- Email: [email protected]
Activation-induced cell death of rat astrocytes
Brain Research 900 (2001) 342–347 www.elsevier.com / locate / bres
Short communication
Activation-induced cell death of rat astrocytes Kyoungho Suk a , *, Jongseok Lee a , Jinyoung Hur a , Yong S. Kim b , Myung-Shik Lee c , Sang-hoon Cha d , Sun Yeou Kim a , Hocheol Kim a a
Department of Herbal Pharmacology, Graduate School of East-West Medical Science, Kyung Hee University, Hoegi-dong, Tongdaemun-ku, Seoul 130 -701, South Korea b Department of Pharmacology, Seoul National University College of Medicine, Seoul, South Korea c Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea d Division of Food Science and Biotechnology, College of Agriculture and Life Sciences, Kangwon National University, Chunchon, South Korea Accepted 20 February 2001
Abstract Inflammatory activation of astrocytes has been implicated in various neurodegenerative diseases. The elimination of activated astrocytes by apoptosis or the deactivation may be the mechanisms for auto-regulation of activated astrocytes. To test the possibility of apoptotic elimination of activated astrocytes, we examined a potential correlation between activation state of astrocytes and their viability using C6 rat glial cells and rat primary astrocyte cultures exposed to a variety of inflammatory stimuli such as lipopolysaccharide, interferon-g, and tumor necrosis factor-a. Nitric oxide production was measured to evaluate inflammatory activation of astrocytes. We found that: (i) the activation of astrocytes by the combination of lipopolysaccharide and inflammatory cytokines, but not by either alone, led to nitric oxide production followed by apoptotic cell death; (ii) the amount of nitric oxide produced by activated astrocytes was inversely proportional to the viability of the cells; (iii) inhibition of nitric oxide synthase by N-monomethyl L-arginine blocked death of activated astrocytes; and (iv) nitric oxide donors induced apoptosis of astrocytes in a caspase-dependent manner. Taken collectively, our results suggest that activated astrocytes produce nitric oxide as an autocrine mediator of caspase-dependent apoptosis, and this type of programmed cell death of astrocytes may be the underlying mechanism for the auto-regulation of inflammatory activation of astrocytes. 2001 Elsevier Science B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Neuroglia and myelin Keywords: Astrocyte; C6 glial cell; Apoptosis; Activation; Nitric oxide; Central nervous system
Glial cells including astrocytes and microglial cells provide mechanical and metabolic support for neurons. Astrocytes play an essential role in maintaining the function of neurons by producing and responding to a variety of growth factors and cytokines [4]. Stimulated astrocytes proliferate and produce diverse inflammatory mediators such as nitric oxide (NO) and TNFa [12,21,22]. There is also growing evidence that toxic mediators produced by activated astrocytes might be involved in the pathogenesis of various neurodegenerative diseases [3,5]. Thus, production of toxic inflammatory mediators by activated and proliferated astrocytes needs to be tightly *Corresponding author. Tel.: 182-2-961-9233; fax: 182-2-961-9215. E-mail address: [email protected] (K. Suk).
regulated. Potential mechanisms for down-regulation of these activated astrocytes are the deactivation or elimination of activated cells. Recently, activated macrophages have been shown to undergo apoptosis [1,2,23]. It has been suggested that the apoptosis of activated macrophages is one mechanism whereby an organism may regulate immune and inflammatory responses involving macrophages [1]. We have previously shown that microglial cells that are closely related to monocytes and macrophages also undergo apoptosis upon inflammatory activation [16]. Astrocytes may not be an exception for auto-regulation by apoptosis. Apoptosis of astrocytes by diverse stimuli has been previously reported. IL-1b induced apoptosis of human astrocytes [10]. S100b, a calcium binding protein expressed primarily by astrocytes, induced apoptotic cell
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02326-5
K. Suk et al. / Brain Research 900 (2001) 342 – 347
death in cultured rat astrocytes via a nitric oxide-dependent pathway [14]. Anti-Fas antibody mediated apoptosis of cultured human glioma cells [24]. Apoptosis of C6 rat astrocytes was induced by ginsenoside Rh2 [15] and staurosporine [13]. These previous studies demonstrated that astrocytes could undergo apoptosis when exposed to certain physiological or non-physiological stimuli. However, the correlation between apoptosis of astrocytes and their activation state has not been thoroughly investigated. In the current work, we hypothesized that activated astrocytes may undergo apoptosis as one way of downregulating their own activation state, and NO that is produced by activated astrocytes may have a role in their own demise. We attempted to test this hypothesis by examining the production of NO and viability of astrocytes exposed to various inflammatory activating agents. Lipopolysaccharide (LPS), N-monomethyl L-arginine (NMMA), sodium nitroprusside (SNP), and S-nitroso-Nacetylpenicillamine (SNAP) were obtained from Sigma (St Louis, MO). A caspase inhibitor (z-VAD-fmk) was purchased from Enzyme Systems (Livermore, CA). Recombinant mouse TNFa, which has been shown to be active on rat C6 cells [19], was purchased from R&D Systems (Minneapolis, MN). Recombinant rat IFNg was generously provided by Dr van der Meide, TNO Primate Center, The Netherlands. The C6 rat glial cells were obtained from American Type Culture Collection. The cell line was maintained in DMEM supplemented with 10% FBS, 2 mM glutamine, and penicillin–streptomycin (Gibco–BRL, Gaithersburg, MD). Rat primary astrocytes were prepared as previously described with minor modifications [17]. In brief, forebrains of newborn Sprague–Dawley rats were chopped and dissociated by trypsinization and mechanical disruption. The cells were seeded into poly-L-lysine-coated culture flasks. After in vitro culture for 10 days, astrocytes were isolated by shaking of the culture flasks and reseeded into multi-well plates for assays. The purity of astrocyte cultures was greater than 90% as determined by glial fibrillary acidic protein (GFAP) immunofluorescence staining. Cytotoxicity was assessed by MTT assays. Briefly, cells (3310 4 cells in 200 ml of medium per well for C6 cells; 2310 4 cells in 200 ml of medium per well for rat primary astrocytes) were seeded in 96-well microtiter plates and treated with various reagents for the indicated time periods. In some experiments, cells were pretreated for 1 h with NMMA or z-VAD-fmk. After various treatments, medium was removed and 3-[4,5-dimethylthiazol-2-yl]2,5-diphenyltetrazolium bromide (MTT, 0.5 mg / ml) was added followed by incubation at 378C for 2 h in a CO 2 incubator. After a brief centrifugation, supernatants were carefully removed and dimethylsulfoxide was added. After insoluble crystals were completely dissolved, absorbance at 540 nm was measured using a Thermomax microplate reader (Molecular Devices). NO 2 2 in culture supernatants was measured to assess NO
343
production in astrocytes. Sample aliquots (50 ml) were mixed with 50 ml of Griess reagent (1% sulphanilamide / 0.1% naphthylethylene diamine dihydrochloride / 2% phosphoric acid) in 96-well plates and incubated at 258C for 10 min. The absorbance at 550 nm was measured on a microplate reader. NaNO 2 was used as standard to calculate NO 22 concentrations. Morphological changes in the nuclear chromatin of cells undergoing apoptosis were detected by staining with 2.5 mg / ml of DNA-binding bisbenzimide Hoechst 33258 fluorochrome (Calbiochem, La Jolla, CA), followed by an examination under a fluorescence microscope. For the detection of oligonucleosomal cleavage of DNA, genomic DNA was isolated from primary astrocytes and electrophoresed on 1.5% agarose gel, then stained with ethidium bromide. For DNA ploidy analysis, cells were suspended in PBS–5 mM EDTA, and fixed by adding 100% ethanol dropwise. RNase A (40 mg / ml) was added to resuspended cells, and incubation was carried out at room temperature for 30 min. Propidium iodide (50 mg / ml) was then added for flow cytometric analyses (FACS Vantage; Becton Dickinson). Western blot analysis was performed as follows. Cells were lysed in triple-detergent lysis buffer (50 mM Tris– HCl, pH 8.0, 150 mM NaCl, 0.02% sodium azide, 0.1% SDS, 1% NP-40, 0.5% sodium deoxycholate, 1 mM PMSF). Protein concentration in cell lysates was determined using a Bio-Rad protein assay kit. An equal amount of protein for each sample was separated by 10% SDSPAGE and transferred to Hybond ECL nitrocellulose membrane (Amersham). The membrane was blocked with 5% skimmed milk and sequentially incubated with polyclonal rabbit anti-mouse / rat iNOS antibody (Transduction Laboratories) or polyclonal goat anti-mouse / rat COX-2 antibody (Santa Cruz Biotechnology) and HRP-conjugated secondary antibodies (anti-rabbit or -goat IgG, Amersham) followed by ECL detection (Amersham). All data were presented as means6S.D. from three or more independent experiments. Statistical comparison between different treatments was done by one-way ANOVA with Dunnett’s multiple comparison test using GraphPad Prism program (GraphPad Software Inc.). Differences with a P-value less than 0.05 were considered statistically significant. In order to investigate a possible correlation between inflammatory activation and the viability of astrocytes, C6 glial cells were treated with LPS, IFNg, and TNFa either alone or in various combinations for 48 h, then the viability of the cells and NO production were assessed. LPS or cytokines alone did not affect viability of C6 cells or produce a significant amount of NO; however, a combination of these stimuli decreased cell viability and induced NO production to varying extents (Fig. 1A,B). Combinations of LPS and cytokines induced cytotoxicity in the potency order LPS1IFNg1TNFa.IFNg1TNFa. LPS1IFNg.LPS1TNFa. Importantly, the amount of NO
344
K. Suk et al. / Brain Research 900 (2001) 342 – 347
Fig. 1. The cause and effect relationship between NO production and cell death of C6 glial cells. (A, B) C6 cells were treated for 48 h with LPS (100 ng / ml), IFNg (100 U / ml), and TNFa (10 ng / ml) either alone or in various combinations indicated, and then cell viability was evaluated by MTT assays (A) or NO production was assessed by Griess reaction (B). Asterisks indicate statistically significant differences from untreated control (P,0.05). (C, D) Alternatively, C6 cells were treated with LPS plus cytokines for the indicated time periods, then cell viability (C) or NO production (D) was similarly assessed. Asterisks indicate statistically significant differences from untreated control, which was set to 100% viability (P,0.05) (C). All treatments for 24, 48, and 72 h resulted in a significant amount of NO production compared to untreated control (P,0.05) (D). (E, F, G) Nuclear morphology of C6 cells after treatment with LPS / IFNg/ TNFa for 48 h (E, the arrows indicate apoptotic cells with nuclear condensation), cell viability after treatment with LPS / IFNg/ TNFa for 48 h in the absence or presence of 0.5 mM NMMA (F, pre-treatment with NMMA showed 63.2% inhibition of the cell death; asterisks indicate statistically significant differences with P,0.05), or iNOS and COX-2 protein expression after treatment with LPS / IFNg/ TNFa for 16 h (G) was analyzed. Viability of untreated cells was set to 100%. The results are mean6S.D. (n53). The following concentrations of activating agents were used throughout all experiments: LPS, 100 ng / ml; IFNg, 100 U / ml; TNFa, 10 ng / ml.
