Neuroscience 117 (2003) 811– 820
5'-AMINOIMIDAZOLE-4-CARBOXAMIDE RIBOSIDE INDUCES APOPTOSIS IN HUMAN NEUROBLASTOMA CELLS M. GARCIA-GIL,a R. PESI,a S. PERNA,a S. ALLEGRINI,b M. GIANNECCHINI,a M. CAMICIa* AND M. G. TOZZIb
festations of syndromes related to purine dismetabolisms. © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved.
Dipartimento di Fisiologia e Biochimica, Universita` di Pisa, Via S. Maria, 55 Pisa 56100, Italy
Key words: neuroblastoma cells, purine dismetabolisms, differentiation, 5ⴕ-nucleotidase, AICA riboside.
Dipartimento di Scienze del Farmaco, Universita` di Sassari, Sassari, Italy
We have demonstrated that 5'-aminoimidazole-4-carboxamide (AICA) riboside is toxic in rat hippocampal HN9 cells and human neuroblastoma SH-SY5Y cells, causing DNA fragmentation (Pesi et al., 2000). Our finding was somehow surprising since AICA riboside (acadesine or Z-riboside) is now intensively studied in many cell types as precursor of the corresponding ribotide, which is a good activator of the AMP-dependent protein kinase, the activation of which is known to regulate glucose and fatty acid metabolism (Winder and Hardie, 1999) and seems to protect cardiomyocytes from ischemia damage (Abbud et al., 2000). AICA ribotide is an intermediate of purine de novo biosynthesis and is usually converted directly into inosine5⬘-monophosphate, the precursor of all purine compounds. AICA riboside which derives from the hydrolysis of the ribotide is undetectable in normal healthy cells. However, the presence of AICA riboside triphosphate has been reported in red cells of Lesch-Nyhan patients and of patients affected by a syndrome caused by a phosphoribosyl-pyrophosphate (PRPP) synthase overactivity (Sidi and Mitchell, 1985). In both syndromes an acceleration of purine de novo synthesis has been described. The accumulation of AICA riboside triphosphate in red cells stems from the phosphorylation of AICA riboside taken up from blood which in turn can be generated from the hydrolysis of AICA ribotide oversynthesized in those cells able to perform purine de novo synthesis (Sidi and Mitchell, 1985). We have recently demonstrated that the ubiquitous cytosolic 5'-nucleotidase catalyzes the hydrolysis of AICA riboside monophosphate and is overactive in red cells of patients affected by Lesch-Nyhan syndrome (Pesi et al., 2000). These observations indicate that pathological conditions occur in which AICA riboside might be present in the blood as a consequence of purine dismetabolisms. Both LeschNyhan and PRPP synthase overactivity syndromes are associated with various degrees of neurological impairments. Furthermore, several neurological diseases are associated with some degree of increase in uric acid excretion, including 20% of autistic population in which a fourfold increase of purine de novo synthesis has been demonstrated (Page and Coleman, 2000). AICA riboside is transported inside the cells through the adenosine carrier and can be phosphorylated by adenosine kinase (Young et
Abstract—5'-Aminoimidazole-4-carboxamide riboside (AICA riboside) has been previously shown to be toxic to two neuronal cell models [Neuroreport 11 (2000) 1827]. In this paper we demonstrate that AICA riboside promotes apoptosis in undifferentiated human neuroblastoma cells (SH-SY5Y), inducing a raise in caspase-3 activity. In order to exert its effect on viability, AICA riboside must enter the cells and be phosphorylated to the ribotide, since both a nucleoside transport inhibitor, and an inhibitor of adenosine kinase produce an enhancement of the viability of AICA riboside-treated cells. Short-term incubations (2 h) with AICA riboside result in five-fold increase in the activity of AMP-dependent protein kinase (AMPK). However, the activity of AMPK is not significantly affected at prolonged incubations (48 h), when the apoptotic effect of AICA riboside is evident. The results demonstrate that when the cell line is induced to differentiate both toward a cholinergic phenotype (with retinoic acid) or a noradrenergic phenotype (with phorbol esters), the toxic effect is significantly reduced, and in the case of the noradrenergic phenotype differentiation, the riboside is completely ineffective in promoting apoptosis. This reduction of effect correlates with an overexpression of Bcl-2 during differentiation. AICA riboside, derived from the hydrolysis of the ribotide, an intermediate of purine de novo synthesis, is absent in normal healthy cells; however it may accumulate in those individuals in which an inborn error of purine metabolism causes an increase in the rate of de novo synthesis and/or an overexpression of cytosolic 5'-nucleotidase, that appears to be the enzyme responsible for AICA ribotide hydrolysis. In fact, 5'-nucleotidase activity has been shown to increase in patients affected by Lesch-Nyhan syndrome in which both acceleration of de novo synthesis and accumulation of AICA ribotide has been described, and also in other neurological disorders of unknown etiology. Our results raise the intriguing clue that the neurotoxic effect of AICA riboside on the developing brain might contribute to the neurological mani*Corresponding author. Tel: ⫹39-050-500-292; fax ⫹39-050-502-583. E-mail address: [email protected]
(M. Camici). Abbreviations: AICA, 5'-aminoimidazole-4-carboxamide; 5'amino5⬘dAdo, 5⬘amino-5'deoxyadenosine; AMPK, AMP-dependent protein kinase; BDNF, brain-derived neurotrophic factor; BSA, bovine serum albumin; DEVD-pNA, Ac-Asp-Glu-Val-Asp-paranitroaniline; DMEM, Dulbecco’s modified Eagle’s medium; DTT, dithiothreitol; FBS, fetal bovine serum; HEPES, HCO(3-)-free N-2-hydroxyethylpiperazine-N'2-ethanesulfonic acid; IgG, immunoglobulin G; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; PBS, phosphate-buffered saline; PRPP, phosphoribosyl-pyrophosphate; RA, all-trans-retinoic acid; SAMS, NH2-His-Met-Arg-Ser-Ala-Met-Ser-Gly-Leu-His-LeuVal-Lys-Arg-Arg-COOH; TPA, 12-O-tetradecanoyl-phorbol-13-acetate; YVAD-pNA, Ac-Tyr-Val-Ala-Asp-paranitroaniline.
