Oxidative metabolites are involved in polyamine-induced microglial cell death

Neuroscience 134 (2005) 1123–1131

OXIDATIVE METABOLITES ARE INVOLVED IN POLYAMINE-INDUCED MICROGLIAL CELL DEATH K. TAKANO,a M. OGURA,a Y. YONEDAa AND Y. NAKAMURAb*

karyotes and eukaryotes (Tabor and Tabor, 1984). In the CNS, where the concentration of SPD and SPM can reach the high micromolar range (Shaw and Pateman, 1973), polyamines are important for the regulation of cell proliferation and differentiation, can serve a protective function (Gilad and Gilad, 1992), and can be toxic (Virgili et al., 1992; Doyle and Shaw, 1994; Sparapani et al., 1997; Segal et al., 1999). Polyamines modulate ligand-receptor interaction and function. For example, polyamines potentiate the effects of N-methyl-D-aspartate (NMDA) receptor activation (Yoneda and Ogita, 1991) and block the channel of the ␣-amino-3hydro-5-methyl-4-isoxazolepropionate (AMPA) receptor (Washburn and Dingledine, 1996). Polyamine toxicity in the CNS has been extensively studied, particularly with regard to its role in NMDA receptor-mediate excitotoxicity. Moreover, polyamines were released by stimulation of the NMDA receptor and a high concentration of potassium (Fage et al., 1992; Harman and Shaw, 1981). Polyamine concentrations increase in the brain of neurodegenerative disorders, such as Alzheimer’s disease and ischemia (Morrison and Kish, 1995; Paschen et al., 1987). Polyamines are metabolized by polyamine oxidases (PAO) after their acetylation and produce aldehydes and reactive oxygen species, such as hydrogen peroxide, that possibly damage proteins, DNA, and lipids (Rao et al., 2000; Segal and Skolnick, 2000). Thus, it is generally considered that increase in extracellular polyamine concentration aggravates neuronal damage. Pathological activation of microglia is not only the result of neuronal damage but also the cause of it in certain CNS diseases including ischemic brain damage and Alzheimer’s diseases (Morioka et al., 1991; McGeer and McGeer, 1995). Microglial activation reflected by cell proliferation was observed shortly after ischemic insult and preceded histologically-detectable neuronal damage in the CA1 area of the hippocampus (Morioka et al., 1991). In Alzheimer’s disease, ␤-amyloid protein, an extracellular deposit in senile plaque, was found to induce neuronal damage by activating microglia (Meda et al., 1995). Activated microglia have been proposed to release deleterious substances to aggravate the neuronal viability under many pathological conditions. Previously, we observed that low concentrations of SPD and SPM induced microglial cell death via apoptotic pathway in the presence of fetal bovine serum (FBS), which is generally used for primary cell culture (Takano et al., 2003). It has been shown that polyamines are toxic in the presence of ruminant serum, but molecular mechanisms underlying this toxicity are still unclear. On the other

a Laboratory of Molecular Pharmacology, Kanazawa University Graduate School of Natural Science and Technology, Kakuma-machi, Kanazawa, 920-1192, Japan b Laboratory of Integrative Physiology in Veterinary Sciences, Osaka Prefecture University, 1-1, Gakuen-cho, Sakai, Osaka 599-8531, Japan

Abstract—Pathological activation of microglia, which reside quiescently in physiological CNS, is associated with various neurodegenerative diseases. Endogenous polyamines, spermidine and spermine, are known to be activators of cell proliferation and differentiation. We previously reported that both spermidine and spermine induce dose-dependent cell death in cultured rat brain microglia at a submicromolar concentration range via apoptotic process, whereas cultured astrocytes were less sensitive to these polyamines [Neuroscience 120 (2003) 961]. These polyamine effects were observed only in the presence of fetal bovine serum. In the present study we examined further the mechanism of polyamine-induced microglial cell death. Amine oxidase in fetal bovine serum produces hydrogen peroxide and an aminoaldehyde from spermine, and the latter generates acrolein spontaneously. Acrolein was found to be much more toxic to microglia than to astrocytes and the effective concentration of acrolein was similar to that of spermine, whereas hydrogen peroxide was marginally toxic. Aminoguanidine, an inhibitor of amine oxidase, blocked the toxic effects of spermine on microglia. Spermine cytotoxicity was also prevented by antioxidant reagents; glutathione (reduced form), cysteine, and N-acetylcysteine. These results suggest that polyamine-induced apoptotic cell death of microglia is triggered by an oxidative stress with acrolein, which is produced by amine oxidase from polyamine. The different toxicities of polyamine between two glial cells may regulate the balance of glial activation in some pathological conditions of CNS. © 2005 Published by Elsevier Ltd on behalf of IBRO. Key words: microglia, astrocyte, polyamines, spermidine, spermine, NO production.

