The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

Contents lists available at ScienceDirect

The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel

Cell proliferation and drug sensitivity of human glioblastoma cells are altered by the stable modulation of cytosolic 5 -nucleotidase II F. Cividini a,∗ , E. Cros-Perrial d,e,f,g,h , R. Pesi a , C. Machon c , S. Allegrini b , M. Camici a , C. Dumontet d,e,f,g,h , L.P. Jordheim d,e,f,g,h,1 , M.G. Tozzi a,1 a

University of Pisa, Department of Biology, Biochemistry Unit, Pisa, Italy University of Sassari, Department of Chemistry and Pharmacology, Sassari, Italy c Hospices Civils de Lyon, Centre Hospitalier Lyon-Sud, Laboratoire de Biochimie et Toxicologie, Lyon, France d Université de Lyon, F-69000 Lyon, France e Université de Lyon 1, F-69622 Lyon, France f Université de Lyon 1, F-69000 Lyon, France g INSERM U1052, Centre de Recherche en Cancérologie de Lyon, F-69000 Lyon, France h CNRS UMR 5286, Centre de Recherche en Cancérologie de Lyon, F-69000 Lyon, France b

a r t i c l e

i n f o

Article history: Received 25 February 2015 Received in revised form 31 May 2015 Accepted 8 June 2015 Available online 14 June 2015 Keywords: Cytosolic 5 -nucleotidase II Stable protein modulation Human glioblastoma ADF cells Cell proliferation Anticancer drugs Nucleotide metabolism

a b s t r a c t Cytosolic 5 -nucleotidase II (cN-II) has been reported to be involved in cell survival, nucleotide metabolism and in the cellular response to anticancer drugs. With the aim to further evaluate the role of this enzyme in cell biology, we stably modulated its expression the human glioblastoma cell ADF in which the transient inhibition of cN-II has been shown to induce cell death. Stable cell lines were obtained both with inhibition, obtained with plasmids coding cN-II-targeting short hairpin RNA, and stimulation, obtained with plasmids coding Green Fluorescence Protein (GFP)-fused wild type cN-II or a GFP-fused hyperactive mutant (GFP-cN-II-R367Q), of cN-II expression. Silenced cells displayed a decreased proliferation rate while the over expressing cell lines displayed an increased proliferation rate as evidenced by impedance measurement using the xCELLigence device. The expression of nucleotide metabolism relevant genes was only slightly different between cell lines, suggesting a compensatory mechanism in transfected cells. Cells with decreased cN-II expression were resistant to the nucleoside analog fludarabine confirming the involvement of cN-II in the metabolism of this drug. Finally, we observed sensitivity to cisplatin in cN-II silenced cells and resistance to this same drug in cN-II over-expressing cells indicating an involvement of cN-II in the mechanism of action of platinum derivatives, and most probably in DNA repair. In summary, our findings confirm some previous data on the role of cN-II in the sensitivity of cancer cells to cancer drugs, and suggest its involvement in other cellular phenomenon such as cell proliferation. Published by Elsevier Ltd.

1. Introduction

Abbreviations: NT5C2, cytosolic 5 -nucleotidase II gene; PRPS1, phosphoribosyl pyrophosphate synthetase 1 gene; HGPRT1, hypoxanthine-guanine phosphoribosyl transferase 1 gene; PNP, purine nucleoside phosphorylase gene; ADA, adenosine deaminase gene; AMPD3, adenosine monophosphate deaminase 3 gene; RRM1, ribonucleotide reductase M1 gene; RRM2, ribonucleotide reductase M2 gene; PPATT, phosphoribosyl pyrophosphate amidotransferase gene; AMP-CP, alpha, beta-methyleneadenosine 5 -diphosphate; H2AX, H2A histone family, member X; BPG, 2,3-bisphosphoglycerate; 6MP, 6-mercaptopurine; MMC, mitomycin C; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; PEI, polyethylenimine; ALL, acute lymphoblastic leukemia. ∗ Corresponding author at: Via San Zeno, 51, 56127, Italy. Tel.: +39 050 2211450; fax: +39 050 2211460. E-mail address: [email protected] (F. Cividini). 1 These authors contributed equally to senior authorship. http://dx.doi.org/10.1016/j.biocel.2015.06.011 1357-2725/Published by Elsevier Ltd.

The IMP/GMP cytosolic 5 -nucleotidase II (cN-II, EC 3.1.3.5), is the most conserved cytosolic nucleotidase through evolution and the one with the highest level of regulation (Tozzi et al., 2013). During the last decades cN-II has been intensively studied and shown to be an allosterically regulated bifunctional enzyme (Pesi et al., 1994, 2010; Allegrini et al., 2001) belonging to the haloacids dehalogenase superfamily (Allegrini et al., 2004; Bretonnet et al., 2005). Its activity relies on the formation of a phospho-enzyme intermediate where the phosphate is either transferred to a nucleoside (phosphotransferase activity) or released by hydrolysis (Allegrini et al., 2001). The structure of cN-II has been described both alone and in presence of various substrates and effectors, as well as with an inhibitor (Walldén et al., 2007; Walldén and Nordlund, 2011;

