Canonical Transient Receptor Potential 3 Channel Triggers Vascular Endothelial Growth Factor-Induced Intracellular Ca 2+ Oscillations in Endothelial Progenitor Cells Isolated from Umbilical Cord Blood

ORIGINAL RESEARCH REPORTS

STEM CELLS AND DEVELOPMENT Volume 22, Number 19, 2013  Mary Ann Liebert, Inc. DOI: 10.1089/scd.2013.0032

Canonical Transient Receptor Potential 3 Channel Triggers Vascular Endothelial Growth Factor-Induced Intracellular Ca2 + Oscillations in Endothelial Progenitor Cells Isolated from Umbilical Cord Blood Silvia Dragoni,1 Umberto Laforenza,2 Elisa Bonetti,3 Francesco Lodola,1 Cinzia Bottino,2 Germano Guerra,4 Alessandro Borghesi,5 Mauro Stronati,6 Vittorio Rosti,3 Franco Tanzi,1 and Francesco Moccia1

Endothelial colony-forming cells (ECFCs) are the only endothelial progenitor cells (EPCs) that are capable of acquiring a mature endothelial phenotype. ECFCs are mainly mobilized from bone marrow to promote vascularization and represent a promising tool for cell-based therapy of severe ischemic diseases. Vascular endothelial growth factor (VEGF) stimulates the proliferation of peripheral blood-derived ECFCs (PB-ECFCs) through oscillations in intracellular Ca2 + concentration ([Ca2 + ]i). VEGF-induced Ca2 + spikes are driven by the interplay between inositol-1,4,5-trisphosphate (InsP3)-dependent Ca2 + release and store-operated Ca2 + entry (SOCE). The therapeutic potential of umbilical cord blood-derived ECFCs (UCB-ECFCs) has also been shown in recent studies. However, VEGF-induced proliferation of UCB-ECFCs is faster compared with their peripheral counterpart. Unlike PB-ECFCs, UCB-ECFCs express canonical transient receptor potential channel 3 (TRPC3) that mediates diacylglycerol-dependent Ca2 + entry. The present study aimed at investigating whether the higher proliferative potential of UCB-ECFCs was associated to any difference in the molecular underpinnings of their Ca2 + response to VEGF. We found that VEGF induces oscillations in [Ca2 + ]i that are patterned by the interaction between InsP3-dependent Ca2 + release and SOCE. Unlike PB-ECFCs, VEGF-evoked Ca2 + oscillations do not arise in the absence of extracellular Ca2 + entry and after pharmacological (with Pyr3 and flufenamic acid) and genetic (by employing selective small interference RNA) suppression of TRPC3. VEGF-induced UCB-ECFC proliferation is abrogated on inhibition of the intracellular Ca2 + spikes. Therefore, the Ca2 + response to VEGF in UCB-ECFCs is shaped by a different Ca2 + machinery as compared with PB-ECFCs, and TRPC3 stands out as a promising target in EPC-based treatment of ischemic pathologies.

Introduction

E

ndothelial progenitor cells (EPCs) are mobilized from bone marrow (BM) in response to an ischemic insult to recapitulate the injured vascular network and restore blood perfusion [1–6]. Moreover, circulating EPCs replace either injured or senescent endothelial cells that are detached from the vascular wall into peripheral blood (PB), thereby preventing the establishment of conditions which are prone to atherosclerosis and other vascular pathologies [6–8]. As a consequence, the transplantation of autologous EPCs holds major promise as an alternative tool in the stem cell therapy of peripheral artery disease (PAD) and myocardial infarction

(MI) [1–3,9–12]. The term EPC fails to refer to a unique population of BM-derived mononuclear cells (MNCs) due to the absence of a cell-specific panel of surface antigens or gene expression profiles [3,6,13]. The so-called endothelial colonyforming cells (ECFCs), which may be mobilized from both BM and arterial vessel walls, stand amid the different subsets of EPCs described in the literature, for the following reasons: (1) They truly belong to an endothelial lineage and display a clonal proliferative potential (ie, may undergo approximately 30 population doublings without senescence and replate into secondary and tertiary colonies); (2) they possess the rare ability to form bidimensional tubular networks in vitro; and (3) they may form patent vessels that anastomize with the

1

Department of Biology and Biotechnology ‘‘Lazzaro Spallanzani,’’ 2Department of Molecular Medicine, University of Pavia, Pavia, Italy. Laboratory of Biotechnology, Center for the Study of Myelofibrosis, Foundation IRCCS Policlinico San Matteo, Pavia, Italy. 4 Department of Health Sciences, University of Molise, Campobasso, Italy. 5 Neonatal Intensive Care, Foundation IRCCS Policlinico San Matteo, Pavia, Italy. 6 Laboratory of Neonatal Immunology, Foundation IRCCS Policlinico San Matteo, Pavia, Italy. 3

2561

2562 systemic vasculature of the host when implanted into immunodeficient mice [1–3,14–17]. Unfortunately, the frequency of circulating ECFCs in PB is scarce, that is, about 0.05–0.2 cells/mL [14], a feature that does not render them amenable for cell therapy purposes. Indeed, the estimated volume of blood required to achieve therapeutically relevant numbers of cells is equal to *12 L [9]. An additional disadvantage is the decreased number and impaired functional capability of adult stem cells in a patient population of advanced age burdened with significant comorbidity [3,9,18]. It has, however, been demonstrated that ECFCs are far more abundant in human umbilical cord blood (UCB), where they attain a concentration of about 2–5 cells/mL [14]. In addition, UCB-derived ECFCs (UCB-ECFCs) exhibit a much higher proliferation potential (approximately 100 population doublings) and a greater telomerase activity than their peripheral counterparts [14]. When considering that UCB-ECFCs give rise to functional capillary networks in vivo [16,17], they represent a more attractive solution for regenerative medicine. The mechanistic differences between UCB- and PB-derived ECFCs have never been elucidated. Vascular endothelial growth factor (VEGF) is the most potent pro-angiogenic cytokine driving ECFC proliferation, recruitment to target organs, and physical assembly into perfused vascular networks. Assessing whether VEGF stimulation activates distinct signal transduction pathways in the two cell types might outline novel molecular targets to develop more efficacious EPCs-based strategies. An increase in intracellular Ca2 + concentration ([Ca2 + ]i) represents a key signal in the activation of mature endothelium [19–21], as well as of human ECFCs [1–3]. A growing body of evidence demonstrated that store-operated Ca2 + entry (SOCE) controls ECFC expansion in vitro [22–24]. The physiological stimulus for the induction of SOCE in PBECFCs is provided by VEGF [24,25], which binds to its cognate tyrosine kinase receptor, VEGFR-2 (KDR/Flk1), to phosphorylate and engage phospholipase C-g (PLCg) [3,26]. PLCg, in turn, cleaves the membrane-bound phosphatidylinositol-4,5-bisphosphate (PIP2) to generate the intracellular second messengers, inositol-1,4,5-trisphosphate (InsP3) and diacylglycerol (DAG). Then, InsP3 diffuses throughout the cytoplasm, where it acts on specific InsP3 receptors (InsP3R) to release Ca2 + from the endoplasmic reticulum (ER) reservoir [3,26]. The following drop in ER Ca2 + content is detected by an intralumenal Ca2 + sensor, termed Stromal Interaction Molecule-1 (Stim1), which communicates the amount of stored Ca2 + to store-operated Ca2 + channels on the plasma membrane (PM) [3,21,27,28]. Stim1 is stimulated by the InsP3-dependent Ca2 + discharge to associate into multimers that rapidly redistribute to peripheral ER sites in close proximity to PM. Here, Stim1 clusters aggregate into multiple puncta, interact with and gate Orai1 and canonical transient receptor potential 1 (TRPC1) channel, the Ca2 + permeable routes that conduct Ca2 + into human ECFCs [3,22,24,25,27–29]. The interplay between InsP3-induced Ca2 + release and SOCE leads to asynchronous oscillations in [Ca2 + ]i, which encode the information necessary to trigger the program of gene expression driving PB-ECFC proliferation and tubulogenesis [2,3,24,25,30,31]. Therefore, Ca2 + entry is not required to initiate the spiking response to VEGF in PB-ECFCs, but is essential to maintain Ca2 + oscillations over time by replenishing the InsP3-sensitive Ca2 + pool [25]. In addition to SOCE, mature endothelial cells are endowed

DRAGONI ET AL. with alternative pathways for Ca2 + influx that are activated by second messengers, such as DAG [20,21]. For instance, TRPC3 and TRPC6 support the pro-angiogenic Ca2 + signal triggered by VEGF in human microvascular endothelial cells [32,33], as well as TRPC6 drives VEGF-induced proliferation and tube formation in human umbilical vein endothelial cells (HUVECs) [34]. Surprisingly, PB-ECFCs lack DAG-operated Ca2 + channels, whereas UCB-express TRPC3 [23]. TRPC3mediated Ca2 + entry has recently been associated to the selective enlisting of a number of Ca2 + -sensitive transcription factors, including nuclear factor of activated T-cells (NFAT) [35,36] and nuclear factor kappaB (NF-kB) [37], as well as to the sustained activation of extracellular signal-regulated kinase (ERK), which may also impact gene expression [38]. The role of TRPC3 in the pro-angiogenic action of VEGF on UCBECFCs is, however, yet to be elucidated. The present investigation aimed at elucidating TRPC3 involvement in the mechanistic differences of VEGF signaling in human PB- and UCB-ECFC by combining Ca2 + imaging measurements with a gene silencing approach. We found that, similar to PB-ECFCs, VEGF triggers repeated [Ca2 + ]i transients that were not synchronous between adjacent cells. However, the oscillatory mechanism was strikingly different as, in UCB-ECFCs, the interplay between InsP3-dependent Ca2 + release and SOCE was triggered by TRPC3-mediated Ca2 + inflow. In addition, TRPC3 opening was necessary throughout the duration of the stimulation to maintain the spiking signal. Consistent with these data, the genetic suppression of TRPC3 prevented VEGF-induced Ca2 + oscillations, as well as the entry of Ca2 + induced by 1oleoyl-2-acetyl-sn-glycerol (OAG), a membrane-permeant analog of DAG. Therefore, the selective recruitment of TRPC3 by VEGF in UCB-ECFCs, but not in PB-ECFCs, represents the first molecular difference ever described between the two cell types and might open a new avenue in the design of alternative strategies for EPCs-based therapies.

Materials and Methods Isolation and cultivation of ECFCs Blood samples (40 mL) were obtained from healthy human volunteers aged from 20 to 48 years old (n = 12). The Institutional Review Board at ‘‘Istituto di Ricovero e Cura a Carattere Scientifico Policlinico San Matteo Foundation’’ in Pavia approved all protocols and specifically approved this study. Informed written consent was obtained according to the Declaration of Helsinki of 1975 as revised in 2008. We focussed on the so-called ECFCs [22,23,25], a subgroup of EPCs that are found in the CD34 + CD45 - fraction of circulating MNCs, exhibit robust proliferative potential, and form capillary-like structures in vitro [13–15]. To isolate ECFCs, MNCs were separated from either PB or UCB by density gradient centrifugation on lymphocyte separation medium for 30 min at 400 g and washed twice in endothelial basal medium (EBM)-2 with 2% fetal calf serum. A median of 36 · 106 MNCs (range 18–66) was plated on collagen-coated culture dishes (BD Biosciences) in the presence of the endothelial cell growth medium EGM-2 MV Bullet Kit (Lonza) containing EBM-2, 5% fetal bovine serum, recombinant human (rh) EGF, rhVEGF, rhFGF-B, rhIGF-1, ascorbic acid, and heparin, and maintained at 37C in 5% CO2 and a humidified atmosphere. Discarding of non-adherent cells was performed

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS after 2 days; thereafter, medium was changed thrice a week. The outgrowth of ECs from adherent MNCs was characterized by the formation of a cluster of cobblestone-appearing cells [13–15,22,23,25]. That ECFC-derived colonies belonged to endothelial lineage was confirmed as described in [23] and [22]. In greater detail, EPC-derived colonies were stained with anti-CD31, anti-CD105, anti-CD144, anti-CD146, antivWf, anti-CD45, and anti-CD14 monoclonal antibodies and by assessment of capillary-like network formation in an in vitro Matrigel assay. For our experiments, we have mainly used endothelial cells obtained from early passage ECFC (P1-3, which roughly encompasses a 15–18 day period) with the purpose of avoiding (or maximally reducing) any potential bias due to cell differentiation. However, in order to make sure that the phenotype of the cells did not change throughout the experiments, in preliminary experiments, we tested the immunophenotype of ECFCs at different passages and we found no differences, as shown in [22]. We also tested whether functional differences occurred when early- (P2) and late- (P6) passage ECFCs were used by testing the in vitro capacity of capillary network formation in a Matrigel assay and found no differences between early- and late-passage ECFC-derived cells.