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved. doi:10.1016/S0306-4522(02)00836-9
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
al., 1996), indicating that it may be considered as an adenosine analog. Furthermore, the corresponding monophosphate exerts its action inside the cells since it interacts with sites specific for AMP (Muoio et al., 1999). Adenosine and AMP are intermediates in ATP metabolism and usually their level increases in the cells during ATP breakdown, so that they play several roles in the cellular response to the stress caused by energy depletion (Muoio et al., 1999). Therefore, it is reasonable to assume that a continuous exposition of cells to AICA riboside simulates cellular signals for metabolic stress. For some of the neurological diseases described above the cause for the metabolic derangement is clarified, but that for the neurological manifestations is not yet understood (Visser et al., 2000). We have addressed this issue, studying whether the increase of AICA riboside levels, which may arise from the concomitant acceleration of purine synthesis, subsequent to some purine dismetabolisms, and/or 5'-nucleotidase hyperactivity, might interfere with normal neuronal differentiation of dopaminergic and cholinergic neurons, and therefore might contribute to the neurological disorders often associated to purine dismetabolisms. For this purpose, we have used SH-SY5Y, an adrenergic clone of the human neuroblastoma cell line SK-N-SH (Biedler et al., 1973) as a model of neuronal differentiation. These cells can be induced to differentiate by several agents; phorbol esters induce differentiation toward a noradrenergic phenotype, whereas retinoic acid does it toward a cholinergic phenotype (Pahlman et al., 1990).
EXPERIMENTAL PROCEDURES Materials Ac-Tyr-Val-Ala-Asp-paranitroaniline (YVAD-pNA), and Ac-AspGlu-Val-Asp-paranitroaniline (DEVD-pNA) were from Calbiochem, La Jolla, CA, USA; brain-derived neurotrophic factor (BDNF) was from Alomone Laboratories, Jerusalem, Israel; fetal bovine serum (FBS) was from Biochrom, Berlin, Germany; all-trans-retinoic acid (RA), Dulbecco’s modified Eagle’s medium (DMEM), RPMI-1640 medium, propidium iodide, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), AICA riboside, mouse monoclonal antibody anti neurofilament 200, antimouse immunoglobulin G (IgG)-tetramethylrhodamineisothiocyanate (TRITC) conjugate, 12-O-tetradecanoylphorbol-13-acetate (TPA), 5'amino-5'deoxyadenosine (5'amino-5'dAdo) were purchased from Sigma, St. Louis, MO, USA. Super Signal West Pico chemiluminescent substrate was from Pierce, Rockford, IL, USA. The NH2-His-Met-ArgSer-Ala-Met-Ser-Gly-Leu-His-Leu-Val-Lys-Arg-Arg-COOH (SAMS) peptide was synthesized by Primm srl (S. Raffaele Biomedical Science-Park, Milano, Italy). [␥32P]ATP (3000 Ci/mmol) was from PerkinElmer Life Sciences, Milano, Italy. Mouse monoclonal antibody anti Bcl-2 was purchased from Boehringer-Mannheim, Monza, Italy. Peroxidase-conjugated rabbit anti mouse immunoglobulins were from Dako, Glostrup, Denmark. Rabbit polyclonal antibodies anti caspase-8 and anti-actin were from Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA. Peroxidase-conjugated chicken antirabbit immunoglobulins were obtained from Chemicon International, Temecula, CA, USA. DE-81 phosphocellulose was from Whatman, Maidstone, England. SH-SY5Y human neuroblastoma cells were a kind gift of Dr. Isabel Varela, CSIC, Madrid, Spain. All other reagents were of analytical grade.