K. Suk et al. / Brain Research 900 (2001) 342 – 347
produced by activated C6 cells was inversely correlated with cell viability. When the viability and NO production were evaluated at different time points after LPS and cytokine treatment, it was revealed that NO production was followed by cell death (Fig. 1C,D). A significant NO production was observed at 24 h after treatment with LPS and cytokines (Fig. 1D), while no significant decrease in cell viability was observed until 48 h (Fig. 1C). These results suggest that NO production representing the inflammatory activation of astrocytes is causally associated with the cell death. A decrease in viability of activated C6 cells was due to apoptosis. Hoechst nuclear staining (Fig. 1E) and DNA ploidy analysis (data not shown) indicated that C6 cells treated with LPS and cytokines underwent apoptosis. Typical apoptotic morphological changes in nuclei (Fig. 1E) and appearance of sub-diploidy cells (data not shown) were induced by LPS / IFNg/ TNFa treatment. Because NO is known to induce apoptosis of a variety of cell types [6], and activated astrocytes produced NO (Fig. 1B,D), we speculated that NO may be a cytotoxic mediator involved in apoptosis of activated astrocytes. Inhibition of NO synthase (NOS) by NMMA significantly blocked death of activated C6 cells (Fig. 1F), again indicating that NO is involved in the apoptosis of the activated astrocytes. LPS and cytokines are known to induce inducible NO synthase (iNOS) expression in astrocytes, and this is believed to be the underlying mechanism for the increased NO production [11]. Western blot analysis revealed that treatment of C6 cells with LPS / IFNg/ TNFa induced iNOS (Fig. 1G), which was in agreement with previous reports [11]. In contrast, the expression of cyclooxygenase-2 (COX-2) that is closely associated with NO action in astrocytes [18] was not affected by LPS / IFNg/ TNFa. We next sought to determine whether the correlation between NO production and apoptosis observed in C6 cells is also applicable to primary astrocytes. Treatment of rat primary astrocytes with LPS and cytokines induced cell death and NO production in a manner similar to C6 cells (Fig. 2A,B). In contrast to C6 cells, primary astrocytes produced a significant amount of NO in response to LPS alone, which was in agreement with previous reports that demonstrated primary astrocytes could be activated by LPS alone [12,20]. However, the degree of NO production by LPS-activated astrocytes was far lower compared to astrocytes activated by the combination of LPS and cytokines, and activation by LPS alone did not induce a significant cell death (Fig. 2A). Although the possibility of microglial cell contamination in the primary astrocyte cultures cannot be ruled out, the major cellular source of NO and a majority of the cells undergoing apoptosis in response to the combination of LPS and cytokines appear to be astrocytes. No matter how pure the primary astrocyte cultures are, possible contamination of a small number of microglial cells cannot be completely excluded. This may be one of the limitations of using primary cells. Our results
345
of the correlation between activation and apoptosis of the primary astrocytes were complemented by the results obtained from homogeneous C6 cell lines (Fig. 1). Induction of oligonucleosomal cleavage of genomic DNA confirmed that the death of activated primary astrocytes is the typical apoptosis (Fig. 2B inset). These results indicate that a similar correlation between activation and apoptosis exists in primary astrocytes and C6 glial cells. An exogenous NO donor, SNP, exhibited a strong cytotoxicity toward primary astrocytes, and this was partially reversed by broad-spectrum caspase inhibitor, z-VAD-fmk (Fig. 2C). A similar cytotoxicity was induced by treatment with SNAP (0.25 mM), another NO donor (data not shown). A similar SNP-mediated cytotoxicity was observed in C6 glial cells: 89% cytotoxicity after treatment with 1 mM SNP for 24 h. Induction of astrocyte death by exogenous NO donors and its inhibition by a caspase inhibitor further support the view that NO is an autocrine mediator in the apoptosis of activated astrocytes. In the current work, we present evidence for activationinduced cell death (AICD) of astrocytes and the involvement of NO in this process. First, activation of astrocytes with LPS and inflammatory cytokines induced apoptotic death. Second, activated astrocytes produced NO as shown previously [11,12,20,22], whose amount was inversely proportional to the cell viability. Third, production of NO by activated astrocytes preceded the cell death. Finally, inhibition of NO production partially prevented the death of activated astrocytes. These results strongly suggest that activated astrocytes go through apoptosis and the production of NO by activated astrocytes is causally associated with the cell death. Together with our previous report on AICD of microglial cells [16], the present work suggests that the elimination of activated astrocytes or microglial cells by apoptosis could be an important mechanism whereby undesirable effects of long-term activation of glial cells can be minimized. Inflammatory mediators that are produced by activated astrocytes or microglial cells in the central nervous system may have harmful effects on neurons or other glial cells. Thus, in various neurodegenerative diseases involving chronic activation of glial cells, these cells appear to play a more significant role in mediating diseases than in the protection of neurons [3,5]. The results presented here suggest the possibility that a failure of auto-regulatory mechanism for activated astrocytes may contribute to the development of such neurodegenerative diseases. The expression of anti-apoptotic Bcl-x L protein has been found in activated microglial cells associated with Alzheimer’s disease and other pathological states [9], supporting the view that blockade of auto-regulatory apoptosis of activated glial cells may lead to the development of neurodegenerative diseases. Macrophages [1,2], B lymphocytes [8], and T lymphocytes [7,25] have been previously shown to undergo apoptosis for the regulation of their own activation state. Now, our present work
346
K. Suk et al. / Brain Research 900 (2001) 342 – 347
Fig. 2. Correlation between NO production and apoptosis of primary astrocytes. (A, B) Rat primary astrocytes were treated with LPS and cytokines alone or in combinations as indicated for 72 h, then cell viability (A) or NO production (B) was measured. Asterisks indicate statistically significant differences from untreated control (P,0.05). LPS / IFNg/ TNFa treatment activated primary astrocytes to produce NO, and induced apoptosis as demonstrated by internucleosomal cleavage of genomic DNA (inset of panel B). Inset: lane 1, untreated; lane 2, treated with LPS / IFNg/ TNFa for 72 h. (C) Treatment of rat primary astrocyte cultures with SNP (1 mM) for 24 h also induced a marked reduction in cell viability. Pretreatment of the astrocytes with z-VAD-fmk (50 mM) significantly blocked cell death (52.3% inhibition of cell death). Viability of untreated cells was set to 100%. The results are mean6S.D. (n53). Asterisks indicate statistically significant differences (P,0.05).
suggests that a similar auto-regulatory mechanism that involves AICD may exist in astrocytes as well. We have demonstrated that NO is involved in AICD of astrocytes. Although the inhibition of NOS by NMMA blocked apoptosis of activated astrocytes, NO may not be the sole mediator of apoptosis of these cells. Activated astrocytes can produce other potential toxic mediators [3], and a recent report indicated that anti-Fas antibody initiated apoptosis of human glioma cells [24], suggesting the possible involvement of Fas–FasL interaction in the apoptosis of activated astrocytes. Moreover, a complete inhibition of apoptosis of activated C6 cells was not achieved by NMMA (63.2% inhibition), suggesting the presence of other auto-regulatory mediators. Further studies are required to identify these regulatory mediators and to assess the relative contribution of NO to AICD of astrocytes.
Acknowledgements This work was supported by grants from the Brain Korea 21 Project (Ministry of Education, South Korea), and by a grant (PF002102-00) from the Plant Diversity Research Center of 21st Frontier Research Program funded by the Ministry of Science and Technology of the Korean government.
References [1] B. Adler, H. Adler, T.W. Jungi, E. Peterhans, Interferon-alpha primes macrophages for lipopolysaccharide-induced apoptosis, Biochem. Biophys. Res. Commun. 215 (1995) 921–927. [2] J.E. Albina, S. Cui, R.B. Mateo, J.S. Reichner, Nitric oxide-mediated apoptosis in murine peritoneal macrophages, J. Immunol. 150 (1993) 5080–5085.
K. Suk et al. / Brain Research 900 (2001) 342 – 347 [3] M. Aschner, Astrocytes as mediators of immune and inflammatory responses in the CNS, Neurotoxicology 19 (1998) 269–281. [4] B.A. Barres, Y. Barde, Neuronal and glial cell biology, Curr. Opin. Neurobiol. 10 (2000) 642–648. [5] B. Becher, A. Prat, J.P. Antel, Brain-immune connection: immunoregulatory properties of CNS-resident cells, Glia 29 (2000) 293– 304. [6] B. Brune, A. von Knethen, K.B. Sandau, Nitric oxide (NO): an effector of apoptosis, Cell Death Differ. 6 (1999) 969–975. [7] I.N. Crispe, Death and destruction of activated T lymphocytes, Immunol. Res. 19 (1999) 143–157. [8] D. Donjerkovic, D.W. Scott, Activation-induced cell death in B lymphocytes, Cell Res. 10 (2000) 179–192. [9] B. Drache, G.E. Diehl, K. Beyreuther, L.S. Perlmutter, G. Konig, Bcl-xl-specific antibody labels activated microglia associated with Alzheimer’s disease and other pathological states, J. Neurosci. Res. 47 (1997) 98–108. [10] L.C. Ehrlich, P.K. Peterson, S. Hu, Interleukin (IL)-1beta-mediated apoptosis of human astrocytes, Neuroreport 10 (1999) 1849–1852. [11] D.L. Feinstein, E. Galea, S. Roberts, H. Berquist, H. Wang, D.J. Reis, Induction of nitric oxide synthase in rat C6 glioma cells, J. Neurochem. 62 (1994) 315–321. [12] E. Galea, D.L. Feinstein, D.J. Reis, Induction of calcium-independent nitric oxide synthase activity in primary rat glial cultures, Proc. Natl. Acad. Sci. USA 89 (1992) 10945–10949. [13] A.M. Gorman, U.A. Hirt, S. Orrenius, S. Ceccatelli, Dexamethasone pre-treatment interferes with apoptotic death in glioma cells, Neuroscience 96 (2000) 417–425. [14] J. Hu, L.J. Van Eldik, S100 beta induces apoptotic cell death in cultured astrocytes via a nitric oxide-dependent pathway, Biochim. Biophys. Acta 1313 (1996) 239–245. [15] H.E. Kim, J.H. Oh, S.K. Lee, Y.J. Oh, Ginsenoside RH-2 induces apoptotic cell death in rat C6 glioma via a reactive oxygen- and caspase-dependent but Bcl-X(L)-independent pathway, Life Sci. 65 (1999) L33–40.
347
[16] P. Lee, J. Lee, S. Kim, H. Yagita, M.S. Lee, S.Y. Kim, H. Kim, K. Suk, NO as an autocrine mediator in the apoptosis of activated microglial cells: correlation between activation and apoptosis of microglial cells, Brain Res. 892 (2001) 380–385. [17] K.D. McCarthy, J. de Vellis, Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol. 85 (1980) 890–902. [18] V. Mollace, M. Colasanti, C. Muscoli, G.M. Lauro, M. Iannone, D. Rotiroti, G. Nistico, The effect of nitric oxide on cytokine-induced release of PGE 2 by human cultured astroglial cells, Br. J. Pharmacol. 124 (1998) 742–746. [19] T. Nishiya, T. Uehara, M. Kaneko, Y. Nomura, Involvement of nuclear factor-kappaB (NF-kappaB) signaling in the expression of inducible nitric oxide synthase (iNOS) gene in rat C6 glioma cells, Biochem. Biophys. Res. Commun. 275 (2000) 268–273. [20] K. Pahan, A.M. Namboodiri, F.G. Sheikh, B.T. Smith, I. Singh, Increasing cAMP attenuates induction of inducible nitric-oxide synthase in rat primary astrocytes, J. Biol. Chem. 272 (1997) 7786–7791. [21] M. Sawada, N. Kondo, A. Suzumura, T. Marunouchi, Production of tumor necrosis factor-alpha by microglia and astrocytes in culture, Brain Res. 491 (1989) 394–397. [22] M.L. Simmons, S. Murphy, Induction of nitric oxide synthase in glial cells, J. Neurochem. 59 (1992) 897–905. [23] A. von Knethen, A. Lotero, B. Brune, Etoposide and cisplatin induced apoptosis in activated RAW 264.7 macrophages is attenuated by cAMP-induced gene expression, Oncogene 17 (1998) 387–394. [24] M. Weller, K. Frei, P. Groscurth, P.H. Krammer, Y. Yonekawa, A. Fontana, Anti-Fas /APO-1 antibody-mediated apoptosis of cultured human glioma cells. Induction and modulation of sensitivity by cytokines, J. Clin. Invest. 94 (1994) 954–964. [25] S. Wesselborg, O. Janssen, D. Kabelitz, Induction of activationdriven death (apoptosis) in activated but not resting peripheral blood T cells, J. Immunol. 150 (1993) 4338–4345.
