whereas others dispute this suggestion by indicating a radical mechanism leading to the formation of nitrosothiols. The latter functionalities have been found in the cardiovascular and pulmonary systems (Stamler et al., 1992). Irrespective of the mechanism, glutathione oxidation is an important detoxification process in the tissue. GSH can protect the tissue by removing H2O2, thus preventing hydroxyl radical formation, by scavenging ONOO- and other RNS formed and by recycling antioxidants like vitamin E. Studies have shown cellular total glutathione levels to be a good indication of cell sensitivity to NO-mediated cytotoxicity (Walker et al., 1995). The current study solidifies the notion that glutathione is an efficient scavenger of ONOO- by protecting against ONOO- -induced synaptosomal membrane protein conformational changes, by inhibiting protein oxidation, by preventing inactivation of GS, and by increasing cell survival.
Oxidative stress is implicated as playing a major role in AD, and there is much evidence that can directly link ONOO- to some of the damage observed in AD, such as the presence of nitrotyrosines (Good et al., 1996 ; Smith et al., 1997). ONOO- has also been shown to cause lipid oxidation and DNA oxidation (reviewed by Beckman, 1996, and references cited therein), both of which have been reported to occur in AD (Hajimohammadreza and Brammer, 1990 ; Subbarao et al., 1990 ; Mecocci et al., 1994 ; Markesbery, 1997). Other indirect evidences for ONOO- involvement in the pathogenesis of AD include stimulation of iNOS by Aβ (1-40), which then leads to ONOO- formation ; Aβ (1-40) and interferon-γ together triggering the production of RNS from microglia (Meda et al., 1995) ; and apolipoprotein E increasing NO production by modulating the activity of NOS (Vitek et al., 1997). Also, S100β an astrocyte-derived neural protein, overexpressed in AD and known to cause neurodegeneration, can activate iNOS in the presence of astrocytes to increase NO formation, and this effect is enhanced in the presence of Aβ (Hu et al., 1997).
ONOO- formation conceivably may be a link between oxidative stress and excitotoxicity, because NO is formed in the neurons only when nNOS binds to calmodulin, following Ca2+ influx into the cytosol, after glutamate-based NMDA receptor stimulation (Garthwaite and Boulton, 1995 ; Good et al., 1996). NMDA receptors are abundant in the hippocampus (Good et al., 1996), one of the regions most severely affected in AD, GS activity is decreased in AD brain (Smith et al., 1991 ; Hensley et al., 1995), and glutamate uptake is shown to be inhibited by Aβ (Harris et al., 1996), potentially making more glutamate available for NMDA receptor stimulation. Therefore, NMDA receptor excitoxicity and ONOO- -induced oxidative stress together may be involved in a whole cascade of events leading to neurodegeneration. Ca2+ -stimulated ATPase and other ATPases, like Na+/K+ - and Mg2+ -ATPases, show diminished activity after treatment of membranes with ONOO-, leading to impaired calcium homeostasis and further ROS production (Denicola et al., 1995 ; Soszynski and Bartosz, 1996). Similarly, decreased glutathione levels are also correlated with increased intracellular calcium levels (Jurma et al., 1997). Decreased ion-motive ATPase activities in AD following treatment of synaptosomal membranes with Aβ have been reported (Mark et al., 1995), and increased calcium levels can lead to further production of ONOO-, thus exacerbating the damage.
Luminol-dependent chemiluminescence studies showed ONOO- to be involved with increased ROS production (Radi et al., 1993 ; Denicola et al., 1995). The ROS formed then can augment ONOO- production, as ROS from glycated tau activate NFκB (Hensley et al., 1997), a transcription factor for iNOS. This finding is in agreement with NFκB being activated under conditions of increased oxyradical or decreased antioxidant levels. One of the reasons for increased ROS production, especially superoxide, is that ONOO- inactivates both Mn and Fe superoxide dismutases, superoxide scavenging enzymes (Ischiropoulos et al., 1992), resulting in less scavenging of superoxide by superoxide dismutase and increased ONOO- production (reviewed by Beckman, 1996). Supporting this notion are experiments with mice overexpressing Cu/Zn superoxide dismutase known to be protected against toxicity induced by 3-nitropropionic acid, an NO generator, and neural cell lines overexpressing Mn superoxide dismutase showing decreased ONOO- formation and lower apoptosis, following exposure to iron, Aβ, and NO-generating agents (Keller et al., 1998).
Thus, there is increasing evidence suggesting that ONOO- serves as an important link in the chain of events leading to oxidative damage observed in AD, which involves both ROS and RNS. ONOO- is detrimental because of its fast reactivity and phospholipid bilayer permeability (Marla et al., 1997), which may explain the widespread damage observed away from its site of production. In the current study physiological levels of glutathione almost completely abrogated ONOO- damage to membrane and cytosolic proteins and neurons in cell culture. Hence, compounds that elevate the level of or mimic the action of glutathione may be potential candidates as ROS and RNS scavengers and therefore potentially useful in therapeutic approaches to AD (Schultz et al., 1995 ; Van Dyke, 1997).