The Cystine/Cysteine Cycle: a Redox Cycle Regulating Susceptibility Versus Resistance to Cell Death

Oncogene (2008) 27, 1618–1628 & 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc ORIGINAL ARTICLE The cystine/cysteine cycle: a redox cycle regulating susceptibility versus resistance to cell death

A Banjac1, T Perisic1, H Sato2,6, A Seiler1, S Bannai2, N Weiss3,PKo¨ lle3, K Tschoep4, RD Issels4, PT Daniel5, M Conrad1,7 and GW Bornkamm1,7

1GSF-Forschungszentrum fu¨r Umwelt und Gesundheit, Institut fu¨r Klinische Molekularbiologie und Tumorgenetik, Mu¨nchen, Germany; 2Department of Biochemistry, Institute of Basic Medical Sciences, University of Tsukuba, Tsukuba, Ibaraki, Japan; 3Medizinische Poliklinik Innenstadt der Ludwig-Maximilians-Universita¨t, Mu¨nchen, Germany; 4Medizinische Klinik III der Ludwig-Maximilians-Universita¨t, GSF-Klinische Kooperationsgruppe Hyperthermie, Mu¨nchen, Germany and 5Medizinische Klinik mit Schwerpunkt Ha¨matologie und Onkologie, Charite´, Humboldt Universita¨t, Berlin, Germany

