E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
ava i l ab l e a t www.sc i enced i r ec t . com
www.e l sev i e r. com/ loca te /yexc r
Research Article
The matrix protein CCN1 (CYR61) promotes proliferation,migration and tube formation of endothelial progenitor cells
Yang Yua,1, Yu Gaob,1, Hong Wanga, Lan Huanga,⁎, Jun Qina, Ruiwei Guoa, Mingbao Songa,Shiyong Yua, Jianfei Chena, Bin Cuia, Pan Gaoa
aInstitute of Cardiovascular Diseases of PLA, Xinqiao Hospital, Third Military Medical University, Chongqing 400037, People's Republic of ChinabDepartment of Rehabilitation, Southwest Hospital, Third Military Medical University, Chongqing 400038, People's Republic of China
Revised version received5 August 2008Accepted 6 August 2008Available online 13 August 2008
Neovascularization and re-endothelialization relies on circulating endothelial progenitor cells(EPCs), but their recruitment and angiogenic roles are subjected to regulation by the vascular
microenvironment, which remains largely unknown. The present study was designed toinvestigate the effects of mature ECs and matrix protein CCN1 on the properties of EPCs. In acoculture system, effects of ECs on proliferation, migration and participation in tube-like formationof EPCs were evaluated, and functional assays were employed to identify the exact role of CCN1 inEPCs vitality and function. We demonstrated that ECs, as an indispensable part of the cellularmilieu, significantly promoted the proliferation, migration and tube formation activities of EPCs,and more importantly, CCN1 was potentially involved in such effects of ECs. Expression of CCN1 inEPCs was significantly increased by serum, VEGF, ECs-cocultivation and ECs conditioned medium.Moreover, Ad-CCN1-mediated overexpression of CCN1 directly enhanced migration and tubeformation of EPCs, whereas silencing of endogenous CCN1 in EPCs inhibits cell functions.Furthermore, CCN1 induced the expressions of chemokines and growth factors, such as MCP-1 and
VEGF, suggesting a complex interaction between those proangiogenic factors. Our data suggestthat matrix protein CCN1 may play an important role in microenvironment-mediated biologicalproperties of EPCs.
A growing body of evidence indicates that bone marrow-derivedendothelial progenitor cells (EPCs) are mobilized, recruited andprominently contributed to neovascularization and re-endotheliali-zation after vascular injury, in contrast to conventional assumptionthat postnatal neovascularization is attributed to the migration andproliferation of preexisting mature endothelial cells (ECs) [1–4]. Inparticular, EPCs not only directly incorporate into blood vessels,replacing the defective or injured mature ECs, but also secrete a
1.).
r Inc. All rights reserved.
variety of cytoprotective or proangiogenic factors in a paracrinemanner to promote the survival and proliferation of ECs. However,up to now, investigations have emphasized a strong impact ofmicroenvironment on the biological properties of EPCs [5–7],though the matrix microenvironment and interactions betweenlocal ECs and EPCs that may control EPCs proliferation and functionsremain poorly understood.
CCN1 is a secreted matrix protein belonging to the emergingCCN family, which also includes CCN2/CTGF, CCN3/Nov, CCN4/WISP-1, CCN5/WISP-2 and CCN6/WISP-3 [8]. CCN1 is expressed by
3199E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
all types of vascular cells, and has been implicated in diversecellular processes such as adhesion, migration, proliferation andsurvival [9,10]. Moreover, noticeably, a potential role of CCN1 inangiogenesis and vascularization has been demonstrated by accu-mulating evidence. Targeted knockout of CCN1 gene in miceresults in embryonic lethality due to placental vascular insuffi-ciency and compromised vessel integrity [11]. Besides its ability toinduce tumor angiogenesis, CCN1 is able to stimulate neovascu-larization in rat cornea [12] as well as rabbit ischemic hindlimb[13]. In addition, CCN1 can be rapidly induced by vascularendothelial cell growth factor (VEGF), and promote proliferation,migration and tube formation of ECs in vitro [14]. However,whether and what impact CCN1 may play on EPCs is largelyunknown at present. Interestingly, more recent observationsshowed that recombinant CCN1 and supernatants from CCN1-stimulated CD34+ cells effectively promoted proliferation of ECs[15], and exogenous CCN1 could promote adhesion and migrationof human CD34+ progenitor cells and mesenchymal stem cells[16], suggesting a potential role of CCN1 in matrix microenviron-ment which may affect properties of EPCs.