Endogenous polyamines such as putrescine (PUT), spermidine (SPD), and spermine (SPM) are required for cell growth, proliferation, and differentiation in both pro*Corresponding author. Tel: ⫹81-72-254-9477; fax: ⫹81-72-254-9478. E-mail address: [email protected] (Y. Nakamura). Abbreviations: cDNA, complementary DNA; Cys, cysteine; DAN, 2,3diaminonaphthalene; DMEM, Dulbecco’s modified Eagle’s medium; FBS, fetal bovine serum; GAPDH, glyceraldehydes 3-phosphate dehydrogenase; GFAP, glial fibrillary acidic protein; LPS, lipopolysaccharide; NAC, N-acetylcysteine; NMDA, N-methyl-D-aspartate; NO, nitric oxide; N-2-MPG, N-2-mercaptopropionyl glycine; PAO, polyamine oxidase; PUT, putrescine; RT-PCR, reverse transcription polymerase chain reaction; SPD, spermidine; SPM, spermine; SSA, sodium sulfosalicylate dihydrate. 0306-4522/05$30.00⫹0.00 © 2005 Published by Elsevier Ltd on behalf of IBRO. doi:10.1016/j.neuroscience.2005.05.014

1123

1124

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

hand, SPM at high concentration protects thymocytes from apoptosis (Brune et al., 1991), while its oxidation leads to necrosis of leukemia cells (Bonneau and Poulin, 2000). Therefore, further study is needed to clarify the molecular mechanisms responsible for polyamine toxicity and to investigate cell type specificity. In the present study, the effects of SPD and SPM in the presence of FBS and that of polyamine metabolites, hydrogen peroxide and acrolein, were examined on primary cultures of rat brain microglia and astrocytes.

EXPERIMENTAL PROCEDURES Materials Lipopolysaccharide (LPS) from Salmonella enteritidis, FBS, PUT, SPD, SPM, DNase I (DN-25) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Dulbecco’s modified Eagle’s medium (DMEM) was obtained from Gibco BRL (Grand Island, NY, USA). Sodium sulfosalicylate dihydrate (SSA), glutathione (reduced form), cysteine (Cys), N-acetylcysteine (NAC) were purchased from Wako (Osaka, Japan). 2,3-Diaminonaphthalene (DAN) was obtained from Dojindo (Kumamoto, Japan). Aminoguanidine and acrolein were obtained from Tokyo Kasei (Tokyo, Japan).

Preparation of microglial culture This study was carried out with an effort to minimize the number of animals used and their suffering in compliance with the Guide for Animal Experimentation at Kanazawa University enacted according to the Japanese governmental recommendation. Microglia were isolated from primary cultures of rat brains by the method previously described (Si et al., 1996) with slight modifications. Three whole brains of neonatal rats (1–2 days old) were used for each flask (SoLo flask 185 cm2; Nunc Inc., Naperville, IL, USA). The tissues were briefly triturated with a Pasteur pipette with 3 ml DMEM then 5 ml of 0.25% trypsin (Difco, Sparks, MD, USA) in Ca2⫹, Mg2⫹-free phosphate-buffered saline containing 5.5 mM glucose was added. Trypsinization for 15 min at 37 °C was stopped by addition of 5 ml horse serum supplemented with 0.1 mg/ml of DNase I and the suspension was centrifuged at 350⫻g for 5 min. The precipitates were triturated completely using a blue-tip-mount pipette with 20 ml DMEM containing 10% FBS, 100 ␮g/ml streptomycin, and 50 unit/ml penicillin and were plated in a flask described above. These mother cultures were maintained in a 5% CO2 humidified incubator at 37 °C for 2–3 weeks with changing medium every week. Microglia growing on the top of the confluent cell monolayer were detached by shaking at 100 r.p.m. for 1 h (Double Shaker NR-3, Titec, Tokyo, Japan) in a 37 °C incubator. Floating cells in the supernatant were collected, centrifuged, and disseminated onto 96-well plates (MS-8096R; for suspension cell culture, Sumitomo, Tokyo, Japan). After 45 min, the medium was changed and the cells were washed twice with DMEM containing 10% FBS to remove non-adherent cells. The remaining microglia (ca. 4⫻104/well) were allowed to stabilize in DMEM containing 10% FBS for 1 day; then, used for experiments. In our culture, more than 90% of the cells were positive to a microglial marker, lectin from Bandeiraea simplicifolia bs-I isolectin b4 (peroxidase-labeled, Sigma).

Preparation of astrocytes culture Astrocytes were prepared as described previously (Si et al., 1997; Murakami et al., 2003). In brief, brain cortices from 20-day-old embryos, which were taken out from pregnant Wistar rats deeply anesthetized with diethyl ether, were cleared of meninges, cut into