Solutions Physiological salt solution (PSS) had the following composition (in mM): 150 NaCl, 6 KCl, 1.5 CaCl2, 1 MgCl2, 10 Glucose, and 10 Hepes. In Ca2 + -free solution (0Ca2 + ), Ca2 + was substituted with 2 mM NaCl, and 0.5 mM EGTA was added. Solutions were titrated to pH 7.4 with NaOH. The high-K + extracellular solution was prepared by replacing 100 mM NaCl with an equimolar amount of KCl. The solution was then titrated to pH 7.4 with KOH. The osmolality of the extracellular solution, as measured with an osmometer (Wescor 5500), was 338 mmol/kg.

[Ca2 + ]i measurements ECFCs were loaded with 4 mM fura-2 acetoxymethyl ester (fura-2/AM; 1 mM stock in dimethyl sulfoxide) in PSS for 1 h at room temperature. After washing in PSS, the coverslip was fixed to the bottom of a Petri dish and the cells were observed by an upright epifluorescence Axiolab microscope (Carl Zeiss), which was usually equipped with a Zeiss · 40 Achroplan objective (water immersion, 2.0 mm working distance, and 0.9 numerical aperture). ECFCs were excited

2563

alternately at 340 and 380 nm, and the emitted light was detected at 510 nm. A first neutral density filter (1 or 0.3 optical density) reduced the overall intensity of the excitation light, and a second neutral density filter (optical density = 0.3) was coupled to the 380 nm filter to approach the intensity of the 340 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, working in the LINUX environment, was used to drive the camera (Extended-ISIS Camera, Photonic Science) and the filter wheel, and to measure and plot online the fluorescence from 10 to100 rectangular ‘‘regions of interest’’ (ROI). Each ROI was identified by a number. Since cell borders were not clearly identifiable, an ROI may not include the whole cell or may include a part of an adjacent cell. Adjacent ROIs never superimposed. [Ca2 + ]i was monitored by measuring, for each ROI, the ratio of the mean fluorescence emitted at 510 nm when exciting alternatively at 340 and 380 nm (shortly termed ‘‘ratio’’). An increase in [Ca2 + ]i causes an increase in the ratio [22,23,25]. Ratio measurements were performed and plotted online every 3 s. The experiments were performed at room temperature (22C).

RNA isolation and real-time RT-PCR (qRT-PCR) Total RNA was extracted from the EPCs using the QIAzol Lysis Reagent (QIAGEN). Single cDNA was synthesized from RNA (1 mg) using random hexamers and M-MLV Reverse Transcriptase (Invitrogen S.R.L.). Reverse transcription was always performed in the presence or absence (negative control) of the reverse transcriptase enzyme. qRT-PCR was performed in triplicate using 1 mg cDNA and specific primers (intron-spanning primers) for InsP3R1, InsP3R2, and InsP3R3 and for TRPC3 (Table 1), as previously described elsewhere [22,23,25]. GoTaq qPCR Mastermix (Promega) was used according to the manufacturer’s instructions, and qRT-PCR was performed using Rotor Gene 6000 (Corbett, Concorde). The conditions were as follows: initial denaturation at 95C for 5 min; 40 cycles of denaturation at 95C for 30 s; annealing at 58C for 30 s; and elongation at 72C for 40 s. The qRT-PCR reactions were normalized using b-actin as a housekeeping gene. Melting curves were generated to detect the melting temperatures of specific products immediately after the PCR run. The triplicate threshold cycles (Ct) values for each sample were averaged, resulting in mean Ct values for both the gene of interest and the housekeeping gene bactin. Relative mRNA levels were determined by comparative quantitation (Corbett), and the results were expressed as

Table 1. Primer Sequences Used for Real-Time Reverse Transcription/Polymerase Chain Reaction Gene InsP3R1 InsP3R2 InsP3R3 TRPC3 b-actin

Primer sequences Forward 5¢-TCAACAAACTGCACCACGCT-3¢ Reverse 5¢-CTCTCATGGCATTCTTCTCC-3¢ Forward 5¢-ACCTTGGG GTTAGTGGATGA-3¢ Reverse 5¢-CCTTGTTTGGCTTGCTTTGC-3¢ Forward 5¢-TGGCTTCATCAGCACTTTGG-3¢ Reverse 5¢-TGTCCTGCTTAGTCTGCTTG-3¢ Forward 5¢-GGAGATCTGGAATCAGCAGA-3¢ Reverse 5¢-AAGCAGACCCAGGAAGATGA-3¢ Hs_ACTB_1_SG, QuantiTect Primer Assay QT00095431, Qiagen

Size (bp)

Accession number

180

ENSG00000150995

158

ENSG00000123104

173

ENSG00000096433

336

NM_001130698.1 variant 1 NM_003305.2 variant 2 NM_001101

146

2564 fold change. PCR products were also separated with agarose gel electrophoresis and stained with ethidium bromide.

Sample preparation and immunoblotting UCB-ECFCs were homogenized by using a Dounce homogenizer in a solution containing 250 mM Sucrose, 1 mM EDTA, 10 mM Tris-HCl, pH 7.6, 0.1 mg/mL PMSF, 100 mM b-mercaptoethanol, and Protease Inhibitor Cocktail (P8340; Sigma). The homogenates were solubilized in Laemmli buffer [22,23], and 20 mg proteins were separated on 10% sodium dodecyl sulfate-polyacrilamide gel electrophoresis and transferred to the Hybond-P polyvinyl difluoride Membrane (GE Healthcare) by electroelution. After 1 h blocking with Tris buffered saline (TBS) containing 3% bovine serum albumin (BSA) and 0.1% Tween (blocking solution), the membranes were incubated for 3 h at room temperature with the TRPC3/6/7 (N-18) (Santa Cruz Biotechnology, Inc.; sc15056) goat polyclonal antibody or TRPC3 rabbit polyclonal antibody (Acris Antibodies GmbH) diluted 1:200 in the TBS and 0.1% Tween. The membranes were washed and incubated for 1 h with peroxidase-conjugated goat anti-rabbit IgG (Chemicon, AP132P) or mouse anti-goat IgG (Santa Cruz Biotechnology, Inc.; sc-2354) diluted 1:120,000 in blocking solution. The bands were detected with ECL Advance western blotting detection system (GE Healthcare Europe GmbH). Control experiments were performed as described in [23] and [22]. Prestained molecular-weight markers (SDS7B2, Sigma) were used to estimate the molecular weight of the bands. Blots were stripped with the method described in [23] and re-probed with anti b-actin rabbit antibody as loading control (Rockland Immunochemicals for Research, U.S.A.; code, 600-401-886). The antibody was diluted 1:2,000 in the TBS and 0.1% Tween. Blots were acquired with the Image Master VDS (Amersham Biosciences Europe). Densitometric analysis of the bands was performed by the Total Lab V 1.11 computer program (Amersham), and the results were expressed as a percentage of the gene/b-actin densitometric ratio.

Protein content Protein contents of all the samples were determined by the Bradford’s method using BSA as standard [22,23].

Gene silencing siRNA targeting TRPC3 was purchased by Thermo scientific-Dharmacon (ON-TARGET plus SMART pool Human TRPC3). Scrambled siRNA were used as a negative control. As recently depicted elsewhere [22], once the monolayer cells had reached 50% confluency, the medium was removed and the cells were added with Opti-MEM I reduced serum medium without antibiotics (Life Technologies). siRNAs (120 nM final concentration) were diluted in Opti-MEM I reduced serum medium and mixed with Lipofectamine RNAiMAX transfection reagent (Invitrogen) pre-diluted in Opti-MEM), according to the manufacturer’s instructions. After 20 min incubation at room temperature, the mixes were added to the cells and incubated at 37C for 5 h. Transfection mixes were then completely removed, and fresh culture media were added. The effectiveness in silencing was determined by immunoblotting, and the KO-cells were used 72 h after transfection.

DRAGONI ET AL.

Proliferation assays As described elsewhere [25], a total of 1 · 105 UCB-ECFCderived cells (first passage) were plated in 30-mm collagentreated dishes in EBM medium (Lonza) in the presence of 10 ng/mL VEGF (PeproTech) and with or without 30 mM 1,2-bis(o-aminophenoxy)ethane-N,N,N¢,N¢-tetracetic acid (BAPTA), 10 mM ethyl-1-(4-(2,3,3-trichloroacrylamide)phenyl)-5-(trifluoromethyl)-1H-pyrazole-4-carboxylate (Pyr3), and 20 mM of N-(4-[3,5-bis(trifluoromethyl)-1H-pyrazol-1-yl] phenyl)-4-methyl-1,2,3-thiadiazole-5-carboxamide (BTP-2). Cultures were incubated at 37C, 5% CO2 and cell growth was assessed every day until confluence was reached in the control dishes, usually after 3–4 days. Cells were then recovered by trypsinization, and their number was assessed by counting in a hemocytometer. The percentage of growth stimulation (by VEGF) or inhibition [by Pyr3, flufenamic acid (FFA), BAPTA, and BTP-2] was calculated by dividing the total number of cells obtained in the presence of the aforementioned compounds by the number of cells in control experiments and by multiplying the ratio by 100.

Statistics All the Ca2 + data have been collected from PB-ECFCs isolated from PB of at least three healthy volunteers (for OAGinduced Ca2 + signals) or from UCB-ECFCs harvested from at least three different cord blood samples. Pooled data are given as mean – SE, and statistical significance (P < 0.05) was evaluated by the Student’s t-test for unpaired observations. As described elsewhere [25,39], the interspike interval (ISI) analyzed in Fig. 2E was defined as the peak-to-peak interval between successive Ca2 + transients. Only cells displaying at least three Ca2 + oscillations were included in the analysis depicted in Fig. 2B–F. The statistical analysis was performed on cells challenged with VEGF for 1 h, which was chosen as an arbitrary time interval to compare the Ca2 + activity of different ECFCs both from the same coverslip and from different batches. As for mRNA analysis, all data are expressed as mean – SE. The significance of the differences of the means was evaluated with Student’s t test. In the proliferation assays, results are expressed as percentage ( – SD) of growth compared with controls (given as 100% growth), obtained from three different sets of experiments, each performed in duplicate. Differences were assessed by the Student’s t-test for unpaired values. All statistical tests were carried out with GraphPad Prism 4.

Chemicals EBM and EGM-2 were purchased from Clonetics (Cell System). Fura-2/AM was obtained from Molecular Probes (Molecular Probes Europe BV). BTP-2 was purchased from Calbiochem. VEGF was provided by PeproTech. All other chemicals were obtained from Sigma Chemical Co.

Results VEGF induces asynchronous Ca2 + oscillations in UCB-ECFCs We have recently shown that 10 ng/mL is the most suitable dose for VEGF to induce oscillations in [Ca2 + ]i in human

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS ECFCs [25]. Therefore, this concentration was utilized to analyze VEGF-elicited Ca2 + signals in ECFCs isolated from human UCB. In the absence of extracellular stimulation, the cells did not exhibit any spontaneous increase in [Ca2 + ]i, albeit the resting Fura2 fluorescence among cells both from the same field of view and from different coverslips is extremely scattered (compare, for instance, the baseline fluorescence of the tracings depicted in Fig. 3A and in Fig. 3B). Ca2 + imaging recordings revealed that 10 ng/mL VEGF— homogeneously perfused over the Petri dish—triggered repetitive increases in [Ca2 + ]i in 458 out of 677 cells (67.7%). The remaining 219 cells (32.3%) displayed only a single Ca2 + transient. These percentages are similar to those reported in PB-ECFCs [25]. Moreover, VEGF-induced Ca2 + oscillations were not synchronized among adjacent/neighboring cells from the same microscopic field, as highlighted in Fig. 1, Supplementary Fig. S1; Supplementary Data are available online at www.liebertpub.com/scd and Supplementary Movie S1. More specifically, Supplementary Fig. S1A depicts a cobblestone monolayer of UCB-ECFCs loaded with Fura-2 and showing their typical spindle-shaped morphology [14]. The fluorescent images presented in Supplementary Fig. S1B

2565

and the Ca2 + tracings in Supplementary Fig. S2C reveal that cells from the same monolayer do not fire in synchrony either when placed close to each other (see, for instance, Roi 2 and Roi 3) or when lying at several micrometers of distance from each other (see, for instance, Roi 1 and Roi 7). Therefore, the pattern of the Ca2 + response varied dramatically from cell to cell. This feature represents a hallmark of the Ca2 + response to growth factors in endothelial committed cells [25,39], as well as other cell types [40,41]. Each baseline Ca2 + spike was preceded by a gradual increase in [Ca2 + ]i that induced the sudden shift from a slow to a more rapid Ca2 + release (Fig. 1). This event, known as pacemaker Ca2 + ramp, is typical of InsP3-dependent Ca2 + discharges [25,39,42,43]. The oscillatory signal was reversible: Intracellular Ca2 + transients returned to the baseline on removal of the growth factor, but they were rapidly recovered on VEGF restoration (Fig. 2A). Similar to PB-ECFCs [25], VEGF-induced Ca2 + oscillations display very irregular and unpredictable patterns in terms of their initiation times and frequency. The statistical analysis was performed on cells exposed to the growth factor for 1 h, which was chosen as an arbitrary time interval to compare the behavior of different cells both from the same coverslip

FIG. 1. Vascular endothelial growth factor (VEGF) induces asynchronous oscillations in intracellular Ca2 + concentration ([Ca2 + ]i) in endothelial colony-forming cells (ECFCs) harvested from umbilical cord blood (UCB). VEGF elicits recurring elevations in [Ca2 + ]i in four distinct cells from the same microscopic field of view. The repetitive Ca2 + transients are not synchronized among the different cells and are heterogeneous in terms of lag time between agonist application and onset of the response, number and frequency of the Ca2 + spikes, and duration of the response. For instance, the cell illustrated by trace (A) displays only two Ca2 + transients; whereas the cells depicted by traces (C) and (D) fire throughout all VEGF stimulation and the cell shown in trace (B) exhibit sporadic Ca2 + spikes. However, the frequency of the Ca2 + waves is dramatically higher in cell (D) as compared with cell (C). VEGF has been administrated at 10 ng/mL during the time period indicated by the black bars placed above the Ca2 + traces.