Cell culture and treatments. SH-SY5Y neuroblastoma cells were grown in DMEM supplemented with 10% FBS, 100 IU/ml penicillin, 100 g/ml streptomycin and 2.5 g/ml amphotericin B (complete medium). Cells were maintained at 37 °C in a saturating humidity atmosphere containing 95% air/5% CO2. In experiments with undifferentiated cells, 15,000 cells were plated in 96-well plates; 24 h after plating, cells were washed with serum-free DMEM and cells were incubated for the indicated times in the presence of AICA riboside. In the experiments performed in the presence of dipyridamole or 5'amino-5'dAdo, these compounds were added 30 min before AICA riboside. For experiments of RA-induced differentiation, 18,000 cells were plated directly on 24-well plates for MTT assay or on coverslips coated with 20 g/ml laminin in 24-well culture dishes for immunohistochemistry; 24 h after plating, cells were washed with serum-free DMEM. Differentiation was induced with 10-M RA. According to the experiments, 200-M AICA riboside was added simultaneously or after 5 days of treatment with RA. For BDNF-induced differentiation, cells were pretreated as above with 10-M RA for 5 days in complete medium and then switched to serum-free medium containing 50 ng/ml BDNF (Encinas et al., 2000). In experiments of differentiation with TPA, cells were grown in RPMI supplemented with 10% FBS, 100 IU/ml penicillin, 100 g/ml streptomycin and 2.5 g/ml amphotericin B and plated (9,000 –18,000 cell/well); 24 h after plating, cells were washed with serum-free RPMI and differentiated in defined medium (RPMI supplemented with 16 g/ml putrescine, 30-nM selenite, 6 ng/ml progesterone and 100 g/ml transferrine). Differentiation was induced with 3.2- or 16-nM TPA. Morphology of the cells was observed with a Zeiss microscope (Axiovert 25) using 40⫻ magnification. MTT assay. Viability was assessed by the reduction of MTT. MTT is a water-soluble tetrazolium salt that is reduced by metabolically viable cells to a colored, water-insoluble formazan salt. Cells have been grown in 96-well or 24-well plates and treated with AICA riboside and other additives as described above. The treatments were performed in order to have six wells per experimental condition. After adding MTT (0.5 mg/ml final concentration) to the culture medium of cells, plates were incubated at 37 °C for 30 min; the assay was stopped by replacement of the MTT-containing medium with 100 l dimethylsulfoxide (DMSO). Formazan salts were dissolved in DMSO by gentle shaking for 10 min at room temperature. Absorbance at 570 nm was read by means of an ELISA plate reader. Each experiment was repeated at least three times. Caspase-3 and caspase-1 activities. Cells (6 million) were grown in 75-cm2 flasks and treated with AICA riboside for 24 –72 h. The culture medium containing the detached cells was added to the trypsinized adherent cells, subjected to centrifugation and the pellet washed twice with phosphate-buffered saline (PBS). Pellets were re-suspended and lysed with 150 l of lysis buffer containing 20-mM Tris–HCl pH 7.4, 150-mM NaCl, 1-mM dithiothreitol (DTT), 5-mM EDTA, 5-mM EGTA, 0.1-mM phenylmethylsulfonylfluoride, 5 g/ml leupeptin, 1 g/ml pepstatin A and 1% Triton X-100 for 30 min at 4 °C. Cell extracts were then recovered after centrifugation and protein concentration was determined according to Bradford (1976). Extracts were stored in 50% glycerol at ⫺70 °C. For the assay of caspase activity, aliquots of protein (30 g) were added to 100-M DEVD-pNA or YVAD-pNA in 50-mM Tris–HCl, pH 7.4 containing 10-mM DTT in a total volume of 0.5 ml. The assay was carried out at 37 °C and the release of p-nitroaniline was recorded spectrophotometrically at 405 nm (Du et al., 1997). Each measurement was carried out at least in triplicate. Immunoblot analysis. For immunoblot analysis, cells (4 – 6 million) were cultured for 48 –72 h both in the absence and in the presence of AICA riboside. Cells were lysed for 30 min at 4 °C in lysis buffer (20-mM Tris–HCl, pH 7.4, 150-mM NaCl, 1-mM EGTA,
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 1. 5'-Aminoimidazole-4-carboxamide (AICA) riboside induces apoptosis in undifferentiated human neuroblastoma cells. (A) Cells have been incubated with different concentrations of AICA riboside for 72 h and viability measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay as described in Experimental Procedures. (B) Cells have been incubated without (a) or with (b) 200-M AICA riboside for 72 h, fixed, and stained with propidium iodide. The arrow indicates one of the cells with apoptotic morphology. Scale bar⫽50 m.
5-mM EDTA, 1-mM DTT, 1-mM sodium orthovanadate, 10-mM NaF, 10-mM sodium pyrophosphate, 0.1-mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml pesptatin, 1% (v/v) Triton X-100). Extracts were centrifuged at 14,000⫻g for 30 min at 4 °C, and the protein content was measured using the method of Bradford (1976). The supernatants were mixed with SDS sample buffer (3% w/v SDS, 0.05% w/v Bromophenol Blue, 10% w/v glycerol, 1% v/v ␤-mercaptoethanol in 0.1-M Tris–HCl, pH 6.8) and placed in a boiling water bath for 5 min. Fifty micrograms of proteins were applied in a 15% (for caspase-8 detection) and 12% (for Bcl-2 detection) SDS-polyacrylamide gel, and then transferred to a polyvinylidene difluoride membrane. The membrane was incubated overnight in 5% nonfat milk at room temperature and then 2 h in the presence of anti Bcl-2 antibody, or anti-caspase-8 antibody. The membrane was then exposed for 2 h to horseradish peroxidase-conjugated secondary antibody and immunoreactive protein was visualized using a chemiluminescence-detection kit. The membrane used for Bcl-2 detection was then stripped and re-used for the detection of actin. For the quantification of Bcl-2, the X-ray films were scanned using an image analysis system (GS-670, Bio-Rad, Milan, Italy). Immunohistochemistry. Cells have been grown on coverslips coated with laminin in the presence or the absence of 200-M AICA riboside for 4 days. Cells were fixed for 15 min with 4% paraformaldehyde and washed three times with PBS, blocked with 1% bovine serum albumin (BSA) for 30 min, and incubated overnight at 4 °C with mouse monoclonal antibody antineurofilament 200 in 1% BSA and 0.05% Triton X-100. Cells were washed three times with PBS and incubated with antimouse IgG-TRITC conjugate for 1 h at room temperature, washed three times with PBS and mounted with Immunofluor mounting medium (ICN, Irvine, CA, USA). Images of NF200 immunofluorescence were acquired using a 60⫻ immersion objective in a Nikon inverted microscope Eclipse TE 300 equipped with a confocal scanning system Radiance-Plus (Bio-Rad). Images were acquired with Laser-Sharp software and analyzed with Adobe Photoshop. Propidium iodide staining. Cells were grown on coverslips as above, fixed with paraformaldehyde, and stained with 3-M
