Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

Contents lists available at ScienceDirect

Journal of Molecular and Cellular Cardiology

j ourna l homepage: www.e lsev ie r.com/ locate /y jmcc

Original article

Carbamylated darbepoetin derivative prevents endothelial progenitor cell damagewith no effect on angiogenesis

Rafael Ramirez a,⁎, Julia Carracedo a, Sonia Nogueras a, Paula Buendia a, Ana Merino a, Sagrario Cañadillas a,Mariano Rodríguez a, Ciro Tetta b, Alejandro Martin-Malo a, Pedro Aljama a

a Unidad de Investigación, Servicio de Nefrología, Hospital Universitario Reina Sofía, Avda. Menéndez Pidal S/N, Cordoba-14004, Spainb Department of Medicine, University of Cordoba, Cordoba, Spain

Abbreviations: EPCs, endothelial progenitor cells; EPOkidney disease; rHuEPO, recombinant human erythrC-darbe, carbamylated darbepoetin; TNBSA, trinitrobenzrating cell nuclear antigen; KMESF,molecular equivalent⁎ Corresponding author. Tel.: +34 957 010452; fax: +

E-mail address: manuelr.ramirez.sspa@juntadeandal

0022-2828/$ – see front matter © 2009 Elsevier Inc. Adoi:10.1016/j.yjmcc.2009.09.005

a b s t r a c t

a r t i c l e i n f o

Article history:Received 27 April 2009Received in revised form 3 September 2009Accepted 9 September 2009Available online 24 September 2009

Keywords:AngiogenesisApoptosisCarbamylated darbepoetinCell senescenceEndothelial progenitor cellsInflammationProliferation

Erythropoietin (EPO) prevents cell apoptosis induced by oxidative stress. Carbamylated EPO maintains thetissue-protective activities of the unmodified EPO but does not stimulate erythropoiesis. This study evaluateswhether carbamylated erythropoietin is as effective as recombinant human erythropoietin in protectingendothelial progenitor cells (EPCs) from apoptosis without stimulating erythropoiesis. Experiments wereperformed in an erythroid cell line (UT-7) and in humanEPCs. Cell signals regulatingproliferation and apoptosis(Jak-2, Akt, Erk1/2, NFκB and Stat-5) were measured by Western blotting. In human EPCs, cell senescence,apoptosis and proliferation were assessed by acidic β-gal and measurement of telomere length, TUNEL andPCNA labeling, respectively. Angiogenesis was evaluated using the endothelial tube formation assay. In UT-7,carbamylated erythropoietin (C-darbe) induced phosphorylation of the anti-apoptotic Jak-2/Akt signal and, asopposed to recombinant human erythropoietin (darbe), did not produce a significant activation of cellproliferating signals. Darbe increased the percent of proliferating EPCs and promoted angiogenesis. By contrast,C-darbe failed to stimulate proliferation of EPCs. Both C-darbe and darbe equally reduced apoptosis andsenescence. Thus, C-darbe protects EPCs from apoptosis and does not increase erythropoiesis.

© 2009 Elsevier Inc. All rights reserved.

1. Introduction

Human erythropoietin (EPO) is regularly used for the treatment ofanemia secondary to chronic diseases and cancer. It is particularlyuseful in correction of anemia in chronic renal failure patients inwhich there is a marked reduction of EPO producing cells [1].Additionally, EPO prevents cell apoptosis induced by oxidative stress[2,3]. Activation of EPO receptor stimulates anti-apoptotic signalsmediated by Akt phosphorylation [4].

During the last decade there have been a number of reportsshowing the key role of endothelial progenitor cells (EPCs) to preventcardiovascular events [5,6]. Patients with chronic inflammation showa decrease in EPCs, which may be a mechanism that links chronicinflammation with cardiovascular disease [7–9]. Several groupsdemonstrated a positive effect of EPO on EPC mobilization [10,11],which may be used as therapeutic tool for the treatment ofcardiovascular diseases. This subject becomesmost relevant in uremicpatients in which chronic inflammation and/or accumulation ofuremic toxins induce EPC damage [8,12,13].

, erythropoietin; CKD, chronicopoietin; darbe, darbepoetin;enesulfonic acid; PCNA, prolife-s of soluble fluorochrome units.34 957 010452.

ucia.es (R. Ramirez).

ll rights reserved.

The administration of EPO may have a beneficial effect on thevascular disease of patients with chronic inflammation. However,concern has arisen over the increased mortality in patients treatedwith EPO, when the hemoglobin concentration reaches levels higherthan those currently recommended [14]. The clinical use of high doseof EPO carries the risk of an unhealthy increase in the hematocrit.Recent studies show that carbamylation of EPO results in a modifiedmolecule that does not stimulate erythropoiesis but maintains thetissue-protective activities of the unmodified EPO [3,15]. Carbamy-lated EPO targets a cell receptor that is different from the classical EPOreceptor [16]. Thus, it may be possible that signaling pathways andcell response to carbamylated EPO differ from the classical response tounmodified EPO.

The aim of this study was to compare the ability of EPO andcarbamilated EPO to protect erythroid cells and EPCs from the damageinduced by inflammation. The study included experiments designedto identify specific cell signaling that may explain the differentialeffect of EPO and carbamylated EPO.

2. Methods

2.1. Study subjects

The study was performed in EPCs from five healthy subjects.Uremic serum was pooled from 20 non-dialysis stage 4–5 CKD

782 R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

patients recruited from the Nephrology Service of Reina SofiaUniversity Hospital, Cordoba, Spain. The patients were neither onanti-inflammatory/immunosuppressive drugs nor erythropoietinduring the 3 months prior to inclusion in the study. None of thepatients had diabetes.

2.2. Experimental design

To study the effect of inflammation, cells were cultured during3 days in the presence of TNF-α (10 ng/ml, Sigma, St. Louis,MO), uremic serum or human autologous serum (AB serum)(Bio-Whittaker, Walkersville, MD). The AB serum is heated at56 °C during 60 min to inactivate complement. In preliminaryexperiments, we found that a final concentration of 10% was themaximal, non-toxic concentration of uremic serum.

2.3. Erythropoietin

The EPO used in this study were darbepoetin-α (darbe, Aranesp,AMGEN, Seattle, WA), C-darbepoetin (C-darbe, prepared as describedbelow), epoetin α (EPREX, Janssen-Cilag, Johnson & Johnson, Beerse,Belgium), and epoetin β (NeoRecormon, Roche Molecular Biochem-icals, Indianapolis, IN). Darbe (100 μg) was carbamylated byincubation with a cyanate solution at 37 °C for 48 h. Excess cyanatewas removed by extensive dialysis at 4 °C with saline solutions withpH 7.4. The extent of carbamylation was monitored by assessing theloss of free amino groups using trinitrobenzenesulfonic acid (TNBSA,Pierce, Rockford, IL). Absorbance was measured at 335 nm against asample blank, and the TNBSA reactivity was expressed as apercentage of the absorbance obtained for the non-carbamylateddarbepoetin.

2.4. UT-7 cell culture

UT-7 cells, a human erythroid cell line [17], were obtained fromEndogen, Piercenet, Rockford. These cells were cultured in 80% alpha-MEM (minimumessential medium fromGIBCO, Invitrogen, San Diego,CA)+20% fetal bovine serum (GIBCO Invitrogen)+5 ng/ml granulo-cyte–macrophage colony-stimulating factor (Roche Diagnostics,Indianapolis, IN) maintained at 0.5×10−6 in 12- or 24-well plate foroptimal growth; split saturated culture 1:2 every 2–3 days.

In subsequent studies of proliferation and cell signals the densityof cells plated was 1×106.

2.5. Characterization and culture of EPCs

A 30-ml sample of venous blood was used for the isolation ofendothelial progenitor cells [20]. Samples were processed within 4h after collection; peripheral-blood mononuclear cells were isolatedby Ficoll density-gradient centrifugation (Lymphoprep, Axis-ShieldPoC AS, Oslo, Norway), washed with PBS (GIBCO, Invitrogen)+20%FBS, and resuspended in Endocult liquid medium kit (StemcellTechnologies, Vancouver, British Columbia, Canada) supplementedwith antibiotics (penicillin 100 U/ml and streptomycin 100 μg/ml)(Invitrogen). Mononuclear cells were plated on fibronectin (SigmaChemical Co., St. Louis, MO) coated six-wells at a density of 5×106

cells per well. After 2 days, non-adherent cells were collected andplated on fibronectin-coated wells at a density of 1×106 cells perwell. On day 5, the number of adherent cell colonies was countedusing an inverted fluorescent microscope (Eclipse Ti-S, NikonInstruments Europe B.V., Badhoevedorp, Netherlands) in threerandom low power fields (×5) by two independent blindedinvestigators.

To characterize the EPC phenotype, cells were detached usingtrypsin–EDTA reagent kit (ReagentPack from Lonza, Walkersville,MD), cell scraping, and gentle mechanical pipetting to disperse cell

clumps. Cells were washed with Endocult Complete Medium. A totalof 2×106 cells were incubated in Endocult Complete Medium for30 min in the dark at 4 °C with the following monoclonal antibodies(mAb) using the concentrations recommended by the manufacturer:fluorescein-labeled (FITC) anti-CD144 mAb (SEROTEC, 22 BanksideStation Approach Kidlington, UK), tricolor-labeled (TC) anti-CD34 mAb (Caltag, Invitrogen, San Diego, CA), KDR-PE anti-VEGFR-2 mAb (RD Systems, Minneapolis), and the correspondingisotype controls. Quantitative analysis was performed on a FACSCa-libur flow cytometer measuring 1×105 cells per sample. Data wereanalyzed using CellQuest software (Becton Dickinson) by sidescatter-fluorescence dot plot analysis. The number of EPCs wasdefined as events triple-positive for CD34, CD144, and KDR with lowcytoplasmic granularity (low sideward scatter, SSC). Using thisconditioning culture medium we consistently obtained more than90% of EPCs.

Cells used for all subsequent experiments were obtained atpassages 3–5 and were plated at a density of 1×106.

2.6. Cell proliferation and cell signals

The effect of the various EPO (100 and 200 ng/ml) on UT-7 cellproliferation (erythropoiesis) was evaluated by PCNA labeling using akit containing PE-PCNA antibody (BD Pharmingen, San Diego, CA) andquantification by flow cytometry (FACSCalibur, Becton Dickinson, SanJose, CA). At a protein concentration of 100 ng/ml all commercial EPOs(darbepoetin α, epoetin α, and epoetin β) showed similar biologicalactivity as illustrated by a comparable increase in cell proliferationafter 24 and 48 h of treatment.

EPC proliferation was also analyzed by PCNA quantification after48 h in culture with darbe and C-darbe (100 ng/ml).

Binding of EPO to its receptor induces phosphorylation of Jak-2which in turn triggers cell proliferating signals (Erk1/2, NFκB, andStat-5) and Pi3K/Akt that prevents cell apoptosis.

Activation of cell signaling systems by darbe or C-darbe wasevaluated in UT-7 cells and EPCs. After 72-h incubation with 100 ng/ml of darbe or C-darbe, UT-7 and EPC viability was N90%. Cells werelysed and proteins were extracted from cytoplasm and nucleus toperform Western blotting following standard protocols [18,19].Cytosol and nuclear extracts (50 μg) were separated in sodiumdodecyl sulfate–10% polyacrylamide gel electrophoresis and thentransferred to nitrocellulose membranes. Antibodies used to identifyconstitutive and phosphorylated (P) forms of signaling moleculeswere the following: JAK2 (SC-278), P-JAK2 (SC-21870), Akt (SC-8312), P-Akt (SC-16646), ERK1/2 (#4695), P-ERK1/2 (SC-7383),nuclear form of NFκB (SC-109), nuclear form of Stat-5 (SC-28463),cytosolic IKB (SC-4094), and actin (SC-8432). Antibodies werepurchased from Santa Cruz Biotechnology (Santa Cruz, CA). Theantibody against the constitutive form of ERK1/2 was obtained fromCell Signaling Technology, Inc. (Danvers, MA). Visualization ofimmune complexes was performed using horseradish peroxidase-conjugated secondary antibodies, and the Luminol reagent detectionsystem (Santa Cruz Biotechnology). Protein levels were quantifiedusing the image analysis software Intelligent Quantifier, version 2.1.1(Bio Image, Ann Arbor, MI). Values were calculated in terms ofintegrated optical density (IOD) and expressed in arbitrary units (AU).

2.7. Senescent cells (assessed by acidic β-galactosidase staining)

Senescent cells were detected by Acidic β-Galactosidase Stainingusing a commercial kit (SA-β-gal Staining cat. no. CBA-230, CellBiolabs, Inc., San Diego, CA) [21,22]. After staining, cells (90%previously characterized as EPC) were washed and counted using alight microscope (Eclipse Ti-S). Results were expressed as percentageof β-gal positive cells (blue-stained cells) on at least 100 countedcells.

783R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

2.8. Assessment of telomere length by fluorescent in situ hybridization(FISH) in flow cytometry (Flow-FISH)

Telomere length was measured by Flow-FISH [23] using thetelomere specific FITC-conjugated probe (FITC-O-CCCATAACTAAA-CAC-NH2, DakoCytomation, Glostrup, Denmark). Results wereexpressed in molecular equivalents of soluble fluorochrome units(kMESF). The telomere length was calculated comparing the kMESFvalues to those obtained using cell lines with different telomerelengths, which served as standard. In order to assess variability, themeasurement of telomere length was determined using a standardcurve in which values were established from three cell lines (K562,U937, and Daudi) that are known to have different telomere lengths.The intra-assay coefficient of variation was 4.6.

2.9. Apoptosis

Apoptosiswasmeasured byTUNEL labeling (RocheApplied Science,Indianapolis, IN), and analyzed by cell cytometry (FACSCalibur).

2.10. In vitro angiogenesis

Angiogenesis was assessed on the basis of endothelial tubeformation using the Cell Biolabs Endothelial Tube formation AssayKit from Cell Biolabs, Inc., San Diego, CA. Cells were resuspended inculture mediumwith darbe or C-darbe (100 ng/ml). A cell suspension(1.5–3×105cells) was added to each well onto the solidified gel. Theplate was incubated for 4–18 h at 37 °C. Measurements of tubeformation were obtained at 8 h. No further increase in tube formationwas observed after 8 h. The endothelial tubes were examined by lightmicroscopy using a high magnification field. The results are expressedas number of tubes per mm2.

2.11. Statistical analysis

Comparisons between several means were performed by ANOVAtest followed by the Bonferroni test. A comparison of two means wasdone by the Student t test for paired or unpaired data. Non-parametricdata were analyzed by theWilcoxon or Mann–Whitney test for pairedand unpaired comparisons, respectively. Differences were consideredsignificant when pb 0.05.

3. Results

3.1. Cell proliferation and cell signals

EPO-α, EPO-β, and darbe (at 100 and 200 ng/ml) induced asignificant increase in UT-7 cell proliferation, assessed as percent ofPCNA positive cells (Table 1). By contrast the same concentration ofC-darbe did not produce a significant increase in the percent of PCNApositive cells (Table 1).

EPO-α, EPO-β, and darbe also increased the number of EPCscolonies in culture. The wells were plated with 106 EPCs; after 48h with 100 ng/ml of EPO-α, the number of EPCs colonies per well

Table 1The effect of different erythropoietins on UT-7 proliferation (% of PCNA positive cells).

Erythropoietins Epoetin α Epoetin β

Treatment(ng/ml)

None 50 100 200 50 100 2

24 h 36.7±1.9 42.6±2.6 42.9±2.1 40.9±3.2 40.5±1.2 40.1±0.3 448 h 57.2±0.9 62.6±2.6 74.5±2a 76.2±2.2a 63.2±3.1 76.1±3.8a 7

Data are the mean±SD from five independent experiments performed under the same cona pb 0.05 vs. none.

increased from 12±3 to 29±4 (pb 0.001); with EPO-β, the numberof colonies increased to 28±6 (pb0.001) and with darbe to 29±7(pb 0.001). By contrast the 100 ng/ml of C-darbe did not induce asignificant increase in the number of EPCs colonies (15±4).Increasing the concentration of the different EPOs, including C-darbe, did not result in a further increase in the number of colonies.

In UT-7 cells, darbe and C-darbe induced phosphorylation of cellsignaling molecules (Fig. 1A and B). Both darbe and C-darbe inducedphosphorylation of Jak-2 and Akt which mediate an anti-apoptoticeffect. Phosphorylation of cell proliferation molecules Erk1/2 andStat-5, and traslocation of NFκB, was significantly increased in cellstreated with darbe as compared with C-darbe. Densitometric valuesare shown in Fig. 1B. In EPCs the phosphorylation of cell signalmolecules induced by darbe and C-darbe showed a pattern similar tothat observed in UT-7 except that NFκB translocation by darbe and C-darbe was similar. Densitometric values for EPCs are shown in Fig. 1B.

3.2. Senescence of EPCs

After 3 days in culture with TNF-α, EPCs demonstrate features ofcellular senescence: 63±9% of cells were β-gal positive and 53±4%of cells showed telomere shortening. In the absence of TNF-α, cellsmaintain the EPC phenotype and the respective values of % of β-galand telomere shortening were 7.6±0.3% and 11.7±0.6% (pb 0.001).The average telomere length (kMESF units) in control cells vs. thosegrown in TNF-α was 11.5±0.7 and 9.3±0.7 (pb 0.001).

Similarly to the observed in EPCs cultured with TNF-α, EPCscultured in medium containing uremic serum for 36 h exhibitedfeatures of cellular senescence; 75±5% of these cells showed β-galstaining and a distinct flat and enlargedmorphology (Fig. 2A and B). Inaddition, 71±6% of EPCs cultured with uremic serum exhibit adecreased length of telomeres (8.2±0.4 vs. 11.5±0.7 kMESF,pb 0.001). Uremic serum, similarly to TNF -α, leads to emergence oftwopopulations of cellswith different telomere lengths (10.9±0.5 and7.6±0.3 kMESF). The addition of darbe andC-darbe to EPCs cultured inuremic serummarkedly reduced the percentage of β-gal positive cellsto 38.3±2.3% and 40.7±2.8%, respectively (pb 0.001 vs. uremicserum) (Fig. 2C and D). Similarly, both darbe and C-darbe markedlyreduced the percentage of EPCs with short telomeres (23.4±4.6% and21.6±7.4%, respectively, pb 0.001 vs. uremic serum) (Fig. 3).

3.3. Apoptosis

Both darbe and C-darbe reduced apoptosis of EPCs induced byTNF-α (Table 2). A similar effect was observed with epoetin α and β(data not shown).

Fifty percent of EPCs underwent apoptosis after 24-h incubationwith uremic serum. Darbe significantly reduced uremic serum-induced apoptosis to 27.1% while C-darbe reduced it to 31.1%.

3.4. EPC proliferation

EPC proliferation was evaluated after 8 h in culture (n= 5) in cellsobtained at passage 3. EPCproliferation in AB serum is shown in Fig. 4A.

Darbepoetin C-darbepoetin

00 50 100 200 50 100 200

1.6±2.3 43.6±1.8 44.6±0.9 43.7±2.4 39.6±0.9 41.9±1.1 41.6±2.24.3±1.9a 66.3±1.1 77.6±2.3a 76.9±0.6a 61±3.1 61.7±1.4 57.1±2.3

ditions.

Fig. 1. (A) Western blots from UT-7 cells and EPCs proteins at baseline and after 72-h incubation with 100 ng/ml of either darbepoetin (darbe) or C-darbepoetin (C-darbe). (B)Densitometric values (mean±SD) of Westerns corresponding to three different experiments. a, pb 0.05 vs. control; b, pb 0.05 vs. darbe.

784 R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

As compared with control medium (no growth factors added), bothdarbe andendothelial growth factors (EGF)produced a similar increasein EPCs (by 8.7- and 7.3-fold, respectively). By contrast, C-darbe did notstimulate proliferation of EPCs. The addition of darbe to the EGFenriched medium resulted in a further increase in EPC proliferation(12.3-fold increase as compared with control) (pb 0.001). C-darbe didnot augment the EGF induced EPC proliferation (Fig. 4A).

The effect of uremic serum on EPC proliferation is shown in Fig 4B.The uremic serum reduced the effect of darbepoetin, EGF, and darbe+EGF on EPC proliferation by approximately half, as compared to theireffect in AB serum. However as compared with uremic control serumEGF and darbe produced a 11.6- and 8.6-fold increase, respectively. Asexpected, C-darbe in the presence of uremic serum did not induce EPCproliferation. In uremic serum the addition of darbe to EGF produced afurther increase EPC proliferation (19.2-fold) which was of greatermagnitude than with C-darbe (14.9-fold).

As compared with AB serum, TNF-α reduced the EPC proliferationinduced by EGF, and darbe or C-darbe alone or in combination withEGF (Fig. 4C).

3.5. In vitro angiogenesis

Tube formation was increased by the addition of darbe to theEndocult liquid medium enriched with angiogenic factors (5.6±2.2vs. 15.4±2.1 tubes per mm2) (pb 0.001). This effect was notobserved with C-darbe (6.3±1.4 tubes per mm2). In the presence ofuremic serum or TNF-α, EPCs failed to form tubular structures.However both darbe and, to a lesser degree, C-darbe were able torestore the ability of EPCs to form tubes in the presence of uremicserum (5.9±1.4 and 3.2±1.1 tubes per mm2, respectively, Fig. 5),and in the presence of TNF-α (5.4±0.9 and 3.6±1.3 tubes permm2, respectively).

4. Discussion

The EPOs epoetin-α, epoetin-β, and darbe stimulated EPCproliferation and the formation of endothelial tube structures. Darbewas able to potentiate theproliferating activity of EPCs inducedbyEGF.In addition, darbe prevented the EPCs apoptosis induced by uremic

Fig. 2. Cellular senescence in endothelial progenitor cells (EPCs). Representative images of β-gal staining (associated with cellular senescence) of EPCs. Cells were cultured withcontrol AB serum (A) or uremic serum (B). In the lower panels, darbepoetin (C) and C-darbepoetin (D) were added to cultures of EPCs with uremic serum. Magnification, ×20.

785R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

serum or TNFα. Chemical modification of darbe by carbamylation (C-darbe) does not alter the anti-apoptotic effect of darbe; however, itfailed to stimulate proliferation. This differential effect can be

Fig. 3. Effect of darbepoetin and C-darbepoetin in the length of EPCs telomeres. In the top pa(Flow-FISH) in EPCs cultured with control AB serum (A) or uremic serum (B). Telomere lengmean fluorescence obtained with the telomere probe. Down appears the representative histC-darbepoetin (D) treatment.

explained by the failure of C-darbe to activate proliferating cell signals(Erk1/2, NFκB, and Stat-5) whereas anti-apoptotic effect through Aktactivation was unaffected.

nels shown representative histograms of telomere length measured by flow cytometryth was calculated by subtracting the mean background telomere fluorescence from theograms of telomere length in EPCs cultured with uremic serum and darbepoetin (C) or

Table 2The effect of darbepoetin and C-darbepoetin on the apoptosis of endothelial progenitorcells induced by uremic serum.

Erythropoietins Control TNF-α Uremic serum

None 16.5±1.2 59.1±5.7a 50.3±1.2a

Darbepoetin 12.4±1.1 31.3±2.4a,b 27.1±1.4a

C-darbepoetin 14.7±0.9 34.7±2.1a,b 31.1±1.4a

Values represent the percent of apoptotic cells relative to the total cells which containmore than 90% of EPCs. Data are the mean±SE from five independent experiments.TNF-α was used at a concentration of 10 ng/ml.

a pb 0.001 vs. corresponding control.b pb 0.001, darbepoetin and C-darbepoetin vs. none.

786 R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

EPO is a hematopoietic cytokine produced by the kidney inresponse to hypoxia. Recombinant EPO is widely used to treatanemia in cancer, surgery, and kidney failure. Besides its rolein hematopoiesis, EPO has pleiomorphic activities since EPO-

Fig. 4. Endothelial progenitor cell (EPC) proliferation evaluated by PCNA expression.The effect of endothelial growth factors (EGF), darbepoetin (darbe), C-darbepoetin(C-darbe), EGF plus darbe, and EGF plus C-darbe on EPC proliferation in the presenceof normal AB serum (A), uremic serum (B), and TNF-α (C). Values were obtained after48 h in culture and represent the percent of PCNA cells relative to the total cells whichcontain more than 90% of EPCs. Data are mean±SE (n=5). ⁎, pb0.001 vs. control;#, pb0.001 vs. EGF; ^, pb0.001 vs. EGF + darbe. a, pb 0.05 vs. the correspondinggroups in AB serum.

receptors are present in a variety of tissues [24]. Several invitro and in vivo studies have proposed that EPO may protectcells from damage associated to increased oxidative stress[2,3].

Recent evidence strongly suggests a role of EPCs in the repairof vascular endothelial damage [5,6]. Humans EPCs are derivedfrom CD34+ hematopoietic stem cells, or from the even moreimmature CD133+ stem cells, co-expressing endothelial markerssuch as vascular endothelial growth factor receptor 2, or vascularendothelial–cadherin. EPCs may also be derived from differentiatedCD14+ mononuclear cells [25]. Mobilization of these cells resultsfrom mechanical injury and ischemic stress, via the generation ofhypoxia-inducible factor-1-regulated release of VEGF, erythropoietin,and stromal cell-derived factor-1, as well as by placental growthfactor, granulocyte colony-stimulating factor, and granulocyte–macrophage colony-stimulating factor [26]. In experimental studies,increased neovascularization by EPCs after myocardial ischemiaimproves cardiac function [5–9,26–28]. In patients with myocardialinfarction, clinical outcome correlates strongly with the number ofmobilized EPCs. Thus, the search for substances able to modulate thenumber and/or function of EPCs is of considerable interest. Forexample, VEGF regulates EPC proliferation and differentiation [18].This study demonstrates that darbe possesses a remarkable effect ininducing EPC proliferation and the stimulation of colony-formingunit. This confirms that EPO is a key molecule in the process ofvascular repair and neo-angiogenesis via the stimulation of EPCs[10,11]. Nevertheless, among clinicians there is a justified concernabout the danger of EPO to increase hematocrit above desirablelevels. In fact mortality of CKD patients treated with EPO is increased,when the hemoglobin concentration reaches levels higher than thosecurrently recommended [14]. More recent publications have shownthat overall mortality of hemodialysis patients is also negativelycorrelated with EPO dose independent of hematocrit [29].

Carbamylated EPO is a structurally modified EPO, which has beendeveloped for tissue protection with no relevant erythropoieticactivity [15–16]. To the best of our knowledge this is the first timethat the biological effect of C-darbe has been evaluated. The additionof C-darbe to UT-7 cells, an EPO dependent human erythroid cellline, enabled cells to survive with no effect on proliferation. Thesefindings are in contrast with the marked proliferative effect ofdarbe and other EPOs on erythroid cell. Our experiments revealedthat C-darbe induced phosphorylation of the anti-apoptotic Jak-2/Akt signal; as opposed to darbe, C-darbe has a marginal effect on cellproliferating signals (Erk1/2, NFκB and Stat-5). Differential effects ofC-darbe and darbe have been also demonstrated at the cell surfacelevel: Carbamylated EPO interacts only with the common beta-subunit of the heteroreceptor for EPO. This has been demonstratedby Brines et al. in a recently published article [16]; these authorssuggested that the lack of ability of C-EPO to bind to the EPOreceptor homodimer may allow the C-EPO to prevent apoptosiswithout stimulating erythropoiesis.

Apoptosis of EPCs was induced by incubation with theinflammatory cytokine TNF-α or uremic serum. Both TNF-α anduremic serum also induced accelerated senescence and impairmentof functional activity of the EPCs (as assessed for the ability toform cell colonies or tubes in matrigel). These findings support thenotion that patients with chronic inflammation are predisposed tovascular damage and subsequent development of prematureatherosclerotic lesions. Both darbe and C-darbe reduced the EPCsapoptosis and senescence, and consequently, angiogenic activitywas improved.

The results obtained in this study demonstrate that, as opposedto darbe, C-darbe does not induce cell proliferation; however, bothC-darbe and darbe were able to antagonize apoptosis associatedwith uremia or inflammation. Therefore the use of C-darbe, bytargeting solely on prevention of EPCs apoptosis, could be a

Fig. 5. Representative images of tube-like three dimensional structures of (endothelial progenitor cells) EPCs in the presence of AB normal serum (A), darbepoetin (B), C-darbepoetin(C), uremic serum (D), uremic serum with darbe (E), and uremic serum with C-darbe (F). Magnification, ×20.

787R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

potential tool to protect against cardiovascular disease. Furtherstudies are needed to evaluate possible strategies to exploit thedifferential properties of darbe and C-darbe, and possibly newfuture derivatives.

Acknowledgments

We are grateful to M.J. Jimenez for technical assistance. J.Carracedo was supported by contract from the Instituto de SaludCarlos III/Fundación Progreso y Salud (Programa de Estabilización eIncentivación de la Investigación 2006). This work was supported bygrants from Fondo de Investigación Sanitaria, Instituto de Salud CarlosIII (FIS PI05/0896, PI06/0724, PI06/0747, PI07/0204, RETICs RedRenal RD06/0016/0007), Junta de Andalucía (CM0008, TCRM0006/2006, P06-CVI-02172, P08-CTS-3797), and Fundación Nefrológica.

References

[1] Winearls CG, Oliver DO, Pippard MJ, Reid C, Downing MR, Cotes PM. Effect ofhuman erythropoietin derived from recombinant DNA on the anaemia of patientsmaintained by chronic haemodialysis. Lancet 1986;22:1175–8.

[2] Ghezzi P, Brines M. Erythropoietin as an antiapoptotic, tissue-protective cytokine.Cell Death Differ 2004;11:S37–44.

[3] Brines M, Cerami A. Erythropoietin-mediated tissue protection: reducingcollateral damage from the primary injury response. J Intern Med 2008;264:405–32.

[4] Bao H, Jacobs-Helber SM, Lawson AE, Penta K, Wickrema A, Sawyer ST. Proteinkinase B (c-Akt), phosphatidylinositol 3-kinase, and STAT5 are activated byerythropoietin (EPO) in HCD57 erythroid cells but are constitutively active in anEPO-independent, apoptosis-resistant subclone (HCD57-SREI cells). Blood 1999;93:3757–73.

[5] Takahashi T, Kalka C, Masuda H, Chen D, Silver M, Kearney M, et al. Ischemia- andcytokine-induced mobilization of bone marrow-derived endothelial progenitorcells for neovascularization. Nat Med 1999;4:434–8.

[6] Geft D, Schwartzenberg S, George J. Circulating endothelial progenitor cells incardiovascular disorders. Expert Rev Cardiovasc Ther 2008;8:1115–21.

[7] Verma S, Kuliszewski MA, Li SH, Szmitko PE, Zucco L, Wang CH, et al. C-reactiveprotein attenuates endothelial progenitor cell survival, differentiation, andfunction: further evidence of a mechanistic link between C-reactive protein andcardiovascular disease. Circulation 2004;109:2058–67.

[8] Ramirez R, Carracedo J, Merino A, Nogueras S, Alvarez-Lara MA, Rodríguez M, et al.Microinflammation induces endothelial damage in hemodialysis patients: the roleof convective transport. Kidney Int 2007;72:108–13.

[9] Tousoulis D, Andreou I, Antoniades C, Tentolouris C, Stefanadis C. Role ofinflammation and oxidative stress in endothelial progenitor cell function andmobilization: therapeutic implications for cardiovascular diseases. Atherosclerosis2008;201:236–47.

[10] Heeschen C, Aicher A, Lehmann R, Fichtlscherer S, Vasa M, Urbich C, et al.Erythropoietin is a potent physiologic stimulus for endothelial progenitor cellmobilization. Blood 2003;102:1340–6.

[11] Bahlmann FH, DeGroot K, Duckert T, Niemczyk E, Bahlmann E, Boehm SM, et al.Endothelial progenitor cell proliferation and differentiation is regulated byerythropoietin. Kidney Int 2003;64:1648–52.

[12] de Groot K, Bahlmann FH, Sowa J, Koenig J, Menne J, Haller H, et al. Uremia causesendothelial progenitor cell deficiency. Kidney Int 2004;66:641–6.

[13] Ramirez R, Martin-Malo A, Aljama P. Inflammation and hemodiafiltration. ContribNephrol 2007;158:210–5.

[14] Drüeke TB, Locatelli F, Clyne N, Eckardt KU, Macdougall IC, Tsakiris D, et al.Normalization of hemoglobin level in patients with chronic kidney disease andanemia. N Engl J Med 2006;355:2071–84.

[15] Leist M, Ghezzi P, Grasso G, Bianchi R, Villa P, Fratelli M, et al. Derivatives oferythropoietin that are tissue protective but not erythropoietic. Science 2004;305:239–42.

[16] Brines M, Grasso G, Fiordaliso F, Sfacteria A, Ghezzi P, Fratelli M, et al.Erythropoietinmediates tissue protection through an erythropoietin and commonbeta-subunit heteroreceptor. Proc Natl Acad Sci U S A 2004;101:14907–12.

[17] Komatsu N, Yamamoto M, Fujita H, Miwa A, Hatake K, Endo T, et al.Establishment and characterization of an erythropoietin-dependentsubline, UT-7/Epo, derived from human leukemia cell line, UT-7. Blood1993;82:456–64.

[18] Ratajczak J, Majka M, Kijowski J, Baj M, Pan ZK, Marquez LA, et al. Biologicalsignificance of MAPK, AKT and JAK-STAT protein activation by various erythro-poietic factors in normal human early erythroid cells. Br J Haematol 2001;115:195–204.

788 R. Ramirez et al. / Journal of Molecular and Cellular Cardiology 47 (2009) 781–788

[19] Andrews NC, Faller DV. A rapid micropreparation technique for extraction of DNA-binding proteins from limiting numbers of mammalian cells. Nucleic Acids Res1991;19:2499–507.

[20] Hill HM, Zalos G, Halcox JPJ, Schenke WH, Wacliwiw MA, Quyyumi AA, et al.Circulating endothelial progenitor cells, vascular function and cardiovascular risk.N Engl J Med 2003;348:593–600.

[21] Teoh ML, Turner PV, Evans DH. Tumorigenic poxviruses up-regulate intracellularsuperoxide to inhibit apoptosis and promote cell proliferation. J Virol 2005;79:799–811.

[22] Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C, et al. A biomarker thatidentifies senescent human cells in culture and in aging skin in vivo. Proc NatlAcad Sci U S A 1995;92:9363–7.

[23] Ramirez R, Carracedo J, Jimenez R, Canela A, Herrera E, Aljama P. Massivetelomere loss is an early event of DNA damage-induced apoptosis. J Biol Chem2003;278:836–42.

[24] Farrell F, Lee A. The erythropoietin receptor and its expression in tumor cells andother tissues. Oncologist 2004;9(Suppl 5):18–30.

[25] Asahara T. Cell therapy and gene therapy using endothelial progenitor cells forvascular regeneration. Handb Exp Pharmacol 2007;180:181–94.

[26] Ceradini DJ, Kulkarni AR, CallaghanMJ, Tepper OM, Bastidas N, KleinmanME, et al.Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1induction of SDF-1. Nat Med 2004;10:858–64.

[27] Asahara T, Takahashi T, Masuda H, Kalka C, Chen D, Iwaguro H, et al. VEGFcontributes to postnatal neovascularization by mobilizing bone marrow-derivedendothelial progenitor cells. EMBO J 1999;18:3964–72.

[28] Jujo K, Ii M, Losordo DW. Endothelial progenitor cells in neovascularization ofinfarcted myocardium. J Mol Cell Cardiol 2008;45:530–44.

[29] López-Gómez JM, Portolés JM, Aljama P. Factors that condition the response toerythropoietin in patients on hemodialysis. and their relation to mortality. KidneyInt 2008;111 Suppl:75–81.