about 1-mm3 blocks, and treated with 0.25% trypsin in Ca2⫹, Mg2⫹-free phosphate-buffered saline containing 5.5 mM glucose for 20 min at 37 °C with gentle shaking. An equal volume of horse serum supplemented with 0.1 mg/ml of DNase I was added to the medium to inactivate the trypsin. Then, the tissues were centrifuged at 350⫻g for 5 min. The tissue sediments were triturated through a Pasteur pipette with DMEM containing 10% FBS, 100 ␮g/ml streptomycin, and 50 unit/ml penicillin. After filtering cell suspensions through a lens-cleaning paper (Fuji Photo Co., Tokyo, Japan), the cells were plated on polyethyleneiminecoated 100-mm-diameter plastic dishes (Nunc) at a density of 0.8 –1.3⫻105 cells/cm2. Cultures were maintained in a humidified atmosphere of 5% CO2 and 95% air at 37 °C with changing medium every 3 days. After 1 week, astrocytes were replated to remove neurons. On days 12–14, they were replated onto 96-well plates (MS-8096F; for tissue culture, Sumitomo) using an ordinary trypsin-treatment technique at a density of 4⫻104 cells/well and stabilized for 1 day, then we used for experiments. Approximately 90% of the cells were immunoreactively positive to an astrocyte marker, glial fibrillary acidic protein (GFAP); using anti-GFAP antibody (Sigma) and FITC-conjugated anti-rabbit IgG antibody. Less than 10% of the cells were positive to a microglial marker, lectin from Bandeiraea simplicifolia bs-I isolectin b4 (peroxidaselabeled, Sigma).

Nitrite assay Nitric oxide (NO) production by microglia was determined by the assay of nitrite, a relatively stable metabolite of NO. The nitrite concentration was assayed by a fluorometric method using DAN as previously described (Si et al., 1997). The supernatants (100 ␮l) of the cultured microglia were collected after 24 h stimulation with 100 ng/ml LPS, and mixed with 20 ␮l of freshly prepared DAN solution (0.05 mg/ml in 0.62 M HCl) at room temperature. After 10 –20 min, 100 ␮l of 0.28 M NaOH was added and formation of 2,3-diaminonaphthotriazole, the fluorescent product, was measured using a fluorescent microplate reader (MTP-100F, Corona Electric Company, Naka-Hitachi, Japan) with excitation at 365 nm and emission at 450 nm. A standard curve of NaNO2 was established in an identical fashion in each assay.

Cell viability To evaluate cell viability, we assayed total mitochondrial activity with Cell Counting Kit-8 (Dojindo). After the cells were incubated with polyamines usually for 24 h, one-tenth volume of the reagent was added and incubated for 1–3 h, then the color development was measured at 450 nm (reference wavelength at 650 nm).

Quantification of glutathione content We assayed cellular glutathione content using Total Glutathione Quantification Kit (Dojindo). After the cells were incubated for 24 h, the cells were collected and triturated with GPBS containing 0.5% SSA. Then, the supernatants of centrifugation at 20,000⫻g for 10 min were collected as samples. Coenzyme working solution (70 ␮l; NADPH in phosphate buffer) was mixed with 10 ␮l of enzyme working solution (glutathione reductase in phosphate buffer), and incubated at 30 °C for 5min. Added 10 ␮l of samples and incubated at 30 °C for 10 min, 10 ␮l of substrate working solution (5,5=-dithiobis(2-nitrobenzoic acid) (DTNB) in phosphate buffer) was added and incubated for 10 –30 min at room temperature, then the absorbance was measured at 405 nm. A standard curve of glutathione was established in an identical fashion in each assay. Protein concentrations were determined using Bradford dye-binding assay (Bio-Rad, Tokyo, Japan), according to the manufacturer’s protocol, with bovine serum albumin as the standard.

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

Reverse transcription polymerase chain reaction (RT-PCR) Cultured microglia were washed with phosphate-buffered saline, followed by extraction of mRNA using mRNA purification kit (Quickprep Micro mRNA Purification Kit, Amersham Pharmacia Biotech, Buckinghamshire, UK) and subsequent synthesis of complementary DNA (cDNA) with 12.5 ␮M random hexamer primers, mixture of dNTP (deoxy nucleotide triphosphate), 10 mM dithiothreitol, RNase inhibitor, First-Strand Buffer and M-MLV Reverse Transcriptase (Invitrogen, Tokyo, Japan). Reverse transcriptase reaction was run at 37 °C for 50 min and an aliquot of synthesized cDNA was directly used for PCR. PCR was performed in buffer containing 10 mM Tris–HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 200 ␮M each of dNTP, 20 pmol of each primer for the corresponding Cystine/Glutamate antiporter (Xc⫺) subunits; xCT and 4F2hc and 2 U TaqDNA polymerase (Takara, Tokyo, Japan) as described previously (Hinoi and Yoneda, 2001). PCR was performed with primers specific for the xCT, 4F2hc and glyceraldehydes 3-phosphate dehydrogenase (GAPDH). Rat xCT have not been cloned yet, so we used the specific primer for mouse xCT. The reaction for GAPDH was carried out for 28 cycles and the reactions for xCT and 4F2hc were carried out for 35 cycles. The conditions of each PCR cycles for these primers were as follows: denaturation at 95 °C for 1 min; annealing at 60 °C for 1 min; and extension at 72 °C for 1 min. Electrophoresis was run for an aliquot of PCR amplification products on a 2% agarose gel, followed by detection of DNA with ethidium bromide. The results showed xCT/GAPDH and 4F2hc/GAPDH ratio of the density of detection bands using Scion image (Scion Corporation, MD, USA). Specific primers; xCT: sense 5=-CCTGGCATTTGGACGCTACAT-3= antisense 5=-TCAGAATTGCTGTGAGCTTGC-3= 4F2hc: sense 5=-CTCCCAGGAAGATTTTAAAGACCTTCT-3= antisense 5=-TTCATTTTGGTGGCTACAATGTCAG-3= GAPDH: sense 5=-GGTGAAGGTCGGTGTCAACGGATT-3= antisense 5=-GATGCCAAAGTTGTCATGGATGACC-3=