In this study, we found that expression levels of CCN1 wereincreased in EPCs and ECs after stimulation with serum andVEGF. Our co-culture experiments revealed that surrounding ECswas involved in the regulation of proliferation, migration andtube formation of EPCs, and what is more, CCN1 was signifi-cantly contributed to the paracrine effects of ECs on EPCs. Forcedexpressed exogenous CCN1 was shown to induce secretion ofVEGF and MCP-1, and enhanced proliferation, migration and tubeformation of EPCs. RNA interference (RNAi)-mediated knock-down of CCN1 expression diminished migration and tube forma-tion of EPCs. These findings suggest that matrix protein CCN1promotes proliferation, recruitment and function of EPCs, andmay play an important role in microenvironment-mediatedbiological properties of EPCs.
Materials and methods
Reagents
All chemicals were purchased from Sigma-Aldrich (St. Louis,MO) unless otherwise specified. All cell culture plates wereobtained from Costar, whereas culture medium and serum wereobtained from Gibco (Grand Island, NY, USA). FITC-conjugatedantibodies against rat CD133, rat CD34, rat VEGFR-2 andcorresponding isotype control IgG were from Bios (Beijing,China), while antibodies against rat CD44 and rat CD45 werefrom BD (BD Biosciences, San Jose, CA). Rabbit polyclonalantibody against rat CCN1 and blocking antibody were fromSanta Cruz Biotechnology, Inc. (Santa Cruz, CA). Recombinanthuman VEGF was from R&D Systems, Inc. (Minneapolis, MN)and Dil-AcLDL was from Biomedical Technologies, Inc.(Stoughton, MA).
Recombinant adenoviral vectors expressing CCN1
To evaluate the role of CCN1, adenovirus vector expressing CCN1was generated using the AdEasy system. Briefly, full-length ratCCN1 cDNA was generated by RT-PCR using total RNA fromSprague–Dawley (SD) rat heart and the following primers: sense
5'-aggagatctatgagctccagcaccatcaag-3'and antisense 5'-gctaagctt-ctttagtccc tgaacttgtgg-3' (nucleotides 186 to 1325, GenBank acces-sion number NM_031327). The CCN1 cDNA was first TA-clonedinto pMD19-T simple vector and then subcloned into pAdTrack-CMV, resulting in pAdTrack-CCN1. The shuttle vector was used togenerate recombinant adenovirus Ad-CCN1 according to themanufacturer's protocol. All PCR-amplified fragments and cloningjunctions were verified by DNA sequencing (Sangon, Shanghai,China). An adenovirus encoding green fluorescent protein (GFP;Ad-GFP) was used as control. All adenoviruses were replicationdeficient and used at 20 multiplicity of infection (mois) for 24 hwithout apparent cytotoxicity.
RNAi-mediated silencing of CCN1 expression
The target sequences were selected by the web-based small inter-fering RNA (siRNA) hairpin engine at www.genscript.com andwww.ambion.com, which showed no homology to any othersequences by a blast search. The synthesized template oligonu-cleotides consisting of sense target sequence, antisense targetsequence, loop structure and a transcription stop signal were thenannealed and subcloned, respectively, into shRNA expressionvector pGenesil1-U6 (Genesil, Wuhai, China) to obtain the finalconstruct pGenesil1-CCN1. According to the preliminary experi-ment on rat renal fibroblasts (line NRK) about CCN1 RNAiefficiency, following sequences were used in this study: CCN1siRNA,5'-gatccgcaactcaacgaggactgcttcaagacggcagtcctcgtt-gagttgcttttttgtcgaca-3' , negative control siRNA 5'-gatccgacttca-taaggcgcatgcttcaagacggcatgcgccttatgaagtcttttttgtcgaca-3' andGAPDH-A siRNA as positive control 5'-gatccgtggatattggttgccatcatt-caagacgtgatggcaacaatatccacttttttgtcgaca-3'. Transfection of thesiRNA was carried out using Lipofectamine 2000 reagent with amolar ratio about DNA: lipid= 1:3 (Invitrogen, CA, USA). 24 h aftertransfection, cells were collected and used for functional assays(see below).