2566

DRAGONI ET AL.

FIG. 2. Characterization of VEGF-induced Ca2 + oscillations in ECFCs harvested from UCB. (A) The spiking response to VEGF (10 ng/mL) is reversibly abrogated on removal of the growth factor from the bath. (B) Variability in the latency of the Ca2 + response to VEGF. A 100 s bin width was used to calculate the response latency histogram. Each point represents the number of cells displaying a delay in the Ca2 + response falling into the corresponding time interval. (C) Number of cells exhibiting a given number of Ca2 + spikes in the presence of VEGF (10 ng/mL) during 1 h recording. (D) Mean – SE of the peak amplitude of each Ca2 + transients induced by VEGF (10 ng/mL) during 1 h recording. For each value, n ranges from 3 to 47. (E) Mean – SE of the interspike interval (ISI) during the oscillatory response to VEGF (10 ng/mL). The interspike period was evaluated by measuring the peak-to-peak interval between two successive transients. (F) Dependence of the standard deviation s of ISI on the average ISI Tav in UCB-ECFCs exposed to VEGF, as described in [45] and [46]. The correlation coefficient r shows that s and Tav are highly correlated. Only cells displaying at least three Ca2 + oscillations were included in the analysis.

and from different batches. The lag time between exposition to the stimulus and onset of the first Ca2 + elevation ranged between 20 and 1480 s (Fig. 2B), with the average latent period being 287.5 – 13.6 s (n = 458). The frequency of the discrete Ca2 + elevations from the baseline arising during 1 h recording varied between 2 and 36 oscillations/h (Fig. 2C) and averaged 9.6 – 0.4 oscillations/h (n = 458). The amplitude of the Ca2 + oscillations did not significantly (P > 0.05) change throughout the Ca2 + train (Fig. 2D). The apparent decrease in the magnitude of the increases in [Ca2 + ]i occurring after the 15th transient did not reach statistical (P < 0.05) significance. There was no statistical correlation between the resting Fura2 fluorescence and both the height of the first Ca2 + transient (r = 0.088) and the overall number of Ca2 + transients elicited by VEGF (r = 0.1541). Despite the cell-to-cell variability, the

average ISI progressively decreased from an initial value of about 300 s to a sort of plateau level at around 130 s after the 16th Ca2 + transient (Fig. 2E). This result indicates that, once established, the Ca2 + burst takes some time before reaching a regular ISI. A number of modeling studies have disclosed that InsP3-mediated Ca2 + spikes occur in a stochastic regime, in which the average ISI (Tav) increases linearly with the standard deviation s of ISI itself [44–46]. Moreover, s is in the same order as Tav [25,45,46]. Figure 2F shows that both these parameters are satisfied by the repetitive Ca2 + spikes elicited by VEGF in UCB-ECFCs. The variability in VEGFinduce Ca2 + dynamics did not depend on differences in cell volume: We could not find any statistically significant correlation between cell volume and height of the first peak (Supplementary Fig. S2A), number of Ca2 + transients/1 h

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS

2567

FIG. 3. Extracellular Ca2 + entry is essential to trigger and maintain VEGF-induced Ca2 + oscillations. (A) No increase in [Ca2 + ]i was detected when the cells were exposed to VEGF (10 ng/mL) in the absence of extracellular Ca2 + (0Ca2 + ). The bursting signal occurred immediately after Ca2 + restitution to the perfusate. (B) Removal of extracellular Ca2 + (0Ca2 + ) during ongoing Ca2 + oscillations caused the rapid interruption of the Ca2 + response to VEGF (10 ng/mL). VEGF-induced Ca2 + oscillations resumed on Ca2 + readmission to the extracellular solution. (C) Pre-incubating the cells with N-(4-[3,5-bis(trifluoromethyl)-1H-pyrazol-1-yl]phenyl)-4-methyl-1,2,3-thiadiazole-5-carboxamide (BTP-2) (20 mM, 20 min) did not prevent the onset of the Ca2 + train ignited by VEGF (10 ng/mL), albeit it dramatically curtailed the number of the [Ca2 + ]i elevations. The statistical analysis (mean – SE) of VEGF-induced Ca2 + oscillations in the presence of BTP-2 has been reported in (D) (latency), (E) (amplitude of the first spike), and (F) (number of the Ca2 + spikes). In each panel, the number of cell analyzed, n, is equal to 45 for control cells and to 57 for BTP-2-trated cells. *Indicates P < 0.05, as described in the Statistics paragraph.

(Supplementary Fig. S2B), and average ISI (Supplementary Fig. S2C). In conclusion, our findings indicate that the mode of the Ca2 + response is similar in ECFCs irrespective of their source; that is, human peripheral or cord blood, and it mainly consists of asynchronous Ca2 + oscillations that are characterized by a large variability in both the onset and the frequency of the Ca2 + spikes.

Ca2 + entry is required to initiate and maintain VEGF-induced Ca2 + oscillations in UCB-ECFCs We and others have previously shown that VEGF-induced Ca2 + oscillations in PB-ECFCs still arise in the absence of extracellular Ca2 + (0Ca2 + ), whereas SOCE is essential to support them over time. The contribution of Ca2 + influx to the spiking response to VEGF in UCB-ECFCs was first as-

sessed by stimulating such cells under 0Ca2 + conditions. The rationale behind this experiment was to probe whether UCBECFCs undergo a rearrangement of the molecular machinery, leading to the generation of the Ca2 + response to VEGF. Figure 3A shows that no cells displayed any detectable elevation in [Ca2 + ]i on Ca2 + removal from the bath, whereas the rhythmic Ca2 + release immediately resumed on restoration of extracellular Ca2 + . This finding is strikingly different from the results obtained on PB-ECFCs, which always generated 1-to-4 Ca2 + spikes when challenged with VEGF under such conditions [25]. In other words, InsP3-dependent intracellular Ca2 + mobilization triggers the Ca2 + response to VEGF in PB-ECFCs; whereas extracellular Ca2 + entry is essential to initiate VEGF-induced repetitive Ca2 + transients in UCB-ECFCs. To the best of our knowledge, this is the first mechanistic difference ever observed in VEGF signaling

2568 between the two cell types. To determine whether the oscillatory dynamics became independent on Ca2 + entry after its onset, extracellular Ca2 + was depleted from the perfusate after the onset of the signal. Figure 3B shows that this maneuver reversibly abolished the Ca2 + train elicited by VEGF in 73 out of 79 cells (92.4%). SOCE is activated by emptying of the InsP3-sensitive Ca2 + pool in UCB-ECFCs [23] and sustains VEGF-induced Ca2 + oscillations in PB-ECFCs [25]. It has been proposed that SOCE may be activated by the initial drop in the ER Ca2 + content that occurs within the more peripheral organelle sheets, which are located in close proximity to PM. Ca2 + entering the cells through such a pathway is sequestered back by Sarco/Endoplasmatic Reticulum Ca2 + -ATPase (SERCA), thereby loading up the store and sensitizing from within the InsP3Rs to initiate the first Ca2 + spike [47,48]. We hypothesized that an InsP3-dependent Ca2 + release event restricted to the sub-plasmalemmal domain might contribute toward activating SOCE and initiating the spiking signal. Therefore, VEGF was administered to UCB-ECFCs pre-incubated in the presence of BTP-2 (20 mM, 20 min), a selective blocker of store-dependent Ca2 + influx in human ECFCs [1,3,22,23,25], as well as all other cell types [31]. This treatment did not inhibit the onset of the intracellular Ca2 + waves (Fig. 3C), and the latency of the Ca2 + signal was not significantly (P < 0.05) different as compared with control cells (Fig. 3D). The magnitude of the first Ca2 + transient elicited by VEGF was slightly (P < 0.05) smaller in the presence of BTP-2 (Fig. 3E); in addition, it was followed by no more than 1–3 spikes (Fig. 3C, F and Supplementary Fig. S3). Collectively, these results suggest that SOCE does not initiate the oscillatory signal, but is required to support it over time. Therefore, a store-independent membrane pathway is responsible for the onset of the Ca2 + response to VEGF in UCB-ECFCs.

The role of PLCc and InsP3 receptors in the oscillatory response to VEGF The activation of VEGFR-2 by VEGF leads to the synthesis of the two intracellular second messengers, InsP3 and DAG, which is mediated by the effector enzyme, PLCg [3,26]. The role served by PLCg in establishing VEGF-induced Ca2 + oscillations was probed by pre-incubating the cells with U73122 (10 mM, 10 min), an aminosteroid that blocks PLCg in human ECFCs [23,25], as well as other cell types [40,49–52]. This treatment prevented the onset of VEGF-elicited Ca2 + oscillations in 44 out of 46 (95.7%) cells (Fig. 4A), whereas the inactive analogue, U73343 (10 mM, 10 min), was without effect in 64 out of 66 cells (96.7%) (Fig. 4B and Supplementary Fig. S4). Therefore, either InsP3 or DAG (or both) were required for the onset of the oscillatory response to VEGF in UCB-ECFCs. The involvement of the InsP3-sensitive Ca2 + pool was evaluated by pre-incubating the cells with 2aminoethoxydiphenyl borate (2-APB) (50 mM, 20 min), a wellknown InsP3R inhibitor [25,49]. As shown in Fig. 4C, 2-APB suppressed the repetitive Ca2 + transients evoked by VEGF in 105 out of 111 cells (94.6%). It should, however, be noted that 2-APB may also affect a number of plasmalemmal TRP channels [3,53], including TRPC3, thereby potentially interfering with the influx of Ca2 + that triggers Ca2 + signals in UCB-ECFCs. In the absence of other membrane-permeable InsP3R blockers devoid of side-effects, we sought to elucidate

DRAGONI ET AL. the contribution of the intracellular reservoir by first assessing the expression of InsP3Rs. Unlike PB-ECFCs [25], UCBECFCs lack InsP3R1, whereas they display both InsP3R2 and InsP3R3 mRNAs (Fig. 4D), the qRT-PCR products being of the expected size. Negative controls were performed by omitting the reverse transcriptase (not shown). The transcript levels of InsP3R2 and InsP3R3 in UCB-ECFCs were not statistically different (P < 0.05). Subsequently, we depleted the InsP3-dependent Ca2 + pool by pre-treating the cells with cyclopiazonic acid (CPA; 10 mM) [23,25], which prevents SERCA from counterbalancing the passive and activity-dependent Ca2 + fluxes from ER to the cytosol. Figure 4E shows that CPA caused an immediate elevation in [Ca2 + ]i due to passive Ca2 + efflux from the intracellular reservoir followed by a decay phase to a plateau level slightly higher than the baseline, due to SOCE activation [23]. When added at 30 min after the exposition to CPA, a time interval that is sufficient to fully empty the InsP3-sensitive Ca2 + pool [23,25], VEGF generated a monotonic elevation in [Ca2 + ]i rather than an oscillatory signal (Fig. 4E). This pattern of Ca2 + signal has been well associated to a VEGF-induced store-independent Ca2 + entry from the extracellular space [32,54,55]. Indeed, under these conditions (ie, after 30 min of stimulation with CPA), the ER Ca2 + store has already been depleted and SOCE has attained a plateau [56,57]. Accordingly, no Ca2 + response occurred when CPA was perfused in the absence of extracellular Ca2 + (Fig. 4F). Overall, these results indicate that (1) the onset of VEGF-elicited Ca2 + oscillations in UCBECFCs lies downstream the signaling pathway triggered by PLCg; (2) the initial Ca2 + response to VEGF consists of the activation of a Ca2 + entry pathway, which is not regulated by the ER Ca2 + content and is likely gated by an intracellular second messenger; and (3) InsP3-evoked intracellular Ca2 + release drives the Ca2 + spikes triggered by VEGF-induced Ca2 + influx. The role of InsP3-dependent Ca2 + mobilization was further investigated by interfering with ER refilling during the ongoing variations in [Ca2 + ]i induced by VEGF [25,39]. The acute addition of CPA (10 mM) rapidly inhibited VEGF-induced Ca2 + oscillations in 54 out of 56 cells (96.4%) (Fig. 5A), whereas thapsigargin (2 mM) blocked the Ca2 + response to VEGF in 45 out of 48 cells (93.7%) (Fig. 5B). These data demonstrate that SERCA-mediated refilling of the InsP3-sensitive Ca2 + is essential to maintain the Ca2 + train.