Data analysis Data values varied from one cell preparation to the other, although trends were entirely consistent. Therefore, we presented a representative result of several times experiments that were done independently and described the numbers of replicate experiments using different cell preparations. For statistical analysis of the data, one-way ANOVA followed by Scheffe’s multiple comparison procedure was used. Differences between treatments were considered statistically significant when P⬍0.05.

RESULTS Effects of polyamines on cultured microglia and astrocytes To examine the effect of polyamines on microglial activation, we used primary cultured microglia prepared from the whole brains of newborn rats. Cultured microglia produced NO by the stimulation with LPS. Previously we observed that in the presence of FBS LPS-induced NO production in cultured microglia was markedly inhibited by SPD and SPM; half effective concentrations (EC50) of SPD and SPM were about 3 and 1 ␮M, respectively (Takano et al., 2003). Cell viability assessed by total mitochondrial activity decreased by the incubation with SPD and SPM for 24 h at similar concentration ranges. The microglial cell death induced by low concentrations of polyamines was that via

1125

apoptotic pathway (Takano et al., 2003). To compare the effects of polyamines on microglia with other cells, we used cultured astrocytes prepared from cortex of 20-days rats embryo. Cultured astrocytes also produced NO by the stimulation with LPS. Astrocytes were stimulated with 100 ng/ml LPS in the presence of 1–10 ␮M SPD and SPM for 24 h. In contrast to microglia, 10 ␮M SPD and SPM caused no significant difference in cell viability of cultured astrocytes (data not shown). LPS-induced NO production in astrocytes was slightly decreased by 10 ␮M SPM (data not shown). Polyamine cytotoxicity in the presence of FBS As mentioned above, in the presence of 10% FBS, LPSinduced NO production in cultured microglia was markedly inhibited by 1–10 ␮M SPD and SPM in a dose-dependent manner and the cell viability assessed by total mitochondrial activity also decreased. We examined the effects of SPM on cultured microglia in the presence of various concentrations of FBS (Fig. 1). The effects of 10 ␮M SPM were observed only when FBS was present. Inhibition of LPS-induced NO production was observed with an increase in FBS concentration, whereas FBS concentration did not affect LPS-induced NO production in the absence of SPM (Fig. 1a). Total mitochondrial activity decreased with an increased in FBS concentration (Fig. 1b). SPM did not show any cytotoxic effect in the absence of FBS at all. The cytotoxic effect of SPM on cultured microglia caused in a dose-dependent manner of more than 0.1% FBS. Inhibition of polyamine cytotoxicity by aminoguanidine and SH compounds To assess whether polyamine-induced microglial cell death depends on the polyamine metabolites, we examined the effect of an inhibitor of amine oxidase, aminoguanidine. Microglial cells were incubated with 10 ␮M SPM in the presence of various concentrations of aminoguanidine, for 24 h and the cell viability was evaluated by the measurement of total mitochondrial activity. Aminoguanidine blocked the effects of SPM in a dose-dependent manner. One millimolar aminoguanidine completely inhibited 10 ␮M SPM-induced microglial cell death (Fig. 2a). The effect of the catalase decomposing hydrogen peroxide was also examined; however, 100 U/ml catalase could not rescue the microglial cell death induced by 10 ␮M SPM (data not shown). The effects of sulfhydryl compounds with ⫺SH groups on primary carbon atoms were also examined; glutathione (reduced form), Cys, and NAC. Microglial cells were incubated with 10 ␮M SPM in the presence of 1 mM glutathione, Cys, and NAC for 24 h and then the cell viability was evaluated. All these antioxidants protected cultured microglia completely against 10 ␮M SPM-induced apoptotic cell death (Fig. 2b). When microglial cells were treated with 100 ␮M SPM in the presence of 1 mM glutathione (reduced form), Cys, and NAC, the cell viability recovered to 33.7⫾1.5, 20.3⫾1.6, and 39.0⫾1.8% of control (without 100 ␮M SPM), respectively (data not shown).

1126

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

Fig. 1. Effect of FCS on the SPM-induced microglial death. Microglia were incubated with 10 ␮M SPM together with 100 ng/ml LPS for 24 h under various concentrations of FCS. The concentration of nitrite in the medium was measured by fluorescent assay with DAN reagent (a). Cell viability was evaluated by the measurement of total mitochondrial activity with Cell Counting Kit (b). Data are mean⫾S.E. of three samples. This result is representative of three replicate experiments. * P⬍0.05, ** P⬍0.01, significantly different from control.