EPCs isolation and characterization
All animal procedures have been approved by the Care of Experi-mental Animals Committee of the Third Military Medical Uni-versity. Culture and characterization of EPCs was performedpreviously in our laboratory [17]. BM was harvested by flushingthe femurs and tibias of SD rats (male, 180 to 220 g, Chongqing).BM-derived mononuclear cells were isolated by density gradientcentrifugation (Lymphoprep 1.083, Tianjing, China) at 400×g for20 min. After purification with three washing steps, cells wereresuspended in low glucose DMEM supplemented with 10% FCSand 10 ng/mL VEGF, plated on gelatin-coated cell culture flasksand incubated at 37 °C under 5% CO2. Twenty-four hours later,nonadherent cells were aspirated and transferred to a newgelatin-coated flask in order to remove rapidly adherent hemato-poietic cells and mature endothelial cells. Another 48 h later,nonattached cells were removed and adherent cells were con-tinuously cultured. Only those adherent cells were used in furtherexperiments.
To confirm the EPCs phenotype, cells were incubated withacLDL-Dil (10 mg/ml) for 4 h, fixed with 4% paraformaldehyde andthen incubated with FITC-labeled lectin (UEA-1, 10 mg/ml) for 1 h.Dual-stained cells positive for both acLDL-DiI and UEA-1 wereidentified as EPCs. Nearly all adherent cells (N95%) were double
3200 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
positive cells. Additionally, flow cytometry (FACS) analysis wasperformed using antibodies against rat CD45, CD44, CD133, CD34,VEGFR-2 and the corresponding isotype control antibodies.
Coculture of EPCs with endothelial cells (ECs)
Endothelial cells from rat thoracic aorta were isolated in parallelwith the endothelial progenitor cells by collagenase digestionusing the standard protocol as described previously [18]. Isolatedendothelial cells were cultivated in DMEM with 10% FCS and10 ng/mL VEGF.
For coculture experiment, passages 2-5 ECs transfected withAd-CCN1, Ad-GFP and non-transfected ECs, together with EPCswere pre-conditioned in serum/VEGF free medium for 24 h inorder to minimize CCN1 basal secretion stimulated by serum orcytokines. Then, EPCs were subcultured in 6-well plates and ECswere placed over the EPCs monolayers in 0.4 mm pores trans-well inserts, allowing the diffusion of soluble factors withoutdirect cell contact. Cells were cocultured in the presence orabsence of anti-CCN1 antibody in a total volume of 2 ml DMEMwith 10% FCS. In some experiments, the coculture medium wasreplaced with 2 ml conditioned medium collected from culti-vated ECs and the transwell inserts without ECs were set up. Ascontrols, ECs and EPCs were cultured alone as descripted forcoculture experiment. After 48 h, the cells and cell supernatantswere collected for further analysis. All experiments were per-formed in triplicate.
Cell proliferation assay
EPCs were harvested from the cultures and replaced into 96-wellplate(2×106 cells/mL) in triplicates. Cell proliferation was mea-sured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) assays according to the protocol of the manu-facturer. Prior to the optical density (490 nm) reading, 15 μL MTTsolution and 200 μL DMSO were sequencely added to each well.All groups of experiments were performed in triplicate.
Cell migration assay
EPCs migration was evaluated using a modified Boyden's chamberassay. EPCs (1×105) in 100 μL serum-free DMEMwere placed in theupper chamber. The lower chamber contained either ECs suspen-sions or ECs conditioned medium. For controls, DMEM plus 10%FCS supplemented with or without 100 ng/mL SDF-1 were filled inthe lower chamber as positive and negative controls, respectively.To test the effects of CCN1 on cell migration, EPCs were transfectedwith either Ad-CCN1 or siRNA-CCN1, and anti-CCN1 antibody wasadded. After 6 h in culture, cells on the upper side of the 8-μmfilters were removed and the filters were washed with PBS. Then,
Fig. 1 – CCN1 expression is up-regulated in EPCs and ECs. (a) FACSDMEM supplemented with 10% FCS and 10 ng/mL VEGF. Cells wereCD44, and CD45, showed as black areas. The white area on each boxdenotes positive gate. Numbers are the percentage of positive cells. (serum, VEGF, ECs-cocultivation, and ECs supernatants (SN) usingWepresented as the percentage of control (serum and VEGF free, null trand VEGF (10 ng/mL). Representative images from semiquantitativethree independent experiments performed in triplicate. ⁎ p<0.05 c
the underside cells were fixed in methanol and stained withhematoxylin stain. Migration activity was evaluated as the meannumber of migrated cells in 5 random high power fields perchamber under a microscope (Leica, Germany).