The glutathione-dependent system is one of the key Introduction systems regulating cellular redox balance, and thus cell fate. Cysteine, typically present in its oxidized form cystine Redox regulation of cell cycle progression and cell death in the extracellular space, is regarded as the rate-limiting has attracted remarkable interest in recent years (Arner substrate for glutathione (GSH) synthesis. Cystine is and Holmgren, 2000). A variety of enzymatic systems transported into cells by the highly specific amino-acid are involved in the maintenance of intracellular redox À antiporter system xc . Since Burkitt’s Lymphoma (BL) homeostasis including the glutathione and thioredoxin- cells display limited uptake capacity for cystine, and are dependent systems. Starting to dissect their functional thus prone to oxidative stress-induced cell death, we stably redundancy, we have created mice with targeted À expressed the substrate-specific subunit of system xc , deficiencies for both cytosolic and mitochondrial thio- xCT, in HH514 BL cells. xCT-overexpressing cells redoxin reductases (Txnrd1 and 2, Conrad et al., 2004; became highly resistant to oxidative stress, particularly Jakupoglu et al., 2005). Thereby, we could demonstrate upon GSH depletion. Contrary to previous predictions, that Txnrd2 indeed efficiently protects cells against the the increase of intracellular cysteine did not affect the detrimental effects of GSH depletion (Conrad et al., cellular GSH pool, but concomitantly boosted extracel- 2004). lular cysteine concentrations. Even though cells were Glutathione is present in cells in millimolar concen- depleted of bulk GSH, xCT overexpression maintained tration and is considered as the major natural antioxi- cellular integrity by protecting against lipid peroxidation, dant, protecting cells from oxidative stress (Meister, a very early event in cell death progression. Our results 1995). Availability of cystine/cysteine is the rate-limiting À show that system xc protects against oxidative stress not step in GSH synthesis (Bannai and Tateishi, 1986; Ishii by elevating intracellular GSH levels, but rather creates a et al., 1987). Cysteine is transported into cells via neutral reducing extracellular environment by driving a highly amino-acid transport systems, whereas cystine, the efficient cystine/cysteine redox cycle. Our findings show predominant form in plasma, extracellular body fluids that the cystine/cysteine redox cycle by itself must be and cell culture medium, is carried by the anionic À viewed as a discrete major regulator of cell survival. amino-acid transport system, system xc (Bannai and À Oncogene (2008) 27, 1618–1628; doi:10.1038/sj.onc.1210796; Tateishi, 1986). Expression of system xc is fairly low in published online 10 September 2007 many cell types as firstly depicted for murine B lymphocytes. Provision of b-mercaptoethanol (2-ME) Keywords: cystine–glutamate exchange; glutathione me- or other sulfhydryl-containing compounds is thus a À tabolism; lipid peroxidation; redox regulation; system xc prerequisite for the survival of those cells in vitro (Broome and Jeng, 1973; Metcalf, 1976). Thiol-containing compounds form mixed disulfides and thereby release cysteine which eventually enters cells via neutral amino-acid transport systems. The mixed disulfides in turn enter the cell via a transport system for bulky amino acids, releasing cysteine intracellularly. Thiol Correspondence: Dr M Conrad, Institute of Clinical Molecular Biology and Tumor Genetics, GSF-Research Centre, Marchioninistr. 25, compounds are, however, not required if B cells are co- Munich, Bavaria 81377, Germany. cultured with irradiated fibroblasts. Fibroblasts have a E-mail: [email protected] high uptake capacity for cystine, and upon intracellular 6Current address: Department of Bioresources, Faculty of Agriculture, reduction, provide cysteine to co-cultured B cells (Falk Yamagata University, Tsuruoka, Yamagata 997-8555, Japan. 7These authors contributed equally to this work. et al., 1998). Received 22 February 2007; revised 4 July 2007; accepted 20 August Limited uptake capacity for cystine is a phenomenon 2007; published online 10 September 2007 neither restricted to murine cells nor to B cells. T cells The cystine/cysteine redox cycle A Banjac et al 1619 are dependent on the supply of cysteine by antigen- transfected cells and hxCT-overexpressing cells for the presenting cells (Droge et al., 1991; Kuppner et al., time intervals of 1, 2 and 3 min. L-cystine uptake was 2003). Proliferation of many human and murine virtually undetectable in untransfected and vector- lymphoma and leukemia cell lines are dependent on transfected cells, whereas xCT-transfected cells showed free thiols or a feeder layer of irradiated fibroblasts an uptake activity of more than 2 nmol minÀ1 per mg (Falk et al., 1993, 1998). Likewise, neurons show limited protein, which was linear at least for the first 3 min. uptake capacity for cystine which may lead to low Uptake of L-cystine was efficiently inhibited in the intracellular GSH concentrations and high susceptibility presence of 2.5 mML-glutamate, indicating that L-cystine to oxidative stress-induced cell death (Murphy et al., uptake is solely mediated by the cystine–glutamate 1989, 1990). exchange transporter (Figure 1d). Seeding BL cells at À The xc cystine/glutamate-exchange transporter is a low cell density generates oxidative stress which can be heterodimer composed of the xCT light chain conferring overcome by the addition of antioxidant supplements the specificity of the amino-acid exchange reaction, and (Falk et al., 1993; Brielmeier et al., 1998). To investigate the 4F2 heavy chain, a ubiquitously expressed cell whether overexpression of human xCT provides a surface component shared with several other amino-acid growth advantage to HH514 cells, a critical cell density transport systems (Sato et al., 1999; Verrey et al., 2000). was determined that discriminates between cell survival/ The xCT light chain consists of 12 putative transmem- proliferation and cell death. Cells were plated in brane domains and is linked to the 4F2 heavy chain 96-well plates in serial dilutions from 10 000 cells per through an extracellular disulfide bond (Sato et al., well (100 000 cells mlÀ1) down to 20 cells per well. À 1999). Transcription of the xCT gene and xc transport For untransfected and vector-transfected control activities are induced by oxidative stress, mediated by cells, the critical cell density was 50 000 cells mlÀ1, electrophilic agents, depletion of cystine and by oxygen whereas overexpression of human xCT supported (Bannai et al., 1989). xCT-deficient mice are viable and cell growth to a density as little as 6000 cells mlÀ1 fertile indicating that the supply of cysteine can be (Figure 1e). compensated by other routes during normal development. By contrast, fibroblasts isolated from xCTÀ/À mice can only be cultivated if the culture medium is xCT-overexpressing HH514 cells are highly resistant to supplemented with 2-ME or N-acetylcysteine (NAC, L-buthionine-sulfoximine- (BSO) mediated cell death Sato et al., 2005). Next, we examined whether xCT-overexpression sup- À To study the contribution of system xc to the redox ports growth of BL cells even under strongly limiting balance in proliferation and prevention of human B cells GSH conditions. To this end, HH514 cells were from oxidative stress-induced cell death, we have cultivated in the presence of various L-buthionine- established a BL cell line that stably expresses xCT light sulfoximine (BSO) concentrations, and the number of chain. Our data demonstrate that elevated expression of viable and dead cells was determined over a period of 8 xCT efficiently protects BL cells from oxidative stress- days (Figures 2a and b). We used BSO in our studies, induced cell death even under conditions of cellular since BSO specifically inhibits g-glutamyl-cysteine GSH depletion. synthetase (g-GCS), the rate-limiting enzyme in GSH anabolism, and thus causes rapid depletion of intracellular GSH (Griffith, 1982). While control cells already died at the lowest BSO concentration within 48 h, Results proliferation of xCT-overexpressing cells was slowed down in a dose-dependent manner and cells remained L-cystine uptake is strongly increased in viable even in the presence of 100 mM BSO (Figure 2b). xCT-overexpressing cells and is sensitive to BSO treatment did not impair the uptake capacity of inhibition by L-glutamate cells for cystine (Figure 2c). Addition of 2.5 mM GSH Human and murine xCT light chain were cloned into the rescued the BSO-mediated decrease in proliferation rate, expression vector p141CAG-3SIP and used for stable again with lesser efficiency at high BSO concentrations, expression in HH514 BL cells. For each gene, three ruling out a toxic side effect of BSO (Figure 2d). independent clones were selected with puromycin and A decrease in the proliferation rate of BSO-treated constantly maintained under selection pressure. Expres- cells might be caused by an increase in the number of sion levels were analysed by northern blotting of total cells exiting the cell cycle and entering G0 or G1,byan RNA, isolated from empty vector-transfected and xCT- increased death rate in the BSO-treated culture that overexpressing cells. By using an xCT-specific cDNA diminishes the number of cells without affecting the cell probe, strong hybridization signals, corresponding to cycle distribution, or both. To discriminate between the bicistronic xCT-IRES-puromycin acetyl transferase these possibilities, the cell cycle distribution of xCT mRNA of about 3.5 kb, were detected (Figure 1b). The overexpressing cells was studied that had been treated endogenous xCT mRNA of 12.5 kb became visible after with 50 mM BSO. As shown in Figure 2e, the ratio of extended exposure (Figure 1b). L-cystine uptake activity cells in G1 to those in G2 remained unaffected, while the was determined in vector- and xCT-transfected cells. number of dead cells (sub-G1) increased from 6.3 to Figure 1c shows a time course of L-cystine uptake 25% in the presence of BSO. The number of cells in S (0.1 mCi per sample) in non-transfected cells, vector- phase decreased to some extent suggesting that cells in

Oncogene The cystine/cysteine redox cycle A Banjac et al 1620 S phase are more prone to oxidative stress-induced cell detectable at any given BSO concentration. Most death. importantly, empty vector-transfected cells treated with To correlate intracellular GSH levels with survival 5 mM BSO died within 48 h (Figure 2a), whereas xCT- and proliferation, cells were treated with different overexpressing cells continued to proliferate under the concentrations of BSO, and GSH concentrations were same conditions (Figure 2b), despite equal low GSH determined after 24 h of incubation, a time point when levels (Figure 3a). Moreover, xCT-overexpressing cells the number of viable control cells had not yet decreased. sustained cell survival even at significantly lower GSH As shown in Figure 3a, no relevant change in GSH levels due to higher BSO concentrations as compared to levels of xCT-overexpressing cells and control cells was control cells treated with only 5 mM BSO (Figure 3a). In addition, intracellular GSH became even undetectable by HPLC analysis in BSO-treated cells irrespective of xCT-expression levels (Figure 3b). These unexpected findings strongly argue against a putative, critical threshold of GSH that discriminates between cell survival and cell death. Our data imply that HH514 cells can survive and proliferate at strongly decreased GSH levels provided that sufficient L-cystine uptake is À guaranteed by system xc .