Effects of polyamine metabolites on cultured microglia and astrocytes We examined the effect of polyamine metabolites, hydrogen peroxide and acrolein, on the cell viability of cultured microglia and astrocytes. The incubation with 10 –100 ␮M of hydrogen peroxide for 24 h inhibited total mitochondrial activity of both microglia (Fig. 3a) and astrocytes (Fig. 3b) in a dose-dependent manner. However, 100 ␮M hydrogen peroxide was not sufficient for the half inhibition of the total mitochondrial activity in both cell cultures. In the microglial culture, the half effective concentration of hydrogen peroxide was more than 100 times higher than that of SPM (about 1 ␮M). Acrolein was much more toxic. The incubation with 0 –10 ␮M acrolein for 24 h inhibited the total mitochondrial activity of microglia in a dose-dependent manner; only 20% activity was remained after incubation with 10 ␮M acrolein (Fig. 4a); the half effective concentration was about 0.3 ␮M. On the other hand, no significant effect on astrocytes was observed by 10 ␮M acrolein (Fig. 4b). Acrolein decreased cell viability of microglia at similar concentration ranges of SPM. Glutathione content and Cys/glutamate antiporter expression SPM-induced microglial cell death was protected by glutathione (reduced form) that is an important endogenous intracellular antioxidant, so we examined the glutathione content of cultured microglia incubated with or without 10 ␮M SPM for 24 h. The method used was for the cellular contents of both reduced and oxidized form of glutathione. Glutathione content of cultured microglia was 45⫾2.9 nmol/mg cell protein, and it increased more than twice to

105⫾13 nmol/mg cell protein by the incubation with 10 ␮M SPM for 24 h. It is known that Cys is the limiting precursor for glutathione synthesis, and that intracellular Cys is converted from cystine. A major pathway of cystine incorporation is mediated by cystine/glutamate antiporter (Xc⫺), which consists of heterodimmer of 4F2hc and xCT. Incubation with 10 ␮M SPM increased glutathione content of microglia, so we assessed the effects of SPM on Xc⫺ expression by semi-quantitative RT-PCR of 4F2hc and xCT in cultured microglia. Cultured microglia expressed both 4F2hc and xCT mRNA (Fig. 5a). The expression level of xCT mRNA enhanced markedly after the incubation with 1–10 ␮M SPM for 24 h (Fig. 5b). On the contrary, no significant difference was detected in the expression of 4F2hc mRNA (Fig. 5c) between before and after the incubation with SPM.

DISCUSSION We previously reported that the addition of very low concentrations (less than 10 ␮M) of polyamines in the medium induced cell death of cultured microglia via apoptotic pathway in the presence of FBS (Takano et al., 2003). Microglia, which reside in the CNS quiescently under physiological conditions, should be immediately activated on the demand of pathological condition and they phagocytose unnecessary debris. When such scavenging function is completed, microglial activation should be terminated without delay. Disturbance of the balance between microglial activation and its termination may cause pathological activation of microglia, consequently leading to progressive neurodegenerative disease. There have been published many reports on the mechanism underlying such activation; however, the sig-

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

1127

Fig. 2. Protection of SPM-induced microglial death by aminoguanidine and antioxidants. Microglia were incubated with various concentration of aminoguanidine (a), 1 mM glutathione (reduced form; GSH), Cys, and NAC (b) together with or without 10 ␮M SPM for 24 h. Then, cell viability was evaluated. Data are mean⫾S.E. of six samples. This result is representative of three replicate experiments. * P⬍0.05, ** P⬍0.01, significantly different from control.

naling mechanism for the termination of microglial activation has not been well elucidated (Nakamura, 2002). Ap-

optotic cell death induced by the endogenous polyamine at very low concentrations might be one such mechanism.

Fig. 3. Effects of hydrogen peroxide on the viability of cultured microglia and astrocytes. Microglia (a) and astrocytes (b) were incubated with 10 –100 ␮M hydrogen peroxide for 24 h. Then, cell viability was evaluated. Data are mean⫾S.E. of four samples. This result is representative of three replicate experiments. * P⬍0.05, significantly different from control.

1128

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

Fig. 4. Effects of acrolein on the viability of cultured microglia and astrocytes. Microglia (a) and astrocytes (b) were incubated with various concentrations of acrolein for 24 h. Then, cell viability was evaluated. Data are mean⫾S.E. of four samples. This result is representative of three replicate experiments. ** P⬍0.01, significantly different from control.

The low concentration of polyamine induced cell death of cultured microglia in the medium containing 10% FBS generally used for cell culture, but not in the FBS-free medium. Some serum factors are suggested to be involved in the mechanism of this cell death. There are several reports that aldehydes and hydrogen peroxide generated from polyamines (SPD and SPM) by serum amine oxidase cause cytotoxicity (Higgins et al., 1969; Gaugas and Dewey, 1979; Averill-Bates et al., 1993; Bonneau and Poulin, 2000). Among such substances, acrolein is the focus of attention as the most cytotoxic polyamine metabolite (Sharmin et al., 2001).