Tube formation assay
Matrigel of the same batch was thawed and laid into 24-wellculture plates at 37 °C for 1 h to allow solidification. EPCstransfected with Ad-CCN1, Ad-GFP, or siRNA-CCN1 were har-vested, resuspended and placed on the matrigel, with non-transfected EPCs as a control. The effect of ECs was studied usinga coculture model. ECs seeded on inserts or ECs-conditionedmedium were added to the wells, in the presence or absence ofCCN1 blocking antibody. After incubation at 37 °C for 18 h, EPCstube formation was observed microscopically and the total lengthof such tube like structures was measured by Leica Qwin V3.1software.
RNA extraction and reverse transcriptase-PCR (RT-PCR)
Total RNA was extracted from EPCs and ECs by using TRIzol(Invitrogen), followed by cDNA synthesis using oligo (dT) and M-MLV reverse transcriptase (Takara). Then, cDNA amplification andsemi-quantitative PCR were performed using the followingprimers: CCN1: 5'gccgtcacccttctccacttg 3'(forward) and 5'-gcccc-ttggtgtggtcgcag-3'(reverse); VEGF: 5'-cctccgaaaccatgaactttctgctc-3'(forward) and 5'-cagcctggctcaccgccttggctt-3'(reverse); MCP-1:5'-ctcagccagatgcagtta-3'(forward) and 5'-tggaagggaatagtgtaat-3'(reverse); GAPDH: 5'-accacagtccatgccatcac-3'(forward) and 5'-tccaccaccctgttgctgta-3'(reverse). For quantitative RT-PCR analyses,the ABI PRISM 7000 Sequence Detection System (Applied Biosys-tems, Foster City, CA) and SYBR Green PCR Master Mix (Toyobo,Japan) were used with specific primers as follows: CCN1: the sameas above; VEGF: 5'-atcctggagcgttcactg-3'(forward) and 5'-tcaccgccttggcttgtc-3'(reverse); MCP-1: 5'-aacttgacccataaatctg-3'(forward) and 5'-tggaagggaatagtgtaat-3'(reverse); GAPDH: 5'-acccatcaccatcttccaggag-3'(forward) and 5'-gaaggggcggatattatgac-3'(reverse). All primers were synthesized by Invitrogen (Shanghai,China) and were of high purity salt-free quality.
Western blot analysis
Culture supernatants were collected and concentrated approxi-mately 20-fold with a centricon (Millipore, Bedford, MA). Proteinsfrom both cell supernatants and lysates were measured using theBradford method. The same amount of proteins were loaded ineach lane, separated by 10–15% SDS-PAGE, and transferred topolyvinylidene difluoride membranes. The membranes wereblocked with 5% nonfat milk, and then the membrane-bound
analysis of primary EPCs cultured for 4–7 days in low glucoselabeled with fluorescent antibodies to CD133, CD34, VEGFR-2,represents corresponding negative control labeling, and the lineb) Protein andmRNA levels of CCN1 in the EPCs stimulated withstern blot and real-time quantitative RT-PCR analysis. Values areeatment). (c) Influence on CCN1 levels of ECs by serum (10% FCS)RT-PCR and Western blots are shown. Data are mean±S.D. of
ompared with the control.
3201E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
proteins were probed with primary antibodies against CCN1 andGAPDH followed by secondary horseradish peroxidase-conjugatedantibodies. Protein bands were visualized by chemiluminescentdetection (ECL) (Amersham Biosciences) and quantified by a gelimage analysis system.