Short communication
Activation-induced cell death of rat astrocytes Kyoungho Suk a , *, Jongseok Lee a , Jinyoung Hur a , Yong S. Kim b , Myung-Shik Lee c , Sang-hoon Cha d , Sun Yeou Kim a , Hocheol Kim a a
Department of Herbal Pharmacology, Graduate School of East-West Medical Science, Kyung Hee University, Hoegi-dong, Tongdaemun-ku, Seoul 130 -701, South Korea b Department of Pharmacology, Seoul National University College of Medicine, Seoul, South Korea c Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea d Division of Food Science and Biotechnology, College of Agriculture and Life Sciences, Kangwon National University, Chunchon, South Korea Accepted 20 February 2001
Abstract Inflammatory activation of astrocytes has been implicated in various neurodegenerative diseases. The elimination of activated astrocytes by apoptosis or the deactivation may be the mechanisms for auto-regulation of activated astrocytes. To test the possibility of apoptotic elimination of activated astrocytes, we examined a potential correlation between activation state of astrocytes and their viability using C6 rat glial cells and rat primary astrocyte cultures exposed to a variety of inflammatory stimuli such as lipopolysaccharide, interferon-g, and tumor necrosis factor-a. Nitric oxide production was measured to evaluate inflammatory activation of astrocytes. We found that: (i) the activation of astrocytes by the combination of lipopolysaccharide and inflammatory cytokines, but not by either alone, led to nitric oxide production followed by apoptotic cell death; (ii) the amount of nitric oxide produced by activated astrocytes was inversely proportional to the viability of the cells; (iii) inhibition of nitric oxide synthase by N-monomethyl L-arginine blocked death of activated astrocytes; and (iv) nitric oxide donors induced apoptosis of astrocytes in a caspase-dependent manner. Taken collectively, our results suggest that activated astrocytes produce nitric oxide as an autocrine mediator of caspase-dependent apoptosis, and this type of programmed cell death of astrocytes may be the underlying mechanism for the auto-regulation of inflammatory activation of astrocytes. 2001 Elsevier Science B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Neuroglia and myelin Keywords: Astrocyte; C6 glial cell; Apoptosis; Activation; Nitric oxide; Central nervous system
Glial cells including astrocytes and microglial cells provide mechanical and metabolic support for neurons. Astrocytes play an essential role in maintaining the function of neurons by producing and responding to a variety of growth factors and cytokines [4]. Stimulated astrocytes proliferate and produce diverse inflammatory mediators such as nitric oxide (NO) and TNFa [12,21,22]. There is also growing evidence that toxic mediators produced by activated astrocytes might be involved in the pathogenesis of various neurodegenerative diseases [3,5]. Thus, production of toxic inflammatory mediators by activated and proliferated astrocytes needs to be tightly *Corresponding author. Tel.: 182-2-961-9233; fax: 182-2-961-9215. E-mail address: [email protected] (K. Suk).
regulated. Potential mechanisms for down-regulation of these activated astrocytes are the deactivation or elimination of activated cells. Recently, activated macrophages have been shown to undergo apoptosis [1,2,23]. It has been suggested that the apoptosis of activated macrophages is one mechanism whereby an organism may regulate immune and inflammatory responses involving macrophages [1]. We have previously shown that microglial cells that are closely related to monocytes and macrophages also undergo apoptosis upon inflammatory activation [16]. Astrocytes may not be an exception for auto-regulation by apoptosis. Apoptosis of astrocytes by diverse stimuli has been previously reported. IL-1b induced apoptosis of human astrocytes [10]. S100b, a calcium binding protein expressed primarily by astrocytes, induced apoptotic cell
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02326-5
K. Suk et al. / Brain Research 900 (2001) 342 – 347
death in cultured rat astrocytes via a nitric oxide-dependent pathway [14]. Anti-Fas antibody mediated apoptosis of cultured human glioma cells [24]. Apoptosis of C6 rat astrocytes was induced by ginsenoside Rh2 [15] and staurosporine [13]. These previous studies demonstrated that astrocytes could undergo apoptosis when exposed to certain physiological or non-physiological stimuli. However, the correlation between apoptosis of astrocytes and their activation state has not been thoroughly investigated. In the current work, we hypothesized that activated astrocytes may undergo apoptosis as one way of downregulating their own activation state, and NO that is produced by activated astrocytes may have a role in their own demise. We attempted to test this hypothesis by examining the production of NO and viability of astrocytes exposed to various inflammatory activating agents. Lipopolysaccharide (LPS), N-monomethyl L-arginine (NMMA), sodium nitroprusside (SNP), and S-nitroso-Nacetylpenicillamine (SNAP) were obtained from Sigma (St Louis, MO). A caspase inhibitor (z-VAD-fmk) was purchased from Enzyme Systems (Livermore, CA). Recombinant mouse TNFa, which has been shown to be active on rat C6 cells [19], was purchased from R&D Systems (Minneapolis, MN). Recombinant rat IFNg was generously provided by Dr van der Meide, TNO Primate Center, The Netherlands. The C6 rat glial cells were obtained from American Type Culture Collection. The cell line was maintained in DMEM supplemented with 10% FBS, 2 mM glutamine, and penicillin–streptomycin (Gibco–BRL, Gaithersburg, MD). Rat primary astrocytes were prepared as previously described with minor modifications [17]. In brief, forebrains of newborn Sprague–Dawley rats were chopped