Fig. 6 shows the major pathway of polyamine metabolism in which two kinds of oxidases are involved. SPM and SPD are direct substrates of serum amine oxidase (SAO) whose reaction products are their aldehydes, ammonia, and hydrogen peroxide (Tabor et al., 1964). The aldehydes are spontaneously decomposed to acrolein (Kimes and Morris, 1971). On the other hand, PAO is considered to prefer N-acetylpolyamines to the plain polyamines; thus, it catalyzes N-acetylpolyamines that have been acetylated by SPD/SPM N-acetyltransferase (SSAT). In this pathway of PAO (designated as PAOa in Fig. 6), the reaction products are hydrogen peroxide and 3-acetoami-

Fig. 5. SPM effect on glutathione content (a), and xCT and 4F2hc mRNA expression by semiquantitative RT-PCR (b, c) in cultured microglia. Microglia were incubated with various concentration of SPM for 24 h. Glutathione content was measured with Total Glutathione Quantification Kit. Data are mean⫾S.E. of three samples. ** P⬍0.01, significantly different from control (a). The expression of mRNAs of xCT (b), 4F2hc (c) and GAPDH was detected by RT-PCR procedure. A typical semiquantitative analysis of RT-PCR for xCT, 4F2hc and GAPDH mRNA was shown in the photograph. The graph shows xCT and 4F2hc/GAPDH ratio of the density of detection bands. Data are mean⫾S.E. of three samples.

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

1129

Fig. 6. Major pathways of SPD and SPM metabolism catalyzed by serum amine oxidase and PAO. See text for details.

dopropanal; the latter is spontaneously decomposed to acrolein and acetamide. Ivanova et al. (1998) reported that cerebral ischemia in animal model of permanent middle cerebral artery occlusion increased activity of PAO. They proposed a possibility that PAO catalyzes directly polyamines to produce 3-aminopropanal (designated as PAOb in Fig. 6), and 3-aminopropanal mediated progressive neuronal necrosis and glial apoptosis. In the present study, aminoguanidine, an inhibitor of amine oxidase, suppressed the SPM-induced microglial cell death completely in a dose-dependent manner. Furthermore, among polyamine metabolites we tested, acrolein was the most toxic; the concentration of acrolein that induced microglial cell death was as low as that of SPM. Therefore, aldehydes such as acrolein rather than hydrogen peroxide should play a more important role in polyamine-induced microglial cell death. On the other hand, hydrogen peroxide inhibited cell viability of both cultured astrocytes and microglia in a similar concentration range, which was much higher than that of acrolein for microglia. Although there are various metabolic products of polyamines by several oxidases (Seiler, 2004), it is suggested that oxidative metabolites of polyamines, such as acrolein rather than hydrogen peroxide, caused microglial cell death induced by polyamines. Ivanova et al. (2002) also reported that sulfhydryl compounds with ⫺SH groups on primary carbon atoms, such as glutathione, Cys, NAC, N-2-mercaptopropionyl glycine

(N-2-MPG), and 2-mercaptoethylamine, protected against cell death and that N-2-MPG attenuated neuronal cell death in the brain ischemia model. Consistently, in our microglial culture, the toxicity of SPM was not only prevented by aminoguanidine but also attenuated by different antioxidants; including glutathione (reduced form), Cys, and NAC. Antioxidant such as glutathione inhibited polyamine cytotoxicity and oxidative metabolites suggested to cause microglial death induced by SPD and SPM, therefore we examined the effects of SPM on glutathione content as an intracellular antioxidant capacity. Incubation with 10 ␮M SPM increased more than twice of the total glutathione content of cultured microglia. Hirrlinger et al. (2000) reported a prominent glutathione system with high activity of glutathione reductase and glutathione peroxidase in microglia-rich cultures from rat brain, which is likely to reflect the necessity for self-protection against reactive oxygen species when being produced by themselves and/or surrounding brain cells. In our study, total glutathione content increased likely due to respond against oxidative stress induced by incubation with SPM. Cystine/glutamate antiporter (Xc⫺) consists of heterodimmer of 4F2hc and xCT and exhibited the Na⫹independent transport of L-cystine and L-glutamate. The expression of xCT was rapidly induced in U87 cells upon an oxidative stress, which was accompanied by the increase in the L-cystine uptake by U87 cells (Kim et al.,