Enzyme-linked immunospecific assay (ELISA)
Protein levels of VEGF and MCP-1 in the cell supernatants weredetermined by ELISA kit (R&D Systems), according to the manu-facturer's instructions. Samples were measured in triplicate and
3202 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
were properly diluted to ensure that measured values were withinthe concentration range of the standard curve.
Statistical analysis
Data from independent experiments were expressed as mean ±S.D. of at least three experiments. Comparisons between groupswere analyzed by two-tailed Student's t test or ANOVA, asappropriate. p values<0.05 was considered statistically significant.
Results
CCN1 expression is up-regulated in EPCs and ECs
After 4–7 days of culture (typical culture period before cocultureand further experiments), adherent EPCs were characterized byimmunofluorescence and flow cytometry (FACS) analysis. Themajority of cells (N90%) stained positive for Dil-AcLDL and lectin,and expressed endothelial/stem cell markers, including CD34,VEGFR-2, and CD133, but not CD44 or CD45 (Fig. 1a). However, ECsonly express the marker of endothelial cells like vWF.
We thereafter investigated the expressions of CCN1 in primaryEPCs and ECs. Fig. 1b showed that CCN1 was present at fairly lowlevels in quiescent EPCs but was rapidly up-regulated upon sti-mulation with serum, either the mRNA by RT-PCR or the proteinby immunoblotting. In addition, coculture with ECs and with ECsconditioned medium also induced a significant increase of CCN1expression in EPCs. For comparison, VEGF, a strong growth factorof EPCs, did not show any predominant effect on the expression ofCCN1, comparable with that cocultured with ECs and ECs-con-ditioned medium. Similar results were obtained from ECs as wellas EPCs (Fig. 1c).
CCN1 is involved in effects of ECs on EPCs proliferation,migration and tube formation
We previously observed that coculture with ECs promoted mesen-chymal stem cells proliferation and differentiation, in a milieu-dependent manner [19]. To determine whether the mature ECscould affect EPCs in the same way, we used a previously describedcoculture system and the proliferation, migration and tube for-mation of EPCs were detected, respectively.
The proliferation of EPCs was significantly accelerated aftercoculture with ECs or ECs-conditioned medium compared withunstimulated groups, and was comparable with that of VEGF(Fig. 2a). In addition, no statistically significant differences werefound in the EPCs treated with ECs and ECs-conditioned medium.
Fig. 2 – Cocultivation with mature ECs and conditioned medium oEPCs, and CCN1 is required for such effects on EPCs by ECs. (a) MTT aof EPCs. Results represented mean of three independent experimenproliferation of each group compared with the control (low glucos⁎, p<0.05 compared with the control. (b) Migration of EPCs was an(ECs-SN) and VEGF(10 ng/mL) as stimulus in a modified Boyden cha(CCN1-Ab, 20 μg/mL) and control rabbit IgG (not shown). Data are mthe control. (c) EPCs tube formation, after cultured with mature ECcontrol IgG (not shown), is expressed as total length of tube like stanalyzed by Leica Qwin software. Values are the mean±S.D. of 27 dConcentrated supernatants from ECs and EPCs were subjected to W
Next, we investigated the ability of mature ECs to promote themigration of EPCs, using in vitro migration assay, with SDF-1 as apositive control. Fig. 2b shows a notable enhancement in migra-tion of EPCs when cocultured with ECs or supernatants fromcultured ECs. Lastly, matrigel angiogenesis assay was performedwith EPCs, and VEGF served as a positive control. The total lengthof tube-like structures after 18-h coculture were measured andcompared with controls. It is noteworthy that ECs and super-natants taken from ECs significantly increased tube formation ofEPCs (Fig. 2c). Taken together, all these results indicated thatmature ECs promoted EPCs proliferation, migration and angiogen-esis in vitro, as expected.