Evidence for TRPC3 expression in UCB-ECFCs The influx of Ca2 + necessary for the onset of the oscillatory response to VEGF in UCB-ECFCs could be provided by Ca2 + -permeable membrane channel that is sensitive to DAG, an intracellular second messenger that is synthesized by PLCg along with InsP3 [3,58]. Unlike PB-ECFCs [23] (see also Fig. 6A), these cells are endowed with a transcript encoding for the DAG-dependent TRPC3 channel [23] (see also Fig. 6A), which provides the pathway for VEGF-induced Ca2 + entry in mature endothelium [32,33]. Therefore, we sought to unveil the role served by TRPC3 in the initiation of VEGFevoked Ca2 + oscillations in UCB-ECFCs. First, we evaluated the translation of TRPC3 mRNA into a functional protein by immunoblotting. We utilized two different antibodies, purchased from two distinct companies, which recognized a single band of about 97 kDa, the expected molecular weight for TRPC3 [59,60] (Fig. 6B, C). The blanks, where the primary

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS

2569

FIG. 4. Role of the phospholipase C-g/InsP3 receptors signaling pathway in VEGF-induced Ca2 + oscillations. VEGF (10 ng/ mL) did not produce any detectable elevation in intracellular Ca2 + levels when ECFCs were pre-incubated in the presence of U73122 (10 mM, 10 min) (A), while the oscillatory Ca2 + signal was not affected by its inactive structural analog, U73343 (10 mM, 10 min) (B). (C) Pre-incubating the cells with 2-APB (50 mM, 20 min) abrogated the onset of the Ca2 + response to VEGF (10 ng/mL) (light gray tracing; Control, Ctl). Control cells, that is, not pre-treated with 2-APB, displayed recurring elevations in [Ca2 + ]i when challenged with VEGF (10 ng/mL) (black tracing). (D) Expression of InsP3R1, InsP3R 2, and InsP3R3 transcripts in ECFCs isolated from UCB. mRNA levels were measured by real-time polymerase chain reaction relative to the b-actin internal standard (see Materials and Methods section), and the values obtained were reported as DCt. Bars represent the mean – SEM of at least four different experiments, each from different RNA extracts. (E) Depletion of the InsP3-sensitive Ca2 + stores with cyclopiazonic acid (CPA; 10 mM) turned VEGF-induced Ca2 + oscillations into a monotonic increase in [Ca2 + ]i. (F) The monotonic increase in [Ca2 + ]i elicited by VEGF in UCB-ECFCs pre-treated with CPA was suppressed in the absence of extracellular Ca2 + (0Ca2 + ). 2-APB, 2-aminoethoxydiphenyl borate.

antibody was omitted, did not show any immunoreactivity (Fig. 6B, C). TRPC3 expression was further assessed by immunocytochemistry, as depicted in Supplementary Fig. S5. Then, UCB-ECFCs were exposed to OAG, a membranepermeable analog of DAG that is widely employed to pharmacologically activate TRPC3 [53,59,61]. Notably, OAG (100 mM) induced a rapid increase in [Ca2 + ]i followed by a sustained plateau only in the presence (Fig. 7A), but not in the absence (Fig. 7B), of extracellular Ca2 + in 296 out of 408 cells (72.5%). The Ca2 + signal was restored on Ca2 + restitution to the extracellular solution (Fig. 7B). As a negative control, we utilized PB-ECFCs, which lack TRPC3 [23] and failed to produce any detectable elevation in intracellular

Ca2 + levels when challenged with the DAG analog (Fig. 7A). In order to assess whether OAG-gated Ca2 + entry was mediated by TRPC3, we exploited Pyr3, a widely employed inhibitor of this pathway [35,38,53,62]. As shown in Fig. 7C, pre-incubating the cells with Pyr3 (10 mM, 10 min) prevented the Ca2 + response to OAG. Similarly, OAG-induced Ca2 + entry was abrogated by pre-treating the cells with FFA (100 mM, 10 min) (Fig. 7D), another well-established inhibitor of TRPC3 [53,62]. Interestingly, OAG-elicited Ca2 + inflow quickly resumed on removal of FFA (Fig. 7D), but not Pyr3 (Fig. 7C), from the bath. BTP-2 has shown to impair TRPC3mediated Ca2 + signals by one single study [63]. Therefore, we probed the effect exerted by pre-incubating the cells with

2570

DRAGONI ET AL.

FIG. 6. Canonical transient receptor potential C 3 (TRPC3) mRNA and protein are expressed in ECFCs isolated from UCB. (A) mRNA levels were measured by real-time polymerase chain reaction relative to the b-actin internal standard (see Materials and Methods section), and the values obtained were reported as DCt. Bars represent the mean – SEM of at least four different experiments each from different RNA extracts. (B, C) TRPC3 protein was detected by immunoblotting, as described in Materials and Methods. Lanes were loaded with 30 mg of proteins and probed with TRPC3 rabbit polyclonal antibody (B) (Acris Antibodies GmbH) or with TRPC3/6/7 (N-18) (Santa Cruz Biotechnology, Inc.; sc15056) goat polyclonal antibody (C). A band of about 97 kDa was observed with both antibodies. Blots were stripped and re-probed with anti b-actin rabbit antibody as loading control (Rockland Immunochemicals for Research, U.S.A.; code, 600-401-886). Blots representative of four were shown. FIG. 5. CPA interrupts VEGF-induced Ca2 + oscillations. The acute addition of either CPA (10 mM) (A) or thapsigargin (2 mM) (B) rapidly blocked the spiking response to VEGF (10 ng/mL). Note the immediate increase in [Ca2 + ]i induced by both drugs, which reflects the passive Ca2 + efflux from the endoplasmic reticulum on SERCA inhibition. BTP-2 (20 mM, 30 min) on the Ca2 + response to OAG. We found that this maneuver did not impair OAG-induced Ca2 + signaling in UCB-ECFCs (Fig. 7E), a finding that, in turn, reinforces the selective action of this drug on SOCE. The bar histogram in Fig. 7F summarizes the percentage of cells responding to OAG in the presence and in the absence of each TRPC3 blocker and of BTP-2. Overall, these data confirm that the mRNA encoding for TRPC3 is translated into a functional Ca2 + -permeable channel in UCB-ECFCs.

Pharmacological evidence that TRPC3 triggers the VEGF-induced Ca2 + oscillations essential for UCB-ECFC proliferation In order to assess the involvement of TRPC3 in the initiation of the spiking response to VEGF in UCB-ECFCs, the cells were pre-incubated with either Pyr3 (10 mM, 10 min) or FFA (100 mM, 10 min) before being challenged with the growth factor. VEGF induced asynchronous Ca2 + oscillations under control conditions (Fig. 8A) (ie, in the absence of any TRPC3 inhibitor), whereas no evident increase in [Ca2 + ]i

was observed in the presence of either Pyr3 (Fig. 8B) or FFA (Fig. 8C) in VEGF-stimulated cells. Notably, the blocking effect of either drug was irreversible, and the Ca2 + transient did not resume on their removal from the bath. Moreover, the acute addition of Pyr3 (10 mM) during an ongoing Ca2 + burst caused the rapid interruption of the signal (Fig. 8D). The bar histogram in Fig. 8E summarizes the fractions of cells responding to VEGF in the absence and in the presence of either Pyr3 or FFA, while Supplementary Fig. S6 illustrates the distribution of the Ca2 + spikes recorded under each condition (ie, control, Pyr3, and FFA). FFA has been reported to activate TRPC6 [64] and to inhibit TRPC4 [65] at the concentrations employed in the present study. UCB-ECFCs do not possess TRPC6, but are endowed with TRPC4 [23]. In the same way, Pyr3 was recently found to inhibit Orai1gated Ca2 + entry in native RBL-2H3 cells [66]. Therefore, we carried out a number of control experiments on VEGFevoked Ca2 + oscillations in PB-ECFCs, which express TRPC4 and Orai1, but not TRPC3. As depicted in Supplementary Fig. S7, neither Pyr3 nor FFA impaired the onset of the Ca2 + response in these cells. The physiological meaning of VEGFevoked intracellular Ca2 + transients was evaluated by investigating the percentage of UCB-ECFC proliferation in the presence of the following drugs: Pyr3 (10 mM), FFA (100 mM), BAPTA (30 mM), a membrane-permeable buffer of intracellular Ca2 + , and BTP-2 (20 mM) [22,23,25]. As shown in Fig. 8F, each treatment reduced UCB-ECFC growth. Collectively, these results suggest that TRPC3 triggers and sustains

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS

2571

FIG. 7. Evidence for TRPC3-mediated Ca2 + entry in ECFCs isolated from UCB. (A) 1-oleoyl-2-acetyl-sn-glycerol (OAG) (100 mM), a membrane-permeable analog of diacylglycerol, induced a rapid increase in [Ca2 + ]i in UCB-ECFCs (black tracing), but not PB-ECFCs (light gray tracing). (B) The Ca2 + response to OAG (100 mM) did not occur in the absence of extracellular Ca2 + (0Ca2 + ), but resumed on Ca2 + re-addition to the bath. (C) Pyr3 (10 mM, 10 min), a selective inhibitor of TRPC3, inhibited OAGinduced Ca2 + entry. OAG was administrated at 100 mM. (D) Flufenamic acid (FFA; 100 mM, 10 min), which may also interfere with TRPC3, reversibly abrogated OAG-induced Ca2 + inflow. OAG was administrated at 100 mM. (E) BTP-2 (20 mM, 30 min) did not prevent the Ca2 + response to OAG (100 mM). (F) Percentage of UCB-ECFCs displaying OAG-induced Ca2 + signals in the absence and in the presence of each TRPC3 blocker and of BTP-2. *Indicates P < 0.05, as described in Statistics paragraph. VEGF-elicited Ca2 + oscillations, which are, in turn, necessary to promote the pro-angiogenic effect of the growth factor.

The genetic disruption of TRPC3 prevents VEGFinduced Ca2 + oscillations in UCB-ECFCs The role served by TRPC3 in igniting the Ca2 + response to VEGF was further assessed by down-regulating its expression in UCB-ECFCs. The cells were transfected with either an siRNA selectively targeting or a scrambled (ie, control) siRNA. Immunoblot analysis revealed that the content of TRPC3 protein decreased by *68.5% 72 h after siRNA transfection (Fig. 9A, B), whereas TRPC1 and TRPC4 proteins were still present in TRPC3-deficient cells (Supplementary Fig. S8). Consistent with this finding, the genetic suppression of TRPC3 dramatically reduced the percentage of UCB-ECFCs responding to OAG (100 mM) with an increase in [Ca2 + ]i as compared with control cells (Fig. 9C, E).

Similarly, the fraction of cells producing a detectable Ca2 + signal when challenged with VEGF dramatically diminished on siTRPC3 transfection (Fig. 9D, E). Under such conditions, the elevation in [Ca2 + ]i consisted of a rapid Ca2 + spike in 9 out of 143 cells (6.3%), whereas VEGF-evoked repetitive Ca2 + spikes were observed only in 1 out of 143 cells (0.7%). Finally, the selective knockdown of TRPC3 significantly (P < 0.05) impaired VEGF-dependent UCB-ECFC proliferation, by reducing the rate of cell growth by about 80% (Fig 9F). Altogether, these results confirm that TRPC3-induced Ca2 + entry represents the triggering event of the oscillatory response to VEGF in ECFCs isolated from UCB.

Discussion Local delivery of autologous EPCs is now regarded as one of the most promising strategies to treat severe ischaemic diseases, such as MI and PAD, by regenerating the impaired

2572

DRAGONI ET AL.