Intra- and extracellular L-cysteine levels are strongly augmented in xCT-overexpressing cells Since increased cystine uptake activity rescued xCT- overexpressing cells from BSO-induced cell death, and GSH did not play the suspected role in maintaining cell survival, we focussed on the fate of intracellular cystine. As expected, a consistent increase of intracellular cysteine by a factor of about 6 was observed in xCT- overexpressing cells (Figure 4a). However, when we studied extracellular total mercaptan levels, we observed remarkably higher levels in xCT-overexpressing cells regardless of BSO treatment (Figure 4b). In contrast, only marginal levels were secreted by control cells. HPLC analysis revealed that secreted mercaptans consist predominantly of cysteine (Figure 4c). These data À support the notion that system xc drives a redox cycle consisting of cystine import, intracellular reduction of cystine to cysteine, cysteine secretion and reoxidation to cystine in the extracellular environment.

Figure 1 Stable overexpression of human xCT in HH514 cells promotes cystine uptake and allows cells to grow at cell densities non-permissive for control cells. (a) A northern blot showing high expression of human xCT (arrow) in two transfected HH514 cell clones (lanes 3 and 4) and low expression in two cell clones transfected with the empty vector (lanes 1 and 2). Panel (b) is a long exposure of (a) to visualize endogenous 12.5 kb xCT-mRNA (arrowhead). (c) Uptake of L-[14C]Cystine was measured in non- transfected cells (’), empty vector-transfected cells (J) and hxCT-overexpressing cells (E) for the indicated time intervals. (d) L-cystine uptake was inhibited by glutamate. L-cystine uptake was measured for 1 min in the absence (ﬁlled bars) or presence (empty bars) of 2.5 mM glutamate (Glu) resulting in 87% inhibition by glutamate. A clear decrease was also noticed in empty vector- transfected cells. (e) Cell survival was monitored three weeks after seeding. The critical cell density for survival and proliferation of non-transfected BL cells (’) and cells transfected with the empty vector (m) was between 100 000 and 50 000 cells mlÀ1.(E) In the presence of 100 mM a-thioglycerol (a-TG), and 3 mM pyruvate non- transfected BL were capable of proliferating at cell densities of 200 cells mlÀ1. Overexpression of human () and murine xCT (x) in BL cells supported growth at densities of about 6000 cells mlÀ1 and higher. Results represent mean values of four independent experiments with duplicate measurements7s.d.

Oncogene The cystine/cysteine redox cycle A Banjac et al 1621

Figure 2 xCT-overexpressing BL cells are highly resistant to cell death induced by GSH-depletion. (a) Vector-transfected control cells died already at concentrations as little as 5 mM. The inset represents a magnification of the cell numbers at days 0, 1 and 2. The scale corresponds to 0.05 Â 106 cells mlÀ1 per bar. The columns in the inset reflect, from left to right, the same increasing BSO concentrations as depicted for the symbols (from top to bottom). (b) Dose-dependent impairment of proliferation of xCT-overexpressing BL cells in the presence of 0–100 mM BSO. (c) To exclude the possibility that BSO interferes with cystine uptake through system xc-, L-cystine uptake activity was measured in the absence (filled bars) and presence (empty bars) of BSO. No significant difference in xCT activity was observed in xCT-overexpressing and vector-transfected cells at a BSO concentration of 5 mM. Results represent mean values of four independent experiments with duplicate measurements7s.d. (d) Glutathione (2.5 mM) supplementation efficiently rescued cell death induced by BSO treatment. (e) Cell cycle analysis of xCT-overexpressing cells revealed that cells treated with 50 mM BSO do not exit the cell cycle but rather die. Note the increase in cells in sub-G1 (M1), while cells in G1 (M2) and S phase (M3) decrease.

xCT-overexpression protects HH514 cells from hydrogen expression of CD95 and of the death receptors DR4 and peroxide induced cell death DR5 was studied by ﬂow cytometry (Figure 5b). To address whether xCT overexpression is able to Expression of neither CD95, nor DR4 and DR5 was protect HH514 cells also from other death-inducing affected by overexpression of xCT. CD95 was expressed stimuli, xCT overexpressing HH514 and control cells were at very low levels on HH514, mock-transfected and treated with increasing concentrations of hydrogen xCT-transfected cells, whereas the majority of cells peroxide. As shown in Figure 5a, xCT-overexpres- stained positive for DR4 and about half of the cells for sing cells were more resistant to hydrogen peroxide DR5 (Figure 5b). Treatment of the cells with CD95/Fas treatment than untransfected or mock-transfected ligand (FasL) did not induce cell death consistent with control cells. the low CD95 expression on the cell surface. Mock- To study susceptibility of xCT-overexpressing cells transfected and xCT-transfected HH514 cells were also and control cells to killing through death receptors, resistant to killing by TRAIL independently of the

Oncogene The cystine/cysteine redox cycle A Banjac et al 1622

Figure 3 Glutathione levels are not altered by xCT overexpression. (a) Total cellular GSH levels were measured (Tietze) in hxCT- overexpressing (ﬁlled bars) and vector-transfected cells (empty bars) after 24 h cultivation in the absence/presence of various BSO concentrations. (b) Semi-quantitative determination of intracellular cysteine and GSH by HPLC in hxCT-overexpressing and control cells in the absence/presence of 10 mM BSO.

expression of xCT (Figure 5c). Of note, xCT-overexpression did not modulate the expression of a variety Figure 4 Intra- and in particular extracellular cysteine levels are of anti- or proapoptotic genes (Supplementary Figure strongly increased in xCT-overexpressing cells. (a) Intracellular 1), and did not protect the cells from epirubicin-induced cysteine levels are approximately six-fold increased in overexpressing versus vector-transfected cells (ﬁlled bars) regardless of BSO cell death (Figure 5c). treatment (50 mM BSO, 24 h; empty bars). (b) Total extracellular mercaptans are highly elevated in the supernatant of xCT-overexpressing cells. The data represent mean values of three independent The cystine/cysteine cycle protects cells from experiments with duplicate measurements7s.d. (c)HPLCanalysis BSO-induced lipid peroxidation conﬁrmed that cysteine is the secreted mercaptan in the culture medium of hxCT-overexpressing cells in the absence (ÀBSO) as well Time-course experiments revealed that cell death be- as in the presence of BSO ( þ BSO). came apparent in control cells 24 h after BSO treatment (Figure 6). To address whether dying cells show hallmarks of classical apoptosis, cells were stained with Annexin V and propidium iodide (PI). Annexin V-positive, (Supplementary Figure 2), indicating that BSO-treated PI-negative apoptotic (pre-necrotic) cells could not be cells do not undergo a classical form of apoptosis detected at any time point after BSO treatment (Nicotera and Melino, 2004; Melino, 2005).

Oncogene The cystine/cysteine redox cycle A Banjac et al 1623

Figure 5 xCT-overexpression confers resistance to hydrogen peroxide, but does not influence receptor-mediated apoptotic pathways. (a) Cells were incubated with increasing concentrations of H2O2 and the number of viable cells was determined 48 h after treatment. (b) Expression of CD95 receptor and TRAIL receptors DR4 and DR5 was determined by flow cytometry on a single cell level. Representative histograms are shown on a log10 logarithmic scale for relative receptor expression of the death receptor (black) or control stained cells (white). The T-ALL Jurkat and the lymphoblastoid line SKW6.4 served as positive controls. Percentages indicate relative numbers of positively stained cells based on marker sets relative to the control stained cells. (c) Mock and hxCT overexpressing cells were exposed for 72 h to epirubicin, or the death ligands TRAIL (TNF-related apoptosis inducing ligand, APO-2L) and Fas ligand (CD95L, APO-1L) at a dose of 100 ng mlÀ1. Cell death was determined by flow cytometric analysis of genomic DNA fragmentation on a single cell level. Means7s.d. (n ¼ 3) from a typical experiment are shown.

We examined additional cell death initiating or should protect cells from BSO-induced cell death. As promoting parameters such as lipid peroxidation, shown in Figure 7a, treatment of control cells with accumulation of intracellular ROS, breakdown of the vitamin E fully protected HH514 cells from lipid mitochondrial membrane potential, caspase activation peroxidation and cell death. xCT-overexpressing cells and DNA fragmentation (Figure 6 and Supplementary but not vector-transfected HH514 cells were protected Figure 3). Lipid peroxidation clearly preceded all other from lipid peroxidation (Figure 7b). In line with the data cellular processes, starting already 8 h after addition of described in Figure 2, lipid peroxidation increased in BSO (Figure 6). Lipid peroxidation became prominent vector-transfected cells with increasing BSO concentra- in the vast majority of cells 12 h of BSO treatment, tion, and likewise, protection from lipid peroxidation by clearly before ROS accumulation and breakdown of the xCT overexpression was less complete at higher BSO mitochondrial membrane potential occurred. An in- concentration (Figure 7b). crease in intracellular ROS was observed 24 h after BSO addition, a time point when lipid peroxidation had progressed in virtually all cells (Figure 6). Breakdown of the mitochondrial membrane potential ensued after lipid Discussion peroxidation and ROS accumulation (Figure 6), suggesting that loss of mitochondrial integrity initiates the To investigate the role of cystine uptake for the execution phase of BSO-induced cell death associated regulation of cell survival and proliferation, we estab- with broad caspase activation and DNA fragmentation lished HH514 BL cell clones that stably overexpress À (Supplementary Figure 3). human xCT light chain of the xc cystine/glutamate If lipid peroxidation and subsequent breakdown of exchange transporter. Our data show that overexpres- membrane integrity trigger BSO-induced cell death, sion of xCT light chain promotes a strong increase in the antioxidants speciﬁcally targeting lipid membranes uptake activity for L-cystine and protects HH514 cells

Oncogene The cystine/cysteine redox cycle A Banjac et al 1624

Figure 6 Lipid peroxidation precedes BSO-induced cell death. BL cells were treated with BSO for the indicated time points. Cell viability was assessed by morphology (FSC/SSC), lipid peroxidation by C11-BODIPY-staining, cellular ROS levels by DCF and mitochondrial membrane potential by JC-1 staining.

from oxidative stress-induced cell death. The protective in intracellular and to an even more pronounced effect of xCT is not associated with changes in the increase in extracellular cysteine levels. These high expression level of various pro- or antiapoptotic cysteine levels are obviously sufficient to sustain cell genes. growth of xCT-overexpressing cells even under condi- Availability of cysteine is regarded as the substrate- tions of strongly reduced GSH levels. Hence, our results limiting step in GSH synthesis (Ishii et al., 1987), and we do not support the idea of a putative threshold of GSH therefore investigated whether increased cystine uptake that decides on cell survival or cell death. In fact, system À impacts the intracellular GSH pool. We included BSO, xc drives a distinct cystine/cysteine cycle that efficiently the most powerful inhibitor of GSH synthesis, in our protects cells from detrimental lipid peroxidation, a very analyses to correlate survival and proliferation of early event in the onset of BSO-induced cell death control and xCT-overexpressing cells with intracellular starting about 8 h after the addition of BSO. Lipid GSH levels. Side effects of BSO could be excluded as peroxidation is initiated much earlier than ROS addition of GSH rescued the effect of BSO on survival accumulation (24 h) and breakdown of the mitochon- and proliferation of HH514 cells. At high BSO drial membrane potential (30 h) suggesting that lipid concentrations, the protective effect of xCT overexpres- peroxidation is the causal and initiating event in the sion on cell survival was less complete suggesting that course of cell death induction. This is strongly supported high BSO concentrations select for high xCT expression by our finding that cell death induced by BSO-mediated and cells with lower xCT expression are killed by BSO in GSH depletion is completely rescued by the lipophilic this experimental setting. Unexpectedly, xCT-over- antioxidant vitamin E. Breakdown of the mitochondrial expression does not fuel intracellular GSH levels either membrane potential is a comparatively late event that in the absence or in the presence of a wide range of presumably initiates the execution phase of cell death different BSO concentrations (Figure 3). In fact, the associated with broad caspase activation, DNA frag- increased cystine uptake leads to about six-fold increase mentation and Annexin V/PIstaining.

Oncogene The cystine/cysteine redox cycle A Banjac et al 1625 with increased glutamate toxicity, neuronal death and seizures. In the hematopoietic system, the particular sensitivity of the cells to ionizing radiation may be related to a limited uptake capacity of cystine. Limited activity of the cystine/cysteine cycle may also be necessary for an orchestrated regulation of apoptosis À in the immune system. Low activity of system xc in B and T cells may be a prerequisite to render the immune system highly dynamic. Redox regulation also appears to be crucially important to fine tune the T helper cell response towards Th1 or Th2 (Peterson et al., 1998). In the present work, we provide strong evidence that À the cystine/cysteine redox cycle, driven by system xc ,is characterized by slightly increased intracellular and exceedingly high extracellular cysteine concentrations, and efficiently protects cells from oxidative stress- induced cell death. The previous paradigm of augmented GSH levels due to increased cystine availability is not supported by our findings. In fact, we provide experimental evidence that increased cysteine uptake does not alter cellular GSH levels to a significant extent; hence, the cystine/cysteine cycle by itself might be regarded as a major redox system regulating cell survival and cell death. Figure 7 BSO-induced lipid peroxidation and cell death can be prevented by vitamin E (a) and xCT overexpression (b). Lipid peroxidation was monitored by C11-BODIPY staining. Materials and methods

Cell line and chemicals The BL cell line HH514 cells was cultured in RPMI1640 The bulk antioxidant function of cellular GSH can medium containing 10% fetal bovine serum (FBS) (Biochrom, thus be substituted by the cystine/cysteine cycle, though Berlin, Germany), 100 U mlÀ1 penicillin, 100 mgmlÀ1 strepto- it cannot be ruled out at present whether small amounts mycin and 2 mM glutamine. L-[14C]Cystine was purchased from of GSH are indispensable for other aspects in cellular Amersham Corp. (Freiburg, Germany) and other chemicals functions. For instance, it was shown recently that GSH from Sigma (Deisenhofen, Germany), Invitrogen (Karlsruhe, is essential for cytosolic Fe–S cluster formation in yeast Germany), and ICN (ICN Biomedicals GmbH, Eschwege, Germany). Restriction and DNA-modifying enzymes and (Sipos et al., 2002). Independent evidence that bulk DNA linkers were obtained from MBIFermentas (Vilnius, GSH is dispensable for cell proliferation at least in vitro Lithunia) and Biolabs GmbH (Schwalbach, Germany). has been provided by genetic means. Targeted inactiva- tion of g-GCS in mice revealed that GSH is essential for embryonic development, but unexpectedly not required Cloning of mouse and human xCT into an expression vector and stable expression in HH514 cells for proliferation of blastocyst-derived cell lines in the The BclIsite of pSPORT-mxCT was converted into an EcoRI presence of N-acetylcysteine (Shi et al., 2000). In vitro site and the HindIII site into a NheIsite. Murine xCT cDNA experiments with CaCo2 cells showed that the extra- was transferred as a HindIII/NheIfragment into the vector cellular redox potential defined by the ratio of extra- p141CAG-3SIP driving expression from a CMV enhancer- cellular cysteine to cystine is critical for cell proliferation chicken-b-actin promoter (Niwa et al., 2002; Okita et al., and independent of GSH (Jonas et al., 2002; Sonoda 2004). For cloning of human xCT, a synthetic linker carrying et al., 2004). EcoRI, BstBI, BglII and NheIsites (MCa: 5 0-ATTTCATTCG The importance of the cystine/cysteine cycle for the AACGG AGATCTTG-30; MCb: 50-CTAGCATGA TCTCC 0 resistance of cancer cells to chemotherapy has been GTTCGAATG-3 ) was cloned in p141CAG-3SIP digested recognized only recently (Okuno et al., 2003). Activation with EcoRIand HindIII. Human xCT cDNA was excised from a pBluescript SK clone as a ClaI/BglII fragment and inserted of the xCT gene in cancer cells by cytotoxic drugs may into the modified p141CAG-3SIP vector between the BstBI be a mechanism contributing to chemotherapy resis- and BglII sites. Transfections of the cell line HH514 with xCT- tance in vivo, a possibility that apparently deserves expressing or empty vectors were performed as described further studies. It is another important issue whether the previously (Brielmeier et al., 1998). Stably transfected clones limitation of cystine uptake, as observed in various cell were selected with puromycin at a final concentration of lines, has any pathophysiological impact. As proposed 2 mgmlÀ1. by Ye and Sontheimer (1999), in the central nervous system, deregulated activity of the cystine/cysteine cycle, Northern blot analysis as observed in human malignant glioma cells, leads to a Total RNA was isolated using Qiagen RNeasy Midi Kit detrimental increase in secreted glutamate associated according to the manufacturer’s instructions (Qiagen, Hilden,

Oncogene The cystine/cysteine redox cycle A Banjac et al 1626 Germany). Total RNA (10 mg) per cell clone were blotted and Determination of the mitochondrial membrane potential detection of human and murine xCT-expression was per- In all, 2 Â 105 cells were collected and stained with 5,50,6,60- formed with radiolabelled human- and mouse-specific cDNA tetrachloro-1,10,3,30-tetraethyl-benzimidazolycarbocyanin iodide probes. (JC-1; Biotium, Inc., Hayward, CA, USA) at a final concentration of 2.5 mgmlÀ1 in 500 ml PBS as described by von Haefen et al. (2004). After incubation for 30 min at 37 1C, cells were Measurement of L-cystine transport activity washed and resuspended in 200 ml PBS. The mitochondrial Uptake of L-cystine was measured as described (Novogrodsky membrane potential was measured by flow cytometry using et al., 1979; Sato et al., 1995). a FACScan (Becton Dickinson, Heidelberg, Germany) equip- ped with CellQuest software. Data are presented as per- Viability assay centage of cells with lowered membrane potential (DCm). Cells were plated in 6-well plates at 1 Â 105 cells mlÀ1 in 3 ml standard medium supplemented with different BSO concentrations. The number of viable cells was determined by trypan Measurement of DNA fragmentation blue exclusion or PIstaining by flow cytometry. To measure genomic DNA fragmentation, the nuclear DNA content was determined by flow cytometry (Gillissen et al., 2003). Overall, 2 Â 105 cells were pelleted in a 96-well U- Glutathione content (Tietze) bottom plate and fixed in 200 mlof2%(vvÀ1) formaldehyde in Total glutathione (GSH and GSSG) was extracted with 5% PBS on ice for 30 min. After fixation, DNA was precipitated trichloroacetic acid from cells grown in 10 or 200 ml flasks. with 100% ethanol for 15 min, pelleted and resuspended in Total glutathione was measured as outlined previously PBS containing RNaseA (40 mgmlÀ1). Following incubation at (Bannai and Ishii, 1982). The method is based on the reduction 37 1C for 30 min, the pellet was resuspended in 200 mlof of 5,50-dithiobis-2-nitrobenzoic acid by GSH which in turn is propidium iodide (PI, 50 mgmlÀ1) and incubated overnight in reduced by glutathione reductase (Tietze, 1969). Intracellular the dark at 4 1C. Nuclear DNA fragmentation was quantified GSH was additionally determined by HPLC analysis (see by flow cytometric determination of hypodiploid DNA. Data below). were collected and analysed using FACScan flow cytometer with CellQuest software. Determination of total thiols in the medium Total mercaptans secreted into the cell culture medium were Immunofluorescence determined as described previously (Bannai and Ishii, 1982). Surface expression of death receptors was determined via GSH was used as a standard. direct immunofluorescence by the use of a FACScan flow cytometer (Becton-Dickinson, Heidelberg, Germany) on a HPLC determination of extra- and intracellular cysteine single cell level. PE-labelled antibodies for CD95/Fas (clone Determination of cysteine in cell culture supernatants was DX2), TRAIL receptor 1/DR4 (clone DJR1), Trail R2/DR5 performed by a HPLC after derivatization of the free (clone DJR2-4) and a PE-labelled non-binding control anti- sulfhydryl group with SBDF (7-fluoro-benzo-2-oxa-1,3-dia- body (clone P3, mouse IgG1 k) were from eBioscence (San zole-4-sulphonate) as described previously with slight mod- Diego, CA, USA). Histograms are shown for 10 000 events ifications (Feussner et al., 1997). A measure of 2 Â 106 cells acquired in a 256-channel resolution. Percentages of positive were washed, resuspended in 2 ml of serum-free medium and cells were determined by setting markers in relation to control incubated for 2 h at 37 1C. Cells were collected by centrifuga- stained cultures. tion and 100 ml of the supernatant were mixed with 100 ml2M borate buffer (containing 5 mM Na2EDTA, pH 10.05 and 100 ml SBDF (1 mg mlÀ1 in borate buffer)) and incubated at Protein extraction and immunoblotting 60 1C for 60 min. The reaction was performed without any Cells were washed twice with PBS and lysed in buffer reducing agent to detect free cysteine only. The reaction was containing 10 mM Tris–HCl pH 7.5, 300 mM NaCl, 1% Triton stopped by incubation on ice for 5 min. The sample was X-100, 2 mM MgCl2,5mM EDTA, 1 mM pepstatin, 1 mM leupeptin and 0.1 mM phenylmethylsulfonyl fluoride. Protein deproteinized by adding 200 ml HClO4 containing 0.5 M EDTA and centrifuged. A volume of 50 ml of the supernatant was concentration was determined using the bicinchoninic acid injected in a HPLC column and analysed as described assay (Pierce, Rockford, IL, USA), and equal amounts of previously. Cysteine was used as a standard. protein (usually 20 mg per lane) were separated by SDS– For the determination of intracellular cysteine levels, cell PAGE. Immunoblotting was performed as described (Wieder pellets were resuspended and sonified in 500 ml borate buffer et al., 2001). For protein visualization the following antibodies (2 M, pH 9.5). After centrifugation, 200 ml of the supernatant were used: the mouse monoclonal antibodies anti-Bax YTH- was reduced by adding 20 ml of tri-n-butylphosphine for 30 min 2D2 (1:10 000, Trevigen, Gaithersburg, MA, USA) and anti- Bcl-2 100/D5 (1:100) (Novocastra, Newcastle upon Tyne, at 4 1C, deproteinized using HClO4, and centrifuged again. A volume of 100 ml of the supernatant was mixed with 250 mlof UK), and the rat monoclonal antibody anti-Bcl-w (16H12, 2 M borate buffer and thiol groups were derivatized using 1:1000) (Alexis). Anti-PUMA (N-19, 1:100), anti-Bid (C-20, SBDF. HPLC analysis was performed as described above. 1:1000) and anti-Nbk/Bik (N-19, 1:1000) were polyclonal goat antibodies from Santa Cruz. All other reagents were polyclonal rabbit antibodies used in a dilution of 1:1000. Anti-Bak Quantification of lipid peroxidation and intracellular reactive antibodies were purchased from Dako, anti-Bcl-x and anti- oxygen species (ROS) Bad antibodies from Transduction Laboratories, anti-Bim and Lipid peroxidation was measured by FACS analysis using anti-p53 from PharMingen, and anti-b-actin antibodies from C11-BODIPY. ROS were detected by flow cytometry using Sigma. Protein bands were detected using the enhanced DCF according to the manufacturer’s instructions (Molecular chemiluminescence system (Amersham Buchler, Braunsch- Probes, Inc., OR, USA). weig, Germany).

Oncogene The cystine/cysteine redox cycle A Banjac et al 1627 Detection of active caspases performed using FACScan and CellQuest analysis software Active caspases were determined by the FLICA apoptosis (Becton Dickinson, Heidelberg, Germany). detection kit (Alexis GmbH, Gruenberg, Germany) according to the manufacturer’s instructions. The methodology is based Statistical analysis on the covalent binding of a fluorochrome-labelled inhibitor of t caspases (FLICA) to the active site (Ekert et al., 1999). These Comparison between groups was made by the Student’s -test. P inhibitors (VDVAD-AFC for caspase 2; DEVD-AFC for A difference between groups with o0.05 was considered caspase 3; IETD-AFC for caspase 8 and LEHD-AFC for significant. caspase 9) are cell permeable and non-cytotoxic. The number of cells with active caspase was determined by FACScan flow Acknowledgements cytometry using the CellQuest software. FITC-labelled FLI- CAs were used for staining and positive cells were measured in We thank A Richter for expert technical assistance, A Klo¨ pfer the FL-1 green channel. for technical advice and J-M Beche´ t for critical reading of the manuscript. We are grateful to T Schroeder for the Measurement of cell death by Annexin-V-FITC and propidium p141CAG-3SIP vector, W Droege and A Roscher for helpful iodide staining discussions. This work was supported by a grant of Deutsche Cell death was determined by staining cells with Annexin-V- Forschungsgemeinschaft (Priority Programme ‘Biology of FITC and counterstaining with PI. Cells were washed twice Selenoproteins’) to GWB and MC, by a short-term fellowship with cold PBS and resuspended in buffer containing 10 mM of Deutsche Akademische Austauschdienst (DAAD), and N-(2-hydroxyethyl)piperazin-N0-3(propansulfonic acid)/NaOH, Japanese Society for Promotion of Science (JSPS) to HS and pH 7.4, 140 mM NaCl, 2.5 mM CaCl2. Next, 5 ml of Annexin-V- MC for working in Munich and in Tsuruoka, respectively. FITC (BD PharMingen, Heidelberg, Germany) and 10 ml GWB was additionally supported by Fonds der Chemischen PI(20 mgmlÀ1, Sigma-Aldrich) were added. Analyses were Industrie.

References

Arner ES, Holmgren A. (2000). Physiological functions of human plasma by isocratic high-performance liquid chro- thioredoxin and thioredoxin reductase. Eur J Biochem 267: matography. Eur J Clin Chem Clin Biochem 35: 687–691. 6102–6109. Gillissen B, Essmann F, Graupner V, Starck L, Radetzki S, Bannai S, Ishii T. (1982). Transport of cystine and cysteine and Dorken B et al. (2003). Induction of cell death by the BH3- cell growth in cultured human diploid fibroblasts: effect of only Bcl-2 homolog Nbk/Bik is mediated by an entirely Bax- glutamate and homocysteate. J Cell Physiol 112: 265–272. dependent mitochondrial pathway. EMBO J 22: 3580–3590. Bannai S, Tateishi N. (1986). Role of membrane transport in Griffith OW. (1982). Mechanism of action, metabolism, and metabolism and function of glutathione in mammals. toxicity of buthionine sulfoximine and its higher homologs, J Membr Biol 89: 1–8. potent inhibitors of glutathione synthesis. J Biol Chem 257: Bannai S, Sato H, Ishii T, Sugita Y. (1989). Induction of 13704–13712. cystine transport activity in human fibroblasts by oxygen. Hattori H, Imai H, Furuhama K, Sato O, Nakagawa Y. J Biol Chem 264: 18480–18484. (2005). Induction of phospholipid hydroperoxide glu- Brielmeier M, Bechet JM, Falk MH, Pawlita M, Polack A, tathione peroxidase in human polymorphonuclear neutro- Bornkamm GW. (1998). Improving stable transfection phils and HL60 cells stimulated with TNF-alpha. Biochem efficiency: antioxidants dramatically improve the outgrowth Biophys Res Commun 337: 464–473. of clones under dominant marker selection. Nucleic Acids Ishii T, Sugita Y, Bannai S. (1987). Regulation of glutathione Res 26: 2082–2085. levels in mouse spleen lymphocytes by transport of cysteine. Broome JD, Jeng MW. (1973). Promotion of replication in J Cell Physiol 133: 330–336. lymphoid cells by specific thiols and disulfides in vitro. Jakupoglu C, Przemeck GK, Schneider M, Moreno SG, Mayr Effects on mouse lymphoma cells in comparison with splenic N, Hatzopoulos AK et al. (2005). Cytoplasmic thioredoxin lymphocytes. J Exp Med 138: 574–592. reductase is essential for embryogenesis but dispensable for Conrad M, Jakupoglu C, Moreno SG, Lippl S, Banjac A, cardiac development. Mol Cell Biol 25: 1980–1988. Schneider M et al. (2004). Essential role for mitochondrial Jonas CR, Ziegler TR, Gu LH, Jones DP. (2002). Extracellular thioredoxin reductase in hematopoiesis, heart development, thiol/disulfide redox state affects proliferation rate in a and heart function. Mol Cell Biol 24: 9414–9423. human colon carcinoma (Caco2) cell line. Free Radic Biol Droge W, Eck HP, Gmunder H, Mihm S. (1991). Modulation Med 33: 1499–1506. of lymphocyte functions and immune responses by cysteine Kuppner MC, Scharner A, Milani V, Von Hesler C, Tschop and cysteine derivatives. Am J Med 91: 140S–144S. KE, Heinz O et al. (2003). Ifosfamide impairs the Ekert PG, Silke J, Vaux DL. (1999). Caspase inhibitors. Cell allostimulatory capacity of human dendritic cells by Death Differ 6: 1081–1086. intracellular glutathione depletion. Blood 102: 3668–3674. Falk MH, Hultner L, Milner A, Gregory CD, Bornkamm GW. Meister A. (1995). Glutathione metabolism. Methods Enzymol (1993). Irradiated fibroblasts protect Burkitt lymphoma cells 251: 3–7. from apoptosis by a mechanism independent of bcl-2. Int J Melino G. (2005). Discovery of the ubiquitin proteasome Cancer 55: 485–491. system and its involvement in apoptosis. Cell Death Differ Falk MH, Meier T, Issels RD, Brielmeier M, Scheffer B, 12: 1155–1157. Bornkamm GW. (1998). Apoptosis in Burkitt lymphoma cells Metcalf D. (1976). Role of mercaptoethanol and endotoxin in is prevented by promotion of cysteine uptake. Int J Cancer stimulating B lymphocyte colony formation in vitro. 75: 620–625. J Immunol 116: 635–638. Feussner A, Rolinski B, Weiss N, Deufel T, Wolfram G, Murphy TH, Miyamoto M, Sastre A, Schnaar RL, Coyle JT. Roscher AA. (1997). Determination of total homocysteine in (1989). Glutamate toxicity in a neuronal cell line involves