It is documented that the extracellular matrix (ECM) and cyto-kines plays important regulations in cells viability and functions.To probe the potential role of CCN1 in the regulation of EPCsproliferation, migration and angiogenesis, we first measured CCN1levels in the conditioned medium and then added neutralizingantibody against CCN1 in the coculture systems. As shown in Fig.2d, CCN1 was present at high levels in conditioned medium of ECsand EPCs. And interestingly, the proliferation, migration and tubeformation augmented by ECs-couple or ECs supernatants wereattenuated significantly in the presence of anti-CCN1 antibody(p<0.01), while no obvious changes were seen in the controlgroups (data with control IgG not shown). These results suggestedthat the effect of mature ECs on EPCs proliferation and functionswas at least partially through the matrix protein CCN1. Moreover,CCN1 might not be restricted to the ECM, but somehow acted asparacrine factors betweenmature and progenitor endothelial cells.
CCN1 induces VEGF and MCP-1 release of ECs and EPCs
In the present study, we observed that the matrix protein CCN1could be secreted by either ECs or EPCs, and that was important forthe paracrine effects of mature ECs on the proliferation andfunctions of EPCs. Furthermore, to explore which factors in res-ponse to CCN1 might be involved in the effects of ECs, weconstructed recombinant adenovirus Ad-CCN1 and transduced itinto ECs for detection the potential downstream factors. Using RT-PCR, we showed that the overexpression of CCN1 caused a 3.7-foldinduction of VEGFmRNA and 4.0-fold induction of MCP-1mRNA inECs compared with the Ad-GFP transfected ECs (Figs. 3a, c). BothVEGF and MCP-1 are important factors during vascular repair andangiogenesis processes, and specifically, this Ad-CCN1-mediatedinduction of VEGF mRNA and MCP-1 mRNA was inhibited by aCCN1-blocking antibody (p<0.05) whereas the control antibodywas not effective at all (data not shown). Then, using ELISAanalysis, we also confirmed the same expression pattern of VEGFand MCP-1 at the protein levels (Figs. 3b, d).
f ECs enhances proliferation, migration and tube formation ofssay was used to assess the effect of mature ECs on proliferationts performed in triplicates. Bar graph illustrates celle DMEM alone). Results are expressed as the mean±S.D.alyzed using ECs cocultivation, conditioned medium of ECsmber, in the absence or presence of CCN1 blocking antibodyean±S.D. of three experiments (n=15), ⁎, p<0.05 compared withs, ECs supernatants(SN), VEGF or, with or without anti-CCN1 orructures per high-power field (HPF) (magnification, ×200)eterminations (⁎p<0.05 vs control; #p value<0.05). (d)estern blot analysis using antibodies against CCN1 and GAPDH.
3203E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
Moreover, the presence of VEGF and MCP-1 in EPCs stimulatedwith ECs supernatant was measured. Supernatant from Ad-CCN1-transfected ECs significantly increased the expression of VEGF andMCP-1 at bothmRNA and protein levels in EPCs comparedwith thecontrol, which was reversed by the CCN1 antibody (p<0.05),indicating a CCN1-dependent mechanism for the paracrineinduction of MCP-1 and VEGF by EPCs.
Overexpression of CCN1 enhances migration and tubeformation of EPCs
Since the matrix protein CCN1 plays an evident role in ECs para-crine effects on EPCs proliferation and functions, it may exertdirect effects on EPCs. To further investigate the role of CCN1 inEPCs proliferation and functions, we constructed an adenoviral
Fig. 3 – CCN1 induces VEGF and MCP-1 release of ECs and EPCs. (a, c) RNA from ECs subjected to Ad-CCN1 transduction andEPCs stimulated with ECs supernatants (ECs-SN) were isolated, reverse transcribed and amplified by PCR using specific primersfor MCP-1, VEGF and GAPDH. Representative images of semi-quantitative PCR are shown in the top panel and the graphicalrepresentation of mRNA levels normalized to those of GAPDH are shown in the bottom panel of a and c. Quantitative evaluationis expressed as a percentage of control (in the absence of stimuli, serum free). (b, d) Supernatants from ECs transfected withAd-CCN1 and EPCs stimulated with ECs-SN were collected and subjected to ELISA analysis, using VEGF and MCP-1 specific kits.Expression of MCP-1 and VEGF was increased by Ad-CCN1 and supernatants of Ad-CCN1 transfected ECs, and the increasewas damped in the presence of 20 μg/mL CCN1 antibody (CCN1 Ab), whereas a control antibody was not effective at all (data notshown). Data are mean±S.D.(n=9), ⁎p value<0.05.