FIG. 8. The pharmacological blockade of TRPC3 suppresses VEGF-induced Ca2 + oscillations in ECFCs harvested from UCB. The spiking response to VEGF (10 ng/mL) arose in the absence (A), but not in the presence of either Pyr3 (10 mM, 10 min; B) or FFA (100 mM, 10 min; C). (D) The acute addition of Pyr3 (10 mM) rapidly inhibited VEGF-induced Ca2 + transients. VEGF was administrated at 10 ng/mL. (E) Percentage of UCB-ECFCs responding to VEGF (10 ng/mL) in the absence and in the presence of Pyr3 and FFA. (F) Fraction of proliferating UCB-ECFCs in the absence (Control; Ctl) and in the presence of the following inhibitors of VEGF-evoked Ca2 + oscillations: Pyr3 (10 mM), FFA (100 mM), and 1,2-bis(o-aminophenoxy) ethane-N,N,N¢,N¢tetraacetic acid (BAPTA) (30 mM; BAP), a well-known membrane permeable buffer of intracellular Ca2 + , and BTP-2 (20 mM). *Indicates P < 0.05, as described in Statistics paragraph. vascular network [1–3,9–12]. In particular, ECFCs stand out as the most suitable EPC subset by virtue of their ability to acquire a truly endothelial phenotype, to assembly into a bidimensional capillary-like network in vitro, and form patent vessels in vivo [1,6,13,15,22,25,67]. An inherent limitation in autologous cell therapy is provided by the scarcity of circulating ECFCs and by their limited angiogenic response to VEGF as compared with UCB-derived cells [14]. These features suggest that the signaling pathways by which VEGF activates ECFCs are different depending on their source, that is, either peripheral or UCB. Highlighting the molecular mediators of VEGF-induced UCB-ECFC proliferation might supply the biological bases required to improve the clinical outcome of cell therapy by a genetic intervention on the underlying regulatory machinery [1,3,9,68,69]. We have recently shown that VEGF stimulates PB-ECFCs to undergo angiogenesis by an oscillatory increase in [Ca2 + ]i, which is tightly patterned by the concerted interplay between InsP3-dependent Ca2 + release and consequent SOCE

activation [25]. More specifically, the Ca2 + response to VEGF is initiated by InsP3-induced intracellular Ca2 + mobilization; whereas store-dependent Ca2 + entry is required to maintain it over time. We, therefore, speculated that the Ca2 + response to VEGF in UCB-ECFCs might involve the activation of either different and/or additional Ca2 + -permeable channels. This hypothesis was sustained by the notion that a different blend of ion channels is expressed in UCB- versus PBderived cells [23] and by the observation that the Ca2 + toolkit expressed by ECFCs is not set in stone, but may be dramatically remodeled, as observed when they are harvested by patients affected by renal cellular carcinoma [22]. Similar to their peripheral counterparts [24,25], VEGF induces UCBECFCs to undergo asynchronous oscillations in [Ca2 + ]i, which stochastically arise in the presence of the growth factor throughout the stimulation. The major difference between the two cell types was observed when UCB-ECFCs were challenged with VEGF in the absence of extracellular Ca2 + : Unlike PB-ECFCs, no evident increase in [Ca2 + ]i occurred, but

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS

2573

FIG. 9. The genetic knockdown of TRPC3 prevents VEGF-induced Ca2 + oscillations and proliferation. (A) TRPC3 short interfering RNA (siRNA) and scrambled siRNA (C) were transfected in UCB-ECFCs as described in Materials and Methods. Lanes were loaded with 30 mg of proteins, probed with affinity-purified antibodies, and processed as described in Materials and Methods. The same blots were stripped and re-probed with anti-b-actin antibody. Bands of the expected molecular weights were shown. Bands were acquired with the Image Master VDS (Amersham Biosciences Europe). Blots representative of three were shown. (B) Densitometric analysis of the bands was performed by Total Lab V 1.11 computer program (Amersham Biosciences Europe), and the results were normalized to the corresponding b-actin. Western blot and densitometry demonstrating a significant decrease in TRPC3 protein expression in silenced (siRNA) UCB-ECFCs compared with controls (scrambled; C) (*P < 0.001; Student’s t test). (C) Ca2 + signals evoked by OAG (100 mM) in UCB-ECFCs treated with scrambled siRNA (black tracing) and siTRPC3 (light gray tracing). (D) Ca2 + signals elicited by VEGF (10 ng/mL) in UCB-ECFCs treated with scrambled siRNA (black tracing) and siTRPC3 (light gray tracing). (E) Percentage of responding cells exposed to either OAG (100 mM) or VEGF (10 ng/mL) under the described conditions. (F) Fraction of proliferating cells after transfection with scrambled siRNA and siTRPC3. the repetitive Ca2 + spikes resumed immediately after Ca2 + restoration to the bath. This result clearly indicates that Ca2 + entry is essential to trigger the spiking response to VEGF in UCB-derived cells, while InsP3-dependent Ca2 + release is sufficient to initiate Ca2 + signaling in circulating ECFCs [24,25]. The following pieces of evidence hinted at TRPC3 as the Ca2 + -permeable route responsible for the initiation of VEGF-induced Ca2 + oscillations in UCB-ECFCs. First, TRPC3 mRNA [23] and protein (present study) are expressed in UCB-ECFCs, but not in their peripheral counterparts. Second, TRPC6 and TRPC7, which may also mediate DAG-sensitive Ca2 + entry, are not present in UCB-ECFCs. Third, OAG, a widely employed membrane-permeable analog of DAG,

evokes Ca2 + inflow in ECFCs isolated from UCB rather than adult PB. This latter feature rules out the contribution of TRPC1, which is present in both cell types and may gate DAG-sensitive Ca2 + -permeable cationic channels in human prostate cancer cells [70]. Notably, OAG-evoked Ca2 + influx is abolished by Pyr3 and FFA, two well-known inhibitors of TRPC3-mediated elevations in intracellular Ca2 + levels pathway [35,38,53,62]. At the same way, Pyr3 and FFA prevent the onset of VEGF-induced Ca2 + signals, an effect that may be mimicked by blocking with PLCg activity and DAG synthesis with U73122 (but not its inactive structural analog, U73343). Fourth, the genetic knockdown of TRPC3 expression via a specific siRNA dramatically impairs UCB-ECFC

2574 sensitivity to OAG and suppresses the oscillatory response to VEGF. Fifth, both the pharmacological (Pyr3 and FFA) and the genetic (siRNA) inhibition of TRPC3-gated Ca2 + entry strongly affects VEGF-induced UCB-ECFC proliferation. Sixth, the blockade of SOCE by pre-incubating the cells with BTP-2 does not affect the onset of the Ca2 + response, albeit it dramatically reduces the number of the following Ca2 + transients. As a consequence, the specific expression of TRPC3 in ECFCs harvested from UCB does not change the kinetics of the Ca2 + response, that is, asynchronous Ca2 + spikes, to VEGF as compared with their peripheral counterparts. Nevertheless, it profoundly alters its underlying mechanisms; whereas TRPC3 is absolutely required both to trigger the signal and to maintain Ca2 + transients over time. Indeed, the acute addition of Pyr3 during an ongoing response causes the rapid interruption of the [Ca2 + ]i elevation. This finding concurs with previous studies, which documented the role served by TRPC3 as a stimulator of intracellular Ca2 + spiking in a variety of cell types. For instance, TRPC3 drives the spontaneous Ca2 + oscillations arising in human umbilical vein-derived endothelial cell line EA.hy926 plated on Matrigel [71] and the periodic increases in [Ca2 + ]i elicited by B-cell receptor engagement [38,72]. In addition, TRPC3-dependent Ca2 + entry triggers thrombin-evoked repetitive Ca2 + spikes in 1321N1 human astrocytoma cells [73], as well as antigen-induced intracellular Ca2 + waves in rat basophilic leukemia and BM-derived rat mast cells [74]. Finally, TRPC3 activation causes cyclic oscillations in [Ca2 + ]i in rat type I astrocytes [59] and is likely to underpin OAGinduced repetitive Ca2 + spikes in human myometrial cells [75]. Conversely, the monotonic Ca2 + entry mediated by TRPC3 in CD133 + progenitor cells isolated from human adipose stroma and exposed to VEGF does not lead to an intracellular Ca2 + spike [55]. A caveat in the interpretation of the data generated by the pharmacological inhibition of TRPC3 is represented by the lack of selectivity of the inhibitors that we employed. FFA may also affect TRPC6, which is however absent in UCB-ECFCs [23], and TRPC4 [65]; while Pyr3 has been recently reported to interfere with Orai1mediated SOCE [66]. Similarly, BTP-2 has previously been reported to affect TRPC3-dependent Ca2 + entry in heterologous expression systems [63]. However, the Ca2 + response to OAG is not affected by BTP-2 in UCB-ECFCs (present study); whereas it is sensitive to both Pyr3 and FFA. These results suggest that the side-effects of Ca2 + channels blockers may be absent in these cells and related to ancillary subunits that are not expressed in ECFCs. Consistently, control experiments conducted on PBECFCs, which express TRPC1, TRPC4, and Orai1 but lack TRPC3, revealed that the Ca2 + response to VEGF still occurred and adopted an oscillatory pattern in the presence of either Pyr3 or FFA. Finally, immunoblotting analysis disclosed that TRPC3-deficient cells do not manifest evident off-target effects on the Ca2 + machinery, such as a drop in TRPC1 and TRPC4 proteins. These data strongly indicate that Pyr3 and FFA are unlikely to affect SOCE in UCB-ECFCs, but caution should be warranted when these compounds are utilized in the absence of gene silencing as a confirmatory approach. The Ca2 + signaling machinery placed downstream TRPC3-mediated Ca2 + entry in ECFCs isolated from UCB is somehow similar to that recruited by VEGF in their peripheral counterparts [22–25,76]. Accordingly, TRPC3-initiated Ca2 + oscillations are shaped by the rhythmic Ca2 + release

DRAGONI ET AL. through InsP3Rs and sustained over time by store-dependent Ca2 + influx accomplished by Stim1, Orai1, and TRPC1. This scenario is corroborated by four pieces of evidence. First, the pharmacological inhibition of InsP3-dependent signaling with either U73122 or 2-APB, a well-known blocker of InsP3Rs, prevents the onset of the Ca2 + response to VEGF in UCB-ECFCs, as well as in their peripheral counterparts [24,25]. Unfortunately, 2-APB is not a selective drug, as it may also interfere with Stim1, Orai1, and TRPC1, all of which contribute to SOCE in human ECFCs [22,23,76], and TRPC3 at the concentration employed in the present study (ie, 50 mM) [3,53]. Since 2-APB could not be probed in the absence of extracellular Ca2 + , which is required for the generation of the Ca2 + oscillations, UCB-ECFCs were exposed to VEGF on depletion of the InsP3-sensitive Ca2 + pool with either CPA or thapsigargin [23,25]. Under these conditions, VEGF causes a monotonic elevation in [Ca2 + ]i, which is a hallmark of VEGF-induced DAG-activated Ca2 + entry in mature endothelium [32–34,54]. This result indicates that TRPC3-mediated Ca2 + entry controls InsP3-dependent Ca2 + mobilization and, as a consequence of the drop in ER Ca2 + levels, it ultimately leads to SOCE activation. Second, UCBECFCs express the transcripts encoding for two of the three hitherto identified InsP3R isoforms, namely InsP3R2 and InsP3R3, whereas they lack InsP3R1. This feature adds to the growing list of molecular differences in the Ca2 + machinery reported in PB- versus UCB-derived ECFCs, as the former present all the three isotypes [25]. It is, however, worth of noting that InsP3R2 is the isoform that is more strictly required for both the onset and the maintenance of Ca2 + oscillations, due to its peculiar sensitivity to both InsP3 and Ca2 + [77]. Third, the slow pacemaker-like increase in [Ca2 + ]i which precedes the Ca2 + spikes, is a hallmark of InsP3dependent Ca2 + release [25,39,42,43], as well as the stochastic sequence of the intracellular Ca2 + spikes elicited by VEGF has been faithfully modeled by the random opening of InsP3R within the ER membrane [44–46]. Fourth, the inhibition of SOCE, which is activated on depletion of the InsP3-dependent Ca2 + store in ECFCs [3,22,23], dramatically curtails the duration of the oscillatory response induced by VEGF in the presence of BTP-2. In particular, SOCE is required to recharge the ER Ca2 + reservoir in an SERCAdependent manner during cell stimulation. This same function, that is, refilling the InsP3-dependent Ca2 + pool in agonist-solicited cells, has long been linked to store-dependent Ca2 + inflow [78], albeit the signaling role served by SOCE in the context of gene expression modulation has been recently unveiled [30,44,78]. SOCE in human ECFCs is mediated by the interaction of Stim1, the ER Ca2 + -sensor, with Orai1 and TRPC1, both of which contribute to the Ca2 + -permeable pore on the PM. It is, however, yet to be elucidated whether Stim1, Orai1, and TRPC1 assemble into a heteromeric ternary complex, as described in human platelets [28], or whether TRPC1 and Orai1 constitute two distinct store-operated pathways, each of which are activated by Stim1 [79]. While TRPC3 is only expressed by UCB-ECFCs [23], Stim1, Orai1, and TRPC1 are present in ECFCs isolated from both UCB and PB [22,23,25]. This feature highlights the central role served by SOCE-related proteins in maintaining the oscillatory response to VEGF over time in different types of ECFCs. The question then arises about the mechanistic link between TRPC3 activation and the ignition of the repetitive