1130

K. Takano et al. / Neuroscience 134 (2005) 1123–1131

2001). It has also been shown that xCT is upregulated on the reduced glutathione depletion by an oxidative stress in the mouse peritoneal macrophages (Sato et al., 1999; Watanabe and Bannai, 1987; Ishii et al., 1991). In brain, it has been reported that system Xc⫺ is upregulated in glial cells upon the oxidative stress and plays an essential roles to protect neurons against oxidative stress (Cho and Bannai, 1990; Kato et al., 1992, 1993). In the present study, the level of xCT mRNA increased markedly in cultured microglia by the stimulation with 10 ␮M SPM for 24 h but no increase in 4F2hc mRNA was detected. These results are consistent with previous reports mentioned above. In contrast to the microglial culture, astrocytes were not sensitive to polyamines. The LPS-induced NO production and total mitochondrial activity of astrocytes were not significantly inhibited by 10 ␮M SPD and SPM regardless of the presence of FBS. Hydrogen peroxide marginally inhibited cell viabilities of both cultured astrocytes and microglia in a dose-dependent manner; the half effective concentrations were more than 100 ␮M. However, acrolein and polyamines were much more toxic to microglia but not to astrocytes. We observed the different sensitivities to polyamines and acrolein between cultured microglia and astrocytes. Hollensworth et al. (2000) also reported the differential susceptibility of central glial cell types to menadione-induced oxidative stress and apoptosis appeared to correlate with increased oxidative mitochondrial DNA damage. They observed the lower susceptibility to oxidative damage and apoptosis in astrocytes than in microglia and oligodendrocytes; however, astrocytes had lower total glutathione content and superoxide dismutase activity than oligodendrocytes and microglia under their culture conditions. They suggest that the differential susceptibility to oxidative damage does not appear related to cellular antioxidant capacity. In our preliminary experiment, on the other hand, astrocytes had higher total glutathione content than microglia under our culture conditions. The extent of the oxidative damage of a certain cell is the consequence of complicated pathways of oxidative stress and anti-oxidative process and they are highly dependent on the kinds of the oxidative stress and the cell conditions.

REFERENCES Averill-Bates DA, Agostinelli E, Przybytkowski E, Mateescu MA, Mondovi B (1993) Cytotoxicity and kinetic analysis of purified bovine serum amine oxidase in the presence of spermine in Chinese hamster ovary cells. Arch Biochem Biophys 300:75–79. Bonneau M-J, Poulin R (2000) Spermine oxidation leads to necrosis with plasma membrane phosphatidylserine redistribution in mouse leukemia cells. Exp Cell Res 259:23–34. Brune B, Hartzell P, Nicotera P, Orrenius S (1991) Spermine prevents endonuclease activation and apoptosis in thymocytes. Exp Cell Res 195:323–329. Cho Y, Bannai S (1990) Uptake of glutamate and cystine in C6 glioma cells and in cultured astrocytes. J Neurochem 55:2091–2097. Doyle KM, Shaw GG (1994) The mechanism of the neurotoxic effects of spermidine (Abstr. ). Biochem Soc Trans 22(Suppl.):386S. Fage D, Voltz C, Scatton B, Carter C (1992) Selective release of spermine and spermidine from the rat striatum by N-methyl-Daspartate receptor activation in vivo. J Neurochem 58:2170 –2175.

Gaugas JM, Dewey DL (1979) Evidence for serum binding of oxidized spermine and its potent G1-phase inhibition of cell proliferation. Br J Cancer 39:548 –557. Gilad GM, Gilad VH (1992) Polyamines in neurotrauma. Ubiquitous molecules in search of a function. Biochem Pharmacol 44: 401– 407. Harman RJ, Shaw GG (1981) The spontaneous and evoked release of spermine from rat brain in vitro. Br J Pharmacol 73:165–174. Higgins ML, Tillman MC, Rupp JP, Leach FR (1969) The effect of polyamines on cell culture cells. J Cell Physiol 74:149 –154. Hinoi E, Yoneda Y (2001) Expression of GluR6/7 subunits of kainate receptors in rat adenohypophysis. Neurochem Int 38:539 –547. Hirrlinger J, Gutterer JM, Kussmaul L, Hamprecht B, Dringen R (2000) Microglial cells in culture express a prominent glutathione system for the defense against reactive oxygen species. Dev Neurosci 22:384 –392. Hollensworth SB, Shen C, Sim JE, Spitz DR, Wilson GL, LeDoux SP (2000) Glial cell type-specific responses to menadione-induced oxidative stress. Free Radic Biol Med 28(8):1161–1174. Ishii T, Nakayama K, Sato H, Miura K, Yamada M, Yamada K, Sugita Y, Bannai S (1991) Expression of the mouse macrophage cystine transporter in Xenopus laevis oocytes. Arch Biochem Biophys 289:71–75. Ivanova S, Botchkina GI, Al-Abed Y, Meistrell M, Batliwalla F, Dubinsky JM, Iadecola C, Wang H, Gregersen PK, Eaton JW, Tracey KJ (1998) Cerebral ischemia enhances polyamine oxidation: Identification of enzymatically formed 3-aminopropanal as an endogenous mediator of neuronal and glial cell death. J Exp Med 188:327–340. Ivanova S, Batliwalla F, Mocco J, Kiss S, Huang J, Mack W, Coon A, Eaton JW, Al-Abed Y, Gregersen PK, Shohami E, Connolly ES Jr, Tracey KJ (2002) Neuroprotection in cerebral ischemia by neutralization of 3-aminopropanal. Proc Natl Acad Sci U S A 99(8): 5579 –5584. Kato S, Ishita S, Sugiwara K, Mawatari K (1993) Cystine/glutamate antiporter expression in retinal Muller glial cells: implications for dl-alpha-aminoadipate toxicity. Neuroscience 57:473– 482. Kato S, Negishi K, Mawatari K, Kuo CH (1992) A mechanism for glutamate toxicity in the C6 glioma cells involving inhibition of cystine uptake leading to glutathione depletion. Neuroscience 48:903–914. Kimes BW, Morris DR (1971) Preparation and stability of oxidized polyamines. Biochim Biophys Acta 228:223–234. Kim JY, Kanai Y, Chairoungdua A, Cha SH, Matsuo H, Kim DK, Inatomi J, Sawa H, Ida Y, Endou H (2001) Human cystine/glutamate transporter: cDNA cloning and upregulation by oxidative stress in glioma cells. Biochim Biophys Acta 1512:335–344. McGeer PL, McGeer EG (1995) The inflammatory response system of brain: implications for therapy of Alzheimer and other neurodegenerative diseases. Brain Res Brain Res Rev 21:195–218. Meda L, Cassatella MA, Szendrei GI, Otvos L Jr, Baron P, Villalba M, Ferrari D, Rossi F (1995) Activation of microglial cells by ␤-amyloid protein and interferon-␥. Nature 374:647– 650. Morioka T, Kalehua AN, Streit WJ (1991) The microglial reaction in the rat dorsal hippocampus following transient forebrain ischemia. J Cereb Blood Flow Metab 11:966 –973. Morrison LD, Kish SJ (1995) Brain polyamine levels are altered in Alzheimer’s disease. Neurosci Lett 197:5– 8. Murakami K, Nakamura Y, Yoneda Y (2003) Potentiation by ATP of lipopolysaccharide-stimulated nitric oxide production in cultured astrocytes. Neuroscience 117:37– 42. Nakamura Y (2002) Regulating factors for microglial activation. Biol Pharm Bull 25(8):945–953. Paschen W, Schmidt-Kastner R, Djuricic B, Meese C, Linn F, Hossmann KA (1987) Polyamine changes in reversible cerebral ischemia. J Neurochem 49:35–37.