propidium iodide for 1 h at room temperature. After two washes with PBS, they were mounted with Immunofluor mounting medium. Images were acquired using a 40⫻ immersion objective in a confocal scanning system as described above. AMPK assay. Cells (4 – 6⫻106/dish) were cultured in the presence or the absence of 200-M AICA riboside. After 2 or 48 h, cells were scraped from plates, and homogenized with a glass/ teflon homogenizer in a buffer containing 50-mM Tris–HCl, pH 7.4, 250-mM mannitol, 1-mM EGTA, 1-mM EDTA, 1-mM DTT, 50-mM NaF, 1-mM sodium pyrophosphate, 0.1-mM phenylmethylsulfonyl fluoride, 0.1-mM benzamidine and 5 g/ml soybean trypsin inhibitor and centrifuged at 14,000⫻g for 3 min. The activity of AMPK was assayed at 37 °C by the phosphorylation of the peptide substrate SAMS, essentially according to Salt et al. (1998). Peptide phosphorylation was assayed in a final volume of 50 l containing 62.5-mM HEPES (sodium salt), pH 7.0, 62.5-mM NaCl, 62.5-mM NaF, 1.25-mM sodium pyrophosphate, 1.25-mM EDTA, 1.25-mM EGTA, 1-mM DTT, 0.1-mM benzamidine, 0.1-mM phenylmethylsulfonyl fluoride, 5 g/ml soybean trypsin inhibitor, 20-mM MgCl2, 200-M SAMS peptide, 200-M AMP, and 200-M [␥32P]ATP (600 d.p.m./pmol). A control was performed in which AMP was omitted from the reaction mixture. Reactions were performed in triplicate using both 1 and 2 g of crude extract protein. At 0, 10 and 20 min incubation, aliquots of 15 l were spotted on phosphocellulose paper treated as described by Davies et al. (1989), and then counted for radioactivity. One unit of enzyme activity is the amount of enzyme which catalyzes the conversion of 1 mol of substrate per minute under the adopted experimental conditions.
RESULTS AICA riboside induces apoptosis in SH-SY5Y human neuroblastoma cells. Human neuroblastoma cells were cultured for 72 h with different concentrations of AICA riboside. Concentrations higher than 100 M reduced cellular viability (Fig. 1A) inducing morphological changes typical of apoptosis (Fig. 1B), since cells lost their neurites
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 2. Caspase-3 activity in cell extracts. Caspase-3 activity was measured in extracts obtained after incubation of cells without (white columns) and with (gray columns) 200-M 5'-aminoimidazole-4-carboxamide (AICA) riboside. Data are mean⫾S.E.M. of at least three measurements. Significance was determined by means of unpaired t-test. Statistical significance versus samples from cells grown in the absence of AICA riboside: ** P⬍0.01; *** P⬍0.001.
and became rounded with shrunken cytoplasm and condensed and fragmented nuclei showing intense fluorescence. Caspase activity after AICA riboside treatment. Cellular extracts obtained after incubation of cells, at different times, with AICA riboside were tested in order to ascertain whether the treatment was able to activate different caspases. Immunoblot analysis with anti caspase-8 antibody revealed that caspase-8 was not proteolytically activated following 48 h of AICA-riboside treatment (data not shown). A significant increase of caspase-3 activity was measured in extracts obtained after 48 h of incubation and was still higher than control by 72 h (Fig. 2). No significant increase of caspase-1 activity was found when measured at 24, 48 or 72 h after the addition of AICA riboside using YVAD-pNA as substrate (data not shown).
Fig. 3. Effect of dipyridamole on 5'-aminoimidazole-4-carboxamide (AICA) riboside toxicity. Cells were incubated for 72 h without (white columns) and with (gray columns) 200-M AICA riboside in the absence or in the presence of 1-M dipyridamole. Viability was measured as percentage of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reduction as described in Experimental Procedures. Data are mean⫾S.E.M. (n⫽7). For this and the following figures, statistical significance was determined by means of Mann-Whitney test. Statistical significance versus cells grown in the absence of additives: ** P⬍0.01; *** P⬍0.001.
Effect of dipyridamole on AICA ribose toxicity. In order to ascertain whether AICA riboside needs to enter the cell to affect viability, cells have been pre-incubated for 30 min with 1-M dipyridamole and then incubated with 200-M AICA riboside and dipyridamole for 72 h. Viability was measured by MTT assay (Fig. 3). In these conditions, dipyridamole slightly decreased viability compared with control cells. However, dipyridamole had a protective effect on AICA riboside-treated cells since viability increased up to 81% of the control (Fig. 3). Effect of 5'amino-5'dAdo on AICA ribose toxicity. In order to ascertain whether AICA riboside needs to be phosphorylated to affect viability, cells have been preincubated for 30 min with 50-M 5'amino-5'dAdo; then, 200-M AICA riboside was added and the incubation was performed for 72 h. Viability was measured by MTT assay (Fig. 4). In these conditions, 5'amino-5'dAdo did not affect viability compared with control cells, but exerted a protective effect on AICA riboside-treated cells, since viability increased up to 83% of that of control cells (Fig. 4).