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS intracellular Ca2 + waves in UCB-ECFCs. Three alternative, albeit not mutually exclusive, scenarios might be envisaged. The first model is based on the well-known increase in InsP3R sensitivity to InsP3 on elevation of both luminal and cytosolic Ca2 + , the latter mechanisms referred to as Ca2 + induced Ca2 + release (CICR) [47,48,58]. If either the luminal Ca2 + load and/or the ambient levels of Ca2 + are too low to enable InsP3-dependent Ca2 + mobilization, TRPC3-gated Ca2 + inflow might provide the bolus of Ca2 + required to sensitize the receptors to their ligand, that is, InsP3, and activate the release. This hypothesis is supported by the finding that extracellular Ca2 + influx may control the frequency of Ca2 + oscillations by triggering InsP3-dependent Ca2 + discharges in a number of cell types [80,81]. Nevertheless, there is no report of direct InsP3R modulation by TRPC3-mediated Ca2 + inflow. Moreover, Pyr3 caused an instantaneous termination of ongoing Ca2 + transients, which suggests that TRPC3-induced Ca2 + influx stimulates the spiking response as long as VEGF is administrated to the cells. Under such conditions, however, the cytosol is turned into an excitable medium by the rhythmic Ca2 + release, and the ER is continuously refurnished with Ca2 + by SOCE, so that it is hard to conceive that InsP3R is not sensitive to InsP3. Therefore, TRPC3 should be more tightly linked to InsP3R activation. According to the second model, TRPC3 might induce Ca2 + release by mediating the selective interaction between InsP3R1 and the receptor for activated C-kinase-1 (RACK1) [82,83]. However, this mechanism implies the formation of a ternary complex formed by TRPC3, InsP3R1 and RACK1 and is not consistent with our observations that Ca2 + inflow through TRPC3, rather than the channel protein itself, is essential to trigger the oscillatory response and that InsP3R1 is absent in UCB-ECFCs. Moreover, the physical recruitment of InsP3R1 by TRPC3 leads to a sustained increase in [Ca2 + ]i rather than to periodic Ca2 + pulses [82], as observed under our conditions. Conversely, according to the third model, TRPC3-induced Ca2 + entry might promote InsP3 synthesis by enhancing the rate of PLCg activation. Indeed, TRPC3 elicits translocation toward the PM and subsequent activation of the Ca2 + -sensitive PLCg in antigen-stimulated DT40 B lymphocytes [72], a process that is suppressed by either Pyr3 [35] or TRPC3 down-regulation by siRNA [72]. We suggest that the amount of InsP3 produced immediately after VEGFR-2 activation is not sufficient to initiate the intracellular Ca2 + spikes, either because of scarce PLCg recruitment to the PM or rapid InsP3 metabolism. However, DAG, which is generated along with InsP3, gates TRPC3 to provide the source of Ca2 + that is necessary to further stimulate PLCg and trigger the first Ca2 + pulse in an InsP3-sensitive manner. The following drop in the ER Ca2 + content will lead to Stim1 heteromerization and translocation beneath the PM, where it recruits Orai1 and TRPC1 to mediate SOCE, thereby recharging the intracellular Ca2 + store. This positive feedback between PLCg, DAG, TRPC3, InsP3R, and SOCE occurs throughout the oscillatory signal, as suggested by its strong sensitivity to Pyr3. This hypothesis is supported by the observation that PLCd, another PLC isoform tightly regulated by sub-membranal Ca2 + levels, is recruited by Ca2 + entry to sustain InsP3-dependent Ca2 + oscillations in damaged endothelium [8,84]. This model implies that VEGF-induced InsP3 synthesis in PB-ECFCs is well beyond the threshold required to ignite a robust intracellular Ca2 + discharge, so

2575

that the spiking response occurs even in the absence of extracellular Ca2 + entry. The Ca2 + responses generated by adjacent/neighboring CB-ECFCs to VEGF differ among each other in terms of latent period, number, and period of the Ca2 + spikes. Variability in the pattern of Ca2 + signals is the hallmark of growth factors-evoked elevations in [Ca2 + ]i in mature endothelial cells [21,39]. Moreover, we have recently found that VEGF-evoked Ca2 + oscillations are highly heterogeneous in PB-ECFCs as well [25]. The molecular underpinnings of the diversity in the spatiotemporal dynamics of VEGF-induced Ca2 + events are far from being fully understood. A combination of experimental and computational studies has disclosed that such a feature is not peculiar to Ca2 + signaling, but is an intrinsic property of tyrosine kinase receptor activation. For instance, the differential activation of jagged1 and DII4 in the Notch signaling pathway determines the distinct fate (tip vs. stalk) of two adjacent endothelial cells exposed to a VEGF gradient during the angiogenic process [85]. A recent study has further demonstrated that the mitogen-activated protein kinase pathway is triggered by VEGF in individual microvascular endothelial cells, whereas their adjoining cells are silent [86]. The dissimilar response of neighboring cells to VEGF might be linked to the wide variation in VEGFR-2 levels, as only about 60% of UCB-ECFCs express it (unpublished data from our group). This feature might drive the observed differences in the signal strength, as originally reported for the epidermal growth factor receptor [87]. The heterogeneous number of VEGFR-2 molecules on the endothelial membrane quantitatively and qualitatively activates distinct intracellular pathways within the same cell monolayer [88]. The heterogeneity in VEGF signaling is exacerbated by the internalization of VEGFR-2 [89], which contributes toward coordinating the complementary functions of adjoining cells during tissue morphogenesis, and its regulation by binding partners, such as VEGFR1 and neuropilin-1, that are differentially distributed within the endothelial network [26]. An additional source of variability is provided by the release of the soluble, ligand-sequestering form of VEGFR1 (sFLT-1), that engenders a difference in the amount of VEGF sensed by individual cells, albeit placed in close proximity to each other [90]. Alternatively, the specific Ca2 + signatures evoked by VEGF in UCB-ECFCs might involve a difference in the sub-cellular distribution of the underlying signaling pathways, such as InsP3Rs [44–46]. The stochastic behavior of the periodic fluctuations in [Ca2 + ]i evoked by VEGF in UCB-ECFCs corroborates this hypothesis. VEGF-dependent Ca2 + oscillations arise randomly, and the standard deviation of their ISI is of the same order of magnitude as the average value. A number of computational studies have revealed the mechanistic link between this feature and the topographical hierarchy of InsP3Rs [25–27]. InsP3Rs are spatially arranged into clusters of 1-to-15 channels, which are scattered within the ER membrane with distances of 1–7 mm. The coupling between adjoining channels and clusters is accomplished by Ca2 + diffusion through the CICR process. As a consequence, InsP3 locally synthesized by PLCg primes all the channels in the cluster for activation by Ca2 + and originates a Ca2 + puff, the elementary Ca2 + event. The probability that a single Ca2 + initiates Ca2 + release from an adjacent site is rather low, as Ca2 + is a poorly diffusible messenger for distances £ 2 mm [45,46,91].

2576 Conversely, if, by chance, a supercritical number of InsP3R clusters open in synchrony, the CICR couples all the unitary Ca2 + puffs and ignites a global Ca2 + spike [25–26]. The probability that such a cell wide event (ie, an intracellular Ca2 + wave) occurs depends on several parameters, that is, strength of the spatial coupling, intracellular Ca2 + buffering, and puff duration, and is responsible for the cell-to-cell variability reported by Ca2 + imaging measurements [44–46,91]. Mathematical modeling unveiled that the heterogeneity in the pattern of Ca2 + oscillations arising within adjacent/neighboring cells is independent of morphological parameters, such as cell volume, ER volume, and shape [91,92]. This finding was supported by lack of any statistically relevant correlation between variability in cell volume and intracellular Ca2 + dynamics in VEGF-stimulated UCB-ECFCs. It is, however, worth noting that earlier computational studies pointed at a cross-talk between the pattern of Ca2 + oscillations and cellular morphology [93]. For instance, a significant enlargement in cell surface area may lead to a decrease in the ISI [93]. More sophisticated experiments should thus be devoted to further addressing this intriguing mechanistic issue. The biological meaning of VEGF-induced Ca2 + oscillations is to drive UCB-ECFC proliferation, as shown by the sensitivity of this process to Pyr3, FFA, BAPTA, BTP-2, and siTRPC3. In addition, our preliminary results indicate that the Ca2 + response to VEGF regulates UCB-ECFCs assembly into capillary-like networks in vitro. These results are similar to our previous report on PB-ECFCs [25]. The intracellular Ca2 + transients are generated by the rhythmic Ca2 + release from InsP3Rs, while SOCE is essential to refill the ER Ca2 + pool during prolonged stimulation. Therefore, InsP3 is likely to mediate the pro-angiogenic effect of VEGF-induced Ca2 + burst. Consistently, InsP3 was found to drive VEGF-elicited migration and proliferation in HUVEC [94,95], in bovine retinal endothelial cells [49], and in human choroidal endothelial cells [96]. More in general, InsP3-dependent Ca2 + mobilization has been linked to cell division, survival, and motility in a variety of cell types [41,48,97], and a remodeling of the InsP3 signaling machinery has been described in highly proliferating and metastatic tumor cells [98,99]. SOCE, however, does not only fulfil the function to replenish emptied stores, but it also controls a number of vital cellular processes, including gene expression, migration, and nitric oxide synthesis, via the physical coupling between Orai1 and a number of Ca2 + -sensitive decoders [3,30,31]. Therefore, the possibility that the biological message encoded by the repetitive Ca2 + spikes, that is, UCB-ECFC proliferation, is not only delivered by InsP3Rs, but is also contributed by SOCE, should not be ruled out. It has long been known that the pro-angiogenic response to VEGF, as evaluated in terms of in vitro proliferation and tubulogenic rates, is dramatically higher in ECFCs isolated from UCB rather than from PB [14]. This notion was not accompanied by the knowledge of any relevant difference in the signal transduction pathways downstream of VEGF in the two cell types. In this context, the involvement of TRPC3 in VEGF-induced Ca2 + oscillations and proliferation in UCBECFCs gains particular interest in the light of the selective coupling between TRPC3 and a number of Ca2 + -sensitive regulators of angiogenesis, such as NF-kB, NFAT, and ERK [100,101]. For instance, in DT40 B cells stimulated with an anti-B cell receptor antibody, TRPC3 recruits to the PM

DRAGONI ET AL. protein kinase Cb (PKCb), which, in turn, phosphorylates and stimulates ERK [38]. Moreover, TRPC3-gated Ca2 + influx controls NF-kB signaling in human coronary artery endothelial cells [37], whereas Ca2 + entry via TRPC3, rather than through L-type (CaV1.2) Ca2 + channels, is selectively conveyed to calcineurin to induce the nuclear translocation of NFAT in HL-1 atrial myocytes [36]. In agreement with these findings, TRPC3 induces the expression of hypertrophy-associated genes in response to hypertrophic stimuli in rat ventricular myocytes, while it does not interfere with CaV1.2-paced cell contraction [102]. Finally, it has been demonstrated that TRPC3-induced Ca2 + influx is essential for endothelial tube formation and proliferation in EA.hy926 cells [71] and for vascular cell adhesion molecule-1 expression in human coronary artery endothelial cells [37]. The tight coupling between TRPC3-mediated Ca2 + inflow and VEGF-induced proliferation is highlighted by the dramatic effect on UCB-ECFC growth exerted by Pyr3, FFA, and siTRPC3. Indeed, all of these treatments blocked the proliferative response to VEGF to a much higher extent than simply buffering intracellular Ca2 + with BAPTA. As outlined in a recent review [103], the weak effect of BAPTA indicates that the Ca2 + -sensitive decoder(s) should be placed at a distance much shorter than 10–100 nm from the channel pore; whereas the inhibition is significantly higher when the Ca2 + source is abrogated by either pharmacological or genetic interventions. It, thus, appears that TRPC3 has the potential to increase the repertoire of Ca2 + -sensitive proangiogenic pathways that are activated during VEGF-induced Ca2 + oscillations. In this view, it is tempting to speculate that the selective expression of a functional TRPC3 protein in UCB-ECFCs contributes to render these cells more sensible to VEGF as compared with PB-derived cells. Experiments are under way in our laboratory to assess whether TRPC3 expression in PB-ECFCs will enhance their angiogenic response to VEGF in vitro.