3204 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
vector that expressed CCN1 (i.e., Ad-CCN1) exogenously using theAdEasy system. Adenovirus-mediated CCN1 expression was con-firmed by Western blot analysis (data not shown). EPCs transfectedwith eitherAd-CCN1or control Ad-GFPwere subsequently subjectedto separate assays to examine their proliferation, migration andangiogenesis activities. As depicted in Fig. 4, overexpression ofexogenous CCN1 extensively improved themigration of EPCs, and infact, the average migrated cell number of the EPCs increased byapproximately 4-fold compared to that of the control cells. More
remarkably, the tube formation ability of EPCs was increased toabout 250% as compared with Ad-GFP transfected EPCs (p< 0.01).In contrast, the proliferation of EPCs was not enhanced, even aminor decrease was observed in cells transfected with Ad-GFP,though it has no significant meaning in compare with untrans-fected control. Nevertheless, the transfection efficacy in our studywas about 50–60% assessed by immunofluorescence, and ourfindings suggested that CCN1 played an important role inregulating the EPCs migration and angiogenesis.
3205E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
EPCs migration and tube formation is inhibited byshRNA-mediated knockdown of CCN1
Although the overexpression of exogenous CCN1 directlyenhanced EPCs migration and tube formation, the actual role ofendogenous CCN1 remained to be fully elucidated. Because thebasal expression of CCN1 in quiescent EPCs was barely detectable,and it was dramatically increased in response to stimulations suchas serum and VEGF, we therefore used a U6 promoter-drivenshRNA (i.e., pGenesil1-CCN1) to silence the CCN1 gene in EPCs,
and culture medium containing 20% FCS and 10 ng/mL VEGF wasused as stimulation. Introduction of pGenesil1-CCN1 was shownto specifically knockdown CCN1 expression as measured byWestern blot and RT-PCR (not shown). To determine whetherthe RNA interference-mediated silencing of CCN1 expressionwould affect serum+VEGF-induced EPCs proliferation, migrationand tube formation, we transduced the pGenesil1-CCN1 intoEPCs. However, introduction of the pGenesil1-CCN1 caused anapproximate 50% loss of CCN1 expression in EPCs. And signifi-cantly, the EPCs exhibited a decrease in cell proliferation,migration and tube formation compared to negative controlsiRNA-transfected cells (p< 0.01) (Fig. 4). The results were repro-ducible in at least three independent batches of experiments.Thus, these results indicated that knockdown of endogenousCCN1 significantly reduced the proliferation, migration and tubeformation of EPCs, suggesting an important role in EPCs of endo-genous CCN1.
Discussion
Accumulating evidence indicates that neovascularization does notexclusively rely on the migration and proliferation of local ECs, butalso includes the recruitment of bone marrow-derived circulatingstem cells. Interest focuses on EPCs which considerably contributeto postnatal neovascularization and re-endothelialization inresponse to vascular injury during physiological and pathologicalprocesses [2]. These EPCs can be recruited and incorporated intoinjured vessel wall and then promote local angiogenesis by directcellular differentiation and/or by releasing growth factors actingin a paracrine fashion [3]. However, it is documented that therecruitment and function of EPCs are tightly regulated by theneovascular microenvironment [5,7,20]. In particular, local vas-
Fig. 4 – Overexpression of CCN1 directly enhances migrationand tube formation of EPCs, whereas silencing of endogenousCCN1 in EPCs inhibits cell proliferation, migration, and tubeformation activities. (a) EPCs were transfected with orwithout Ad-GFP, Ad-CCN1, negative control siRNA orCCN1-siRNA, in the presence or absence of anti-CCN1 antibody(CCN1-Ab), and then followed by MTT assays. Three separateexperiments were done in triplicates. (b) EPCs migration inresponse to Ad-GFP, Ad-CCN1, negative control siRNA orCCN1-siRNA, in the presence or absence of anti-CCN1 antibodywas detected using the Boyden Chamber Model. EPCscultured in 10% low glucose DMEM without any treatmentwere used as control for overexpression experiment, while EPCscultured in low glucose DMEM supplemented with 20%FCS and10 ng/mL VEGF were used as control for RNA interferenceexperiment. (c) