B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te /b ra i n res

Research Report

Erythropoietin-induced neuroprotection requires cystineglutamate exchanger activity

Brian Simsa,b,d, Melinda Clarkea, Wilfred Njaha,E'lana Shuford Hopkinsa, Harald Sontheimerc,d,⁎aDepartment of Pediatrics, University of Alabama at Birmingham, USAbDepartment of Cell Biology, University of Alabama at Birmingham, USAcDepartment of Neurobiology, University of Alabama at Birmingham, USAdCenter for Glial Biology in Medicine, University of Alabama at Birmingham, USA

A R T I C L E I N F O

⁎ Corresponding author. Department of Neuro35294, USA. Fax: +1 205 934 3100.

E-mail address: sontheimer@nrc.uab.eduAbbreviations: Epo, Erythropoietin; CUB,

Ciliary neurotrophic factor; LIF, Leukemia inhexchanger; xCT, Catalytic subunit of system

0006-8993/$ – see front matter © 2010 Elsevidoi:10.1016/j.brainres.2010.01.040

A B S T R A C T

Article history:Accepted 13 January 2010Available online 25 January 2010

Erythropoietin (Epo) has been used for many years in neonates for the treatment of anemiaof prematurity. Epo has also been proposed for treatment of neonatal brain injury, asmounting evidence suggests neuroprotective properties for Epo. However, Epo'sneuroprotective mechanism of action is poorly understood. In this study we hypothesizedthat Epo may confer neuroprotection by enhancing cellular redox defense brought about bycellular glutathione (GSH). This was examined in cultures of differentiated cortical neuralstem cells and using the B104 cell line as model systems. Our data shows that Epo causes atime- and dose-dependent increase in expression and activity of systemXc−, the transporterresponsible for uptake of cystine for the production of glutathione. Cystine uptake increases3–5 fold in differentiated neural stem cells and B104 cells treated with Epo. Exposure of cellsto 100 μMkainate suppressed cellular GSH and caused excitotoxicity, but GSH levels and cellviability were completely restored by Epo in the continued presence of kainate. This rescueeffect of Epo vanished if system Xc− was inhibited pharmacologically using S4-CPG in thepresence of Epo leading to marked cell death of B104 cells and cultured mouse corticalneural stem cells. This could also be achieved using xCT siRNA to decrease xCT expression.This data suggests that system Xc− activity and protein expression are positively regulatedby Epo directly explaining its neuroprotective effect.

© 2010 Elsevier B.V. All rights reserved.

Keywords:ErythropoietinCystine glutamate exchangerxCTNeuroprotection

1. Introduction

Erythropoietin (Epo) is a glycoprotein hormone whose majorfunction is to regulate erythropoiesis (Krantz, 1995). It is a

biology, Center for Glial B

(H. Sontheimer).Cystine uptake buffer; Gibitory factor; NO, Nitric oXc−; GSH, Glutathione; PF

er B.V. All rights reserved

member of the cytokine superfamily which also includesgranulocyte-macrophage colony-stimulating factor (GM-CSF),ciliary neurotrophic factor (CNTF), interleukin 3 and 6,prolactin, cardiotrophin 1 (CT1), leukemia inhibitory factor

iology in Medicine, UAB, 1719 6th Avenue South, Birmingham, AL

M-CSF, Granulocyte–monocyte colony-stimulating factor; CNTF,xide; NMDA, N-methyl d-aspartate; System Xc, Cystine glutamateA, Paraformaldehyde

.

89B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

(LIF) and thrombopoietin (Seko et al., 1998; Uppenkamp et al.,1998) in addition to its role in erythropoiesis, it appears to havean important role in the central nervous system.

Epo's essential role in development became apparent inknockout studies since animals with the targeted deletion ofthe Epo receptor died prematurely at embryonic day 13.5(E13.5) from severe anemia (Yu et al., 2002). Although theseand other studies demonstrate an absolute requirement forEpo signaling in hematopoiesis, it appears to exert a secondaryneuroprotective role during brain development that is lesswell understood (Liu et al., 2005). Epo prevents neural apo-ptosis in the embryonic brain and mice lacking EpoR showsigns of severe neuronal apoptosis. In EpoR knockout animals,neural progenitor cells were markedly reduced and culturingneurons with a functional receptor showed amarked increasein cell viability compared to knockouts.

Epo-induced neuroprotection has been associated withprocesses such as anti-inflammation, anti-apoptosis andneuroregeneration (Sola et al., 2005). These processes requiremultiple genes and signaling pathways. The pleiotropic effectsof Epo are exerted through the EpoR by three distinct path-ways. The common link in these three pathways is the phos-phorylation of Janus Kinase (JAK) 2. Afterwards, there isactivation of signal transducer and activator of transcrip-tion (STAT) 5, the phosphoinositide 3-kinase (PI3K)-serine-threonine kinase AKT and nuclear factor kappa B (NFκB)pathway (Sola et al., 2005). Currently, it is not known how thecell determines which pathway it utilizes to exert the effect ofEpo. Speculatively, there may be an interplay between theEpo receptor and other receptors that ultimately dictates thecumulative effect. Cell type may also play a major role in thisprocess.

Epo is produced in the brain under conditions of nitrosativeand oxidative stress and the Epo gene is upregulated underhypoxic conditions (Buemi et al., 2002). Oxygen and glucosedeprivation (OGD) of neurons causes cell death in vitro. Theneuronal death is attenuated by adding supernatant fromastrocytes exposed to the same conditions. The effect wasblocked by soluble EpoR demonstrating that Epo acts as aparacrine hormone (Ruscher et al., 2002). Excitotoxic condi-tions, for example exposure of cells to N-methyl-D-aspartate(NMDA) or nitric oxide (NO), which lead to the formation offree radicals which are attenuated in the presence of Epo(Digicaylioglu and Lipton, 2001) (Yamamoto et al., 2004).Calapai et al. (2000) demonstrated that Epo decreases NOsynthase activity in human vascular endothelial cells andprotects neurons against ischemia-induced cell death. Down-stream effectors of Epo are poorly understood. However, onegene that is upregulated after Epo exposure is glutathioneperoxidase, an enzyme involved in regenerating free glutathi-one (Genc et al., 2002).

In this study we set out to test the hypothesis that gluta-thione levels are increased in the presence of Epo due to anupregulation of the cystine glutamate exchanger, system Xc−,which is responsible for the cellular import of cystine for thesynthesis of glutathione. System Xc− is a dimeric amino-acidtransporter which couples to equimolar exchange of cystinewith the release of glutamate. The transporter exists as adimer composed of a regulatory subunit (4F2hc-CD98) anda catalytic subunit (xCT). The latter belongs to a family of

amino-acid transporters which has 12 transmembranedomains (Sato et al., 2000, 2002). The heavy chain 4F2hcis responsible for membrane translocation (Wells et al., 1992).Following uptake, cystine is reduced to cysteine, which isconjugated with glutamate and glycine, to form glutathione(L-γ-glutamyl-L-cysteinylglycine=GSH). GSH is a ubiquitoustripeptide involved in the regulation of the cells redox state,and protects cells from reactive oxygen (ROS) and nitrogenspecies. Potential reactive oxygen species include nitric oxide,hydrogen peroxide, superoxide anion, and hydroxyl radicals,many of which are generated in hypoxia–ischemia, excito-toxicity and infection. Thus, it is plausible that dysfunction ordevelopmental expression of system Xc− could be responsiblefor decrease ability of precursor cells to defend against reac-tive oxygen species.

2. Results

To examine the potential role of Epo in protecting neural cells,we used two model systems from mouse and human thathave been previously demonstrated to express Epo receptors(EpoR)were themouse neural stem cells (NSCs) (Yu et al., 2002)and B104 human neuroblastoma cells (Ribatti et al., 2007)(Sartelet et al., 2007). We validated that in our hands, bothexpress the EpoR at the mRNA and protein levels (Figs. 1Aand B) along with expression of xCT, the catalytic subunit ofsystemXc (Figs. 1A and B). Importantly, xCT and EpoR proteinscolocalize as demonstrated by immunohistochemistry inFigs. 1C–H for both cell systems.

Glutamate toxicity data had not been established in our cellline so it was established in Fig. 2A (NSCs) and Fig. 2B (B104s).Both cell lines had an increase in apoptosis with increasingconcentrations of glutamate. To demonstrate that Epo indeedconfers neuroprotection, we challenged differentiated NSCswith 3 mM glutamate yielding a significant (p≤0.01), >75% lossof viable cells (Fig. 3A). Cultures that received cotreatmentwith 10 ng/ml Epo and glutamate challenge were essentiallyunaffected by 3 mMGlu showing a small, 10% reduction in cellviability. However, if Epo treated NSCs were also treated with250 μM of the system Xc− inhibitor S-4CPG, 72% of cells died(p≤0.01). These experiments suggest that Epo is protectingNSCs from Glu toxicity in a manner that engages system Xcactivity. Since the latter is responsible for the cellular importof cystine to generate glutathione (GSH) we sought to examinethe effect of these treatments on cystine uptake directly.Indeed, as illustrated in Fig. 3B, cystine uptake into differen-tiated NSCs was elevated 5-fold in the presence of Epoindicating an upregulation of systemXc expression or activity.Similar data was obtained in B104 cells (Fig. 3C) with a 75%reduction in cell viability (p≤0.01) and >70% (p≤0.01) decreasein cell viability when exposed to S4-CPG. In Fig. 3D, B104 cellshad a 3-fold (p≤0.01) increase in cystine uptake in the presenceof Epo.

To examine whether the increase cystine uptake resultedfrom an increase in system Xc− expression, we next studiedthe time-dependence of xCT protein levels following exposureof cells to 10 ng/ml Epo for 4–24 h along with actin used as aloading control. Similar blots were obtained from 3 differentexperiments in both cell lines and in each case the protein

Fig. 1 – EpoR and xCT expression in B104 cells. (A.) RT-PCR of 2 μg of total RNA collected from B104 cells (B.) Immunoblotof EpoR, xCT and Actin. B104 cells (C.) xCT (TRITC), (E.) EpoR (FITC), (G.) combined. Differentiated neural stem cells(D.) xCT (TRITC), (F.) EpoR (FITC) and (H.) combined.

Fig. 2 – Glutamate toxicity in NSCs and B104.(A.) Immunoblot of NSC protein extract probed for cleavedcaspase 3 and actin at increasing concentrations of glutamate(1 — 0 µM glu, 2 — 100 μM glu, 3 — 250 μM glu, 4 — 500 μMglu, 5— 1 mMglu, 6— 3 mMglu). (B.) Immunoblot of B104 cellextract probed for cleaved caspase 3 and actin at increasingconcentrations of glutamate (1 — 0 µM glu, 2 — 100 μM glu,3 — 250 μM glu, 4 — 500 μM glu, 5 — 1 mM glu,6 — 3 mM glu).

90 B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

concentration was normalized to actin levels allowing evalu-ation of changes in protein concentration. NSCs treated with10 ng/ml of Epo (Fig. 4A) showed a statistical increase inprotein level at 12 and 24 h, >1.2 fold (p≤0.05) increase and>1.6 fold (p≤0.01) increase. The resulting data, shown in Fig. 4Bshow a significant increase in xCT expression as early as 4 hafter addition of Epo. The xCT/actin ratios in Fig. 4B, expressedas fold increase were significant at 4, 8, 12 and 24 h, respec-tively. The Epo concentration of 10 ng/ml was chosen becauseit was closest to the half-maximal effect of B104 cells. B104cells would be used for all experiments including the siRNAexperiments.

To examine whether Epo showed a dose-dependence in itseffect on system Xc− expression we administered Epo to NSCsandB104 cells at increasing concentrations, using 0, 5, 10, 25and50 ng/ml Epo respectively. In Fig. 5A, a representative Westernblot shows the corresponding increases in xCT expressionover 24 h.Normalizing xCT expression to actin expression (xCT/actin) shows a statistically significant increase in xCT expres-sion at all concentrations tested. In NSCs, fold increase overcontrol was statistically significant at 5, 10, 25 and 50 ng/ml butonly statistically significant at 25 and 50 ng/ml in B104 cells.

Since system Xc− provides the cystine as substrate forglutathione production synthesis, these data suggest thatglutathione levels may actually increase following Epo expo-sure. To examine this directly, we measured total GSH levelsin cells treated with 100 μM kainate acid in the presence orabsence of Epo (Fig. 6). Kainate was used in these experimentsto negate the role of glutamate in the formation of glutathione.