Conclusion The present investigation described the molecular underpinnings of VEGF-induced proliferation in ECFCs, the only EPC subset committed to acquire a truly mature endothelial phenotype, isolated from UCB. We found that the VEGF activates UCB-ECFCs by inducing oscillatory changes in [Ca2 + ]i which are not synchronized between adjacent cells and are shaped by the concerted interaction between InsP3dependent Ca2 + release and SOCE activation. These features are similar to those described in their peripheral counterparts, but a major difference is represented by the absolute requirement for TRPC3-mediated Ca2 + entry to initiate and sustain the spiking response. It has long been known that EPCs are more sensitive to VEGF stimulation when harvested from UCB rather than from PB. This property also applies to ECFCs; as a consequence, UCB-ECFCs appear more amenable for the purposes of cell-based therapy of ischemic diseases than PB-derived cells. The selective contribution of TRPC3 to VEGF-induced Ca2 + signals in UCBECFCs hints at this channel as a potential candidate to achieve a genetic improvement of PB-ECFC functionality by virtue of its selective coupling to a number of Ca2 + -sensitive signaling pathways involved in endothelial proliferation, migration, and differentiation.

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS

Author Disclosure statement The authors declare no conflict of interest.

References 1. Moccia F, F Lodola, S Dragoni, E Bonetti, C Bottino, G Guerra, U Laforezan, V Rosti and F Tanzi. (2012). Ca2 + signalling in endothelial progenitor cells: a novel means to improve cell-based therapy and impair tumour vascularisation. Curr Vasc Pharmacol, under press. 2. Moccia F, E Bonetti, S Dragoni, J Fontana, F Lodola, R Berra Romani, U Laforenza, V Rosti and F Tanzi. (2012). Hematopoietic progenitor and stem cells circulate by surfing on intracellular Ca2 + waves: a novel target for cell-based therapy and anti-cancer treatment? Curr Signal Transd Ther 7:161–176. 3. Moccia F, S Dragoni, F Lodola, E Bonetti, C Bottino, G Guerra, U Laforenza, V Rosti and F Tanzi. (2012). Storedependent Ca2 + entry in endothelial progenitor cells as a perspective tool to enhance cell-based therapy and adverse tumour vascularization. Curr Med Chem 19:5802–2818. 4. Yoder MC. (2010). Defining human endothelial progenitor cells. Exp Hematol 38:S111–S111. 5. Critser PJ and MC Yoder. (2010). Endothelial colonyforming cell role in neoangiogenesis and tissue repair. Curr Opin Organ Transplant 15:68–72. 6. Yoder MC. (2012). Human endothelial progenitor cells. Cold Spring Harb Perspect Med 2:a006692. 7. Yoder M and S Rounds. (2011). Bad blood, bad endothelium: ill fate? Blood 117:3479–3480. 8. Moccia F, JE Avelino-Cruz, Y Sa´nchez-Herna´ndez and F Tanzi. (2010). Ca2 + signalling in damaged endothelium: do connexin hemichannels aid in filling the gap? Curr Drug Ther 5:277–287. 9. Asahara T. (2007). Cell therapy and gene therapy using endothelial progenitor cells for vascular regeneration. Handb Exp Pharmacol 180:181–194. 10. Pompilio G, MC Capogrossi, M Pesce, F Alamanni, C DiCampli, F Achilli, A Germani and P Biglioli. (2009). Endothelial progenitor cells and cardiovascular homeostasis: clinical implications. Int J Cardiol 131:156–167. 11. Germani A, C Di Campli, G Pompilio, P Biglioli and MC Capogrossi. (2009). Regenerative therapy in peripheral artery disease. Cardiovasc Ther 27:289–304. 12. Passier R, LW van Laake and CL Mummery. (2008). Stemcell-based therapy and lessons from the heart. Nature 453: 322–329. 13. Hirschi KK, DA Ingram and MC Yoder. (2008). Assessing identity, phenotype, and fate of endothelial progenitor cells. Arterioscler Thromb Vasc Biol 28:1484–1595. 14. Ingram DA, LE Mead, H Tanaka, V Meade, A Fenoglio, K Mortell, K Pollok, MJ Ferkowicz, D Gilley and MC Yoder. (2004). Identification of a novel hierarchy of endothelial progenitor cells using human peripheral and umbilical cord blood. Blood 104:2752–2760. 15. Mead LE, D Prater, MC Yoder and DA Ingram. (2008). Isolation and characterization of endothelial progenitor cells from human blood. Curr Protoc Stem Cell Biol Chapter 2:Unit 2C.1. 16. Melero-Martin JM, ME De Obaldia, S-Y Kang, ZA Khan, L Yuan, P Oettgen and J Bischoff. (2008). Engineering robust and functional vascular networks in vivo with human adult and cord blood-derived progenitor cells. Circ Res 103:194–202.

2577

17. Melero-Martin JM, ZA Khan, A Picard, X Wu, S Paruchuri and J Bischoff. (2007). In vivo vasculogenic potential of human blood-derived endothelial progenitor cells. Blood 109:4761–4768. 18. Ben-Shoshan J and J George. (2007). Endothelial progenitor cells as therapeutic vectors in cardiovascular disorders: from experimental models to human trials. Pharmacol Ther 115:25–36. 19. Munaron L and AF Pla. (2009). Endothelial calcium machinery and angiogenesis: understanding physiology to interfere with pathology. Curr Med Chem 16:4691–4703. 20. Pla AF, D Avanzato, L Munaron and IS Ambudkar. (2012). Ion channels and transporters in cancer. 6. Vascularizing the tumor: TRP channels as molecular targets. Am J Physiol Cell Physiol 302:C9–C15. 21. Moccia F, R Berra-Romani and F Tanzi. (2012). Update on vascular endothelial Ca2 + signalling: a tale of ion channels, pumps and transporters. World J Biol Chem 3:127–158. 22. Lodola F, U Laforenza, E Bonetti, D Lim, S Dragoni, C Bottino, HL Ong, G Guerra, C Ganini, et al. (2012). Storeoperated Ca2 + entry is remodelled and controls in vitro angiogenesis in endothelial progenitor cells isolated from tumoral patients. PLoS One 7:e42541. 23. Sanchez-Hernandez Y, U Laforenza, E Bonetti, J Fontana, S Dragoni, M Russo, JE Avelino-Cruz, S Schinelli, D Testa, et al. (2010). Store-operated Ca2 + entry is expressed in human endothelial progenitor cells. Stem Cells Dev 19: 1967–1981. 24. Li J, RM Cubbon, LA Wilson, MS Amer, L McKeown, B Hou, Y Majeed, S Tumova, VAL Seymour, et al. (2011). Orai1 and CRAC channel dependence of VEGF-activated Ca2 + entry and endothelial tube formation. Circ Res 108: 1190–1198. 25. Dragoni S, U Laforenza, E Bonetti, F Lodola, C Bottino, R Berra-Romani, GC Bongio, MP Cinelli, G Guerra, et al. (2011). Vascular endothelial growth factor stimulates endothelial colony forming cells proliferation and tubulogenesis by inducing oscillations in intracellular Ca2 + concentration. Stem Cells 29:1898–1907. 26. Koch S and L Claesson-Welsh. (2012). Signal transduction by vascular endothelial growth factor receptors. Cold Spring Harb Perspect Med 2:a006502. 27. Salido GM, SO Sage and JA Rosado. (2009). Biochemical and functional properties of the store-operated Ca2 + channels. Cell Signal 21:457–461. 28. Jardin I, JJ Lopez, GM Salido and JA Rosado. (2008). Orai1 mediates the interaction between STIM1 and hTRPC1 and regulates the mode of activation of hTRPC1-forming Ca2 + channels. J Biol Chem 283:25296–25304. 29. Ong HL and IS Ambudkar. (2011). The dynamic complexity of the TRPC1 channelosome. Channels 5:424–431. 30. Parekh AB. (2011). Decoding cytosolic Ca2 + oscillations. Trends Biochem Sci 36:78–87. 31. Parekh AB. (2010). Store-operated CRAC channels: function in health and disease. Nat Rev Drug Discov 9:399–410. 32. Cheng HW, AF James, RR Foster, JC Hancox and DO Bates. (2006). VEGF activates receptor-operated cation channels in human microvascular endothelial cells. Arterioscler Thromb Vasc Biol 26:1768–1776. 33. Hamdollah Zadeh MA, CA Glass, A Magnussen, JC Hancox and DO Bates. (2008). VEGF-mediated elevated intracellular calcium and angiogenesis in human microvascular endothelial cells in vitro are inhibited by dominant negative TRPC6. Microcirculation 15:605–614.

2578 34. Ge RL, YL Tai, YY Sun, KC Zhou, SL Yang, TL Cheng, QF Zou, F Shen and YZ Wang. (2009). Critical role of TRPC6 channels in VEGF-mediated angiogenesis. Cancer Lett 283:43–51. 35. Kiyonaka S, K Kato, M Nishida, K Mio, T Numaga, Y Sawaguchi, T Yoshida, M Wakamori, E Mori, et al. (2009). Selective and direct inhibition of TRPC3 channels underlies biological activities of a pyrazole compound. Proc Natl Acad Sci U S A 106:5400–5405. 36. Poteser M, H Schleifer, M Lichtenegger, M Schernthaner, T Stockner, CO Kappe, TN Glasnov, C Romanin and K Groschner. (2011). PKC-dependent coupling of calcium permeation through transient receptor potential canonical 3 (TRPC3) to calcineurin signaling in HL-1 myocytes. Proc Natl Acad Sci U S A 108:13876–13878. 37. Smedlund K, J-Y Tano and G Vazquez. (2010). The constitutive function of native TRPC3 channels modulates vascular cell adhesion molecule-1 expression in coronary endothelial cells through nuclear factor kappa B signaling. Circ Res 106:1479–1488. 38. Numaga T, M Nishida, S Kiyonaka, K Kato, M Katano, E Mori, T Kurosaki, R Inoue, M Hikida, JW Putney, Jr. and Y Mori. (2010). Ca2 + influx and protein scaffolding via TRPC3 sustain PKC beta and ERK activation in B cells. J Cell Sci 123:927–938. 39. Moccia F, R Berra-Romani, S Tritto, S Signorelli, V Taglietti and F Tanzi. (2003). Epidermal growth factor induces intracellular Ca2 + oscillations in microvascular endothelial cells. J Cell Physiol 194:139–150. 40. Foreman MA, J Smith and SJ Publicover. (2006). Characterisation of serum-induced intracellular Ca2 + oscillations in primary bone marrow stromal cells. J Cell Physiol 206:664–671. 41. Lipskaia L and AM Lompre. (2004). Alteration in temporal kinetics of Ca2 + signaling and control of growth and proliferation. Biol Cell 96:55–68. 42. Tovey SC, P de Smet, P Lipp, D Thomas, KW Young, L Missiaen, H De Smedt, JB Parys, MJ Berridge, J Thuring, A Holmes and MD Bootman. (2001). Calcium puffs are generic InsP(3)-activated elementary calcium signals and are downregulated by prolonged hormonal stimulation to inhibit cellular calcium responses. J Cell Sci 114:3979–3989. 43. Bootman MD, MJ Berridge and P Lipp. (1997). Cooking with calcium: the recipes for composing global signals from elementary events. Cell 91:367–373. 44. Dupont G, L Combettes, GS Bird and JW Putney. (2011). Calcium oscillations. Cold Spring Harb Perspect Biol 3:a004226. 45. Skupin A, H Kettenmann, U Winkler, M Wartenberg, H Sauer, SC Tovey, CW Taylor and M Falcke. (2008). How does intracellular Ca2 + oscillate: by chance or by the clock? Biophys J 94:2404–2411. 46. Skupin A and M Falcke. (2007). Statistical properties and information content of calcium oscillations. Genome Inform 18:44–53. 47. Berridge MJ. (2007). Inositol trisphosphate and calcium oscillations. In: Cell Biology of Inositol Lipids and Phosphates. Wakelam MJO, ed. Portland Press Ltd, London. pp 1–7. 48. Berridge MJ. (2009). Inositol trisphosphate and calcium signalling mechanisms. Biochim Biophys Acta 1793:933– 940. 49. Banumathi E, A O’Connor, S Gurunathan, DA Simpson, JG McGeown and TM Curtis. (2011). VEGF-induced retinal angiogenic signaling is critically dependent on Ca2 + signaling

DRAGONI ET AL.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

by Ca2 + /Calmodulin-dependent protein kinase II. Invest Ophthalm Vis Sci 52:3103–3111. Berra-Romani R, A Raqeeb, JE Avelino-Cruz, F Moccia, A Oldani, F Speroni, V Taglietti and F Tanzi. (2008). Ca2 + signaling in injured in situ endothelium of rat aorta. Cell Calcium 44:298–309. Romani RB, A Raqeeb, U Laforenza, MF Scaffino, F Moccia, JE Avelino-Cruz, A Oldani, D Coltrini, V Milesi, V Taglietti and F Tanzi. (2009). Cardiac microvascular endothelial cells express a functional Ca2 + -sensing receptor. J Vasc Res 46:73–82. Moccia F, GA Nusco, D Lim, K Kyozuka and L Santella. (2006). NAADP and InsP3 play distinct roles at fertilization in starfish oocytes. Dev Biol 294:24–38. Harteneck C and M Gollasch. (2011). Pharmacological modulation of diacylglycerol-sensitive TRPC3/6/7 channels. Curr Pharm Biotechnol 12:35–41. Pupo E, AF Pla, D Avanzato, F Moccia, J-EA Cruz, F Tanzi, A Merlino, D Mancardi and L Munaron. (2011). Hydrogen sulfide promotes calcium signals and migration in tumorderived endothelial cells. Free Rad Biol Med 51:1765–1773. Poteser M, A Graziani, P Eder, A Yates, H Maechler, C Romanin and K Groschner. (2008). Identification of a rare subset of adipose tissue-resident progenitor cells, which express CD133 and TRPC3 as a VEGF-regulated Ca2 + entry channel. FEBS Lett 582:2696–2702. Kwan H-Y, K-T Cheng, Y Ma, Y Huang, NL Tang, S Yu and X Yao. (2010). CNGA2 contributes to ATP-induced noncapacitative Ca2 + influx in vascular endothelial cells. J Vasc Res 47:148–156. Leung P-C, K-T Cheng, C Liu, W-T Cheung, H-Y Kwan, K-L Lau, Y Huang and X Yao. (2006). Mechanism of noncapacitative Ca2 + influx in response to bradykinin in vascular endothelial cells. J Vasc Res 43:367–376. Berridge MJ, MD Bootman and HL Roderick. (2003). Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Cell Biol 4:517–529. Grimaldi M, M Maratos and A Verma. (2003). Transient receptor potential channel activation causes a novel form of Ca2 + i oscillations and is not involved in capacitative Ca2 + entry in glial cells. J Neurosci 23:4737–4745. Smedlund K and G Vazquez. (2008). Involvement of native TRPC3 proteins in ATP-dependent expression of VCAM-1 and monocyte adherence in coronary artery endothelial cells. Arterioscler Thromb Vasc Biol 28:2049–2055. Vazquez G, J-Y Tano and K Smedlund. (2010). On the potential role of source and species of diacylglycerol in phospholipase-dependent regulation of TRPC3 channels. Channels 4:232–240. Smedlund K, M Bah and G Vazquez. (2012). On the role of endothelial TRPC3 channels in endothelial dysfunction and cardiovascular disease. Cardiovasc Hematol Agents Med Chem 10:265–274. He LP, T Hewavitharana, J Soboloff, MA Spassova and DL Gill. (2005). A functional link between store-operated and TRPC channels revealed by the 3,5-bis(trifluoromethyl)pyrazole derivative, BTP2. J Biol Chem 280:10997–11006. Foster RR, GI Welsh, SC Satchell, RD Marlow, MD Wherlock, D Pons, PW Mathieson, DO Bates and MA Saleem. (2010). Functional distinctions in cytosolic calcium regulation between cells of the glomerular filtration barrier. Cell Calcium 48:44–53. Jiang H, B Zeng, G-L Chen, D Bot, S Eastmond, SE Elsenussi, SL Atkin, AN Boa and S-Z Xu. (2012). Effect of

VEGF-INDUCED CA21 OSCILLATIONS IN CORD BLOOD-DERIVED EPCS

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76. 77.

78. 79.

80.

81.

non-steroidal anti-inflammatory drugs and new fenamate analogues on TRPC4 and TRPC5 channels. Biochem Pharmacol 83:923–931. Schleifer H, B Doleschal, M Lichtenegger, R Oppenrieder, I Derler, I Frischauf, TN Glasnov, CO Kappe, C Romanin and K Groschner. (2012). Novel pyrazole compounds for pharmacological discrimination between receptor-operated and store-operated Ca2 + entry pathways. Brit J Pharmacol 167:1712–1722. Critser PJ, SL Voytik-Harbin and MC Yoder. (2011). Isolating and defining cells to engineer human blood vessels. Cell Prolif 44:15–21. Kalka C and I Baumgartner. (2008). Gene and stem cell therapy in peripheral arterial occlusive disease. Vasc Med 13:157–172. Penn MS and AA Mangi. (2008). Genetic enhancement of stem cell engraftment, survival, and efficacy. Circ Res 102: 1471–1482. Sydorenko V, Y Shuba, S Thebault, M Roudbaraki, G Lepage, N Prevarskaya and R Skryma. (2003). Receptor-coupled, DAG-gated Ca2 + -permeable cationic channels in LNCaP human prostate cancer epithelial cells. J Physiol 548:823–836. Antigny F, N Girardin and M Frieden. (2012). Transient receptor potential canonical channels are required for in vitro endothelial tube formation. J Biol Chem 287:5917– 5927. Nishida M, K Sugimoto, Y Hara, E Mori, T Morii, T Kurosaki and Y Mori. (2003). Amplification of receptor signalling by Ca2 + entry-mediated translocation and activation of PLC gamma 2 in B lymphocytes. EMBO J 22: 4677–4688. Nakao K, H Shirakawa, A Sugishita, I Matsutani, T Niidome, T Nakagawa and S Kaneko. (2008). Ca2 + mobilization mediated by transient receptor potential canonical 3 is associated with thrombin-induced morphological changes in 1321N1 human astrocytoma cells. J Neurosci Res 86: 2722–2732. Cohen R, A Torres, H-T Ma, D Holowka and B Baird. (2009). Ca2 + waves initiate antigen-stimulated Ca2 + responses in mast cells. J Immunol 183:6478–6488. Shlykov SG and BA Sanborn. (2004). Stimulation of intracellular Ca2 + oscillations by diacylglycerol in human myometrial cells. Cell Calcium 36:157–164. Beech DJ. (2012). Orai1 calcium channels in the vasculature. Pflugers Arch 463:635–647. Taylor CW and SC Tovey. (2010). IP(3) receptors: toward understanding their activation. Cold Spring Harb Perspect Biol 2:a004010. Parekh AB and JW Putney. (2005). Store-operated calcium channels. Physiol Rev 85:757–810. Cheng KT, X Liu, HL Ong, W Swaim and IS Ambudkar. (2011). Local Ca2 + entry via Orai1 regulates plasma membrane recruitment of TRPC1 and controls cytosolic Ca2 + signals required for specific cell functions. PLos Biol 9: e1001025. Bootman MD, KW Young, JM Young, RB Moreton and MJ Berridge. (1996). Extracellular calcium concentration controls the frequency of intracellular calcium spiking independently of inositol 1,4,5-trisphosphate production in HeLa cells. Biochem J 314:347–354. Shuttleworth TJ and JL Thompson. (1996). Ca2 + entry modulates oscillation frequency by triggering Ca2 + release. Biochem J 313:815–819.

2579

82. Bandyopadhyay BC, HL Ong, TP Lockwich, X Liu, BC Paria, BB Singh and IS Ambudkar. (2008). TRPC3 controls agonist-stimulated intracellular Ca2 + release by mediating the interaction between inositol 1,4,5-trisphosphate receptor and RACK1. J Biol Chem 283:32821–32830. 83. Woodard GE, JJ Lopez, I Jardin, GM Salido and JA Rosado. (2010). TRPC3 regulates agonist-stimulated Ca2 + mobilization by mediating the interaction between type I inositol 1,4,5-trisphosphate receptor, RACK1, and Orai1. J Biol Chem 285:8045–8053. 84. Mi LY, DS Ettenson and ER Edelman. (2008). Phospholipase C-delta extends intercellular signalling range and responses to injury-released growth factors in non-excitable cells. Cell Prolif 41:671–690. 85. Benedito R, C Roca, I Sorensen, S Adams, A Gossler, M Fruttiger and RH Adams. (2009). The Notch ligands Dll4 and Jagged1 have opposing effects on angiogenesis. Cell 137:1124–1135. 86. Akeson A, A Herman, D Wiginton and J Greenberg. (2010). Endothelial cell activation in a VEGF-A gradient: relevance to cell fate decisions. Microvasc Res 80:65–74. 87. Cheyette TE and DJ Gross. (1991). Epidermal growth factor-stimulated calcium ion transients in individual A431 cells: initiation kinetics and ligand concentration dependence. Cell Regul 2:827–840. 88. Mac Gabhann F, MT Yang and AS Popel. (2005). Monte Carlo simulations of VEGF binding to cell surface receptors in vitro. Biochim Biophys Acta 1746:95–107. 89. Nakayama M, A Nakayama, M van Lessen, H Yamamoto, S Hoffmann, HCA Drexler, N Itoh, T Hirose, G Breier, et al. (2013). Spatial regulation of VEGF receptor endocytosis in angiogenesis. Nat Cell Biol 15:249–260. 90. Kappas NC, G Zeng, JC Chappell, JB Kearney, S Hazarika, KG Kallianos, C Patterson, BH Annex and VL Bautch. (2008). The VEGF receptor Flt-1 spatially modulates Flk-1 signaling and blood vessel branching. J Cell Biol 181:847– 858. 91. Thurley K, A Skupin, R Thul and M Falcke. (2012). Fundamental properties of Ca2 + signals. Biochim Biophys Acta 1820:1185–1194. 92. Skupin A, H Kettenmann and M Falcke. (2010). Calcium signals driven by single channel noise. PLoS Comput Biol 6:e1000870. 93. Kraus M and B Wolf. (1996). Crosstalk between cellular morphology and calcium oscillation patterns—insights from a stochastic computer model. Cell Calcium 19:461– 472. 94. Faehling M, J Kroll, KJ Fohr, G Fellbrich, U Mayr, G Trischler and J Waltenberger. (2002). Essential role of calcium in vascular endothelial growth factor A-induced signaling: mechanism of the antiangiogenic effect of carboxyamidotriazole. FASEB J 16:1805–1807. 95. Bhattacharya R, J Kwon, X Li, E Wang, S Patra, JP Bida, Z Bajzer, L Claesson-Welsh and D Mukhopadhyay. (2009). Distinct role of PLC beta 3 in VEGF-mediated directional migration and vascular sprouting. J Cell Sci 122:1025–1034. 96. McLaughlin AP and GW De Vries. (2001). Role of PLC gamma and Ca2 + in VEGF- and FGF-induced choroidal endothelial cell proliferation. Am J Physiol Cell Physiol 281:C1448–C1456. 97. Wilkerson MK, TJ Heppner, AD Bonev and MT Nelson. (2006). Inositol trisphosphate receptor calcium release is required for cerebral artery smooth muscle cell proliferation. Am J Physiol Heart Circ Physiol 290:H240–H247.

2580 98. Shibao K, MJ Fiedler, J Nagata, N Minagawa, K Hirata, Y Nakayama, Y Iwakiri, MH Nathanson and K Yamaguchi. (2010). The type III inositol 1,4,5-trisphosphate receptor is associated with aggressiveness of colorectal carcinoma. Cell Calcium 48:315–323. 99. Szatkowski C, JB Parys, H Ouadid-Ahidouch and F Matifat. (2010). Inositol 1,4,5-trisphosphate-induced Ca2 + signalling is involved in estradiol-induced breast cancer epithelial cell growth. Mol Cancer 9:156. 100. Hofer E and B Schweighofer. (2007). Signal transduction induced in endothelial cells by growth factor receptors involved in angiogenesis. Thromb Haemost 97:355–363. 101. Mancini M and A Toker. (2009). NFAT proteins: emerging roles in cancer progression. Nat Rev Cancer 9:810–820. 102. Brenner JS and RE Dolmetsch. (2007). TrpC3 regulates hypertrophy-associated gene expression without affecting myocyte beating or cell size. PLoS One 2:e802.

DRAGONI ET AL. 103. Eggermann E, I Bucurenciu, SP Goswami and P Jonas. (2012). Nanodomain coupling between Ca2 + channels and sensors of exocytosis at fast mammalian synapses. Nat Rev Neurosci 13:7–21.

Address correspondence to: Dr. Francesco Moccia Department of Biology and Biotechnology ‘‘Lazzaro Spallanzani’’ University of Pavia Via Forlanini 6, 27100 Pavia Italy E-mail: [email protected] Received for publication January 15, 2013 Accepted after revision May 17, 2013 Prepublished on Liebert Instant Online May 17, 2013