and dissociated by trypsinization and mechanical disruption. The cells were seeded into poly-L-lysine-coated culture flasks. After in vitro culture for 10 days, astrocytes were isolated by shaking of the culture flasks and reseeded into multi-well plates for assays. The purity of astrocyte cultures was greater than 90% as determined by glial fibrillary acidic protein (GFAP) immunofluorescence staining. Cytotoxicity was assessed by MTT assays. Briefly, cells (3310 4 cells in 200 ml of medium per well for C6 cells; 2310 4 cells in 200 ml of medium per well for rat primary astrocytes) were seeded in 96-well microtiter plates and treated with various reagents for the indicated time periods. In some experiments, cells were pretreated for 1 h with NMMA or z-VAD-fmk. After various treatments, medium was removed and 3-[4,5-dimethylthiazol-2-yl]2,5-diphenyltetrazolium bromide (MTT, 0.5 mg / ml) was added followed by incubation at 378C for 2 h in a CO 2 incubator. After a brief centrifugation, supernatants were carefully removed and dimethylsulfoxide was added. After insoluble crystals were completely dissolved, absorbance at 540 nm was measured using a Thermomax microplate reader (Molecular Devices). NO 2 2 in culture supernatants was measured to assess NO
343
production in astrocytes. Sample aliquots (50 ml) were mixed with 50 ml of Griess reagent (1% sulphanilamide / 0.1% naphthylethylene diamine dihydrochloride / 2% phosphoric acid) in 96-well plates and incubated at 258C for 10 min. The absorbance at 550 nm was measured on a microplate reader. NaNO 2 was used as standard to calculate NO 22 concentrations. Morphological changes in the nuclear chromatin of cells undergoing apoptosis were detected by staining with 2.5 mg / ml of DNA-binding bisbenzimide Hoechst 33258 fluorochrome (Calbiochem, La Jolla, CA), followed by an examination under a fluorescence microscope. For the detection of oligonucleosomal cleavage of DNA, genomic DNA was isolated from primary astrocytes and electrophoresed on 1.5% agarose gel, then stained with ethidium bromide. For DNA ploidy analysis, cells were suspended in PBS–5 mM EDTA, and fixed by adding 100% ethanol dropwise. RNase A (40 mg / ml) was added to resuspended cells, and incubation was carried out at room temperature for 30 min. Propidium iodide (50 mg / ml) was then added for flow cytometric analyses (FACS Vantage; Becton Dickinson). Western blot analysis was performed as follows. Cells were lysed in triple-detergent lysis buffer (50 mM Tris– HCl, pH 8.0, 150 mM NaCl, 0.02% sodium azide, 0.1% SDS, 1% NP-40, 0.5% sodium deoxycholate, 1 mM PMSF). Protein concentration in cell lysates was determined using a Bio-Rad protein assay kit. An equal amount of protein for each sample was separated by 10% SDSPAGE and transferred to Hybond ECL nitrocellulose membrane (Amersham). The membrane was blocked with 5% skimmed milk and sequentially incubated with polyclonal rabbit anti-mouse / rat iNOS antibody (Transduction Laboratories) or polyclonal goat anti-mouse / rat COX-2 antibody (Santa Cruz Biotechnology) and HRP-conjugated secondary antibodies (anti-rabbit or -goat IgG, Amersham) followed by ECL detection (Amersham). All data were presented as means6S.D. from three or more independent experiments. Statistical comparison between different treatments was done by one-way ANOVA with Dunnett’s multiple comparison test using GraphPad Prism program (GraphPad Software Inc.). Differences with a P-value less than 0.05 were considered statistically significant. In order to investigate a possible correlation between inflammatory activation and the viability of astrocytes, C6 glial cells were treated with LPS, IFNg, and TNFa either alone or in various combinations for 48 h, then the viability of the cells and NO production were assessed. LPS or cytokines alone did not affect viability of C6 cells or produce a significant amount of NO; however, a combination of these stimuli decreased cell viability and induced NO production to varying extents (Fig. 1A,B). Combinations of LPS and cytokines induced cytotoxicity in the potency order LPS1IFNg1TNFa.IFNg1TNFa. LPS1IFNg.LPS1TNFa. Importantly, the amount of NO
344
K. Suk et al. / Brain Research 900 (2001) 342 – 347
Fig. 1. The cause and effect relationship between NO production and cell death of C6 glial cells. (A, B) C6 cells were treated for 48 h with LPS (100 ng / ml), IFNg (100 U / ml), and TNFa (10 ng / ml) either alone or in various combinations indicated, and then cell viability was evaluated by MTT assays (A) or NO production was assessed by Griess reaction (B). Asterisks indicate statistically significant differences from untreated control (P,0.05). (C, D) Alternatively, C6 cells were treated with LPS plus cytokines for the indicated time periods, then cell viability (C) or NO production (D) was similarly assessed. Asterisks indicate statistically significant differences from untreated control, which was set to 100% viability (P,0.05) (C). All treatments for 24, 48, and 72 h resulted in a significant amount of NO production compared to untreated control (P,0.05) (D). (E, F, G) Nuclear morphology of C6 cells after treatment with LPS / IFNg/ TNFa for 48 h (E, the arrows indicate apoptotic cells with nuclear condensation), cell viability after treatment with LPS / IFNg/ TNFa for 48 h in the absence or presence of 0.5 mM NMMA (F, pre-treatment with NMMA showed 63.2% inhibition of the cell death; asterisks indicate statistically significant differences with P,0.05), or iNOS and COX-2 protein expression after treatment with LPS / IFNg/ TNFa for 16 h (G) was analyzed. Viability of untreated cells was set to 100%. The results are mean6S.D. (n53). The following concentrations of activating agents were used throughout all experiments: LPS, 100 ng / ml; IFNg, 100 U / ml; TNFa, 10 ng / ml.