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

Jordheim et al., 2013; Cividini et al., 2015a). The expression of cN-II is ubiquitous and there is a very low variability of enzymatic activity between different cells, organs and individuals (Spychala et al., 1988; Itoh, 1981), except for high proliferative tissues where cNII activity is high (Itoh et al., 1992). The physiological role of this enzyme is to contribute to the homeostasis of IMP and more globally of intracellular purine derivatives (Supplemental Fig. 1). Other enzymes involved in the same pathway, such as purine nucleoside phosphorylase (PNP), are not as regulated, conserved and stably expressed as cN-II (Tozzi et al., 2013). A growing amount of clinical and preclinical data has shown that cN-II is a prognostic factor and predictive of outcome for chemotherapy-treated patients with neoplasia (Jordheim and Chaloin, 2013). One approach to gain a better understanding of the role of cNII in physiological and pathological conditions is the development of cellular models where cN-II is silenced or overexpressed. The enzyme was inducibly overexpressed in Human Embrionic Kidney 293 cells both alone and fused to the C-terminal end of GFP (Gazziola et al., 1999). Overexpression induced in both cases a decrease of nucleotide triphosphates and an increase of duplication time (Rampazzo et al., 1999). When stably inhibited in human cancer cells from hematologic origin, no difference in proliferation or distribution in cell cycle was observed (Jordheim et al., 2015, in press), whereas these cells were more sensitive to nucleoside analogs such as fludarabine (Supplemental Fig. 1). When expressed in Saccharomyces cerevisiae, cN-II caused a decrease of energy charge and an impairment in DNA repair (Allegrini et al., 2013). In 2008, we reported on the knock-down of cN-II using shRNA and the analysis of associated cellular modifications using human astrocytoma cells (ADF) (Careddu et al., 2008). These cells were chosen since alterations of purine metabolism have been associated with severe neurological defects, thus pointing out a fundamental role played by purine compounds in central nervous system (Camici et al., 2010; Micheli et al., 2011). cN-II reduction in ADF cells was associated with a decrease of cell viability that was dependent on the activation of the apoptotic pathway with both modifications of cell morphology and increased caspase-3 activity. However, no significant changes in nucleotide pools were reported. It was therefore evident that the cell death was not a consequence of a gross alteration of intracellular purine content. ADF cell is a model of glioblastoma, a devastating brain tumor particularly resistant to chemotherapy and to apoptosis since they express a mutated inactive caspase-9 (Ceruti et al., 2005). As a consequence any alteration inducing apoptosis in these cells is of particular interest for the design of an efficient therapy. Overall, available information indicates that cN-II plays a fundamental role in nucleotide and drug metabolism that can, in part, be due to its role in determining intracellular concentrations of IMP and PRPP, but indicates that the enzyme can also be involved in some other central regulatory pathway. With the aim to continue investigating the alteration of cN-II expression in human glioblastoma ADF cells, we generated a series of models with stable modulation of cN-II expression. These cell models were used to investigate cell proliferation, the expression of nucleotides and nucleosides metabolism relevant genes, nucleotides and nucleosides content and sensitivity to different chemotherapic drugs.

2. Material and methods 2.1. Cell culture Cells were grown in RPMI medium supplemented with 10% FBS, 2 mM l-glutamine, 100 U/ml penicillin, 100 ␮g/ml streptomycin and, when transfected, once per week with 0.8 mg/ml geneticin. Cells were kept at 37 ◦ C in a humidified 5% CO2 /95% air atmosphere.

223

2.2. Cell transfection Plasmids were prepared as described elsewhere (Jordheim et al., 2015, in press), expanded from Escherichia coli DH5␣ and purified using the The PureLink® HiPure Plasmid Filter Maxiprep Kit (Invitrogen). ADF cells were transfected with pScN-II, pScont, pcDNA3, pcDNA3-GFP, pcDNA3- GFP/cN-II, pcDNA3-GFP/cN-II-R367Q using lipofectin. Briefly, cells (200,000 cells per flask) were incubated for 5 h with 5 ␮g of plasmid in presence of 12.5 ␮g of lipofectin (Invitrogen). After 72 h of growth in complete medium stably transfected cells were selected with 0.8 mg/ml neomycin. Transgene expression and western and immune blot analyses was performed approximately after one month of selection with neomycin. 2.3. cN-II phosphatase assay Cells were harvested and proteins were extracted with 250 ␮l 100 mM Tris HCl pH 7.4 and transferred in a 1.5 ml tube. Crude extracts were obtained by 3 freeze/thaw cycles followed by centrifugation at 10,000 × g at 4 ◦ C for 40 min to pellet the cell debris. Supernatants were analyzed for cN-II activity following the method described earlier (Tozzi et al., 1991) with slight modifications. Briefly, assays were performed at 37 ◦ C in a medium containing 20 or 40 ␮g of proteins, 5 mM BPG, 20 mM MgCl2 , 5 mM 2-14 C UMP, 0.5 mM AMP-CP, 100 mM Tris–HCl, pH 7.4 in a total volume of 40 ␮l. Reactions were stopped every 10 min by spotting 8 ␮l of the incubation medium on PEI-cellulose plates with 30 nmol of standard uridine. The plates were developed with water and successively cut in correspondence of visualized uridine. Cut disks were then incubated with 8 ml of Optiphase Highphase scintillator liquid (Perkin Elmer) and radioactivity was estimated using a Tri-Carb Liquid Scintillation Counter (Perkin Elmer). 2.4. HPLC analysis of intracellular nucleotides and nucleosides Intracellular nucleotide pools were determined as described before (Machon et al., 2014) using 2 × 106 cells per condition. Results are expressed as pmol of nucleotides per mg of proteins. Intracellular energy charge (EC) was calculated using the following formula: EC = ([ATP] + 1/2[ADP])/([ATP] + [ADP] + [AMP]). 2.5. RT-PCR analysis Total RNA was extracted from cells using RNeasy mini kit (Qiagen) as described by manufacturer. Reverse transcription was performed with Moloney leukemia virus reverse transcriptase and quantitative PCR on a LightCycler thermal cycler (Roche) in conditions and with specific primers for cN-II described previously (Jordheim et al., 2005). Expression of GFP or GFP-cN-II was determined in the same conditions using the forward primer 5 -ACTTCAAGATCCGCCACAAC-3 together with 5 -TGGGTGGACAGGTAGTGGTT-3 for GFP expression or with 5 -TTTTCAGGGCATGCTTATCC-3 for GFP-cN-II expression. Other primers used were: PRPS1: fwd 5 -CTAAGAGAGTGACCTCCATTGC3 , rev 5 -TAGCAGGACCGGAGAAGATT-3 ; PNP: fwd 5 -TGAAAT CCCCAACTTTCCCC-3 , rev 5 -AATGTCACCTTCCAGAGTGGG-3 ; fwd 5 -TGGCGTCGTGATTAGTGATG-3 , rev 5 HGPRT1: GTAATCCAGCAGGTCAGCAA-3 ; PPATT: fwd 5 -ATCACACAAG GGAATGGGTC-3 , rev 5 -ACAGACCAATACCATGACGC-3 ; AMPD3: fwd 5 -CACATCCTGGCTCTCATCAC-3 , rev 5 -GGATGTGTGT   GTCCACCTTT-3 ; ADA: fwd 5 -AGCCCAAAGTAGAACTGCAT-3 , rev 5 -CAAACTTGGCCAGGAAGTCT-3 ; RRM1: fwd 5 -GCAGCT GAGAGAGGTGCTTT-3 , rev 5 -AATGGTTGTAGAATTAAGAATAGC3 ; RRM2: fwd 5 -GAGTTCCTCACTGAGGCC-3 , rev GCTTAAGC TTATTTAGAAGTCAGCATCCAAG-3 . Relative quantification was

224

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

Fig. 1. cN-II modulation in human glioblastoma ADF cells stably transfected with corresponding vectors. Relative NT5C2 transgene expression levels to pSCont (A) and pcDNA3/pcDNA3-GFP (B). Western and immunoblotting against cN-II and ␤-actin (C) and against GFP (D). M, molecular weight marker; kDa, kiloDalton. Figure reported is representative of three independent quantitative RT-PCR and western and immune blot analysis.

calculated using the CT-method and human ribosomal 18S RNA, 28S RNA and GAPDH as housekeeping genes.

2.6. Western and immune blot analysis Proteins were extracted from cell pellets as described elsewhere (Jordheim et al., 2005). Proteins (100 ␮g or more per sample) were separated by SDS-PAGE using 10% acrylamide and transferred onto nitrocellulose membrane using iBlot® system (Life Technologies). Membranes were incubated with specific antibodies for cN-II (clone 3C1, 1/500, Abnova), GFP (ab290, 1/2000, Abcam), beta-actine (clone AC-15, 1/5000, Sigma), and anti-murin antibody (IRdye® 800CW, 1/5000, LI-COR Biosciences) or anti-rabbit antibody (IRdye® 680, 1/5000, LI-COR Biosciences). Protein expression was visualized using the Odyssey infrared system (LI-COR Biosciences).

2.7. Cell proliferation The growth characteristics of stably transfected ADF cells were monitored with the xCELLigence DP system (ACEA Biosciences Inc.). Complete medium (200 ␮l) containing 1 × 103 ADF cells per well were seeded and growth was monitored for 180 h. Electrical inputs were registered and converted to a cell index every 15 min. Cell index profiles were normalized at 12 h and the slope of growth curves was determined and quantified between 12 h and 150 h.

2.8. Cytotoxicity assay The sensitivity to 6MP, gemcitabine, fludarabine, cisplatin, taxol and mitomycin C was assessed using the protocol described by Mosman (1983) with slight modifications. Briefly, cells (10,000 per well) were seeded in 96-well plate and after 24 h were incubated with different concentration of drugs for 72 h at 37 ◦ C, 5% CO2 /95% air atmosphere. Then, 0.5 mg/ml of tetrazolium salt previously dissolved in PBS and filtered was added and cells were incubated another 2 h before medium was removed and 100 ␮l of HCl 0.04 M in isopropanol were added. Spectrophotometric determination of absorbance was done using a microplate reader (Labsystem Multiskanner RC). Half maximal effective concentrations (EC50 ) were calculated with CompuSyn software 1.0 (ComboSyn, Inc., USA).

2.9. H2AX staining Cells were seeded (60,000 cells per well) in 6-well plates containing cover slips and incubated for 24 h at 37 ◦ C, then exposed to Cisplatin (1 ␮M) or fresh media for 48 h. Cells were then washed with PBS, fixed with paraformaldehyde 4% (v/v) for 15 min at room temperature, washed again with PBS and blocked for 3 min with a buffer containing 300 mM sucrose, 3 mM MgCl2 , 20 mM Hepes pH 7, 50 mM NaCl and 0.5% (v/v) Triton X-100. After wash with PBS and incubation with 1:800 dilution of antiphospho-H2AXser139 antibody (05636, Millipore) for 40 min at 37 ◦ C, cells were rinsed and incubated with a 1:100 dilution of a FITC-conjugated secondary anti-mouse antibody (F0232, Dako) for 20 minutes at 37 ◦ C. Finally,