K. Takano et al. / Neuroscience 134 (2005) 1123–1131 Rao AM, Hatcher JF, Dogan A, Dempsey RJ (2000) Elevated N1acetylspermidine levels in gerbil and rat brains after CNS injury. J Neurochem 74:1106 –1111. Sato H, Tamba M, Ishii T, Bannai S (1999) Cloning and expression of a plasma membrane cystine/glutamate exchange transporter composed of two distinct proteins. J Biol Chem 274:11455–11458. Segal JA, Harris BD, Kustova Y, Basile A, Skolnick P (1999) Aminoglycoside neurotoxicity involves NMDA receptor activation. Brain Res 815:270 –277. Segal JA, Skolnick P (2000) Spermine-induced toxicity in cerebellar granule neurons is independent of its actions at NMDA receptors. J Neurochem 74:60 – 69. Seiler N (2004) Catabolism of polyamines. Amino Acids 26:217–233. Sharmin S, Sakata K, Kashiwagi K, Ueda S, Iwasaki S, Shirahata A, Igarashi K (2001) Polyamine cytotoxicity in the presence of bovine serum amine oxidase. Biochem Biophys Res Commun 282: 228 –235. Shaw GG, Pateman AJ (1973) The regional distribution of the polyamines spermidine and spermine in brain. J Neurochem 20: 1225–1230. Si QS, Nakamura Y, Schubert P, Rudolphi K, Kataoka K (1996) Adenosine and propentofylline inhibit the proliferation of cultured microglial cells. Exp Neurol 137:345–349.

1131

Si QS, Nakamura Y, Kataoka K (1997) Hypothermic suppression of microglial activation in culture: inhibition of cell proliferation and production of nitric oxide and superoxide. Neuroscience 81: 223–229. Sparapani M, Dall’Olio R, Gandolfi O, Ciani E, Contestabile A (1997) Neurotoxicity of polyamines and pharmacological neuoprotection in cultures of rat cerebellar granule cells. Exp Neurol 148:157–166. Tabor CW, Tabor H (1984) Polyamines. Annu Rev Biochem 53: 749 –790. Tabor CW, Tabor H, Bachrach U (1964) Identification of the aminoaldehyde produced by the oxidation of spermine with purified plasma amine oxidase. J Biol Chem 239:2194 –2203. Takano K, Nakamura Y, Yoneda Y (2003) Microglial cell death induced by a low concentration of polyamines. Neuroscience 120:961–967. Virgili M, Sparapani M, Facchinetti F, Ciani E, Dall’Olio R, Contestabile A (1992) In-vitro and in-vivo evidence for spermine and spermidine neurotoxicity. Neurodegeneration 1:273–279. Washburn MS, Dingledine R (1996) Block of ␣-amino-3-hydroxy-5methyl-4-isoxazolepropionic acid (AMPA) receptors by polyamines and polyamine toxins. J Pharmacol Exp Ther 278:669 – 678. Watanabe H, Bannai S (1987) Induction of cystine transport activity in mouse peritoneal macrophages. J Exp Med 165:628 – 640. Yoneda Y, Ogita K (1991) Neurochemical aspects of the N-methyl-Daspartate receptor complex. Neurosci Res 10:1–33.

(Accepted 11 May 2005) (Available online 12 July 2005)