Fig. 3 – Cell viability and cystine uptake in differentiated neural stem cells. (A.) Total cell count in control, 3 mM glutamate,3 mM glutamate+Epo (10 ng/ml), 3 mM glutamate+Epo (10 ng/ml), 3 mM glutamate+Epo (10 ng/ml)+S4-carboxyphenylglycine(250 μM). (B.) Cystine uptake in differentiated NSCs — control, 3 mM glutamate and 3 mM glutamate+Epo (10 ng/ml).Cell viability and cystine uptake in B104 cells (C.) Total cell count in control, 3 mM glutamate, 3 mM glutamate+Epo (10 ng/ml),3 mM glutamate+Epo (10 ng/ml), 3 mM glutamate+Epo (10 ng/ml)+S4-carboxyphenylglycine (250 μM). (D.) Cystine uptake indifferentiated B104 cells — control, 3 mM glutamate and 3mM glutamate+Epo (10 ng/ml). *p≤0.05 using the paired studentt-test.

91B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

Total glutathione levels were normalized to protein concen-trations and expressed as fluorescence/mg protein. Kainatealone reduced GSH levels by 74% but GSH levels were restoredto control values with 10 ng/ml Epo and to 67% of control with5 ng/ml Epo supplementation.

Fig. 4 – Erythropoietin time course. (A.) Representative Western10 ng/ml of Erythropoietin. (B.) Relative xCT concentration (normprotein was added to each well. *p≤0.05 and **p≤0.01 using the p

These data suggest that in neural stem cells Epo protectsthrough the regulation of system Xc−. As all the above datapoint to system Xc− being an important downstream target ofEpo and the likely candidate that confers cellular protection inthe presence of Epo, we employed the use of siRNA to suppress

blot of system Xc− and actin in B104 cells treated withalized to actin) at 0, 4, 8, 12 and 24 hr time points. 25 μg ofaired student t-test.

Fig. 5 – Epo dose response. (A.) Representative Western blot of system Xc− in B104 cells with 0, 5, 10, 25, 50 ng/ml Epo.(B.) Relative xCT concentration (normalized to actin) at same concentrations listed in (A.). *p≤0.05 using the paired student t-test.

92 B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

system Xc− expression in B104 cells. Β104 cells were trans-fectedwith xCT siRNA at 5 ng, 10 ng and 20 ng/ml and samplescompared to a control and scrambled siRNA. Cells were platedat 500,000 per 6 well plates. In cells transfectedwith scrambledsiRNA there was an 18% decrease in cell number. xCT siRNAtransfected cells showed a 38%, 66% and 74% decrease in cellnumber after 24 h of transfection for 10 ng, 25 ng and 50 ng/mlxCT siRNA, p≤0.001 (Fig. 7A). The relative decrease in cellnumber for each sample was 1.2, 1.67, 2.88 and 3.84 fold forscrambled, 5 ng/ml, 10 ng/ml and 20 ng/ml xCT siRNA. All cellswere treated with 10 ng/ml of Epo. A representative Westernblot is shown to demonstrate the inhibition of xCT expression(Fig. 7B).

3. Discussion

Epo has been shown to be protective inmany systems includingliver (Joyeux-Faure, 2007), pancreatic islet cells (Fenjves et al.,2004), intestine (Guneli et al., 2007), pheochromocyctoma cells

Fig. 6 – Epo effect on total glutathione. B104 cells treatedwithkainate in the presence (5 ng/ml and 10 ng/ml) or absence ofEpo. Glutathione is expressed as fluorescence/mg protein.*p≤0.05 using the paired student t-test.

(Wu et al., 2007), vascular injury (Chong et al., 2003) and heartfailure (Latini et al., 2008). In these systems and others Epois known to be cytoprotective and its mechanism leads todecreased oxidative stress and decreased lipid peroxidation(Szczepanski et al., 2006). Mechanisms of Epo-induced

Fig. 7 – xCT siRNA effect on cell viability. Lanes 1–5 include(1) Control, (2) scrambled siRNA 50 ng/ml, (3) xCT siRNA10 ng/ml, (4) xCT siRNA 25 ng/ml and (5) xCT siRNA 50 ng/ml.(A) Total cell count. *p<0.05 and **p<0.01. (B) xCT and actinimmunoblot, 25 μg protein loaded to each well.

93B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

cytoprotection involvemultiple pathways including endothelialnitric oxide synthase (d'Uscio et al., 2007) and the induction ofheme oxygenase 1 (Katavetin et al., 2007).

In brain, multiple pathways are involved in Epo-inducedneuroprotection including the modulation of antiapoptoticfactors (Chong et al., 2007) such as caspases 1, 3, and 8 (Ghezziand Brines, 2004). Other mechanisms involve reduced lipidperoxidation and increased glutathione peroxidase (Katavetinet al., 2007) (Genc et al., 2002). These studies link glutathionemodulationas a potential pathway of protection but no studieshave examined the rate limiting step of glutathione synthesisas potential regulatory target of Epo.

This study was motivated by prior suggestions that Epo,in addition to its pleiotropic effects on a variety of celltypes in the body may also confer some neuroprotection(Hasselblatt et al., 2006). While previous studies could notdelineate the sole mechanism by which Epo confersneuroprotective effects, our findings propose that it maybe the result of a positive regulation of the cystineglutamate exchanger, system Xc−. Its upregulation causesan increase in cystine uptake and much enhanced produc-tion of the cellular antioxidant GSH. Addition of Epo in vitroshowed a dramatic increase in xCT protein expression in atime dependent manner and we found an increase in xCTafter addition of Epo in vitro at all time points studied.Importantly, Epo effects occurred relatively fast. Of courseone may argue that in order for Epo to have a protectiveeffect it must be quick to upregulate expression of genesthat could be protective. We did find a 2-fold increase insystem Xc− expression as early as 4 h after the insult(Fig. 4B) in B104 cells and this increase was maintained forat least 24 h. The time course is clinically relevant. Epo, inthese experiments, was administered concomitantly withglutamate. Clinically, this is still a useful strategy. Thereare a fair number of neonates who suffer episodes such assevere hypoxemia which could benefit from an agent likeEpo with neuroprotection as early as 4 h.

In earlier work cystine deprivation was shown to be toxic tooligodendroglial precursor cells (Back et al., 1998) which sug-gests that system Xc− activity may be vital for neural precursorcell survival and cellular protection. Other studies also confirmthat inhibition of cystine uptake is neurotoxic by a receptorindependent pathway (Rosin et al., 2004)which includes systemXc−. Glia which overexpressed Nrf2, a regulator of system Xc−,protected neurons from oxidative stress and showed that evenat a density of <1% these glia in mixed culture could protectneurons from oxidative stress (Shih et al., 2003). Conversely,in microglia, increased glutamate release was seen by systemXc− and its inhibition prevented neurotoxicity caused bymicroglia. Taken together, this may suggest that system Xc−

under certain circumstances may be neuroprotective and mayhave a cell specific role in glutamate metabolism.

We used B104 cells as a convenient model system for studysince EpoR is highly expressed in these cells and being a cellline, biochemical studies are readily feasible. However, themajor findings of our study were confirmed in differentiatedneural stem cells, the cell type most likely at risk of injury inpremature infants who routinely receive Epo. An importantquestion is the developmental expression of system Xc− andhow it may alter Epo's ability to confer neuroprotection. The

use of the mouse cortical neural stem cells clearly suggeststhat the immature brain will benefit from Epo, yet it remainsto be seenwhethermature neurons, harmed for example aftera stroke, would also be positively influenced by Epo. Suchstudies are underway.

Inhibition of xCT can be accomplished pharmacologicallyusing for example S4-CPG at a 250 μM concentration. WhenxCT was inhibited we found a mild decrease in cell viability inthe absence of an excitotoxic insult. xCT has been reported tobe neuroprotective when overexpressed (Shih et al., 2006) andits inhibition causes apoptosis in tumors (Chung et al., 2005).The fact that Epo loses any neuroprotective effects when xCTfunction is inhibited pharmacologically or using siRNA sug-gests very strongly that the observed upregulation of systemXc− expression by Epo is a necessary step in the observedneuroprotection. The scrambled siRNAused in the experimentshowed a 17% decrease in viability but was used at the highestconcentration of 50 ng/ml. This concentration of siRNA couldhave a toxic effect on cells. The decrease in total glutathionewhen S4-CPG is administered is supporting evidence thatthere is a substantial contribution to glutathione biosynthesisby system Xc−. Indeed, the knockdown experiments providedirect evidence for the molecular regulation of Epo-inducedneuroprotection by xCT.

Taken together, data presented suggest that Epo is a posi-tive regulator of system Xc−, which in turn confers neuropro-tection. A logical next step is to examine this activity acrossmultiple neural cells and if possible use in vivo model systemsthat examine excitotoxic injury in neonatal brain. If success-ful, these may provide strong rationale for the clinical use ofEpo in prenatal cerebral insults. Moreover, other synergisticdrugs may be identified that could further enhance the posi-tive transcriptional effects on system Xc−.

4. Experimental procedures

All equipment used was from Bio-Rad and chemicals fromSigma Chemical Company (St. Louis, MO) unless stated other-wise. RT-PCR primers were ordered from IDTDNA.

4.1. B104 neuroblastoma cells and cortical neuralstem cells

B104 cells were ordered fromATCC. Total cell count was deter-mined using a standard hemocytometer and Trypan Blue in allexperiments. B104 media consists of 0.75 g sodium bicarbon-ate, 10% newborn calf serum placed in Dulbecco's ModifiedEssential Media (DMEM). B104 cells were grown to 80% con-fluence and then split 1:6 every 3 days and plated in T 25 cmflasks. Epo treatment was given at the time of kainate/glutamate administration with equal cell counts. Treatmentfor all experiments was 24 h.

Neural stem cells were ordered from Chemicon as Cryo-preserved Mouse Cortical Neural Stem Cells (Catalog NumberSCR029). Polyornithine–laminin coated plates were madeusing concentrations of polyornithine (10 mg/ml) and laminin(5 μg/ml). Plates were incubated with polyornithine overnightand then washed and laminin solution added and allowed toincubate overnight then used directly or stored at −80 °C.

94 B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

Protocol for culturing neural stem cells was followed perprotocol. In brief, cells were grown in Neural Stem CellExpansion Medium (Chemicon Neural Stem Cell Medium,20 ng/ml Fibroblast Growth Factor 2 (FGF-2), 20 ng/ml Epider-mal Growth Factor (EGF) and 2 μg/ml heparin) and would bepassaged every 3–5 days. Cells were allowed to grow to 80%confluence then split 1:3 in T 25 cm flasks. Differentiation wasdone after passage 3. Differentiation was induced by with-drawing EGF, FGF-2 and heparin.

4.2. Cystine uptake

Sodium-independent cystine uptake was done using condi-tions described in Chung et al. (2005). In brief, plate 1×106 cells/well and allow 2 hours for cell adherence prior to experimentalconditions. Cells were treated with experimental conditionsand incubated overnight. After centrifugation, cells wereplaced in cystine uptake buffer (CUB) (122 mM choline-chloride, 1.8 mM KCl, 1.3 mM CaCl2, 1.2 mM potassium phos-phate, 25 mMTriethyammonium bicarbonate, 10 mM glucose,and 0.4 mM magnesium sulfate adjusted to pH 7.4) andwarmed to 37 °C. 500 μl of CUB is added to the pellet. 100 μl ofsample is used for protein determination. S35/C14-Cystine isadded to the remaining 400 μl volumeof CUBand incubated for10min. Equal volume of acid (0.3 N HCl) and base (0.3 N NaOH)are added for 10min with a wash between steps. The samplesare allowed to equilibrate for 10min. Radioactivity wasdetected using a S35/C14 detection program in a scintillationcounter.

4.3. Protein concentration assay

The Bio-Rad DC Protein Assay was used in order to determinethe protein concentrations of the samples. The absorbencieswere read on a spectrophotometer at 750 nm.

4.4. Western blot

Cells were triturated and pelleted then lysed with RIPA buffer.50 μg of protein was collected and ran on a Bio-Rad Ready GelTris 10% HCL. Samples were run at 120 V for 45min on ice. Gelwas transferred to nitrocellulose using the Trans Blot Semi-Dryapparatus for 15min at 15 V. Blots were washed three timeswith TBS–0.1% Tween (TBST). Membranes were blocked for30min with 3% milk solution. Primary antibodies were used at1:1000 dilutions in 3%milk. Primary antibodies were allowed toincubate overnight on a rotator. Primary antibodies used wereCleaved Caspase 3 (Asp175) (Cell Signaling), Actin (MAB1501-Chemicon), and System Xc (Novus Biological). Secondaryantibody was used at a 1:1000 dilution for 1 h on a rotator.After washing, blots were developed using the Pierce SuperSignal detection kit and developed on film with the Kodakdeveloper.

4.5. siRNA production

Primers designed from the Block-It siRNA program (Invitrogen)forxCTsiRNA (S1—5′-GGATCCTAATACGACTCACTATAGAT-3′,S2 — 5′-AAGTTATTCTGTCTGATAATCTATAGT-3′, S3 — 5′-GGATCCTAATACGACTCACTATAGTT-3′, and S4 — 5′-AAGA-

TTATCAGACAGAATAACTATAG-3′) were made using theRibomax T7 RNA Polymerase System. Cells were then trans-fected with siRNA using the CodeBreaker siRNA TransfectionReagent protocol. Cells were grown for 24–48 h then cells lysedfor Western blot analysis after total cell count. RNA wasnormalized using the Quant-it Ribogreen RNA quantificationkit.

4.6. Reverse transcriptase polymerase chain reaction

Total RNAwas prepared using the Trizol reagent and protocol.1 μg of total RNA was used for each experiment. The QiagenOne Step RT-PCR kit was used. Actin, EpoR and xCT primerswere ordered from IDTDNA with the following sequence:Actin, forward 5′-ATCATGTTTGAGACCTTCAACAC-3′, reverse5′-TCTGCGCAAGTTAGGTTTTGTC-3′; EpoR, forward 5′-GTCTGACTTGGACTCAAAGC-3′, reverse 5′-CTCAGTCTG-GGACCTTCTGC-3′ and xCT, forward 5′-GCAGTCGCAGGAC-TGATTTA-3′, reverse 5′-GAGAGGCACCTTGAAAGGAC-3′. Theconditions were optimized and ran on a Thermacycler. 100 bpEZ load molecular ruler was used for size determination.

Acknowledgments

Funding for this project was provided by the Robert WoodJohnson Harold Amos Medical Faculty Development Awardand the CHRC K12 NICHD Training Grant.

R E F E R E N C E S

Back, S.A., et al., 1998. Maturation-dependent vulnerability ofoligodendrocytes to oxidative stress-induced death caused byglutathione depletion. J. Neurosci. 18, 6241–6253.

Buemi, M., et al., 2002. Erythropoietin and the brain: fromneurodevelopment to neuroprotection. Clin. Sci. (Lond). 103,275–282.

Calapai, G., et al., 2000. Erythropoietin protects against brainischemic injury by inhibition of nitric oxide formation. Eur. J.Pharmacol. 401, 349–356.

Chong, Z.Z., et al., 2003. Erythropoietin prevents early and lateneuronal demise through modulation of Akt1 and induction ofcaspase 1, 3, and 8. J. Neurosci. Res. 71, 659–669.

Chong, Z.Z., et al., 2007. Vascular injury during elevated glucosecan be mitigated by erythropoietin and Wnt signaling. Curr.Neurovasc. Res. 4, 194–204.

Chung, W.J., et al., 2005. Inhibition of cystine uptake disrupts thegrowth of primary brain tumors. J. Neurosci. 25,7101–7110.

d'Uscio, L.V., et al., 2007. Essential role of endothelial nitric oxidesynthase in vascular effects of erythropoietin. Hypertension49, 1142–1148.

Digicaylioglu, M., Lipton, S.A., 2001. Erythropoietin-mediatedneuroprotection involves cross-talk between Jak2 andNF-kappaB signalling cascades. Nature 412, 641–647.

Fenjves, E.S., et al., 2004. Adenoviral gene transfer oferythropoietin confers cytoprotection to isolated pancreaticislets. Transplantation 77, 13–18.

Genc, S., et al., 2002. Erythropoietin restores glutathioneperoxidase activity in 1-methyl-4-phenyl-1, 2, 5,6-tetrahydropyridine-induced neurotoxicity in C57BLmice andstimulates murine astroglial glutathione peroxidaseproduction in vitro. Neurosci. Lett. 321, 73–76.

95B R A I N R E S E A R C H 1 3 2 1 ( 2 0 1 0 ) 8 8 – 9 5

Ghezzi, P., Brines, M., 2004. Erythropoietin as an antiapoptotic,tissue-protective cytokine. Cell Death Differ. 11 (Suppl 1), S37–S44.

Guneli, E., et al., 2007. Erythropoietin protects the intestine againstischemia/reperfusion injury in rats. Mol. Med. 13, 509–517.

Hasselblatt, M., et al., 2006. The brain erythropoietin system andits potential for therapeutic exploitation in brain disease.J. Neurosurg. Anesthesiol. 18, 132–138.

Joyeux-Faure, M., 2007. Cellular protection by erythropoietin: newtherapeutic implications? J. Pharmacol. Exp. Ther. 323, 759–762.

Katavetin, P., et al., 2007. Antioxidative effects of erythropoietin.Kidney Int. Suppl. S10–S15.

Krantz, S.B., 1995. Erythropoietin and the anaemia of chronicdisease. Nephrol. Dial. Transplant. 10 (Suppl 2), 10–17.

Latini, R., et al., 2008. Do non-hemopoietic effects of erythropoietinplay a beneficial role in heart failure? Heart Fail. Rev. 13, 415–423.

Liu, J., et al., 2005. Neuroprotection by hypoxic preconditioninginvolves oxidative stress-mediated expression ofhypoxia-inducible factor and erythropoietin. Stroke 36, 1264–1269.

Ribatti, D., et al., 2007. Erythropoietin/erythropoietin receptorsystem is involved in angiogenesis in human neuroblastoma.Histopathology 50, 636–641.

Rosin, C., et al., 2004. Excitatory amino acid inducedoligodendrocyte cell death in vitro: receptor-dependent and-independent mechanisms. J. Neurochem. 90, 1173–1185.

Ruscher, K., et al., 2002. Erythropoietin is a paracrine mediator ofischemic tolerance in the brain: evidence from an in vitromodel. J. Neurosci. 22, 10291–10301.

Sartelet, H., et al., 2007. Expression of erythropoietin and itsreceptor in neuroblastomas. Cancer 110, 1096–1106.

Sato, H., et al., 2000. Molecular cloning and expression of humanxCT, the light chain of amino acid transport system Xc.Antioxid. Redox Signal. 2, 665–671.

Sato, H., et al., 2002. Distribution of cystine/glutamate exchangetransporter, system X(c)−, in the mouse brain. J. Neurosci. 22,8028–8033.

Seko, Y., et al., 1998. Vascular endothelial growth factor (VEGF)activates Raf-1, mitogen-activated protein (MAP) kinases, andS6 kinase (p90rsk) in cultured rat cardiac myocytes. J. Cell.Physiol. 175, 239–246.

Shih, A.Y., et al., 2003. Coordinate regulation of glutathionebiosynthesis and release by Nrf2-expressing glia potentlyprotects neurons from oxidative stress. J. Neurosci. 23,3394–3406.

Shih, A.Y., et al., 2006. Cystine/glutamate exchange modulatesglutathione supply for neuroprotection from oxidative stressand cell proliferation. J. Neurosci. 26, 10514–10523.

Sola, A., et al., 2005. Erythropoietin after focal cerebral ischemiaactivates the Janus kinase-signal transducer and activator oftranscription signaling pathway and improves brain injury inpostnatal day 7 rats. Pediatr. Res. 57, 481–487.

Szczepanski, M., et al., 2006. Erythropoietin prevents oxidativestress and lipid peroxidation process in human umbilical veinendothelial cells induced by tumor necrosis factor alpha. Pol.Merkur. Lekarski. 21, 534–539.

Uppenkamp, M., et al., 1998. Thrombopoietin serum concentrationin patients with reactive and myeloproliferativethrombocytosis. Ann. Hematol. 77, 217–223.

Wells, R.G., et al., 1992. The 4F2 antigen heavy chain inducesuptake of neutral and dibasic amino acids in Xenopus oocytes.J. Biol. Chem. 267, 15285–15288.

Wu, Y., et al., 2007. Erythropoietin prevents PC12 cells from1-methyl-4-phenylpyridinium ion-induced apoptosis via theAkt/GSK-3beta/caspase-3 mediated signaling pathway.Apoptosis 12, 1365–1375.

Yamamoto, M., et al., 2004. Effect of erythropoietin on nitric oxideproduction in the rat hippocampus using in vivo brainmicrodialysis. Neuroscience 128, 163–168.

Yu, X., et al., 2002. Erythropoietin receptor signalling is requiredfor normal brain development. Development 129,505–516.