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

225

cells were rinsed with PBS and mounted with one drop of Vectashield (Vector Laboratories). Foci were counted using a Leica DMI3000 microscope in at least 100 nuclei for each condition by a person blinded for experimental conditions. 2.10. Statistical analyses All statistical analyses were performed using the software InStat (ver. 3.05, GraphPad Software, Inc., La Jolla, CA 92037, USA). 3. Results 3.1. cN-II modulation Cells with stable cN-II silencing (pScN-II) or overexpression (pcDNA3-GFP/cN-II and pcDNA3-GFP/cN-II-R367Q), as well as corresponding control cell lines (pScont, pcDNA3 and pcDNA3GFP) were obtained after transfection with plasmids described in material and methods Gene expression analysis and Western/immunoblotting showed decreased cN-II expression in pScN-II cells as compared to pScont cells (Fig. 1A and C), and the presence of the GFP-cN-II fusion protein and its corresponding mRNA in pcDNA3-GFP/cN-II and pcDNA3-GFP-cN-II-R267Q cells (Fig. 1B and D). These modifications in cN-II expression were associated with a 50% decrease of cN-II activity in pScNII cells (4.09 ± 0.41 vs 2.21 ± 0.11), and 40–80% increase of cN-II activity in pcDNA3-GFP/cN-II and pcDNA3-GFP-cN-II-R267Q cells (2.80 ± 0.08 vs 3.89 ± 0.10 and 5.15 ± 0.25), as determined by the phosphatase activity of cN-II in absence and presence of ATP (Fig. 2). The latter results were 0% and 50% when compared to pcDNA3-GFP cells and not to pcDNA3 cells (Fig. 2B). 3.2. Effects of cN-II modulation on nucleotide content, gene expression and cell proliferation Since an inducible increase of cN-II activity has been associated with a depletion of nucleotides triphosphates in HEK 293 cells with consequent depletion of intracellular energy charge (Cividini et al., 2015b), we quantified deoxyribo- and ribonucleotides content in our stably transfected cell lines. Interestingly, no appreciable difference was observed concerning the content of these nucleotides (Supplemental Table 1) or with the respect of intracellular energy charge (pScont = 0.82 ± 0.08; pScN-II = 0.77 ± 0.06; pcDNA3 = 0.76 ± 0.04; pcDNA3-GFP = 0.78 ± 0.09; pcDNA3GFP/cN-II = 0.72 ± 0.08; pcDNA3-GFP/cN-II-R367Q = 0.75 ± 0.06). We further performed transcriptional analysis of genes involved in nucleotide metabolism to ascertain if any metabolic adaptation had occurred to compensate the alteration of cN-II expression in these stably transfected models. Gene expression was calculated with the respect to three different housekeeping genes (18S, GAPDH and 28S), giving similar results, and those related to 18S are reported in Fig. 3. A slight up-regulation of all investigated genes was observed in pcDNA3-GFP/cN-II cells and PPATT, PRPS1, PNP, ADA and AMPD3 were also upregulated in pcDNA3-GFP/cNII-R367Q cells. On the other hand in pScN-II cells only PPATT and RRM2 were found slightly up-regulated, while RRM1, HGPRT1, PRPS1, PNP, ADA and AMPD3 were found either similar to control or slightly down-regulated. All these differences were however not statistically significant. Cell proliferation was evaluated by in the real-time monitoring of cell index with the xCELLigence technology. Cell index is measured by the electrical impedance at the bottom of the well and corresponds to the strength of cell adhesion and cell number. An optimal seeding density was determined and 1000 cells per well were seeded in E-16 plates (corresponding to 96-well plate). pScN-II cells displayed a decreased proliferation rate quantified by a slope of 0.34 ± 0.04

Fig. 2. cN-II phosphatase activity in (A) pSup.neo and (B) pcDNA3 transfected cells. Variances were analyzed by ONE-WAY ANOVA and means were compared by Dunnet’s test. ***p < 0.001; **p < 0.01; *p < 0.05.

as compared to pScont cells (0.74 ± 0.04, p = 0.0021) (Fig. 4A). On the contrary, cells transfected with pcDNA3-GFP/cN-II (slope of 0.64 ± 0.13) and pcDNA3-GFP/cN-II-R367Q (slope of 0.79 ± 0.03) showed an increased proliferation rate as compared to controls (pcDNA3 empty = 0.31 ± 0.07; pcDNA3-GFP = 0.28 ± 0.03) (Fig. 4B). Proliferation slopes for pcDNA3-GFP/cN-II and pcDNA3-GFP/cNII-R367Q were statistically significantly different from the slope of pcDNA3-GFP cells (p = 0.005 and 0.002, respectively). These results suggest a nice correlation between the cN-II activity and cell proliferation in these ADF cell models. 3.3. Drugs sensitivity The sensitivity of our cell models to different anticancer drugs was determined after 72 h of exposure, using the MTT assay. Ratios of EC50 from pScN-II, pcDNA3-GFP/cN-II and pcDNA3GFP/cN-II-R367Q cells and their corresponding control cells are reported in Fig. 5. pScN-II cells were resistant to both fludarabine (EC50 pScont = 26 ␮M ± 3 vs pScN-II = 50 ␮M ± 12, p = 0.004) and

226

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

Fig. 4. Cell proliferation profiles obtained with the xCELLigence technology. (A) Cells transfected with pSup.neo vector: pScont (), pScN-II (䊏). (B) Cells transfected with pcDNA3 vector: pcDNA3 empty (X), pcDNA3-GFP (), pcDNA3-GFP/cN-II (), pcDNA3-GFP/cN-II-R367Q (䊏). Curves are representative of three independent experiments.

cisplatin (EC50 pScont = 14 ␮M ± 1 vs pScN-II = 23 ␮M ± 2, p = 0.008) (Fig. 5A). Cells overexpressing cN-II were only sensitized to cisplatin (EC50 pcDNA3 empty = 19 ␮M ± 3 vs pcDNA3-GFP/cNII = 10 ␮M ± 2 and pcDNA3-GFP/cN-II-R367Q = 11 ␮M ± 3, p = 0.006 and 0.009, respectively) (Fig. 5B). There was a tendency to sensitization of these cells to cisplatin when compared to pcDNA3-GFP cells (EC50 = 15.2 ␮M ± 0.4), but this difference was not statistically significant (Fig. 5C). In order to further investigate this relation between cN-II expression and cisplatin cytotoxicity, we assessed phosphorylated H2AX in cells exposed to cisplatin as a marker of DNA repair. Number of foci/cell was comparable in all cell lines at basal conditions (Fig. 6). pScN-II cells exposed to cisplatin showed more cells without ␥H2AX-staining than pScont cells (79% ± 9 vs 48% ± 8, p = 0.04). This was also the case for cells overexpressing cNII (pcDNA3-GFP/cN-II and pcDNA3-GFP/cN-IIR367Q vs pcDNA3 and pcDNA3-GFP = 81 ± 4 and 76 ± 5 vs 46 ± 15 and 58 ± 2, p = 0.012, 0.02, 0.009 and 0.029 respectively). 4. Discussion 4.1. cN-II modulation

Fig. 3. Transcriptional analysis of nucleotide metabolism relevant genes. Relative gene expressions was calculated using 18S as a housekeeping gene. pScN-II genes expression has been compared to pScont (A), pcDNA3-GFP/cN-II (B) and pcDNA3-GFP/cN-II-R367Q (C) topcDNA3-GFP. Data represent the mean ± SEM of three independent experiments (RNA extraction, retro-transcription and cDNA amplification).

The development of cell models in which the expression of a particular protein is stably altered is a good approach to unravel the physiological function of the given protein. In our case, the protein of interest was the cytosolic 5 -nucleotidase cN-II which is implicated in the regulation of intracellular nucleotide pools and in prodrugs metabolism, and we used a glioblastoma cell line of astrocyte origin particularly resistant to chemotherapy as model. In the same cell line a knock down of cN-II using the RNAi technique, demonstrated that a decrease of about 50% of protein expression was accompanied by activation of caspase 3 and a massive cell death, apparently without a significant alteration of intracellular nucleotide pool (Careddu et al., 2008). In order to continue the evaluation of the role of cN-II in this brain tumor model, and

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

227

Fig. 6. Staining of phosphorylated H2AX in stably transfected ADF cells upon 1 ␮M cisplatin (+) or fresh medium (−) exposure. Foci corresponding to phosphorylated H2AX were counted and grouped in three categories (0/cell; 1–9/cell; >10/cell) and the percentages over at least 100 cells were plotted. Data represent the mean of three independent experiments.

Fig. 7. Purine mononucleotide degradation, interconversion and nucleobase salvage. 1: Adenine phosphoribosyl transferase; 2: AMP deaminase; 3: adenylsuccinate lyase; 4: adenylsuccinate synthetase; 5: IMP dehydrogenase; 6: XMP-glutamine amidotransferase; 7: 5 -nucleotidase cN-I; 8: 5 -nucleotidase cN-II; 9: adenosine kinase; 10: hypoxanthine-guanine phosphoribosyl transferase; 11: purine nucleoside phosphorylase; 12: uridine kinase; 13: uridine phosphorylase; 14: phosphoribomutae; 15: PRPP synthetase; 16: PPATT (phosphoribosyl pyrophosphate amidotransferase); 17: IMP cyclohydrolase. Enzymes 8, 10 and 11 are involved in the purine cycle.

Fig. 5. Sensitivity of stably transfected ADF cells to clinically used anticancer drugs. Results show ratios of EC50 from pScN-II over pScont (A), pcDNA3-GFP/cN-II (white bars) and pcDNA3-GFP/cN-II-R367Q cells (black bars) over pcDNA3 (B) and pcDNA3GFP (C) respectively. Mean values of EC50 (␮M) are from 5 independent experiments. Variances were analyzed by ONE-WAY ANOVA and means were compared by Dunnet’s test. ***p < 0.001; **p < 0.01; *p < 0.05.

in particular to find out if stable knockdown of cN-II could be obtained in these cells, we developed a stably silenced ADF clone evidently expressing about 50% of cN-II with respect to control cells as demonstrated by both western blot analysis (Fig. 1) and phosphatase activity (Fig. 2). The cells that survived to the transfection were chronically adapted to a lower level of cN-II. A decrease of cN-II activity was expected to be associated with an increase of IMP and AMP as observed earlier in human myotubes (Sala-Newby et al., 2000). In this model of differentiated unproliferating muscle cells, AMPK was activated as a consequence of cN-II silencing. Our results however, demonstrated that in stably transfected ADF cells both nucleotide concentrations and energy charge were unaffected by cN-II silencing. To compensate a low rate of IMP hydrolysis without IMP, AMP and GMP accumulation, a decrease of purine de novo and/or salvage could be expected (see Fig. 7). To asses this issue we