K. Suk et al. / Brain Research 900 (2001) 342 – 347
produced by activated C6 cells was inversely correlated with cell viability. When the viability and NO production were evaluated at different time points after LPS and cytokine treatment, it was revealed that NO production was followed by cell death (Fig. 1C,D). A significant NO production was observed at 24 h after treatment with LPS and cytokines (Fig. 1D), while no significant decrease in cell viability was observed until 48 h (Fig. 1C). These results suggest that NO production representing the inflammatory activation of astrocytes is causally associated with the cell death. A decrease in viability of activated C6 cells was due to apoptosis. Hoechst nuclear staining (Fig. 1E) and DNA ploidy analysis (data not shown) indicated that C6 cells treated with LPS and cytokines underwent apoptosis. Typical apoptotic morphological changes in nuclei (Fig. 1E) and appearance of sub-diploidy cells (data not shown) were induced by LPS / IFNg/ TNFa treatment. Because NO is known to induce apoptosis of a variety of cell types [6], and activated astrocytes produced NO (Fig. 1B,D), we speculated that NO may be a cytotoxic mediator involved in apoptosis of activated astrocytes. Inhibition of NO synthase (NOS) by NMMA significantly blocked death of activated C6 cells (Fig. 1F), again indicating that NO is involved in the apoptosis of the activated astrocytes. LPS and cytokines are known to induce inducible NO synthase (iNOS) expression in astrocytes, and this is believed to be the underlying mechanism for the increased NO production [11]. Western blot analysis revealed that treatment of C6 cells with LPS / IFNg/ TNFa induced iNOS (Fig. 1G), which was in agreement with previous reports [11]. In contrast, the expression of cyclooxygenase-2 (COX-2) that is closely associated with NO action in astrocytes [18] was not affected by LPS / IFNg/ TNFa. We next sought to determine whether the correlation between NO production and apoptosis observed in C6 cells is also applicable to primary astrocytes. Treatment of rat primary astrocytes with LPS and cytokines induced cell death and NO production in a manner similar to C6 cells (Fig. 2A,B). In contrast to C6 cells, primary astrocytes produced a significant amount of NO in response to LPS alone, which was in agreement with previous reports that demonstrated primary astrocytes could be activated by LPS alone [12,20]. However, the degree of NO production by LPS-activated astrocytes was far lower compared to astrocytes activated by the combination of LPS and cytokines, and activation by LPS alone did not induce a significant cell death (Fig. 2A). Although the possibility of microglial cell contamination in the primary astrocyte cultures cannot be ruled out, the major cellular source of NO and a majority of the cells undergoing apoptosis in response to the combination of LPS and cytokines appear to be astrocytes. No matter how pure the primary astrocyte cultures are, possible contamination of a small number of microglial cells cannot be completely excluded. This may be one of the limitations of using primary cells. Our results
345
of the correlation between activation and apoptosis of the primary astrocytes were complemented by the results obtained from homogeneous C6 cell lines (Fig. 1). Induction of oligonucleosomal cleavage of genomic DNA confirmed that the death of activated primary astrocytes is the typical apoptosis (Fig. 2B inset). These results indicate that a similar correlation between activation and apoptosis exists in primary astrocytes and C6 glial cells. An exogenous NO donor, SNP, exhibited a strong cytotoxicity toward primary astrocytes, and this was partially reversed by broad-spectrum caspase inhibitor, z-VAD-fmk (Fig. 2C). A similar cytotoxicity was induced by treatment with SNAP (0.25 mM), another NO donor (data not shown). A similar SNP-mediated cytotoxicity was observed in C6 glial cells: 89% cytotoxicity after treatment with 1 mM SNP for 24 h. Induction of astrocyte death by exogenous NO donors and its inhibition by a caspase inhibitor further support the view that NO is an autocrine mediator in the apoptosis of activated astrocytes. In the current work, we present evidence for activationinduced cell death (AICD) of astrocytes and the involvement of NO in this process. First, activation of astrocytes with LPS and inflammatory cytokines induced apoptotic death. Second, activated astrocytes produced NO as shown previously [11,12,20,22], whose amount was inversely proportional to the cell viability. Third, production of NO by activated astrocytes preceded the cell death. Finally, inhibition of NO production partially prevented the death of activated astrocytes. These results strongly suggest that activated astrocytes go through apoptosis and the production of NO by activated astrocytes is causally associated with the cell death. Together with our previous report on AICD of microglial cells [16], the present work suggests that the elimination of activated astrocytes or microglial cells by apoptosis could be an important mechanism whereby undesirable effects of long-term activation of glial cells can be minimized. Inflammatory mediators that are produced by activated astrocytes or microglial cells in the central nervous system may have harmful effects on neurons or other glial cells. Thus, in various neurodegenerative diseases involving chronic activation of glial cells, these cells appear to play a more significant role in mediating diseases than in the protection of neurons [3,5]. The results presented here suggest the possibility that a failure of auto-regulatory mechanism for activated astrocytes may contribute to the development of such neurodegenerative diseases. The expression of anti-apoptotic Bcl-x L protein has been found in activated microglial cells associated with Alzheimer’s disease and other pathological states [9], supporting the view that blockade of auto-regulatory apoptosis of activated glial cells may lead to the development of neurodegenerative diseases. Macrophages [1,2], B lymphocytes [8], and T lymphocytes [7,25] have been previously shown to undergo apoptosis for the regulation of their own activation state. Now, our present work
346
K. Suk et al. / Brain Research 900 (2001) 342 – 347
Fig. 2. Correlation between NO production and apoptosis of primary astrocytes. (A, B) Rat primary astrocytes were treated with LPS and cytokines alone or in combinations as indicated for 72 h, then cell viability (A) or NO production (B) was measured. Asterisks indicate statistically significant differences from untreated control (P,0.05). LPS / IFNg/ TNFa treatment activated primary astrocytes to produce NO, and induced apoptosis as demonstrated by internucleosomal cleavage of genomic DNA (inset of panel B). Inset: lane 1, untreated; lane 2, treated with LPS / IFNg/ TNFa for 72 h. (C) Treatment of rat primary astrocyte cultures with SNP (1 mM) for 24 h also induced a marked reduction in cell viability. Pretreatment of the astrocytes with z-VAD-fmk (50 mM) significantly blocked cell death (52.3% inhibition of cell death). Viability of untreated cells was set to 100%. The results are mean6S.D. (n53). Asterisks indicate statistically significant differences (P,0.05).
suggests that a similar auto-regulatory mechanism that involves AICD may exist in astrocytes as well. We have demonstrated that NO is involved in AICD of astrocytes. Although the inhibition of NOS by NMMA blocked apoptosis of activated astrocytes, NO may not be the sole mediator of apoptosis of these cells. Activated astrocytes can produce other potential toxic mediators [3], and a recent report indicated that anti-Fas antibody initiated apoptosis of human glioma cells [24], suggesting the possible involvement of Fas–FasL interaction in the apoptosis of activated astrocytes. Moreover, a complete inhibition of apoptosis of activated C6 cells was not achieved by NMMA (63.2% inhibition), suggesting the presence of other auto-regulatory mediators. Further studies are required to identify these regulatory mediators and to assess the relative contribution of NO to AICD of astrocytes.