Fig. 4. Effect of 5'amino-5'deoxyadenosine (5'amino-5'dAdo) on 5'aminoimidazole-4-carboxamide (AICA) riboside toxicity. Cells were incubated for 72 h without (white columns) and with (gray columns) 200-M AICA riboside in the absence or in the presence of 50-M 5'amino-5'dAdo. Viability was measured as percentage of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reduction as described in Experimental Procedures. Data are average⫾S.E.M. (n ⫽3). Significance versus cells grown without additives: **P⬍0.01; *** P⬍0.001. The significance of AICA riboside plus 5'amino-5'dAdotreated cells versus AICA riboside-treated cells was P⬍0.001.
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 5. Effect of 5'-aminoimidazole-4-carboxamide (AICA)-riboside on 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced differentiation. (A) Micrographs of cultures stained with propidium iodide after 3 days of treatment with no additives (a), 200-M AICA riboside (b); 16-nM TPA (c); AICA riboside and TPA (d). Scale bar⫽50 m. (B) Cells were incubated for 72 h without (white columns) and with (gray columns) 200-M AICA riboside in the absence and in the presence of 16-nM TPA. Viability was measured as percentage of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reduction as described in Experimental Procedures. Data are average⫾S.E.M. (n⫽3). Significance versus cells grown without additives: * P⬍0.05; *** P⬍0.001.
AMPK activity. Since AICA riboside is a well-characterized activator of AMPK (Corton et al., 1995), we measured AMPK activity in cultures treated with 200-M AICA riboside either for short incubation (2 h) or long incubation periods (48 h). AMPK activity in control cells was 0.025⫾0.010 mU/mg and that of AICA riboside-treated cells for 2 h was approximately five-fold higher, 0.123⫾0.016 (mean⫾S.E.M., n⫽3, P⬍0.01). However, no statistical significant difference between control and AICA riboside-treated samples was found when AMPK activity was measured after 48 h. Effect of AICA riboside on TPA-induced differentiation. After 3 days of treatment with 16-nM TPA, density of cells was higher than control, and viability measured as MTT reduction increased by 23% (Fig. 5B). Morphology of cells changed showing shrinkage of cell body and extension of neurites (Fig. 5A). When TPA was added simultaneously with AICA riboside, a complete protection against apoptosis was observed (Fig. 5A, B), as shown by MTT reduction assay, since viability increased from 52% (AICA riboside in the absence of TPA) to 122% (in the presence of TPA) and by propidium iodide stain of the cells. Lower concentrations of TPA (3.2 nM) partially protected cells from death because co-treatment of cells with AICA riboside induced a slight decrease of cell viability by 30% (data not shown) when compared with TPA alone.
Effect of AICA riboside on RA-induced differentiation. To study whether AICA riboside was able to interfere with the induction of differentiation, it was added simultaneously with RA, one of the inductors of differentiation in SH-SY5Y human neuroblastoma cells (Pahlman et al., 1990). The percentage of MTT reduction of cells treated with RA alone was 64⫾3% of control when measured after 72 h (Fig. 6). Phase-contrast microscopy demonstrated that this decrease in viability in RA-treated cells was mostly due to decrease in proliferation and not to apoptosis (data not shown). Viability of cells incubated for 72 h with AICA riboside was similar in both control and RA-treated cells (Fig. 6). In fact, AICA riboside reduced viability to 27⫾2% in control cells, while viability of AICA riboside⫹RA-treated cells was 39⫾3% that of cells treated with RA alone. Therefore, when AICA riboside was added together with RA, its effect on cell viability was only slightly decreased by the presence of the differentiating agent. Effect of AICA riboside on RA-differentiated cells. Cells were induced to differentiate with 10-M RA for 5 days and then treated with 200-M AICA riboside for 72 h in the presence of RA. The viability was measured by MTT assay and morphology was observed after immunohistochemistry (Fig. 7). Viability of AICA riboside-treated cells was 83⫾4% that of RA-treated cells. Therefore AICA riboside had a lesser effect on differentiated cells than on undifferentiated cells (compare Fig. 6 and Fig. 7). Under our
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 6. Effect of 5'-aminoimidazole-4-carboxamide (AICA) riboside on all-trans-retinoic acid (RA)-induced differentiation. Cells were incubated for 72 h without (white columns) and with (gray columns) 200-M AICA riboside in the absence or in the presence of 10-M RA. Viability was measured as percentage of 3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide reduction as described in Experimental Procedures. Data are average⫾S.E.M. (n⫽3). Significance versus cells grown without additives: *** P ⬍0.001.
experimental conditions, the population of SH-SY5Y was composed mainly of the neuroblastic cells (N-type, arrowhead in Fig. 7c) and a low proportion of substrate adherent cells (S-type, arrow in Fig. 7c). RA treatment induced S-cells to proliferate and N-cells to differentiate toward a more neuronal phenotype by extending neurite processes. In other experiments, SH-SY5Y neuroblastoma cells were pretreated for 5 days with RA in complete medium and then switched to serum-free medium containing 50 ng/ml BDNF (Encinas et al., 2000) and AICA riboside. BDNF induced extension of neurites, that formed a network on the plate by 24 h (Fig. 8A). Treatment with AICA riboside in the presence of BDNF for 72 h decreased viability by 20%, without a significant effect on cellular morphology (Fig. 8). Longer incubation (up to 7 days) with BDNF and AICA riboside did not modify neurite extension (data not shown). Bcl-2 levels in differentiated and undifferentiated cells. The expression of the antiapoptotic protein Bcl-2 was determined by immunoblot analysis in undifferentiated and differentiated cells, incubated both in the absence and in the presence of 200-M AICAR for 72 h. Differentiation induced an overexpression of Bcl-2; indeed, the densitometric analysis of the X-ray films indicated a two-fold increase in the intensity of the band (n⫽3, P ⬍0.01) (compare lane one and lane three of Fig. 9). The incubation with AICA riboside did not decrease the level of expression of Bcl-2, both in undifferentiated and differentiated cells.