Oncogene The cystine/cysteine redox cycle A Banjac et al 1628 inhibition of cystine transport leading to oxidative stress. glutamate transporter-deficient mice. J Biol Chem 280: Neuron 2: 1547–1558. 37423–37429. Murphy TH, Schnaar RL, Coyle JT. (1990). Immature cortical Sato H, Tamba M, Ishii T, Bannai S. (1999). Cloning and neurons are uniquely sensitive to glutamate toxicity by expression of a plasma membrane cystine/glutamate ex- inhibition of cystine uptake. FASEB J 4: 1624–1633. change transporter composed of two distinct proteins. J Biol Nicotera P, Melino G. (2004). Regulation of the apoptosis- Chem 274: 11455–11458. necrosis switch. Oncogene 23: 2757–2765. Shi ZZ, Osei-Frimpong J, Kala G, Kala SV, Barrios RJ, Niwa H, Masui S, Chambers I, Smith AG, Miyazaki J. (2002). Habib GM et al. (2000). Glutathione synthesis is essential Phenotypic complementation establishes requirements for for mouse development but not for cell growth in culture. specific POU domain and generic transactivation function Proc Natl Acad Sci USA 97: 5101–5106. of Oct-3/4 in embryonic stem cells. Mol Cell Biol 22: Sipos K, Lange H, Fekete Z, Ullmann P, Lill R, Kispal G. 1526–1536. (2002). Maturation of cytosolic iron–sulfur proteins requires Nonn L, Williams RR, Erickson RP, Powis G. (2003). The glutathione. J Biol Chem 277: 26944–26949. absence of mitochondrial thioredoxin 2 causes massive Sonoda Y, Aiba N, Utsubo R, Koguchi E, Hasegawa M, apoptosis, exencephaly, and early embryonic lethality in Kasahara T. (2004). Induction of antioxidant enzymes by homozygous mice. Mol Cell Biol 23: 916–922. FAK in a human leukemic cell line, HL-60. Biochim Biophys Novogrodsky A, Nehring Jr RE, Meister A. (1979). Inhibition Acta 1683: 22–32. of amino acid transport into lymphoid cells by the glutamine Tietze F. (1969). Enzymic method for quantitative determina- analog L-2-amino-4-oxo-5-chloropentanoate. Proc Natl tion of nanogram amounts of total and oxidized glutathione: Acad Sci USA 76: 4932–4935. applications to mammalian blood and other tissues. Anal Okita C, Sato M, Schroeder T. (2004). Generation of Biochem 27: 502–522. optimized yellow and red fluorescent proteins with distinct Verrey F, Meier C, Rossier G, Kuhn LC. (2000). Glycopro- subcellular localization. Biotechniques 36: 418–422, 424. tein-associated amino acid exchangers: broadening the range Okuno S, Sato H, Kuriyama-Matsumura K, Tamba M, Wang of transport specificity. Pflugers Arch 440: 503–512. H, Sohda S et al. (2003). Role of cystine transport in von Haefen C, Gillissen B, Hemmati PG, Wendt J, Guner D, intracellular glutathione level and cisplatin resistance in Mrozek A et al. (2004). Multidomain Bcl-2 homolog Bax but human ovarian cancer cell lines. Br J Cancer 88: 951–956. not Bak mediates synergistic induction of apoptosis by Peterson JD, Herzenberg LA, Vasquez K, Waltenbaugh C. TRAIL and 5-FU through the mitochondrial apoptosis (1998). Glutathione levels in antigen-presenting cells mod- pathway. Oncogene 23: 8320–8332. ulate Th1 versus Th2 response patterns. Proc Natl Acad Sci Wieder T, Essmann F, Prokop A, Schmelz K, Schulze-Osthoff USA 95: 3071–3076. K, Beyaert R et al. (2001). Activation of caspase-8 in drug- Sato H, Fujiwara K, Sagara J, Bannai S. (1995). Induction of induced apoptosis of B-lymphoid cells is independent of cystine transport activity in mouse peritoneal macrophages CD95/Fas receptor-ligand interaction and occurs down- by bacterial lipopolysaccharide. Biochem J 310: 547–551. stream of caspase-3. Blood 97: 1378–1387. Sato H, Shiiya A, Kimata M, Maebara K, Tamba M, Ye ZC, Sontheimer H. (1999). Glioma cells release excitotoxic Sakakura Y et al. (2005). Redox imbalance in cystine/ concentrations of glutamate. Cancer Res 59: 4383–4391.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene