store-operated ca entry is expressed in human endothelial ...royanaward.com › files12 ›...

ORIGINAL RESEARCH REPORT

Store-Operated Ca2þ Entry Is Expressedin Human Endothelial Progenitor Cells

Yuly Sanchez-Hernandez,1,* Umberto Laforenza,1,* Elisa Bonetti,2 Jacopo Fontana,1 Silvia Dragoni,1

Marika Russo,3 Jose Everardo Avelino-Cruz,1 Sergio Schinelli,3 Domenico Testa,4 Germano Guerra,5

Vittorio Rosti,2 Franco Tanzi,1 and Francesco Moccia1

Endothelial progenitor cells (EPCs) may be recruited from the bone marrow to sites of tissue regeneration tosustain neovascularization and reendothelialization after acute vascular injury. This feature makes them par-ticularly suitable for cell-based therapy. In mature endothelium, store-operated Ca2+ entry (SOCE) is activatedfollowing emptying of inositol-1,4,5-trisphosphate–sensitive stores, which controls a wide number of functions,including proliferation, nitric oxide synthesis, and vascular permeability. The present work aimed at investi-gating SOCE expression in EPCs harvested from both peripheral blood (PB-EPCs) and umbilical cord blood(UCB-EPCs) by employing both Ca2+ imaging and molecular biology techniques. SOCE was induced upon eitherpharmacological (ie, cyclopiazonic acid) or physiological (ie, ATP) depletion of the intracellular Ca2+ pool.Further, store-dependent Ca2+ entry was inhibited by the SOCE inhibitor,AU1 c BTP-2. Real-time reverse transcription–polymerase chain reaction and western blot analyses showed that both PB-EPCs and UCB-EPCs express all themolecular candidates to mediate SOCE in differentiated cells, including TRPC1, TRPC4, Orai1, and Stim1.Moreover, pharmacological maneuvers demonstrated that, as well as in differentiated endothelial cells, thesignal transduction pathway leading to depletion of the intracellular Ca2+ pool impinged on the phospholipaseC=inositol-1,4,5-trisphosphate pathway. Finally, blockage of SOCE withAU1 c BTP-2 impaired PB-EPC proliferation.These findings provide the first evidence that EPCs express SOCE, which might thus be regarded as a noveltarget to enhance the regenerative outcome of cell-based therapy.

Introduction

Endothelial progenitor cells (EPCs) comprise a sub-population of bone marrow–derived mononuclear cells

(MNCs) that are capable of differentiating into mature en-dothelial cells (ECs) and contribute to adult blood vessel for-mation and repair [1]. Immunophenotyping of circulatingEPCs has long been a matter for debate. Such uncertaintystemmed from the observation that these cells share a profileof surface antigens (CD133, CD34, c-kit=117, VEGFR2=KDR,CXCR4) similar to that of hematopoietic stem cells [1]. How-ever, there is now a general agreement that ‘‘true’’ endothelialprogenitors may be grown in vitro from the CD45�=CD34þ

fraction of peripheral MNCs, which are also named endo-thelial colony-forming cells [ECFCs], may achieve 20–30population doublings without obvious signs of senescence,and may retain the ability to form capillary-like structures in

Matrigel [2,3]. ECFCs differ from the colony-forming unit-ECs(CFU-ECs), which were initially regarded as endothelial inorigin [2,3]. Indeed, CFU-ECs belong to the hematopoieticlineage, retain some myeloid progenitor activity, differentiateinto phagocytic macrophages and not ECs, and lack vessel-forming capacity in vivo [3]. The same protocol has beenemployed to harvest ECFCs from the cord blood: these cellsdisplay a high proliferation potential and replate into second-ary and tertiary colonies [2].

The finding that circulating EPCs traffic to sites of neo-vascularization, proliferate, and differentiate into ECs isconsistent with the notion of adult vasculogenesis [4].More specifically, EPCs are mobilized from bone marrow inresponse to local tissue ischemia or damage to replace dys-functional ECs at the onset of severe cardiovascular disor-ders, such as atherosclerosis and arterial stenosis secondaryto injury [5,6]. As a consequence, EPCs play a pivotal role in

1Department of Physiology, University of Pavia, Pavia, Italy.2Laboratory of Clinical Epidemiology, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy.3Department of Experimental and Applied Pharmacology, University of Pavia, Pavia, Italy.4Institute of Otolaryngology-Head and Neck Surgery, Second University of Naples, Naples, Italy.5Department of Health Sciences, University of Molise, Campobasso, Italy.*These authors contributed equally to this work.

STEM CELLS AND DEVELOPMENT

Volume 00, Number 00, 2010

ª Mary Ann Liebert, Inc.

DOI: 10.1089=scd.2010.0047

1

SCD-2010-0047-ver9-Sanchez_1P.3D 08/23/10 3:06pm

SCD-2010-0047-ver9-Sanchez-Hernandez_1P

Type: research-article

restoring the supply of oxygen and nutrients to infarctedmyocardium [7], ischemic limbs [8], and growing tumors [9].Their regenerative potential renders EPCs particularly ame-nable for cell-based therapy, as discussed in several recentreviews [1,5,10–12]. However, because of their paucity insamples collected from peripheral blood (PB), which may betoo low to exert beneficial effects, and their uncertain dif-ferentiation fate, which may be harmful for the recipient,unveiling the signal transduction pathway regulating EPCbehavior is essential to improve the outcome of cell-basedtherapy [10,11]. As well as PB-derived EPCs (PB-EPCs), thetherapeutic potential of umbilical cord blood–derived EPCs(UCB-EPCs) has been shown in a recent study, in which theauthors demonstrated that transplantation of expandedUCB-EPCs favored neovascularization in a mouse ischemichind-limb model [13]. Similar to their peripheral counter-parts, the cellular mechanisms driving UCB-EPC function areyet to be elucidated.

An increase in intracellular Ca2þ concentration ([Ca2þ]i) isthe key signal in the regulation of a myriad of functions inmature ECs [14–18]. The elevation in [Ca2þ]i displayed bymacrovascular ECs upon stimulation is generally biphasicand consists in an initial rise that rapidly decays to a plateaulevel of lower amplitude until agonist removal [18]. In moredetail, ligand binding to Gq=11 protein–coupled receptors(GPCRs) results in the activation of phospholipase C beta(PLC-b), which, in turn, cleaves phosphatidylinositol-4,5-bisphosphate to inositol 1,4,5-trisphosphate (InsP3) and dia-cylglycerol (DAG) [16,18]. DAG engages both protein kinaseC and nonselective cation channels on the cellular mem-brane, whereas InsP3 rapidly diffuses within the cytosol tobind to and activate the InsP3 receptors (InsP3Rs) in the en-doplasmic reticulum (ER). InsP3Rs, in turn, serve as Ca2þ-permeable channels to release luminal stored Ca2þ and shapethe initial Ca2þ transient [16,18]. In ECs, as well as in mostother nonexcitable (ie, not endowed with voltage-gated ionchannels) cells, ER depletion causes the activation of Ca2þ-permeable ion channels in the plasma membrane, a mecha-nism that has been termed store-operated Ca2þ entry (SOCE),and sustains the following plateau phase [19,20]. The tightlink between SOCE and ER Ca2þ content is underscored bythe observation that inhibitors of the sarco-ER Ca2þ-ATPase(SERCA), such as cyclopiazonic acid (CPA), activate Ca2þ

entry by preventing Ca2þ-reuptake into the stores, a ma-neuver leading to ER depletion [21–23]. SOCE contributes tothe global increase in [Ca2þ]i, which triggers the release of anumber of vasoactive=inflammatory mediators, such as ni-tric oxide, prostacyclin, von Willebrand factor, platelet acti-vating factor, tissue plasminogen activator, and tissue factorpathway inhibitor [14]. Moreover, SOCE controls the geneticprogram underlying EC proliferation by engaging Ca2þ-sensitive transcription factors, including the nuclear factor ofactivated T-cells (NFAT) and the nuclear factor kappaB (NF-kB) [19,24,25]. Finally, store-dependent Ca2þ inflow governsthe paracellular permeability of endothelial monolayers bytriggering cytoskeleton reorganization (through myosin lightchain–dependent contraction) and consequent disassemblyof VE-cadherins at adherens junctions [14]. Unveiling themolecular nature of SOCE in mature ECs has engenderedremarkable controversy [26]. Several lines of evidence hintedat members of the canonical transient receptor potential(TRPC) family of cation channels as prime candidates for

SOCE in vascular endothelium [18]. Indeed, store-dependentCa2þ inflow was strongly dampened in cells lacking eitherTRPC1 [27] or TRPC4 [28,29]. Conversely, a recent reportproposed the stromal interaction molecule-1 (Stim1)=Orai1pathway to mediate SOCE in ECs [19]. More specifically, Stim1is a transmembrane ER Ca2þ sensor that translocates in closeproximity to the plasma membrane upon store depletion andactivates Orai1 Ca2þ channel to gate SOCE [22]. AlthoughEPCs incorporate into neovessels to recapitulate endothelialfunctions following tissue injury or ischemia [4], at the presenttime, it is unknown whether these cells are endowed withsuch a critical signaling pathway, that is, store-dependentCa2þ inflow, before differentiating into mature ECs.

The aim of this article was to investigate SOCE in EPCsharvested from both PB and UCB. We provide the first evi-dence that SOCE occurs also in EPCs upon emptying of InsP3-sensitive intracellular Ca2þ stores. Further, we show that EPCsexpress all the molecular candidates to mediate SOCE inmature ECs, such as TRPC1, TRPC4, Stim1, and Orai1. Finally,we demonstrate that SOCE inhibition blocks EPC prolifera-tion. These findings gain crucial insights into the molecularmechanisms that govern EPC signaling and provide novelputative targets for genetic manipulation aimed at enhancingEPC response to angiogenic stimuli.

Materials and Methods

Isolation and cultivation of EPCs

Blood samples (40 mL) were obtained from healthy humanvolunteers aged from 22 to 28 years. The Institutional ReviewBoard at ‘‘Istituto di Ricovero e Cura a Carattere ScientificoPoliclinico San Matteo Foundation’’ in Pavia approved allprotocols. Informed consent was obtained according to theDeclaration of Helsinki. To isolate ECFCs, MNCs were sep-arated from either PB or UCB by density gradient centrifu-gation on lymphocyte separation medium for 30 min at 400 gand washed twice in endothelial basal medium with 2% FCS(EBM-2). A median of 36�106 MNCs (range: 18–66) from PBand 16�106 MNCs (range: 7–23) from UCB were plated oncollagen-coated culture dishes (BD Biosciences) in the pres-ence of the EC growth medium EGM-2 MV Bullet Kit(Lonza) containing EBM-2, 5% fetal bovine serum, recombi-nant human (rh) EGF, rhVEGF, rhFGF-B, rhIGF-1, ascorbicacid, and heparin and maintained at 378C in 5% CO2 andhumidified atmosphere. Discard of nonadherent cells wasperformed after 2 days; thereafter, medium was changed 3times a week. The outgrowth of ECs from adherent MNCswas characterized by the formation of a cluster of cobble-stone-appearing cells, as previously described [3] (see also

b SF1Supplemental Fig. S1a, b in Supplemental Information,available online at www.liebertonline.com b AU2). Whether ECFC-derived colonies belonged to endothelial lineage was con-firmed by staining with anti-CD31, anti-CD105, anti-CD144,anti-CD146, anti-vWf, anti-CD45, and anti-CD14 monoclonalantibodies (see b ST1Supplemental Table S1 in Supplemental In-formation, available online at www.liebertonline.com) andby assessment of capillary-like network formation in an invitro matrigel assay (see Supplemental Fig. S1c, d in Supple-mental Information, available online at www.liebertonline.com), as previously described [2,30]. More specifically, im-munophenotyping of ECFC-derived cells isolated from bothUCB and PB showed substantially overlapping surface pro-

2 SANCHEZ-HERNANDEZ ET AL.


tein expression. Both UCB- and PB-derived cells showedconsistent expression of endothelial proteins, whereas negli-gible expression of the pan-leukocyte marker CD45 wasobserved, confirming their belonging to the endothelial line-age (see Supplemental Table S1 in Supplemental Information).

Solutions

PSSAU3 c had the following composition (in mM): 150 NaCl, 6 KCl,1.5 CaCl2, 1 MgCl2, 10 glucose, 10 Hepes. In Ca2þ-free solution(0Ca2þ), Ca2þ was substituted with 2 mM NaCl, and 0.5 mMEGTA was added. Solutions were titrated to pH 7.4 withNaOH. The osmolality of the extracellular solution, as measuredwith an osmometer (Wescor 5500), was 300–310 mmol=kg.

[Ca2þ]i measurements

EPCs were loaded with 4mM fura-2 acetoxymethyl ester(fura-2=AM; 1 mM stock in dimethyl sulfoxide) in PSS forAU4 c 1 hmin at room temperature. After washing in PSS, the cover-slip was fixed to the bottom of a Petri dish and the cells wereobserved using an upright epifluorescence Axiolab micro-scope (Carl Zeiss), usually equipped with a Zeiss X63Achroplan objective (water-immersion, 2.0 mm workingdistance, 0.9 numerical aperture). EPCs were excited alter-nately at 340 and 380 nm, and the emitted light was detectedat 510 nm. A first neutral density filter (1 or 0.3 opticaldensity) reduced the overall intensity of the excitation lightand a second neutral density filter (optical density¼ 0.3) wascoupled to the 380 nm filter to approach the intensity of the340 nm light. A round diaphragm was used to increase the

contrast. The excitation filters were mounted on a filter wheel(Lambda 10; Sutter Instrument). Custom software, workingin the LINUX environment, was used to drive the camera(Extended-ISIS Camera; Photonic Science) and the filterwheel and to measure and plot on-line the fluorescence from10 to 15 rectangular ‘‘regions of interest’’ (ROI) enclosing 10–15 single cells. Each ROI was identified by a number. As cellborders were not clearly identifiable, an ROI may not includethe whole EPC or may include part of an adjacent EPC.Adjacent ROIs never superimposed. [Ca2þ]i was monitoredby measuring, for each ROI, the ratio of the mean fluores-cence emitted at 510 nm when exciting alternatively at 340and 380 nm (shortly termed ‘‘ratio’’). An increase in [Ca2þ]i

causes an increase in the ratio [31]. Ratio measurements wereperformed and plotted on-line every 3–5 s. The experimentswere performed at room temperature. The human umbilicalcord rings were loaded with 4mM fura-2=AM for 4 h at roomtemperature and analyzed using the same equipment de-scribed above.

RNA isolation and real-time reversetranscription–polymerase chain reaction

Total RNA was extracted from the EPCs using the QIAzolLysis Reagent (Qiagen). Single cDNA was synthesized fromRNA (1mg) using random hexamers and M-MLV ReverseTranscriptase (Invitrogen S.R.L.). Reverse transcription wasalways performed in the presence or absence (negative con-trol) of the reverse transcriptase enzyme. b AU5q-polymerase chainreaction (PCR) was performed in triplicate using 1mg cDNAand specific primers (intron-spanning primers; b T1Table 1),

Table 1. Primer Sequences Used for Real-Time Reverse Transcription–Polymerase Chain Reaction

Gene Primer sequences Size (bp) Accession number

TRPC1 Forward 50-ATCCTACACTGGTGGCAGAA-30 307 NM_003304.4Reverse 50-AACAAAGCAAAGCAGGTGCC-30

TRPC3 Forward 50-GGAGATCTGGAATCAGCAGA-30 336 NM_001130698.1 variant 1Reverse 50-AAGCAGACCCAGGAAGATGA-30 NM_003305.2 variant 2

TRPC4 Forward 50-ACCTGGGACCTCTGCAAATA-30 300 NM_016179.2 variant alphaReverse 50-ACATGGTGGCACCAACAAAC-30 NM_001135955.1 variant beta

NM_001135956.1 variant gammaNM_001135957.1 variant deltaNM_003306.1 variant epsilonNM_001135958.1 variant zeta

TRPC5 Forward 50-GAGATGACCACAGTGAAGAG-30 221 NM_012471.2Reverse 50-AGACAGCATGGGAAACAGGA-30

TRPC6 Forward 50-AGCTGTTCCAGGGCCATAAA-30 341 NM_004621.5Reverse 50-AAGGAGTTCATAGCGGAGAC-30

TRPC7 Forward 50-CACTTGTGGAACCTGCTAGA-30 387 NM_020389.1Reverse 50-CATCCCAATCATGAAGGCCA-30

Orai1 Forward 50-AGTTACTCCGAGGTGATGAG-30 257 NM_032790.3Reverse 50-ATGCAGGTGCTGATCATGAG-30

Orai2 Forward 50-CCATAAGGGCATGGATTACC-30 334 NM_001126340.1 variant 1Reverse 50-CAGGTTGTGGATGTTGCTCA-30 NM_032831.2 variant 2

Orai3 Forward 50-CCAAGCTCAAAGCTTCCAGCC-30 159 NM_152288.2Reverse 50-CAAAGAGGTGCACAGCCACCA-30

Stim1 Forward 50-CCTCAGTATGAGGAGACCTT-30 347 NM_003156.3Reverse 50-TCCTGAAGGTCATGCAGACT-30

Stim2 Forward 50-AAACACAGCCATCTGCACAG-30 186 NM_020860.2Reverse 50-GGGAAGTGTCGTTCCTTTGA-30

b-Actin Hs_ACTB_1_SG, QuantiTect Primer Assay QT00095431, Qiagen 146 NM_001101

SOCE IN ENDOTHELIAL PROGENITOR CELLS 3


as previously described elsewhere [32]. MESA GREENqPCR MasterMix Plus (Eurogentec) was used according tothe manufacturer’s instruction and qPCR was performed us-ing Rotor Gene 6000 (Corbett). The conditions used wereinitial denaturation at 958C for 5 min, 40 cycles of denatur-ation at 958C for 30 s, annealing at 588C for 30 s, and elonga-tion at 728C for 40 s. The q-reverse transcription (RT)-PCRreactions were normalized using b-actin as a housekeep-ing gene. Melting curves were generated to detect the melt-ing temperatures of specific products immediately after thePCR run.

The triplicate threshold cycle (Ct) values for each samplewere averaged, resulting in mean Ct values for both the geneof interest and the housekeeping gene b-actin. The gene Ctvalues were then normalized to the housekeeping gene bytaking the following difference: DCt¼Ct[gene]�Ct[b-actin],with high DCt values reflecting low mRNA expression levels.Relative gene expression quantitation was determined by the2�DDCt method [33]. Thus, to compare the relative levels ofgene expression of UCB-EPC versusAU3 c BM-EPC, DDCt valueswere first calculated: DDCt¼DCt (UCB-EPC)�DCt (BM-EPC), and then fold changes were determined using trans-formation: fold increase¼ 2�DDCt; fold decrease¼�2DDCt.

The sequences of the bands were checked by using the BigDye Terminator Cycle Sequencing kit (Applied Biosystem).PCR products were also separated by agarose gel electro-phoresis, stained with ethidium bromide, and acquired withthe Image Master VDS (Amersham Biosciences Europe). Themolecular weight of the PCR products was compared withthe DNA molecular weight marker VIII (Roche MolecularBiochemicals).

Sample preparation and immunoblotting

PB-EPCs and UCB-EPCs were homogenized by using aDounce homogenizer in a solution containing 250 mM su-crose, 1 mM ethylenediaminetetraacetic acid, 10 mM Tris-HCl (pH 7.6), 0.1 mg=mL PMSFAU1 c , 100 mM b-mercaptoethanol,and Protease Inhibitor Cocktail (P8340; Sigma). Thehomogenates were solubilized in Laemmli buffer [34] and20–40 mg proteins were separated on 10% sodium dodecylsulfate–polyacrylamide gel electrophoresis and transferredto the Hybond-P PVDF Membrane (GE Healthcare) byelectroelution. After 1 h blocking with Tris-buffered salinecontaining 3% bovine serum albumin and 0.1% Tween(blocking solution), the membranes were incubated for 3 hat room temperature with affinity-purified antibodiesdiluted 1:200 in the Tris-buffered saline and 0.1% Tween.Anti-Stim1 (sc-166840) and anti-TRPC4 (sc-15063) wereobtained from Santa Cruz Biotechnology, anti-Orai1 wasfrom Alomone Labs, and anti-TRPC1 (T8276) from Sigma-Aldrich.