DISCUSSION In this paper, we confirm and extend our previous observation on the toxic effect of AICA riboside on neural cells
(Pesi et al., 2000) using, as a model of neuronal differentiation, SH-SY5Y, an adrenergic clone of the human neuroblastoma cell line SK-N-SH (Biedler et al., 1973). The analysis of the cells treated with AICA riboside and stained with propidium iodide showed the characteristic morphology associated to apoptosis, such as nuclear condensation and cell shrinkage. This confirms our previous finding of DNA fragmentation in AICA riboside-treated cells (Pesi et al., 2000). The involvement of the apoptotic program in the neuronal death induced by AICA riboside is also sustained by the activation of caspase-3, which reached maximum activation after 48 h. Caspase-3 is an effector caspase that is activated in most, if not all, caspase pathways. To test whether other caspases were affected by AICA riboside treatment, the activation of caspase-1 and caspase-8 was also evaluated. Caspase-1 activation occurs in models of cerebral ischemia and trauma, amyotrophic lateral sclerosis and Huntington disease specimens (Friedlander, 2000). The experiments performed indicate that caspase-1 and 8 were not activated in SH-SY5Y cells treated with AICA riboside. In particular, the lack of activation of caspase-8 seems to exclude an involvement of death receptors in the mechanism of AICA riboside-induced apoptosis. Although the mechanism must be further investigated, our results suggest that AICA riboside, to exert its neurotoxic action, must enter the cells, since dipyridamole, a well-known inhibitor of nucleoside membrane carriers (Miras-Portugal et al., 1986; Wang et al., 1992), significantly protects from apoptosis. In addition, the significant reduction in the cytotoxic action of AICA riboside when the compound is added with 5'amino-5'dAdo, a known inhibitor of adenosine kinase (Miller et al., 1979), implies that the riboside must be phosphorylated in order to trigger the apoptotic machinery. Since AICA ribotide is a well-known activator of the AMPK (Corton et al., 1995), we have measured AMPK activity after AICA riboside treatment. A significant increase of AMPK activity was measured 2 h after AICA riboside incubation, whereas no difference compared with control was obtained after 48-h incubation, when apoptotic markers, such as morphological features and caspase-3 activation, were evident. In contrast with this report, other authors have found a protection of AICA riboside against apoptosis induced by different stimuli. For instance, AICA riboside prevents apoptosis induced by glucose deprivation in hippocampal neurons (Culmsee et al., 2001), through AMPK activation. However, concentrations 1–100 M of AICA riboside are protective against glucose deprivation, while higher concentrations are not. Stefanelli et al. (1998) reported that AICA riboside activated AMPK in thymocytes, and protected from apoptosis induced by dexamethasone, but not by staurosporine. A protective effect of AICA riboside has been also reported against apoptosis induced by fatty acids in astrocytes (Blazquez et al., 2001), and by hyperglycemia in human umbilical-vein endothelial cells (Ido et al., 2002). The molecular basis of the relationship, if any, between apoptosis and AMPK activation is presently unclear. In our model, whereas an early activation of AMPK was observed, pro-
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 7. Effect of 5'-aminoimidazole-4-carboxamide (AICA) riboside on all-trans-retinoic acid (RA)-differentiated cells. (A) Cells were induced to differentiate with 10-M RA for 5 days and then treated without (panel a) or with (panel b) 200-M AICA riboside for 72 h in the presence of RA. Immunohistochemistry was performed as described in Experimental Procedures. Panel c: phase-contrast micrograph of a field containing S-cells (arrow) and N-cells (arrowhead), that have been induced to differentiate with 10-M RA for 5 days. Scale bars: b (same as a), 10 m; c, 50 m. (B) Cells were induced to differentiate with 10-M RA for 5 days and then treated without (white column) or with (gray column) 200-M AICA riboside for 72 h in the presence of RA. Viability was measured as percentage of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reduction as described in Experimental Procedures. Data are average⫾S.E.M. (n ⫽3). Significance: *** P⬍0.001.
longed incubation with AICA riboside, at a concentration able to induce apoptosis, did not result in a significant activation of AMPK. Although the mechanism underlying the difference observed between our and other models needs further investigation, it is clear that the effect of AICA riboside depends on its concentration, on the time of exposure, on apoptotic stimuli and on the cellular type. We have found that AICA riboside still affects the viability of cells induced to differentiate by RA, while its effect is significantly reduced when the compound is added to cells differentiated toward a cholinergic phenotype by a previous addition of RA. Although derived from a neuroblastic subclone of the SK-N-SH cell line, SH-SY5Y neuroblastoma cells contain a low proportion of substrate adherent (S-type) cells, with morphological and biochemical characteristics different from neuroblastic cells (N-type) (Arcangeli et al., 1999; Encinas et al., 2000). Under our experimental conditions, the population of SH-SY5Y was mostly of the N-type. RA induced extension of neurites in N-cells and proliferation of S-cells. When treatment with RA was prolonged in order to investigate whether the time
course of the effect of AICA riboside was somehow delayed in differentiating compared with undifferentiated cells, the number of S-cells increased, making difficult the analysis of the effect of AICA riboside on viability of neuronal cells. To overcome this problem, we exposed SHSY5Y cells sequentially to RA and BDNF in serum-free medium, a procedure which, as described by Encinas et al. (2000) yields nearly pure populations of human neuron-like cells from the SH-SY5Y neuroblastoma cell line. Indeed, in our experimental conditions, BDNF induced extension of neurites and the treatment with AICA riboside in the presence of BDNF for 72 h reduced viability to 80% of control value. This decrease in MTT reduction was not accompanied by any evident neurite retraction. Further investigation is needed to verify whether prolonged treatment with AICA riboside might involve changes in events related to the synaptic function such as dopamine content in the vesicles or changes in dopamine release after stimulation. The fact that AICA riboside has a minor effect on differentiating cells compared with undifferentiated ones could be explained, at least in part, by the raise in anti-
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 8. Effect of 5'-aminoimidazole-4-carboxamide (AICA) riboside on brain-derived neurotrophic factor (BDNF)-mediated differentiation of all-transretinoic acid (RA)-pretreated cells. (A) Phase-contrast micrographs of cultures exposed for 5 days to 10-M RA in complete medium followed by 3 days in serum-free medium containing 50 ng/ml BDNF (left) or BDNF plus 200-M AICA riboside (right). Scale bar⫽50 m. (B) Cells were treated as described in A and viability measured as percentage of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide reduction as described in Experimental Procedures. Data are average⫾S.E.M. (n⫽3). Significance: *** P⬍0.001.
apoptotic protein expression occurring in SH-SY5Y cells during differentiation. In order to test it, we have measured the expression of the antiapoptotic protein Bcl-2 before and after differentiation. Differentiated cells, in accordance with previous reports (Hanada et al., 1993; Lasorella et al., 1995), show an increase in Bcl-2 expression, which correlates with the resistance against the apoptotic effect induced by AICA riboside. However, the treatment with AICA riboside did not result in any significant effect on the level of Bcl-2, both in undifferentiated and differentiated cells. AICA riboside reduced cellular viability when differentiation was induced by RA but not by TPA. The different effect of AICA riboside on cell viability of RA- and TPAtreated cells can be correlated to the different degree of differentiation induced by these agents. It is known that cells differentiated with RA have a less mature phenotype compared with TPA (Pahlman et al., 1990). In addition,
Hanada et al. (1993) reported a raise of Bcl-2 expression detectable after 1 day of addition of TPA in SH-SY5Y cells, when no increase in the length of neurites is visible. On the other hand, in RA-treated cells, a 20 – 40-fold increase in the level of Bcl-2 has been measured after 6 days (Lasorella et al., 1995). Our results are compatible with a slower time course of Bcl-2 expression in RA-induced differentiation compared with TPA. Overall, our results seem to indicate that AICA riboside exerts a neurotoxic action on undifferentiated cells as a trigger of the apoptotic program. On the other hand, the compound has a smaller effect on the neuronal cells differentiated toward cholinergic phenotype and does not exert an appreciable action on noradrenergic-differentiated cells. Our results could be relevant in the understanding of the basis of the neurological impairments of Lesch-Nyhan
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Fig. 9. Effect of 5'-aminoimidazole-4-carboxamide (AICA) riboside on Bcl-2 expression. Undifferentiated and differentiated cells (preincubated with all-trans-retinoic acid for 5 days) were treated with 200-M AICA riboside for 72 h as described in Experimental Procedures. Proteins were subjected to electrophoresis and Bcl-2 expression was evaluated by immunoblotting. The figure shows a representative immunoblotting. Lanes one and two: Bcl-2 in undifferentiated cells, before and after treatment with AICA riboside, respectively. Lanes three and four: Bcl-2 in differentiated cells, before and after treatment with AICA riboside, respectively. Western blotting of actin was performed as a loading control.
syndrome and/or other neurological disorders of unknown etiology, where a high 5'-nucleotidase activity has been found (Pesi et al., 2000; Page et al., 1997). The accumulation of AICA riboside at the early stages of brain development, due to an accelerated rate of purine de novo biosynthesis and/or an overexpression of cytosolic 5'-nucleotidase might produce devastating effects. Although results from cell lines of tumoral origin might not correspond to the effects of AICA riboside in the whole brain, our results are intriguing and it will be interesting to test the effect of AICA riboside during brain development. These experiments could provide some light on the basis of neurological manifestations of syndromes caused by purine dismetabolisms. Acknowledgements—This work was supported by grants from Italian CNR (Target Project “Biotecnologie”) and from Italian MURST.
REFERENCES Abbud W, Habinowsky S, Zhang JZ, Kendrew J, Elkairi FS, Kemp BE, Witters LA, Ismail-Beigi F (2000) Stimulation of AMP-activated protein kinase is associated with enhancement of Glut1-mediated glucose transport. Arch Biochem Biophys 380:347–352. Arcangeli A, Rosati B, Crociani O, Cherubini A, Fontana L, Passani B, Wanke E, Olivotto M (1999) Modulation of HERG current and herg gene expression during retinoic acid treatment of human neuroblastoma cells: potentiating effects of BDNF. J Neurobiol 40:214 – 255. Biedler JL, Helson L, Spengler BA (1973) Morphology and growth, tumorigenicity, and cytogenetics of human neuroblastoma cells in continuous culture. Cancer Res 33:2643–2652. Blazquez C, Geelen MJH, Velasco G, Guzman M (2001) The AMPactivated protein kinase prevents ceramide synthesis de novo and apoptosis in astrocytes. FEBS Lett 489:149 –153. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 72:248 –254. Corton JM, Gillespie JG, Hawley SA, Hardie DG (1995) 5-Aminoimidazole-4-carboxamide ribonucleoside: a specific method for activating AMP-activated protein kinase in intact cells? Eur J Biochem 229:558 –565.
Culmsee C, Monnig J, Kemp BE, Mattson MP (2001) AMP-activated protein kinase is highly expressed in neurons in the developing rat brain and promotes neuronal survival following glucose deprivation. J Mol Neurosci 17:45–58. Davies SP, Carling D, Hardie DG (1989) Tissue distribution of the AMP-activated protein kinase, and lack of activation by cyclic-AMPdependent protein kinase, studied using a specific and sensitive peptide assay. Eur J Biochem 186:123–128. Du Y, Bales KR, Dodel RC, Hamilton-Byrd E, Horn JW, Czilli DL, Simmons LK, Ni B, Paul SM (1997) Activation of a caspase 3-related cysteine protease is required for glutamate-mediated apoptosis of cultured cerebellar granule neurons. Proc Natl Acad Sci USA 94:11657–11662. Encinas M, Iglesia M, Liu Y, Wang H, Muhaisen A, Cen˜a V, Gallego C, Comella JX (2000) Sequential treatment of SH-SY5Y cells with retinoic acid and brain-derived neurotrophic factors gives rise to fully differentiated, neurotrophic-factor-dependent human neuronlike cells. J Neurochem 75:991–1003. Friedlander RM (2000) Role of caspase 1 in neurologic disease. Arch Neurol 57:1273–1276. Hanada M, Krajewski S, Tanaka S, Cazals-Hatem D, Spengler BA, Ross RA, Biedler JL, Reed JC (1993) Regulation of Bcl-2 oncoprotein levels with differentiation of human neuroblastoma cells. Cancer Res 53:4978 –4986. Ido Y, Carling D, Ruderman N (2002) Hyperglycemia-induced apoptosis in human umbilical vein endothelial cells: inhibition by the AMPactivated protein kinase activation. Diabetes 51:159 –167. Lasorella A, Iavarone A, Israel MA (1995) Differentiation of neuroblastoma enhances Bcl-2 expression and induces alterations of apoptosis and drug resistance. Cancer Res 55:4711–4716. Miller RL, Adamczyk DL, Miller WH, Koszalka GW, Rideout JL, Beacham LM III, Chao EY, Haggerty JJ, Krenitsky TA, Elion GB (1979) Adenosine kinase from rabbit liver: II. Substrate and inhibitor specificity. J Biol Chem 254:2346 –2352. Miras-Portugal MT, Torres M, Rotllan P, Aunis D (1986) Adenosine transport in bovine chromaffin cells in culture. J Biol Chem 261: 1712–1719. Muoio DM, Seefeld K, Witters LA, Coleman RA (1999) AMP-activated kinase reciprocally regulates triacylglycerol synthesis and fatty acid oxidation in liver and muscle: evidence that sn-glycerol-3-phosphate acyltransferase is a novel target. Biochem J 338:783–791. Page T, Coleman M (2000) Purine metabolism abnormalities in a hyperuricosuric subclass of autism. Biochim Biophys Acta 1500: 291–296. Page T, Yu A, Fontanesi J, Nyhan WL (1997) Developmental disorder associated with increased cellular nucleotidase activity. Proc Natl Acad Sci USA 94:11601–11606. Pahlman S, Mamaeva S, Meyerson G, Mattsson MEK, Bjelfman C, Ortoft E, Hammerling U (1990) Human neuroblastoma cells in culture: a model for neuronal cell differentiation and function. Acta Physiol Scand 140:25–37. Pesi R, Micheli V, Jacomelli G, Peruzzi L, Camici M, Garcia-Gil M, Allegrini S, Tozzi MG (2000) Cytosolic 5'-nucleotidase hyperactivity in erythrocytes of Lesch-Nyhan syndrome patients. Neuroreport 11:1827–1831. Salt IP, Johnson G, Ashcroft SJH, Hardie DG (1998) AMP-activated protein kinase is activated by low glucose in cell lines derived from pancreatic ␤ cells, and may regulate insulin release. Biochem J 335:533–539. Sidi Y, Mitchell BS (1985) Z-nucleotide accumulation in erythrocytes from Lesch-Nyhan patients. J Clin Invest 76:2416 –2419. Stefanelli C, Stanic I, Bonavita F, Flamigni F, Pignatti C, Guarnieri C, Caldarera CM (1998) Inhibition of glucocorticoid-induced apoptosis with 5-aminoimidazole-4-carboxamide ribonucleoside, a cell-permeable activator of AMP-activated protein kinase. Biochem Biophys Res Comm 243:821–826.
M. Garcia-Gil et al. / Neuroscience 117 (2003) 811– 820
Visser JE, Ba`r PR, Jinnah HA (2000) Lesch-Nyhan disease and the basal ganglia. Brain Res Rev 32:449 –475. Wang T, Mentzer RM, van Wylen DGL (1992) Interstitial adenosine with dipyridamole: effect of adenosine receptor blockade and adenosine deaminase. Am J Physiol 263:H552–H558. Winder WW, Hardie DG (1999) AMP-activated protein kinase, a met-
abolic master switch: possible roles in type 2 diabetes. Am J Physiol 277:E1–E10. Young ME, Radda GK, Leighton B (1996) Activation of glycogen phosphorylase and glycogenolysis in rat skeletal muscle by AICAR, an activator of AMP-activated protein kinase. FEBS Lett 382:43–47.
(Accepted 11 October 2002)