Acknowledgements This work was supported by grants from the Brain Korea 21 Project (Ministry of Education, South Korea), and by a grant (PF002102-00) from the Plant Diversity Research Center of 21st Frontier Research Program funded by the Ministry of Science and Technology of the Korean government.
References [1] B. Adler, H. Adler, T.W. Jungi, E. Peterhans, Interferon-alpha primes macrophages for lipopolysaccharide-induced apoptosis, Biochem. Biophys. Res. Commun. 215 (1995) 921–927. [2] J.E. Albina, S. Cui, R.B. Mateo, J.S. Reichner, Nitric oxide-mediated apoptosis in murine peritoneal macrophages, J. Immunol. 150 (1993) 5080–5085.
K. Suk et al. / Brain Research 900 (2001) 342 – 347 [3] M. Aschner, Astrocytes as mediators of immune and inflammatory responses in the CNS, Neurotoxicology 19 (1998) 269–281. [4] B.A. Barres, Y. Barde, Neuronal and glial cell biology, Curr. Opin. Neurobiol. 10 (2000) 642–648. [5] B. Becher, A. Prat, J.P. Antel, Brain-immune connection: immunoregulatory properties of CNS-resident cells, Glia 29 (2000) 293– 304. [6] B. Brune, A. von Knethen, K.B. Sandau, Nitric oxide (NO): an effector of apoptosis, Cell Death Differ. 6 (1999) 969–975. [7] I.N. Crispe, Death and destruction of activated T lymphocytes, Immunol. Res. 19 (1999) 143–157. [8] D. Donjerkovic, D.W. Scott, Activation-induced cell death in B lymphocytes, Cell Res. 10 (2000) 179–192. [9] B. Drache, G.E. Diehl, K. Beyreuther, L.S. Perlmutter, G. Konig, Bcl-xl-specific antibody labels activated microglia associated with Alzheimer’s disease and other pathological states, J. Neurosci. Res. 47 (1997) 98–108. [10] L.C. Ehrlich, P.K. Peterson, S. Hu, Interleukin (IL)-1beta-mediated apoptosis of human astrocytes, Neuroreport 10 (1999) 1849–1852. [11] D.L. Feinstein, E. Galea, S. Roberts, H. Berquist, H. Wang, D.J. Reis, Induction of nitric oxide synthase in rat C6 glioma cells, J. Neurochem. 62 (1994) 315–321. [12] E. Galea, D.L. Feinstein, D.J. Reis, Induction of calcium-independent nitric oxide synthase activity in primary rat glial cultures, Proc. Natl. Acad. Sci. USA 89 (1992) 10945–10949. [13] A.M. Gorman, U.A. Hirt, S. Orrenius, S. Ceccatelli, Dexamethasone pre-treatment interferes with apoptotic death in glioma cells, Neuroscience 96 (2000) 417–425. [14] J. Hu, L.J. Van Eldik, S100 beta induces apoptotic cell death in cultured astrocytes via a nitric oxide-dependent pathway, Biochim. Biophys. Acta 1313 (1996) 239–245. [15] H.E. Kim, J.H. Oh, S.K. Lee, Y.J. Oh, Ginsenoside RH-2 induces apoptotic cell death in rat C6 glioma via a reactive oxygen- and caspase-dependent but Bcl-X(L)-independent pathway, Life Sci. 65 (1999) L33–40.
347
[16] P. Lee, J. Lee, S. Kim, H. Yagita, M.S. Lee, S.Y. Kim, H. Kim, K. Suk, NO as an autocrine mediator in the apoptosis of activated microglial cells: correlation between activation and apoptosis of microglial cells, Brain Res. 892 (2001) 380–385. [17] K.D. McCarthy, J. de Vellis, Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell Biol. 85 (1980) 890–902. [18] V. Mollace, M. Colasanti, C. Muscoli, G.M. Lauro, M. Iannone, D. Rotiroti, G. Nistico, The effect of nitric oxide on cytokine-induced release of PGE 2 by human cultured astroglial cells, Br. J. Pharmacol. 124 (1998) 742–746. [19] T. Nishiya, T. Uehara, M. Kaneko, Y. Nomura, Involvement of nuclear factor-kappaB (NF-kappaB) signaling in the expression of inducible nitric oxide synthase (iNOS) gene in rat C6 glioma cells, Biochem. Biophys. Res. Commun. 275 (2000) 268–273. [20] K. Pahan, A.M. Namboodiri, F.G. Sheikh, B.T. Smith, I. Singh, Increasing cAMP attenuates induction of inducible nitric-oxide synthase in rat primary astrocytes, J. Biol. Chem. 272 (1997) 7786–7791. [21] M. Sawada, N. Kondo, A. Suzumura, T. Marunouchi, Production of tumor necrosis factor-alpha by microglia and astrocytes in culture, Brain Res. 491 (1989) 394–397. [22] M.L. Simmons, S. Murphy, Induction of nitric oxide synthase in glial cells, J. Neurochem. 59 (1992) 897–905. [23] A. von Knethen, A. Lotero, B. Brune, Etoposide and cisplatin induced apoptosis in activated RAW 264.7 macrophages is attenuated by cAMP-induced gene expression, Oncogene 17 (1998) 387–394. [24] M. Weller, K. Frei, P. Groscurth, P.H. Krammer, Y. Yonekawa, A. Fontana, Anti-Fas /APO-1 antibody-mediated apoptosis of cultured human glioma cells. Induction and modulation of sensitivity by cytokines, J. Clin. Invest. 94 (1994) 954–964. [25] S. Wesselborg, O. Janssen, D. Kabelitz, Induction of activationdriven death (apoptosis) in activated but not resting peripheral blood T cells, J. Immunol. 150 (1993) 4338–4345.