The membranes were washed and incubated for 1 h withperoxidase-conjugated mouse, rabbit, and goat IgG(1:120,000 in blocking solution) (QED Bioscience). The bandswere detected with the ECL� Advance western blottingdetection system (GE Healthcare Europe GmbH). Controlexperiments were performed by using the antibody pre-adsorbed with a 20-fold molar excess of the immunizingpeptide and by incubating the blots with nonimmune serum.BenchMark prestained protein ladders (Invitrogen) wereused to estimate the molecular weights.

Protein content

Protein contents of all the samples were determined by theBradford’s method, using bovine serum albumin as a stan-dard [35].

Proliferation assays

A total of 1�105 ECFC-derived cells (first passage) wereplated in 30-mm collagen-treated dishes in EGM-2 MV me-dium with or without 2–20mM b AU3BTP-2. We chose this rangeafter preliminary experiments, which showed no or unspe-cific toxic effect for lower or higher concentrations of BTP-2,respectively. Cultures were incubated at 378C (in 5% CO2

and humidified atmosphere) and cell growth was assessedevery day until confluence was reached in the control cul-tures (0 mM BTP-2). At this point, cells were recovered bytrypsinization from all dishes and the cell number was as-sessed by counting in a hemocytometer. The percentage ofgrowth inhibition by the drug was calculated by dividingthe total number of cells obtained in presence of BTP-2 by thenumber of cells in control experiments and multiplying theratio by 100.

Statistics

All the data have been collected from EPCs isolated fromthe PB of at least 3 healthy volunteers or 3 different samples ofUCB. The amplitude of the peak Ca2þ response was measuredas the difference between the ratio at the peak and the meanratio of1 min baseline before the peak. Plateau amplitude wasmeasured by averaging 60 s of signal at 10 min after the peak.Pooled data are given as mean� SE and statistical significance(P< 0.05) was evaluated by the Student’s t-test for unpairedobservations.

As to mRNA analysis, all data are expressed as meansSEM. Statistical analysis of qRT-PCR experiments was pri-marily performed on DCt values. The significance of thedifferences of the means was evaluated with 1-way ANOVAfollowed by Newman–Keuls’s Q test or Student’s t-test. Allstatistical tests were carried out with GraphPad Prism 4.

Chemicals

EBM and EGM-2 were purchased from Clonetics (CellSystem). Fura-2=AM was obtained from Molecular ProbesEurope BV. N-(4-[3,5-Bis(trifluoromethyl)-1H-pyrazol-1-yl]phenyl)-4-methyl-1,2,3-thiadiazole-5-carboxamide (BTP-2)was purchased from Calbiochem. All other chemicals wereobtained from Sigma Chemical.

Results

Stimulation of SOCE by CPA-induced storeemptying in EPCs harvested from PB

To assess the presence of a functional store-operated Ca2þ

inflow, we carried out fluorescence measurements of EPCsloaded with the Ca2þ-sensitive fluorochrome, Fura-2, andexposed to CPA. CPA, a widely employed SERCA blocker[21], prevents Ca2þ sequestration into the stores, thus leadingto their depletion and to SOCE activation [21–23]. As de-picted in b F1Fig. 1A, when PB-EPCs were bathed in absence ofextracellular Ca2þ (0Ca2þ), CPA (10mM) induced a transient



rise in [Ca2þ]i because of passive emptying of Ca2þ storesthrough leakage channels in ER membrane. Thereafter, Ca2þ

levels decayed to the baseline as the plasma membranetransporters (ie, plasma membrane Ca2þ-ATPase andNaþ=Ca2þ exchanger) extruded Ca2þ from the cytosol. Sub-sequent addition of Ca2þ (1.5 mM) to the extracellular me-dium (the classic ‘‘Ca2þ add-back’’ protocol) induced Ca2þ

inflow through opened store-operated channels, which re-sulted in a robust increase in intracellular Ca2þ levels (Fig.1A, C). In control experiments, replenishment of extracellularCa2þ in absence of CPA did not elicit a detectable elevation in[Ca2þ]i (data not shown). Removal of extracellular Ca2þ re-sulted in a drop in intracellular Ca2þ levels in a fraction ofPB-EPCs (77 of 211 cells; 36.5%) (Fig. 1B). This result hinted atthe presence of a resting Ca2þ permeability that was not in-vestigated. The store-dependent nature of CPA-elicited Ca2þ

entry into EPCs was further assessed by employing BTP-2, aspecific inhibitor of SOCE [36]. Acute application of BTP-2(20mM) strongly inhibited CPA-dependent Ca2þ inflow inPB-EPCs (n¼156) (Fig. 1A). Accordingly, 30-min preincuba-tion with BTP-2 (20mM) prevented SOCE in PB-EPCs (Fig.1B, C), whereas it did not affect the ER Ca2þ content (Fig. 1B,C). Altogether, these data indicate that a functional SOCE isexpressed in both PB-EPCs. Conversely, such cells lackfunctional voltage-gated Ca2þ channels (VGCCs), whichpromote Ca2þ inrush into excitable cells. Consistently, whenPB-EPCs were exposed to high KCl extracellular solution, aprotocol that may activate voltage-dependent Ca2þ transientsin ECs obtained from bovine adrenal medulla [37], no Ca2þ

signal could be detected (n¼ 59) (data not shown).

SOCE may be engaged by physiologicalstimulation in PB-EPCs

SOCE engagement following activation of GPCRs wasinvestigated by first testing the effect of a number of vaso-active agonists that mobilize Ca2þ from InsP3-sensitive in-tracellular stores [18]. b F2Figure 2 shows that neitherl-epinephrine (50mM; n¼ 51) nor acetylcholine (50mM; n¼ 76)evoked a detectable Ca2þ signal in PB-EPCs, whereas bothbradykinin (1–5mM; n¼198) and ATP (100mM) elicited a largeincrease in [Ca2þ]i. Notably, all the ligands were applied atconcentrations known to exert maximal effects on mature ECs[38–40]. ATP, but not bradykinin, evoked a clear plateau(0.069� 0.001; n¼ 600) following the initial Ca2þ peak (0.143�0.003; n¼ 593) (Fig. 2C, D), a feature that hinted at SOCEactivation. As reported in mature endothelium [18], the [Ca2þ]i

plateau rapidly returned to prestimulation level on agonistremoval (Fig. 2D). Therefore, we decided to employ ATP tostudy the physiological activation of SOCE in EPCs. The roleof Ca2þ influx during agonist stimulation was investigated byremoving extracellular Ca2þ (0Ca2þ) during the plateau phase.This maneuver caused Ca2þ to decline even in the continuedpresence of ATP in PB-EPCs (n¼128) ( b F3Fig. 3A). In accordancewith these results, the plateau phase disappeared when ATPwas applied in the absence of extracellular Ca2þ in PB-EPCs(n¼165) (Fig. 3B). Under these conditions, the peak of theCa2þ increase was significantly (P< 0.05) higher than in thepresence of external Ca2þ (0.143� 0.003, n¼ 413, PSS, vs.0.180� 0.005, n¼ 635, 0Ca2þ). The role served by SOCE in theCa2þ response to ATP was investigated by first separatingthe 2 components of the signal, namely intracellular Ca2þ

FIG. 1. SOCE is expressed in EPCs harvested from PB. (A)Intracellular Ca2þ pools were depleted by exposing the cellsto CPA (10 mM) in 0Ca2þ solution. Readdition of extracellularCa2þ led to an increase in [Ca2þ]i, which was indicative ofSOCE. Acute application ofAU3 c BTP-2 (50 mM) inhibited CPA-elicited Ca2þ inflow, thus confirming the store-dependent na-ture of the Ca2þ entry pathway. (B) Twenty-minute preincuba-tion with BTP-2 (20mM) prevented CPA-induced Ca2þ signal onCa2þ readdition to the extracellular solution in PB-EPCs. (C)Statistical evaluation of the effect exerted by BTP-2 on the peakamplitudes of both Ca2þ release and Ca2þ entry stimulated byCPA. The asterisk indicates a level of significant difference of<0.05AU7 c . EPC, endothelial progenitor cell; SOCE, store-operatedCa2þ entry; PB, peripheral blood; CPA, cyclopiazonic acid.



release and external Ca2þ inflow. PB-EPCs were exposed toATP (100mM) in 0Ca2þ to deplete intracellular stores and ac-tivate the SOCE mechanism (F4 c Fig. 4A). Thereafter, Ca2þ wasreintroduced in the absence of the agonist to rule out the en-gagement of receptor-dependent Ca2þ permeable channels[41], such as P2X ionotropic receptors [42], or second messen-ger–operated Ca2þ channels [43]. Such a maneuver induced arobust increase in [Ca2þ]i in PB-EPCs (Fig. 4A), thus hinting atSOCE recruitment by GPCRs. These data were supported bythe BTP-2 sensitivity of Ca2þ entry. Indeed, 30 min pretreat-ment with BTP-2 (20mM) prevented ATP-dependent SOCE inPB-EPCs, whereas it did not affect the intracellular Ca2þ re-lease (Fig. 4B, C). Moreover, acute perfusion of BTP-2 (50mM)blocked ATP-induced sustained component (n¼112) (Fig.4D). Finally, 30-min preincubation with BTP-2 (20mM)strongly dampened the plateau phase of the Ca2þ signal in PB-EPCs (n¼ 265) (Fig. 4E), which thus exhibited a transient Ca2þ

response. Consistent with the 0Ca2þ experiments, where Ca2þ

influx across the plasma membrane was prevented, inhibit-ing Ca2þ entry resulted in a significant elevation in the initialpeak (0.086� 0.004, n¼165, PSS, vs. 0.170� 0.00554, n¼ 265,BTP-2). The higher Ca2þ release observed in absence of Ca2þ

inflow might depend on the well-known biphasic regula-tion of InsP3Rs by Ca2þ. Indeed, local Ca2þ in the sub-micromolar range exerts an excitatory effect on InsP3Rchannels, whereas Ca2þ >10mM may inhibit the gating [44]. Itis, therefore, conceivable that Ca2þ entry partially hinderedInsP3-dependent Ca2þ mobilization in PB-EPCs. This hypoth-esis is supported by the fact that CPA-induced Ca2þ release,

which is not mediated by InsP3Rs, is not influenced by re-moval of extracellular Ca2þ.

GPCRs-induced Ca2þ store depletion requiresthe PLC=InsP3 pathway

In mature ECs, SOCE recruitment by GPCRs is mediatedby the PLC-b=InsP3 signaling pathway. The following ex-periments were carried out in 0Ca2þ to prevent any con-taminating signal from Ca2þ entry and focus on themolecular mechanisms underlying ER Ca2þ store depletion.To assess the effects of the selective antagonists of the PLC-b=InsP3 pathway on the same cells, ATP was bath-applied 3times (control, experiments, washout) for 100 s at 30-min(U73122) and 10-min (caffeine) intervals. U73122 (10mM), awidely employed PLC-b blocker [31,45], prevented ATP-induced [Ca2þ]i elevation in all 96 PB-EPCs ( b F5Fig. 5A).Washout of the drug led to the recovery of Ca2þ response toATP (Fig. 5A). Subsequently, we assessed the effect of caf-feine on intracellular Ca2þ homeostasis. Depending on theconcentration, in vascular endothelium, caffeine may bothinhibit InsP3Rs [31,45] and activate ryanodine receptors(RyRs) through the process of Ca2þ-induced Ca2þ release(CICR) [46]. When applied at 20 mM, a dose able to stimulateRyRs in rabbit aortic ECs [46], caffeine did not elicit anydetectable Ca2þ transient (n¼ 83) (Fig. 5B), a feature thatsuggested the absence of functional RyRs in EPCs. Notably,lower concentrations of the drug (100 mM), which selectivelyblock InsP3Rs, reversibly hindered the Ca2þ response to ATP

FIG. 2. Different effects of endothelial agonists on [Ca2þ]i in EPCs harvested from PB. l-Epinephrine (50mM) (A) andacetylcholine (50 mM) (B) did not elicit any detectable elevation in [Ca2þ]i. (C) Bradykinin (5mM) caused a transient Ca2þ

signal in PB-EPCs. (D) ATP (100 mM) evoked an initial rise in [Ca2þ]i followed by a plateau of lesser magnitude, which wassustained while the agonist was present in PB-EPCs.



(Fig. 5C). The statistical analysis of these data has beensummarized in Fig. 5D. Finally, depletion of InsP3Rs-sensi-tive store by previous addition of the SERCA inhibitor, CPA(10mM), in absence of extracellular Ca2þ prevented ATP-in-duced elevation of [Ca2þ]i (n¼ 50) (not shown). Collectively,these data indicate that, as in mature ECs, the PLC-b=InsP3

signaling pathway causes Ca2þ store depletion, which leadsto SOCE. The latter route is engaged by metabotropic P2Y

receptors when the incoming stimulus is provided by ATP.Consistently, the nonselective purinergic antagonist, suramin(300 mM) [47], reversibly inhibited the Ca2þ response to ATPin PB-EPCs (Fig. 5E). Indeed, the magnitudes of the Ca2þ

peak were 0.314� 0.012, n¼119, and 0.059� 0.005, n¼119, inabsence and presence, respectively, of the drug. Suramin is afluorescent molecule and its application rapidly decreasesthe basal fluorescence ratio (Fig. 5E) (see also 31AU6 c ). However,the presence of the drug still allows the fluorimetric systemto monitor intracellular Ca2þ signals, as indicated by expos-ing the cells to bradykinin (5mM) after suramin addition (notshown).

The mRNA encoding for putative mediatorsof SOCE is present in PB-EPCs

The expression of the molecular entities that have beenproposed to underpin store-dependent Ca2þ entry in matureendothelium has been investigated in PB-EPCs by employing

qRT-PCR. The specific primers depicted in Table 1 have beenemployed to evaluate the expression of mRNAs encoding forall TRPC (TRPC1 and TRPC3–7) channels expressed in hu-mans [48] and for the recently cloned Orai (Orai1-3) and Stim(Stim1 and Stim2) genes [22,34]. TRPC1, TRPC4, Orai1, Orai2,Orai3, Stim1, and Stim2 transcripts were readily detectable inPB-EPCs ( b F6Fig. 6A), whereas there were no significant amountof mRNA for TRPC3, TRPC5, TRPC6, and TRPC7. Singlebands of the expected size of cDNA fragments were ampli-fied and the results of agarose gel electrophoresis of repre-sentative PCR products are shown in Fig. 6A. Negativecontrols were performed by omitting the reverse transcrip-tase (not shown). The comparison of DCt values of mRNAobtained by qRT-PCR in PB-EPCs showed that Orai1 tran-script was significantly higher than Orai2, Orai3, Stim1,Stim2, and TRPC1 and that TRPC4 transcript was the lowest(P< 0.05, ANOVA followed by Newman–Keuls’ Q test)(Fig. 6B).

SOCE is expressed in EPCs isolated from UCB

To assess SOCE expression in UCB-EPCs, the latter un-derwent the Ca2þ add-back protocol upon exposure to CPA(10 mM) in 0Ca2þ. As illustrated in b F7Fig. 7A, CPA induced aBTP-2–sensitive Ca2þ entry, which is the hallmark of SOCE,in 179 cells. Similar to PB-EPCs, removal of extracellular Ca2þ

caused a decrease in intracellular Ca2þ levels in a minorpercentage of UCB-EPCs (29 of 97 cells, 29.8%) (Fig. 7A). Asobserved in their peripheral counterparts, UCB-EPCs dis-played a biphasic Ca2þ response to ATP (100mM) (Fig. 7B). Inthese cells, the amplitudes of the transient Ca2þ rise and thesustained components were, respectively, 0.162� 0.004(n¼ 224) and 0.064� 0.002 (n¼ 216). Both values are signifi-cantly (P< 0.05) smaller than those (see above) measured inPB-EPCs. As observed in the latter, when external Ca2þ wasomitted during the plateau phase, [Ca2þ]i returned to theresting levels in all 130 cells (Fig. 7B). Consistent with thisfinding, when ATP was administered in absence of extra-cellular Ca2þ, the cells exhibited only the transient peakwithout the sustained component (Fig. 7C). As in PB-EPCs,the height of the Ca2þ response to ATP was significantly(P< 0.05) higher in 0Ca2þ (0.164� 0.006; n¼132) than in PSS(0.207� 0.007; n¼103). Notably, the plateau phase of ATP-elicited Ca2þ signal was strongly inhibited by BTP-2 (50 mM)in 131 cells (Fig. 7D). Collectively, these results indicate thatalso UCB-EPCs express a SOCE pathway that may be en-gaged following physiological stimulation of GPCRs. TheqRT-PCR analysis revealed that, in addition to the transcriptsfound in PB-EPCs, UCB-EPCs express TRPC3 ( b F8Fig. 8A).Single bands of the expected size of cDNA fragments wereamplified and the results of agarose gel electrophoresis ofrepresentative PCR reaction products are shown in Fig. 8A.Negative controls were performed by omitting the reversetranscriptase (not shown). The comparison of DCt valuesof the Ca2þ signal pathway mRNA obtained by qRT-PCRshowed that TRPC1, TRPC3, TRPC4, Orai2, and Stim1 tran-scripts were significantly lower than Orai1, Orai3, and Stim2(P< 0.01, ANOVA followed by Newman–Keuls’s Q test)(Fig. 8B). Finally, when comparing the Ca2þ channel geneexpression in the 2 cell populations, we found that Orai1,Orai3, and Stim2 levels were 6-, 22-, and 46-fold higher inUCB-EPCs, respectively ( b F9Fig. 9). Conversely, Orai2 and Stim1

FIG. 3. ATP-induced sustained elevation in [Ca2þ]i requiresCa2þ entry. (A) ATP (100mM) induced a biphasic Ca2þ signalin PB-EPCs. Removal of extracellular Ca2þ (0Ca2þ) causedthe rapid recovery of [Ca2þ]i to resting levels. (B) When ATPwas applied in the absence of extracellular Ca2þ, the intra-cellular Ca2þ signal displayed transient kinetics.



levels were 18- and 12-fold higher in PB-EPCs, respectively,whereas TRPC1 and TRPC4 were similar (Fig. 9).

SOCE-related proteins are expressedin both PB- and UCB-EPCs

Western blot analysis of Orai1, Stim1, TRPC1, and TRPC4expression was conducted using affinity-purified antibodies.As observed for the RNA transcripts, all these proteins weredetected in both PB-EPCs and UCB-EPCs (F10 c Fig. 10). Im-munoblots showed major bands of about 33, 110, 110, and100 kDa for Orai1, Stim1, TRPC1, and TRPC4, respectively.The molecular sizes were similar to those reported in theliterature [19,49]. When the antibodies were preadsorbedwith large amounts of the immunizing peptide or the anti-bodies substituted with nonimmune serum, the proteinbands completely disappeared, thus indicating the specificityof the reaction. Along with the qRT-PCR analysis, these datademonstrate that both PB- and UCB-EPCs possess the mo-lecular machinery that has been associated with SOCE inmature ECs.

SOCE drives EPC proliferation

To assess the physiological outcome of SOCE, we focusedon its involvement in EPC replication, as such a pathwaymay regulate the growth of mature ECs [19]. The prolifera-tion assay was conducted on PB-EPCs. As shown in b T2Table 2,BTP-2 inhibited cell growth in a dose-dependent manner (2–20 mM). None of the cultures carried on in the presence ofBTP-2 reached confluence at the time (3 days) reached bycontrol cultures. Confluence was never reached by BTP-2–exposed cells at longer incubation and cells usually begandetaching and dying after 5–6 days of culture. Therefore,SOCE stands as a novel player in the signaling networkunderlying EPC proliferation and provides an additionaltarget to enhance the outcome of cell-based therapy.

Discussion

EPCs are mobilized from BM to promote neovasculariza-tion and favor reendothelialization following vascular dam-age [1,4,11]. This observation prompted the injection of

FIG. 4. ATP engages SOCE in EPCs isolated from PB. (A) PB-EPCs chal-lenged with ATP (100mM) in the absence of extracellular Ca2þ exhibited atransient elevation in [Ca2þ]i. Restoration of external Ca2þ in the absence ofagonist caused a sustained Ca2þ signal, which was indicative of SOCE. (B)Twenty-minute pretreatment with BTP-2 (20mM) prevented ATP-inducedCa2þ inflow on replenishment of extracellular Ca2þ. (C) Amplitudes of ATP-dependent intracellular Ca2þ mobilization and Ca2þ entry in PB-EPCs. Theasterisk indicates a level of significant difference of <0.05. (D) Acuteapplication of BTP-2 (50mM) inhibited the plateau phase of ATP-elicited in-crease in [Ca2þ]i in PB-EPCs. (E) Twenty-minute pretreatment with BTP-2(20mM) resulted in a transient Ca2þ signal when ATP (100mM) was applied toPB-EPCs.



exogenous EPCs, harvested from both peripheral and UCB,as a novel therapeutic strategy to treat ischemic diseases[1,13]. Indeed, EPCs home to sites of tissue regeneration andcontribute to neovascularization by acquiring an endothelialphenotype and being incorporated into neovessels [1,4]. Inmature ECs lining the lumen of blood vessels, SOCE acti-vation by GPCRs is central to the control of a wide number offunctions, including proliferation, secretion, vascular per-meability, and regulation of blood pressure [14–19,24]. In theview of such a relevant role, it was worth investigatingwhether EPCs possess the SOCE pathway as well. Our re-sults provide the first evidence that store-dependent Ca2þ

inflow is present in both EPCs flowing in PB and EPCsharvested from UCB. Further, both PB-EPCs and UCB-EPCsexpress all the molecular candidates to mediate SOCE inmature ECs, that is, TRPC1, TRPC4, and Stim1=Orai1. Wefinally show that the signal transduction cascade involved inSOCE opening in differentiated endothelium, that is, ER Ca2þ

store depletion by InsP3Rs activation, is also present in EPCs.Several lines of evidence indicate that GPCRs recruit

SOCE by emptying the ER reservoir via PLC-b=InsP3 sig-naling in EPCs. First, ligand (ie, ATP)-induced depletion of

the intracellular Ca2þ pool in 0Ca2þ is a signal sufficient toactivate Ca2þ inflow on Ca2þ readdition in the absence ofagonist. As widely reported in the literature [21,23,41], thisfeature implies that Ca2þ entry occurs independently on re-ceptor occupancy and does not rely on intracellular secondmessengers, such as InsP3 and DAG. When the agonist iswithdrawn in the continuous presence of Ca2þ (Fig. 2D),[Ca2þ]i recovers to the baseline as InsP3 synthesis terminatessuch that InsP3Rs close. As a consequence of the rapid in-terruption of Ca2þ efflux from ER, the intracellular Ca2þ poolrefills and SOCE is inactivated [39,50,51]. On the other hand,exposure to ATP in the absence of extracellular Ca2þ leads toa transient Ca2þ signal (Fig. 4A), as most of Ca2þ releasedfrom InsP3Rs is extruded by plasma membrane transporters,such as Naþ=Ca2þ exchanger and plasma membrane Ca2þ-ATPase. The ensuing depletion of the intracellular Ca2þ

reservoir results in SOCE activation and in the bump in[Ca2þ]i observed on Ca2þ replenishment to the bath even inthe absence of agonist [21,23,41,50,51]. Accordingly, passivedepletion of the ER Ca2þ content with CPA led to theopening of Ca2þ-permeable channels on the plasma mem-brane, which is the hallmark of SOCE [21,23], in both

FIG. 5. The phospholipase C=inositol-1,4,5-trisphosphate signaling pathwayleads to endoplasmic reticulum Ca2þ store depletion in EPCs obtained fromPB. (A) Ten-minute preincubation with U73122 (10mM) prevents the Ca2þ

response to ATP (100mM). ATP-induced Ca2þ signaling resumes upon drugwashout. (B) High doses (20 mM) of caffeine did not elicit any detectable Ca2þ

transients. (C) Low doses (100 mM) of caffeine reversibly dampen ATP-evokedincrease in [Ca2þ]i. (D) Effect of low concentrations (100mM) of caffeine on thepeak amplitude of ATP-induced [Ca2þ]i elevation. The single asterisk indi-cates a level of significant difference of <0.05. (E) The purinergic antagonist,suramin (300mM), reversibly hinders the Ca2þ response to ATP (100 mM) inPB-EPCs.



PB-EPCs and UCB-EPCs. Second, BTP-2, the hitherto mostpowerful and selective inhibitor of store-dependent Ca2þ

inflow [36], abrogated CPA- and P2Y receptor-induced Ca2þ

inflow in both cell types. Third, U73122, a widely employedpharmacological PLC blocker [31], strongly hinderedATP-elicited depletion of the ER Ca2þ pool. This finding issupported by a recent study that reported the expression ofPLC-b2 in mice UCB-EPCs at both mRNA and protein levels[52]. Fourth, low doses of caffeine, which have long beenknown to antagonize InsP3-gated channels in several celltypes, including rat cardiac microvascular ECs [31,45 andreferences therein], significantly reduced ATP-dependentCa2þ mobilization from ER. Further, higher doses of caffeine,which stimulate RyRs to release Ca2þ via the so-called CICR,did not trigger any Ca2þ increase under control conditions.This result suggests either that functional RyRs are not ex-pressed in EPCs or that resting Ca2þ levels are too low toenable the drug to sensitize RyRs [53]. Therefore, althoughour data indicate that RyRs are unlikely to participate in thesignal transduction pathway leading to SOCE, further ana-lyses are necessary to rule out their expression in EPCs.Notably, we found that bradykinin, which may promoteCa2þ signals via the G-protein–coupled B2 receptor [54],

elicited a robust InsP3-dependent Ca2þ release, but not store-dependent Ca2þ entry (Fig. 3C). This feature, however, is notsurprising when considering that partial dissociation be-tween Ca2þ mobilization and SOCE activation may occur inseveral cells, including rat basophilic leukemia cells [55]. Asreviewed in ref. [23], it has been suggested that 2 distinctInsP3-regulated Ca2þ stores exist within the ER: one mainlyresponsible for Ca2þ release and another tightly coupled toSOCE. According to Parekh and Penner [56], the InsP3-sen-sitive subcompartment devoted to the gating of store-oper-ated channels is likely to lie in close proximity to the plasmamembrane. It is, therefore, conceivable that, in human EPCs,ATP, but not bradykinin, engages such a pool and leads tostore-dependent Ca2þ inflow. The same feature is likely toexplain why SOCE amplitude is significantly higher in CPA-rather than ATP-treated cells (compare Fig. 1C with Fig. 4C).Indeed, CPA might deplete all the InsP3-dependent pool thatis tightly coupled to plasmalemmal channels, whereas only afraction of such a reservoir may be recruited by ATP [23,56].

The report of a functional SOCE pathway in both PB-EPCsand UCB-EPCs has 2 remarkable implications: (1) unlike othercomponents of the endothelial signaling toolkit, such asmuscarinic receptors or RyRs, this route is already expressedin immature cells, and consequently, (2) store-dependent Ca2þ

inflow stands as a novel candidate among the mechanismsregulating EPC mobilization, proliferation, and homing.

The molecular underpinnings of SOCE in mature humanendothelium, as well as in other cell types, include severalcandidates, such as TRPC1, TRPC4, and Stim1=Orai1[19,20,27–29]. Earlier studies provided the evidence that themolecular make-up of the Ca2þ channels gated by a drop inER Ca2þ content in ECs comprised both TRPC1 and TRPC4,the latter being the subunit essential for coupling store de-pletion to channel activation [20,27–29]. These data havebeen recently challenged by Trebak’s group, who showedthat small interference RNAs (siRNAs) targeted againsteither Stim1 or Orai1, but not against TRPC1 and TRPC4,inhibited SOCE in human umbilical vein ECs (HUVECs) [19].This apparent discrepancy might arise from the well-knownheterogeneity in the endothelial phenotype [57], so that ECslocated in different vascular beds might mediate SOCEthrough distinct mechanisms [57,58]. Notably, qRT-PCR hasrevealed mRNA encoding for TRPC1=TRPC4 and Stim1=Orai1 in PB-EPCs and UCB-EPCs. The abundance of relativegene expression for these proteins was Orai1> Stim1¼TRPC1>TRPC4 in PB-EPCs and Stim1¼TRPC4>TRPC1Orai1 in UCB-EPCs. When comparing the expression ofsuch transcripts between the 2 cell types, we found thatOrai1 and Stim1 are more abundant in PB-EPCs and UCB-EPCs, respectively, whereas TRPC1 and TRPC4 are ex-pressed at the same level in both EPC types. Such a resultcannot explain the fluorimetric measurements (ie, SOCEwas higher in PB-EPCs) because of the tight stoichiometriccoupling among the molecular candidates of SOCE [22,36].Western blot experiments confirmed that all the transcriptsencoding TRPC1=TRPC4 and Stim1=Orai1 are translatedinto proteins of the expected molecular mass in both PB-EPCs and UCB-EPCs. These data are consistent with thepresence of all the putative mediators of store-operatedCa2þ channels in mature ECs. Therefore, siRNA experi-ments are required to unveil the molecular identity of store-dependent Ca2þ inflow in EPCs. The latter expressed also

FIG. 6. Expression of classical transient receptor potential(TRPC) channels, Orai, and STIM proteins in EPCs harvestedfrom PB. (A) Gel electrophoresis of the PCR products. ThePCR products were of the expected size: TRPC1, 307 bp;TRPC4, 300 bp; Orai1, 257 bp; Orai2, 334 bp; Orai3, 159 bp;Stim1, 347bp; Stim2, 186 bp. No signal was observed forTRPC3, TRPC5, TRPC6, and TRPC7. MW: molecular weightmarker. Blank: reaction without template. (B) mRNA levelswere measured by real-time PCR relative to the b-actin in-ternal standard (see Materials and Methods section) and thevalues obtained were reported as DCt. Bars represent themean� SEM of at least 4 different experiments each fromdifferent RNA extracts. **P< 0.001 versus TRPC1, Orai1,Orai2, Orai3, Stim1, Stim2; *P< 0.05 versus TRPC1 (1-wayANOVA followed by Newman–Keuls’s Q test). PCR, poly-merase chain reaction.



Stim2 and Orai2-3, which have been involved in SOCE indifferentiated cells [22,36]. Stim2 has been reported to con-trol the basal Ca2þ leak in several cell types [59], includingHUVECs, and might regulate the resting Ca2þ inflow thatwe observed upon removal of external Ca2þ in a fraction ofEPCs. As to Orai2 and Orai3, they have been also found inHUVECs [19] and might contribute to SOCE by formingheteromultimers with Orai1 in heterologous expressionsystems [22,36]. Nevertheless, their participation to GPCR-induced store-dependent Ca2þ entry in vascular endothe-lium is unlikely [19]. It is worth noting that we could notdetect any transcript for other members of the TRPC familythat are known to modulate endothelial function [17,60].Indeed, unlike ECs from human arteries [60] and humanmicrovessels [61,62], no transcripts of TRPC2, TRPC3,TRPC5, TRPC6, and TRPC7 were found by either quanti-tative or qualitative RT-PCR in PB-EPCs, whereas onlyTRPC3 was detected in UCB-EPCs. Preliminary experi-ments confirmed that the latter channel is also expressed atprotein level. This feature is rather relevant when consid-ering that peripheral EPCs may differentiate into matureECs on mobilization from bone marrow. For instance,TRPC3, TRPC6, and TRPC7 mediate VEGF-triggered Ca2þ

influx following an increase in the levels of the intracellularsecond messenger, DAG [17,60], in ECs isolated from sev-eral vascular beds [61,62]. It is, therefore, conceivable toassume that EPCs express the mRNA encoding for thesechannels when they are incorporated within the foci ofneovascularization, where they acquire a mature phenotypeand are exposed to site-specific environmental cues [57]. Inagreement with this hypothesis, human cardiac progenitorcells do not express RyRs, although the latter drives Ca2þ

cycling in differentiated cardiomyocytes [63]. Further, EPCsharvested from both peripheral and cord blood manifest anumber of endothelial markers, such as CD31 and vascularcell adhesion molecule-1, and the ability to incorporate DiI-acetylated low-density lipoprotein, only following in vitrodifferentiation [2]. The same feature is likely to apply alsofor muscarinic and catecholamine receptors, which triggerintracellular Ca2þ elevations in mature ECs [18,39,40], butnot in PB-EPCs. Similarly, voltage-dependent Ca2þ tran-sients have been detected in bovine adrenal medulla cap-illary ECs [37], whereas they are absent in PB-EPCs. In thisview, it is predictable that the reported effect of benidipine,a dihydropyridine Ca2þ channel blocker, on the differ-entiative outcome of mice PB-EPCs is not due to selective

FIG. 7. SOCE is expressed in EPCs isolated from UCB. (A) UCB-EPCs were challenged by incubation with CPA (10 mM) inthe absence of extracellular Ca2þ to deplete endoplasmic reticulum Ca2þ stores. SOCE was induced by readding Ca2þ to theextracellular solution. Acute application of BTP-2 (50 mM) inhibited SOCE-mediated Ca2þ inflow. (B) ATP (100mM) induced aCa2þ signal comprising an initial Ca2þ peak followed by a plateau phase of lower amplitude. The plateau phase was rapidlyabolished by removal of extracellular Ca2þ. (C) The Ca2þ response to ATP (100 mM) measured in the absence of extracellularCa2þ lacked the sustained component. (D) Acute application of BTP-2 (50mM) exerted an inhibitory effect on the plateauphase similar to that observed upon removal of external Ca2þ. UCB, umbilical cord blood.



inhibition of Ca2þ entry [64]. This feature concurs with thenotion that benidipine stimulates EPC differentiation byactivating the phosphoinositide-3 kinase (PI3K)=Akt path-way [64].

It has long been known that Ca2þ entry is a key regulatorof endothelial proliferation and migration [16–19]. For in-stance, knocking-down either Stim1 or Orai1, which mediatesSOCE in HUVECs, inhibited EC proliferation and caused cellcycle arrest at S and G2=M phases [19]. In addition, eitherreduction of extracellular Ca2þ or inhibition of plasma-lemmal Ca2þ channels have been repeatedly shown to pre-vent cell growth and replication [16,65]. This featurealso applies to immature cells, as blockage of SOCE dra-matically impairs PB-EPC growth. Ca2þ entry may betranslated into a mitogenic signal by direct binding of the ionto Ca2þ-dependent targets or by Ca2þ binding to intracellulardecoders such as calmodulin (CaM) and its downstreamtargets, including CaM kinase (CaMK) and calcineurin[16,17]. The recruitment of the CaM=CaMK pathway is cru-cial to cell cycle progression through G1 and mitosis byregulating the activation of several cyclin-dependent kinases(cdk), such as cdk2 and cdk4, whereas calcineurin controlsthe transcription factors that promote the G1=S transition,including cAMP-responsive element binding protein 1 and

NFAT [16,65]. The finding that SOCE governs EPC prolifer-ation highlights a novel alternative target to enhance theregenerative outcome of cell-based therapy. This latter ap-proach mainly relies on 2 critical parameters: (1) EPC homingto ischemic organs and (2) EPC incorporation into newlygrowing vessels, proliferation, and differentiation into ma-ture ECs [10–12]. Therefore, once established the molecularmachinery underlying SOCE in EPCs, transfecting autolo-gous cells with genes encoding for the identified channelscould be instrumental to augment EPC number and improveneovascularization. Consistent with this hypothesis, recruit-ment of PLC-b2 by the insulin-like growth factor 2 (IGF2)receptor is essential to ensure proper UCB-EPC homing tosites of hind-limb ischemia [52]. Further, CD133þ hemato-poietic stem and progenitor cells isolated from PB andexposed to SDF-1a displayed a significantly reduced motilityin a 3-dimensional collagen matrix migration assay uponpretreatment with BTP-2 [66]. In addition, SOCE-dependentCa2þ signaling has been shown to drive the differentiationprocess of a number of both embryonic and adult stem cells[63,67–69]. More specifically, SOCE inhibition preventedH19-7 hippocampal neuronal cells and osteoclast precursorcells from adopting their mature phenotype [68,69]. Indeed,store-dependent Ca2þ inflow may regulate the pattern ofgene expression by selectively engaging a number of Ca2þ-sensitive transcription factors, such as the nuclear factor ofactivated T-cells and nuclear factor-kappa B [16,17]. It will beinteresting to assess whether downregulating SOCE willprevent EPC differentiation into mature ECs.

In conclusion, the present study provides the first evi-dence that both PB-EPCs and UCB-EPCs express store-dependent Ca2þ channels. In addition, PB-EPCs are endowedwith both TRPC1=TRPC4 and STIM1=Orai1, which representthe most plausible molecular candidates to mediate SOCE invascular endothelium. Finally, SOCE inhibition blocks PB-EPC proliferation, a finding that places such a pathway inthe signaling network that controls EPC behavior [70]. Fu-ture work will be devoted to unveil the molecular under-pinnings of SOCE to outline novel targets for EPC-basedtherapy.

FIG. 9. Relative expression of canonical transient receptorpotential (TRPC) channels, Orai, and Stim genes in UCB-EPCs versus PB-EPCs. Fold change values for each genewere calculated versus PB-EPCs, as detailed in the Materialsand Methods section. Statistical analysis was performed onbCt values and the significance of the differences of themeans evaluated with Student’s t-test b AU8.

FIG. 8. The mRNA encoding for canonical transient re-ceptor potential (TRPC) channels, Orai, and Stim is ex-pressed in UCB-EPCs. (A) Gel electrophoresis of the PCRproducts. The PCR products were of the expected size:TRPC1, 307 bp; TRPC3, 336 bp; TRPC4, 300 bp; Orai1, 257 bp;Orai2, 334 bp; Orai3, 159 bp; Stim1, 347 bp; Stim2, 186 bp. Nosignal was observed for TRPC5, TRPC6, and TRPC7. MW:molecular weight marker. Blank: reaction without template.(B) mRNA levels were measured by real-time PCR relative tothe b-actin internal standard (see Materials and Methodssection) and the values obtained were reported as DCt. Barsrepresent the mean� SEM of at least 4 different experimentseach from different RNA extracts. *P< 0.01 versus TRPC1,TRPC3, TRPC4, Orai2, Stim1; **P< 0.05 versus TRPC4 (1-way ANOVA followed by Newman–Keuls’s Q test).



Acknowledgments

The work was supported by FAR (Fondo Ateneo per laRicerca) and CONACYT (Consejo Nacional de Ciencia yTecnolog’ya, Mexico) (grant no. 207991 to Y.S.-H.). Theauthors gratefully acknowledge Dr. Alessandra FiorioPla (University of Turin), Prof. Roberto Berra-Romani(University of Puebla), and Amparo Spezzia-Mazzoccofor critical comments on western blot experiments, andMs. Cinzia Bottino for performing some of the experimentsdescribed in Fig. 10.

Author Disclosure Statement

No competing financial interests exist.

References

1. Pompilio G, MC Capogrossi, M Pesce, F Alamanni, C Di-Campli, F Achilli, A Germani and P Biglioli. (2009). En-dothelial progenitor cells and cardiovascular homeostasis:clinical implications. Int J Cardiol 131:156–167.

2. Ingram DA, LE Mead, H Tanaka, V Meade, A Fenoglio, KMortell, K Pollok, MJ Ferkowicz, D Gilley and MC Yoder.(2004). Identification of a novel hierarchy of endothelialprogenitor cells using human peripheral and umbilical cordblood. Blood 104:2752–2760.

3. Yoder MC, LE Mead, D Prater, TR Krier, KN Mroueh, F Li, RKrasich, CJ Temm, JT Prchal and DA Ingram. (2007). Re-defining endothelial progenitor cells via clonal analysis andhematopoietic stem=progenitor cell principals. Blood 109:1801–1809.

4. Fischer C, M Schneider and P Carmeliet. (2006). Principlesand therapeutic implications of angiogenesis, vasculogenesisand arteriogenesis. Handb Exp Pharmacol 176:157–212.

5. Hristov M and C Weber. (2008). Adult progenitor cells invascular remodeling during atherosclerosis. Biol Chem 389:837–844.

6. Khakoo AY and T Finkel. (2005). Endothelial progenitorcells. Annu Rev Med 56:79–101.

7. Shintani S, T Murohara, H Ikeda, T Ueno, T Honma, AKatoh, K Sasaki, T Shimada, Y Oike and T Imaizumi. (2001).Mobilization of endothelial progenitor cells in patients withacute myocardial infarction. Circulation 103:2776–2779.

FIG. 10. Expression of the proteinsassociated with store-operated Ca2þ

influx in PB- and UCB-EPCs. Blots forOrai1, Stim1, TRPC1, and TRPC4 rep-resentative of 4 separate experimentsare shown. Lanes were loaded with20mg of proteins for Orai1 detectionand 40 mg of proteins for Stim1, TRPC1,and TRPC4 detection. The immuno-blots were probed with affinity-purified antibodies diluted 1:200 andprocessed as described in the Materialsand Methods section. Controls wereperformed as indicated in the Materialsand Methods section. Major bands ofthe expected molecular weights wereobserved in both cell types. Conversely,no bands were observed in controls.

Table 2. Effect of b AU1BTP-2 on Endothelial Colony Forming Cell–Derived Cell Growth In Vitro

Experiment 1(%) Experiment 2(%) Experiment 3(%) Mean (n¼ 3)� SEM (%) Pa

Control (noAU3 c BTP-2) 100 100 100 100BTP-2 2 nM 69 81 77 75.7� 3.5 0.06BTP-2 10 nM 29 48 52 43� 7.1 0.045BTP-2 20 nM 27 31 38 32� 3.2 0.006

Results are expressed as percentage of growth compared with control (given as 100% growth).aCompared with control and after Bonferroni’s correction (t-test for paired samples).



8. Gill M, S Dias, K Hattori, ML Rivera, D Hicklin, L Witte, LGirardi, R Yurt, H Himel and S Rafii. (2001). Vascular traumainduces rapid but transient mobilization of VEGFR2(þ)AC133(þ) endothelial precursor cells. Circ Res 88:167–174.

9. Gao D, DJ Nolan, AS Mellick, K Bambino, K McDonnell, and VMittal. (2008). Endothelial progenitor cells control the angio-genic switch in mouse lung metastasis. Science 319:195–198.

10. Ben-Shoshan J and J George. (2007). Endothelial progenitor cellsas therapeutic vectors in cardiovascular disorders: from exper-imental models to human trials. Pharmacol Ther 115:25–36.

11. Murasawa S and T Asahara. (2005). Endothelial progenitorcells for vasculogenesis. Physiology (Bethesda) 20:36–42.

12. Urbich C and S Dimmeler. (2004). Endothelial progenitorcells functional characterization. Trends Cardiovasc Med 14:318–322.

13. Zhang L, R Yang and ZC Han. (2006). Transplantation ofumbilical cord blood-derived endothelial progenitor cells:a promising method of therapeutic revascularisation. EurJ Haematol 76:1–8.

14. Ahmmed GU and AB Malik. (2005). Functional role of TRPCchannels in the regulation of endothelial permeability.Pflugers Arch 451:131–142.

15. Cioffi DL and Stevens T. (2006). Regulation of endothelialcell barrier function by store-operated calcium entry. Micro-circulation 13:709–723.

16. Munaron L. (2006). Intracellular calcium, endothelial cells andangiogenesis. Recent Pat Anticancer Drug Discov 1:105–119.

17. Munaron L and A Fiorio Pla. (2009). Endothelial calciummachinery and angiogenesis: understanding physiology tointerfere with pathology. Curr Med Chem 16:4691–4703.

18. Tran QK and H Watanabe. (2006). Calcium signalling in theendothelium. Handb Exp Pharmacol 176:145–187.

19. Abdullaev IF, JM Bisaillon, M Potier, JC Gonzalez, RK Mo-tiani and M Trebak. (2008). Stim1 and Orai1 mediate CRACcurrents and store-operated calcium entry important forendothelial cell proliferation. Circ Res 103:1289–1299.

20. Cioffi DL, S Wu and T Stevens (2003). On the endothelialISOC. Cell Calcium 33:323–336.

21. Bird GS, WI DeHaven, JT Smyth and JW Putney Jr. (2008).Methods for studying store-operated calcium entry. Meth-ods 46:204–212.

22. Lewis RS. (2007). The molecular choreography of a store-operated calcium channel. Nature 446:284–287.

23. Parekh AB and JW Putney Jr. (2005). Store-operated calciumchannels. Physiol Rev 85:757–810.

24. Rinne A, K Banach and LA Blatter. (2009). Regulation ofnuclear factor of activated T cells (NFAT) in vascular endo-thelial cells. J Mol Cell Cardiol 47:400–410.

25. Scharbrodt W, Y Abdallah, SA Kasseckert, D Gligorievski,HM Piper, Dk Boker, W Deinsberger and MF Oertel. (2009).Cytosolic Ca2þ oscillations in human cerebrovascular endo-thelial cells after subarachnoid hemorrhage. J Cereb BloodFlow Metab 29:57–65.

26. Beech DJ. (2009). Harmony and discord in endothelial cal-cium entry. Circ Res 104:e22–e23.

27. Paria BC, SM Vogel, GU Ahmmed, S Alamgir, J Shroff, ABMalik and C Tiruppathi. (2004). Tumor necrosis factor-alpha-induced TRPC1 expression amplifies store-operatedCa2þ influx and endothelial permeability. Am J Physiol LungCell Mol Physiol 287:L1303–L1313.

28. Cioffi DL, Wu S, Alexeyev M, Goodman SR, Zhu MX andStevens T. (2005). Activation of the endothelial store-oper-ated ISOC Ca2þ channel requires interaction of protein 4.1with TRPC4. Circ Res 97:1164–1172.

29. Freichel M, SH Suh, A Pfeifer, U Schweig, C Trost, P Weiss-gerber, M Biel, S Philipp, D Freise, G Droogmans, F Hofmann,V Flockerzi and B Nilius. (2001). Lack of an endothelial store-operated Ca2þ current impairs agonist-dependent vasor-elaxation in TRP4-=- mice. Nat Cell Biol 3:121–127.

30. Piaggio G, V Rosti, M Corselli, F Bertolotti, G Bergamaschi, SPozzi, D Imperiale, B Chiavarina, E Bonetti, F Novara, MSessarego, L Villani, A Garuti, M Massa, R Ghio, R Cam-panelli, A Bacigalupo, A Pecci, G Viarengo, O Zuffardi, FFrassoni and G Barosi. (2009). Endothelial colony-formingcells from patients with chronic myeloproliferative disorderslack the disease-specific molecular clonality marker. Blood114:3127–3130.

31. Berra Romani R, A Raqeeb, U Laforenza, MF Scaffino, FMoccia, JE Avelino-Cruz, A Oldani, D Coltrini, V Milesi, VTaglietti and F Tanzi. (2009). Cardiac microvascular endo-thelial cells express a functional Ca2þ-sensing receptor.J Vasc Res 46:73–82.

32. Laforenza U, G Gastaldi, M Polimeni, S Tritto, M Tosco,U Ventura, MF Scaffino and M Yasui. (2009). Aquaporin-6is expressed along the rat gastrointestinal tract and upre-gulated by feeding in the small intestine. BMC Physiol 7:9–18.

33. Livak KJ and TD Schmittgen. (2001). Analysis of relativegene expression data using real-time quantitative PCR andthe 2(-Delta Delta C(T)) Method. Methods 25:402–408.

34. Laemmli UK. (1970). Cleavage of structural proteins duringthe assembly of the head of bacteriophage T4. Nature 227:680–685.

35. Bradford MM. (1976). A rapid and sensitive method forthe quantitation of microgram quantities of protein utiliz-ing the principle of protein-dye binding. Anal Biochem72:248–254.

36. Varnai P, L Hunyady and T Balla. (2009). STIM and Orai: thelong-awaited constituents of store-operated calcium entry.Trends Pharmacol Sci 30:118–128.

37. Bossu JL, A Feltz, JL Rodeau and F Tanzi. (1989). Voltage-dependent transient calcium currents in freshly dissociatedcapillary endothelial cells. FEBS Lett 255:377–380.

38. Moccia F, S Baruffi, S Spaggiari, D Coltrini, R Berra-Romani,S Signorelli, L Castelli, V Taglietti and F Tanzi. (2001). P2y1and P2y2 receptor-operated Ca2þ signals in primary culturesof cardiac microvascular endothelial cells. Microvasc Res61:240–252.

39. Moccia F, C Frost, R Berra-Romani, F Tanzi and DJ Adams.(2004). Expression and function of neuronal nicotinic AChreceptors in rat microvascular endothelial cells. Am J PhysiolHeart Circ Physiol 286:H486–H491.

40. Vinet R, F Rojas, M Luxoro, F Vargas, M Cortes. (2000).Catecholamines-evoked cytosolic Ca2þ rise in endothelialcells from bovine adrenal medulla. Mol Cell Biochem 203:53–58.

41. Murata T, MI Lin, RV Stan, PM Bauer, J Yu and WC Sessa.(2007). Genetic evidence supporting caveolae microdomainregulation of calcium entry in endothelial cells. J Biol Chem282:16631–16643.

42. Yamamoto K, R Korenaga, A Kamiya, Z Qi, M Sokabe and JAndo. (2000). P2X(4) receptors mediate ATP-induced cal-cium influx in human vascular endothelial cells. Am J Phy-siol Heart Circ Physiol 279:H285–H292.

43. Yamazaki D, S Ohya, K Asai and Y Imaizumi. (2007).Characteristics of the ATP-induced Ca2þ-entry pathway inthe t-BBEC 117 cell line derived from bovine brain endo-thelial cells. J Pharmacol Sci 104:103–107.



44. Foskett JK, C White, KH Cheung and DO Mak (2007). In-ositol trisphosphate receptor Ca2þ release channels. PhysiolRev 87:593–658.

45. Moccia F, R Berra-Romani, S Tritto, S Signorelli, V Tagliettiand F Tanzi. (2003). Epidermal growth factor induces in-tracellular Ca2þ oscillations in microvascular endothelialcells. J Cell Physiol 194:139–150.

46. Rusko J, X Wang and C van Breemen. (1995). Regenerativecaffeine-induced responses in native rabbit aortic endothelialcells. Br J Pharmacol 115:811–821.

47. Abbracchio MP, G Burnstock, JM Boeynaems, EA Barnard,JL Boyer, C Kennedy, GE Knight, M Fumagalli, C Gachet,KA Jacobson and GA Weisman. (2006). International Unionof Pharmacology LVIII: update on the P2Y G protein-cou-pled nucleotide receptors: from molecular mechanisms andpathophysiology to therapy. Pharmacol Rev 58:281–341.

48. Yao X and CJ Garland. (2005). Recent developments invascular endothelial cell transient receptor potential chan-nels. Circ Res 97:853–863.

49. Zhou Y, P Meraner, HT Kwon, D Machnes, M Oh-hora, JZimmer, Y Huang, A Stura, A Rao and PG Hogan (2010).STIM1 gates the store-operated calcium channel ORA1in vitro. Nat Struct Mol Biol 17:112–116.

50. Madge L, IC Marshall and CW Taylor. (1997). Delayed au-toregulation of the Ca2þ signals resulting from capacitativeCa2þ entry in bovine pulmonary artery endothelial cells. JPhysiol 498:351–369.

51. Sedova M, A Klishin, J Huser and LA Blatter. (2000). Ca-pacitative Ca2þ entry is graded with degree of intracellularCa2þ store depletion in bovine vascular endothelial cells. JPhysiol 523:549–559.

52. Maeng YS, HJ Choi, JK Kwon, YW Park, KS Choi, JK Min, YHKim, PG Suh, KS Kang, MH Won, YM Kim and YG Kwon.(2009). Endothelial progenitor cell homing: prominent role ofthe IGF2-IGF2R-PLCbeta2 axis. Blood 113:233–243.

53. Mozhayeva MG and GN Mozhayeva. (1996). Evidence forthe existence of inositol (1,4,5)-trisphosphate- and ryano-dine-sensitive pools in bovine endothelial cells. Ca2þ releasesin cells with different basal level of intracellular Ca2þ. Pflu-gers Arch 432:614–622.

54. Leung PC, KT Cheng, C Liu, WT Cheung, HY Kwan, KLLau, Y Huang and Yao X. (2006). Mechanism of non-capacitative Ca2þ influx in response to bradykinin in vas-cular endothelial cells. J Vasc Res 43:367–376.

55. Parekh AB, A Fleig and R Penner. (1997). The store-operatedcalcium current I(CRAC): nonlinear activation by InsP3 anddissociation from calcium release. Cell 89:973–980.

56. Parekh AB and R Penner. (1995). Activation of store-oper-ated calcium influx at resting InsP3 levels by sensitization ofthe InsP3 receptor in rat basophilic leukaemia cells. J Physiol489:377–382.

57. Aird WC. (2007). Phenotypic heterogeneity of the endothelium:I. Structure, function, and mechanisms. Circ Res 100:158–173.

58. Droogmans G and B Nilius. (2001). Ion channels and theirfunctional role in vascular endothelium. Physiol Rev 81:1415–1459.

59. Brandman O, J Liou, WS Park and T Meyer. (2007). STIM2 isa feedback regulator that stabilizes basal cytosolic and endo-plasmic reticulum Ca2þ levels. Cell 131:1327–1339.

60. Yip H, WY Chan, PC Leung, HY Kwan, C Liu, Y Huang, VMichel, DT Yew and X Yao. (2004). Expression of TRPC

homologs in endothelial cells and smooth muscle layers ofhuman arteries. Histochem Cell Biol 122:553–561.

61. Cheng HW, AF James, RR Foster, JC Hancox and DO Bates.(2006). VEGF activates receptor-operated cation channels inhuman microvascular endothelial cells. Arterioscler ThrombVasc Biol 26:1768–1776.

62. Hamdollah Zadeh MA, CA Glass, A Magnussen, JC Hancoxand DO Bates. (2008). VEGF-mediated elevated intracellularcalcium and angiogenesis in human microvascular endo-thelial cells in vitro are inhibited by dominant negativeTRPC6. Microcirculation 15:605–614.

63. Ferreira-Martins J, C Rondon-Clavo, D Tugal, JA Korn, RRizzi, ME Padin-Iruegas, S Ottolenghi, A De Angelis, KUrbanek, N Ide-Iwata, D D’Amario, T Hosoda, A Leri, JKajstura, P Anversa and M Rota. (2009). Spontaneous cal-cium oscillations regulate human cardiac progenitor cellgrowth. Circ Res 105:764–774.

64. Ando H, K Nakanishi, M Shibata, K Hasegawa, K Yao and HMiyaji. (2006). Benidipine, a dihydropyridine-Ca2þ channelblocker, increases the endothelial differentiation of endothe-lial progenitor cells in vitro. Hypertens Res 29:1047–1054.

65. Kahl CR and AR Means. (2003). Regulation of cell cycleprogression by calcium=calmodulin-dependent pathways.Endocr Rev 24:719–736.

66. Seidel J, B Niggemann, M Punzel, J Fischer, KS Zanker and TDittmar. (2007). The neurotransmitter GABA is a potent in-hibitor of the stromal cell-derived factor-1alpha inducedmigration of adult CD133þ hematopoietic stem and pro-genitor cells. Stem Cells Dev 16:827–836.

67. Kawano S, K Otsu, A Kuruma, S Shoji, E Yanagida, Y Muto,F Yoshikawa, Y Hirayama, K Mikoshiba and T Furuichi.(2006). ATP autocrine=paracrine signaling induces calciumoscillations and NFAT activation in human mesenchymalstem cells. Cell Calcium 39:313–324.

68. Kim MS, YM Yang, A Son, YS Tian, SI Lee, SW Kang, SMuallem and DM Shin. (2010). RANKL-mediated reactiveoxygen species pathway that induces long lasting Ca2þ os-cillations essential for osteoclastogenesis. J Biol Chem 285:6913–6921.

69. Wu X, TK Zagranichnaya, GT Gurda, EM Eves and MLVillereal. (2004). A TRPC1=TRPC3-mediated increase instore-operated calcium entry is required for differentiationof H19-7 hippocampal neuronal cells. J Biol Chem 279:43392–43402.

70. Weidt C, B Niggemann, B Kasenda, TL Drell, KS Zanker andT Dittmar. (2007). Stem cell migration: a quintessential step-ping stone to successful therapy. Curr Stem Cell Res Ther2:89–103.

Address correspondence to:Dr. Francesco Moccia

Department of PhysiologyUniversity of Pavia

Via Forlanini 6, Pavia 27100Italy

E-mail: [email protected]

Received for publication January 26, 2010Accepted after revision August 2, 2010

Prepublished on Liebert Instant Online Month 00, 0000



SUPPLEMENTAL INFORMATION

Materials and Methods Supplement

Cell culture

Endothelial cell monolayers were obtained from first-passage–expanded endothelial colony-forming cells, fromeither umbilical cord blood or peripheral blood. Cells wererecovered by trypsinization and incubated with the primaryor isotypic control antibodies for 20 min at 48C. After

washing, cells were analyzed by fluorescence activated cellsorting (FACS; Becton Dickinson). The following antibodieswere used (all from BD Biosciences, unless differently spec-ified): FITC-conjugated anti-CD31; PE-conjugated anti-CD45;FITC-conjugated anti-CD105; FITC-conjugated anti-CD144;PE-conjugated anti-CD146; PerCP-conjugated anti-VEGFR-2;FITC-conjugated anti-CD31.

SUPPLEMENTAL FIG. S1. Representative photomicrograph of (a) UCB- and (b) PB-derived endothelial colony-formingcells observed in vitro after 11 and 15 days, respectively. Panels c (CB) and d (PB) show representative examples of in vitroformation of capillary-like structures after plating endothelial colony-forming cell–derived cells in Matrigel. No differenceswere observed between experiments performed with UCB- or PB-derived cells, with the exception of the time of appearanceof colonies, which was earlier in cultures performed with UCB-derived cells than in those performed with PB-derived cells(9 vs. 11 days, respectively). UCB, umbilical cord blood; PB, peripheral blood.

SCD-2010-0047-ver9-Sanchez-Suppl_1P.3D 08/23/10 2:11pm Page 1

Supplemental Table S1. Phenotypic Profile

ofAU1 c CB- and Peripheral Blood–Derived Endothelial

Colony-Forming Cells

Cord blood Peripheral blood(n¼ 4) (n¼ 6)

CD31 (PECAM) 98.1%� 1.8% 98.5%� 1.1%CD45 0.7%� 0.5% 0.9%� 0.8%CD14 0.5%� 0.2% 0.4%� 0.3%CD105 99.8%� 0.2% 97.7%� 1.2%CD144 (VE-cadherin) 96.9%� 2.3% 98.9%� 0.7%CD146 99.8%� 0.1% 97.5%� 2.2%VEGFR-2 62.1%� 11.9% 58.6%� 18.3%von Willebrand factor ND 99.0%� 0.9%

Data are expressed as means� SD.

SCD-2010-0047-ver9-Sanchez-Suppl_1P.3D 08/23/10 2:12pm Page 2

AUTHOR QUERY FOR SCD-2010-0047-VER9-SANCHEZ_1P

AU1: Please expand BTP, PMSF, and CB.AU2: Please let us know whether you want Supplemental Fig. S1 and Supplemental Table S1 in the online

version only.AU3: Please define PSS, BM, and BTP.AU4: The text ‘‘1 h min’’ is not clear. Please correct.AU5: Please define ‘‘q’’ in ‘‘q-polymerase chain reaction.’’AU6: Please check whether ‘‘(see also 31)’’ is a ref. citation. If so, enclose ‘‘31’’ in square brackets.AU7: Asterisk is not found in Fig. 1B, but its description is given in the legend. Please check.AU8: ‘‘*’’ is found in Fig. 9. Please describe it in the legend.

SCD-2010-0047-ver9-Sanchez_1P.3D 08/23/10 3:07pm Page 17

store-operated ca entry is expressed in human endothelial ...royanaward.com › files12 ›...

Documents