228

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

measured the gene expression level of enzymes involved in both de novo and salvage pathways, and showed that some of the genes referred to enzymes involved in the cycle of Fig. 7 are slightly under expressed in silenced cells (Fig. 3). We also developed cell models with an increased expression of cN-II. We expressed both wild type enzyme and a mutated protein which was previously demonstrated to be more active than wild type and that was found expressed in relapsed and resistant ALL patients (Meyer et al., 2013; Tzoneva et al., 2013). Immunoblot analysis showed that the fusion proteins GFP/cN-II and GFP/cNII-R367Q were both expressed in transfected cells. Activity assays showed an increase of phosphatase activity in both cases with respect to the empty control, however, no significant difference were found between GFP/cN-II activity and the control expressing GFP. In addition, applying this protocol of stable transfection, we obtained only a modest increase in cN-II activity in ADF cells, as compared to the inducible overexpression earlier obtained in Human Embrionic Kidney 293 cells (Rampazzo et al., 1999). This result might be due to the different approach (inducible and stable) and is in line with the observation that cells have a very low tolerance to extensive alteration of cN-II expression. On the other hand an adverse effect of GFP linked at the N-terminus of cN-II on the enzyme activity cannot be excluded even if this has been suggested not to be relevant (Gazziola et al., 1999). Indeed, the sole transfection of GFP in our ADF cells was however associated with a slight increase in cN-II activity. A potential bias of our methodology is the risk of isolating cells with other modifications that the cN-II expression during the selection of transfected cells. As we worked with batches of transfected cells and not with isolated clones, and the fact that we observed a “dose-dependent” response to some parameters, we do not believe our observations are cN-IIindependent. The analysis of the expression level of genes referred to enzymes involved in nucleotide metabolism indicates a general, albeit slight, overexpression of salvage and synthesis transcripts to compensate the increase of IMP and GMP catabolism most probably catalyzed by overexpressed or hyperactive cN-II. In fact, an inducible, not compensated increase of cN-II activity in HEK 293 cells and also in yeast caused an important decrease of ATP and energy charge due to an imbalance of the catabolic part with the respect to the anabolic part of the cycle described in Fig. 7 (Allegrini et al., 2013; Cividini et al., 2015b). Surprisingly, the most evident effect of cN-II silencing in our ADF cells was a significant decrease of cell proliferation (Fig. 4), while its overexpression caused an increase of cell proliferation. This effect on cell proliferation cannot be ascribed to an alteration of adenylate or guanylate pools as this was not found in these cells. Therefore, we might hypothesize that cN-II interfere with the pathways regulating proliferation and survival following intracellular and extracellular signals. Indeed, we recently demonstrated that cN-II can interact with an intracellular PRR receptor (Cividini et al., 2015c). Works are in progress to elucidate the physiological function of this interaction and the possible existence of other interactors regulating cell cycle and cell proliferation. 4.2. Sensitivity to cancer drugs Concerning the sensitivity of our cell models to cancer drugs, we decided to test different drugs categories. Gemcitabine, 6MP and fludarabine are antimetabolites and analogs of natural purine and pyrimidine compounds which exert their cytotoxicity by blocking physiological processes such as DNA replication after membrane transport and intracellular metabolism by specific proteins (Nelson et al., 1975; Plunkett et al., 1996; Van Den Neste et al., 2005; Sampath et al., 2003). Taxol is an antiproliferative drugs which interfere with tubulin formation (Horwitz, 1994), whereas cisplatin and MMC are DNA crosslinkers which trigger

apoptosis (Lippert, 1999; Verweij and Pinedo, 1990). The apparent fludarabine-resistance in pScN-II cells is surprising as compared to our recent results on other cell models (Jordheim et al., 2015, in press). Indeed, we observed a high sensitivity to purine nucleoside analogs in cN-II deficient hematological cell linesand both the nucleoside fludarabine or the monophosphate nucleoside f-ara-MP were demonstrated to be cN-II inhibitor and substrate respectively (Cividini et al., 2015a; Jordheim et al., 2006). The fact that cNII deficient brain tumor ADF cells are resistant to this molecule suggests the implication of cN-II in the metabolism or activity of fludarabine, and that this might be tissue or cell specific. Further investigations, as for example the quantification of phosphorylated metabolites of fludarabine, are needed in order to explain this phenomenon. As for cisplatin, we observed a clear resistance in pScN-II and a sensitization in pcDNA3-GFP/cN-II and pcDNA3-GFP/cN-IIR367Q cells. The MTT assay is measuring the metabolic status of surviving cells. As we observed a difference in proliferation between our transfected cells and their respective controls, modifications in drug sensitivity could be linked to this. However, our finding showing that for example pScN-II cells are resistant to fludarabine are contrary to this hypothesis as these cells also proliferate less than their control cells. Cisplatin sensitivity is dependent on the phosphorylation status of several proteins involved in cell survival and apoptosis (Siddik, 2003). As cN-II modulation is potentially associated with a modification in phosphorylation abilities of certain enzymes due to modifications in phosphate donors, this could partly explain the variability in sensitivity to cisplatin in cN-II modulated cells. Both overexpressing and underexpressing cells had a lower percentage of cells with more than 10 foci per cells after cisplatin exposure. This indicates modified kinetics in DNA repair in both kinds of models, even though it is not necessarily in the same way. Indeed, H2AX phosphorylation occurs at double strand brakes during repair of cisplatin adducts (Horwitz, 1994). It could be possible that there are less such brakes in pScN-II cells exposed to 1 ␮M cisplatin as these appear to be resistant, whereas in overexpressing models (GFP/cN-II and GFP/cN-II-R367Q), repair has already occurred and there are less brakes left than in control cells. This hypothesis is concordant with both results from MTT assays and on H2AXphosphorylation. 4.3. Conclusions The cell models presented in this paper offer the advantage to study the metabolic consequences of chronically altered expression of cN-II avoiding the gross alteration of nucleotide pools. This is most probably the context in which a tumor cell expressing a high or low level of cN-II would be in. In fact, a loss of adenylic compounds, a drop of energy charge, AMP and or IMP accumulation, were previously observed in models in which a strong alteration of cN-II expression was obtained with a protocol of transitory silencing or inducible overexpression. The currently reported results complete our previous observations of transcient modification of cN-II in ADF cells. They indicate that cN-II is directly or indirectly involved in the pathway regulating survival, apoptosis and proliferation and also its effects in the sensitization or resistance to antitumor drugs might be ascribed to these mechanisms. Authors contribution L.P.J., M.G.T., F.C. and C.D. designed research; F.C., E.C.P., R.P. performed experiments; C.M. performed HPLC analysis; C.D., L.P.J., M.G.T. and S.A. obtained funding; F.C., M.G.T., L.P.J., C.D. and M.C. wrote the paper.

F. Cividini et al. / The International Journal of Biochemistry & Cell Biology 65 (2015) 222–229

Funding CD’s laboratory receives funding through ANR grant 11-BS07032-01 “cN-II Focus”. MGT’s laboratory was financially supported by a grant from the REGIONE AUTONOMA DELLA SARDEGNA, L.R. 07/08/2007, CRP 3360 (TITLE: Modulation of expression by inducible silencing; heterologous expression in S. cerevisiae; twohybrid system. Three models to make clear the physiological role of cytosolic 5 nucleotidase II). Acknowledgements Authors are thankful to Tuscany Region for providing travel funding for FC to visit CD’s laboratory. L.P.J. acknowledges Olav Raagholt og Gerd Meidel Raagholts stiftelse for forskning. MGT’s laboratory was financially supported by a grant from the REGIONE AUTONOMA DELLA SARDEGNA, L.R. 07/08/2007, CRP 3360 (TITLE: Modulation of expression by inducible silencing; heterologous expression in S. cerevisiae; two-hybrid system. Three models to make clear the physiological role of cytosolic 5 -nucleotidase II). CD’s laboratory received funding through ANR grant 11-BS07-032 “cN-II Focus”. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biocel.2015.06. 011 References Allegrini, S., Scaloni, A., Ferrara, L., Pesi, R., Pinna, P., Sgarrella, F., et al., 2001. Bovine cytosolic 5 -nucleotidase acts through the formation of an aspartate 52-phosphoenzyme intermediate. J. Biol. Chem. 276, 33526– 33532. Allegrini, S., Scaloni, A., Careddu, M.G., Cuccu, G., D’Ambrosio, C., Pesi, R., et al., 2004. Mechanistic studies on bovine cytosolic 5 -nucleotidase II, an enzyme belonging to the HAD superfamily. Eur. J. Biochem. 271, 4881–4891. Allegrini, S., Filoni, D.N., Galli, A., Collavoli, A., Pesi, R., Camici, M., et al., 2013. Expression of bovine cytosolic 5 -nucleotidase (cN-II) in yeast: nucleotide pools disturbance and its consequences on growth and homologous recombination. PLoS ONE 8, e63914. Bretonnet, A.S., Jordheim, L.P., Dumontet, C., Lancelin, J.M., 2005. Regulation and activity of cytosolic 50-nucleotidase II. A bifunctional allosteric enzyme of the Haloacid Dehalogenase superfamily involved in cellular metabolism. FEBS Lett. 579, 3363–3368. Camici, M., Micheli, V., Ipata, P.L., Tozzi, M.G., 2010. Pediatric neurological syndromes and inborn errors of purine metabolism. Neurochem. Int. 56, 367–378. Careddu, M.G., Allegrini, S., Pesi, R., Camici, M., Garcia-Gil, M., Tozzi, M.G., 2008. Knockdown of cytosolic 5 -nucleotidase II (cN-II) reveals that its activity is essential for survival in astrocytoma cells. Biochim. Biophys. Acta 1783, 529–535. Ceruti, S., Mazzola, A., Abbracchio, M.P., 2005. Resistance of human astrocytoma cells to apoptosis induced by mitochondria-damaging agents: possible implications for anticancer therapy. J. Pharmacol. 314, 825–837. Cividini, F., Pesi, R., Chaloin, L., Allegrini, S., Camici, M., Cros-Perrial, E., et al., 2015a. The purine analog fludarabine acts as a cytosolic 5 -nucleotidase II inhibitor. Biochem Pharmacol. 94, 63–68. Cividini, F., Filoni, D.N., Pesi, R., Allegrini, S., Camici, M., Tozzi, M.G., 2015b. IMP-GMP specific 5 -nucleotidase regulates nucleotide pool and prodrug metabolism. Biochim Biophys. Acta. 1850, 1354–1361. Cividini, F., Tozzi, M.G., Galli, A., Pesi, R., Camici, M., Dumontet, C., et al., 2015c. Cytosolic 5 -nucleotidase II interacts with the leucine rich repeat of NLR family member Ipaf. PLoS One 10 (3), e0121525. Gazziola, C., Moras, M., Ferrero, P., Gallinaro, L., Verin, V., Rampazzo, C., et al., 1999. Induction of human high KM 5 -nucleotidase in cultured 293 cells. Exp. Cell. Res. 253, 474–482. Horwitz, S.B., 1994. Taxol (paclitaxel): mechanisms of action. Ann. Oncol. 5, 3–6. Itoh, R., 1981. Purification and some properties of cytosol 5 -nucleotidase from rat liver. Biochim. Biophys. Acta 657, 402–410.

229

Itoh, R., Echizen, H., Higuchi, M., Oka, J., Yamada, K., 1992. A comparative study on tissue distribution and metabolic adaptation of IMP-GMP 5 -nucleotidase. Comp. Biochem. Physiol. B: Comp. Biochem. 103, 153–159. Jordheim, L.P., Chaloin, L., 2013. Therapeutic perspectives for cN-II in cancer. Curr. Med. Chem. 20, 4292–4303. Jordheim, L.P., Guittet, O., Lepoivre, M., Galmarini, C.M., Dumontet, C., 2005. Increased expression of the large subunit of ribonucleotide reductase is involved in resistance to gemcitabine in human mammary adenocarcinoma cells. Mol. Cancer Ther. 4, 1268–1276. Jordheim, L.P., Cros, E., Galmarini, C.M., Dumontet, C., Bretonnet, A.S., Krimm, I., Lancelin, J.M., et al., 2006. F-ara-MP is a substrate of cytoplasmaic 5 -nucleotidase II (cN-II): HPLC and NMR studies of enzymatic dephosphorylation. Nucleoside. Nucleoside. Nucl. Acids 25, 289–297. Jordheim, L.P., Marton, Z., Rhimi, M., Cross-Perrial, E., Lionne, C., Peyrottes, S., et al., 2013. Identification and characterization of inhibitors of cytoplasmic 5 -nucleotidase cN-II issued from virtual screening. Biochem. Pharmacol. 85, 497–506. Jordheim, L.P., Puy, J.Y., Cros-Perrial, E., Peyrottes, S., Lefebvre, I., Périgaud, C., et al., 2015. Determination of the enzymatic activity of cytosolic 5 -nucleotidase cN-II in cancer cells: development of a simple analytical method and related cell line models. Anal. Bioanal. Chem., http://dx.doi.org/10.1007/s00216-0158757-4 (in press). Lippert, B., 1999. The mechanism of action of cisplatin: from adducts to apoptosis. In: Verlag Helvetica Chimica Acta (Ed.), Cisplatin: Chemistry and Biochemistry of a Leading Anticancer Drug. Postfach, Switzerland (Chapter 4). Machon, C., Jordheim, L.P., Puy, J.Y., Lefebvre, I., Dumontet, C., Guitton, J., 2014. Fully validated assay for the quantification of endogenous nucleoside monoand triphosphates using online extraction coupled with liquid chromatography–tandem mass spectrometry. Anal. Bioanal. Chem. 406, 2925–2941. Meyer, J.A., Wang, J., Hogan, L.E., Yang, J.J., Dandekar, S., Patel, J.P., et al., 2013. Relapse-specific mutations in NT5C2 in childhood acute lymphoblastic leukemia. Nat. Gen. 4, 290–294. Micheli, V., Camici, M., Tozzi, M.G., Ipata, P.L., Sestini, S., Bertelli, M., et al., 2011. Neurological disorders of purine and pyrimidine metabolism. Curr. Top. Med. Chem. 11, 923–947. Mosman, T., 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J. Immunol. Methods 65, 9. Nelson, J.A., Carpenter, J.W., Rose, L.M., Adamson, D.J., 1975. Mechanisms of action of 6-thioguanine, 6-mercaptopurine, and 8-azaguanine. Cancer Res. 35, 2872. Pesi, R., Turriani, M., Allegrini, S., Scolozzi, C., Camici, M., Ipata, P.L., et al., 1994. The bifunctional cytosolic 5 -nucleotidase: regulation of the phosphotransferase and nucleotidase activities. Arch. Biochem. Biophys. 312, 75–80. Pesi, R., Allegrini, S., Careddu, M.G., Filoni, D.N., Camici, M., Tozzi, M.G., 2010. Active and regulatory sites of cytosolic 5 -nucleotidase. FEBS J. 277, 4863–4872. Plunkett, W., Huang, P., Searcy, C.E., Gandhi, V., 1996. Gemcitabine: preclinical pharmacology and mechanisms of action. Semin. Oncol. 23, 3–15. Rampazzo, C., Gazziola, C., Ferraro, P., Gallinaro, L., Johansson, M., Reichard, P., et al., 1999. Human high-Km 5 -nucleotidase effects of overexpression of the cloned cDNA in cultured human cells. Eur. J. Biochem. 261, 689–697. Sala-Newby, G.B., Freeman, N.V.E., Skladanowski, A.C., Newby, A.C., 2000. Distinct roles for recombinant cytosolic 5 -nucleotidase-I and -II in AMP and IMP catabolism in COS-7 and H9c2 rat myoblast cell lines. J. Biol. Chem. 275, 11666–11671. Sampath, D., Rao, V.A., Plunkett, W., 2003. Mechanisms of apoptosis induction by nucleoside analogs. Oncogene 22, 9063–9074. Siddik, Z.H., 2003. Cisplatin: mode of cytotoxic action and molecular basis of resistance. Oncogene 22, 7265–7279. Spychala, A.J., Madrid-Marina, V., Fox, I.H., 1988. High Km soluble 5 -nucleotidase from human placenta. Properties and allosteric regulation by IMP and ATP. J. Biol. Chem. 263, 18759–18765. Tozzi, M.G., Camici, M., Allegrini, S., Pesi, R., Turriani, M., Del Corso, A., et al., 1991. Cytosolic 5 -nucleotidase/phosphotransferase of human colon carcinoma. Adv. Exp. Med. Biol. 309B, 173–176. Tozzi, M.G., Pesi, R., Allegrini, S., 2013. On the physiological role of cytosolic 5 -nucleotidase II (cN-II): pathological and therapeutical implications. Curr. Med. Chem. 34, 4285–4291. Tzoneva, G., Perez-Garcia, A., Carpenter, Z., Khiabanian, H., Tosello, V., Allegretta, M., et al., 2013. Activating mutations in the NT5C2 nucleotidase gene drive chemotherapy resistance in relapsed ALL. Nat. Med. 19, 368–371. Van Den Neste, E., Cardoen, S., Offner, F., Bontemps, F., 2005. Old and new insights into the mechanisms of action of two nucleoside analogs active in lymphoid malignancies: fludarabine and cladribine. Int. J. Oncol. 27, 1113–1124. Verweij, J., Pinedo, H.M., 1990. Mitomycin C: mechanism of action, usefulness and limitations. Anticancer Drugs 1, 5–13. Walldén, K., Stenmark, P., Nyman, T., Flodin, S., Graslund, S., Loppnau, P., et al., 2007. Crystal structure of human cytosolic 5 -nucleotidase II. Insights into allosteric regulation and substrate recognition. J Biol Chem. 282, 17828–17836. Walldén, K., Nordlund, P., 2011. Structural basis for the allosteric regulation and substrate recognition of human cytosolic 5 -nucleotidase II. J. Mol. Biol. 408, 684–696.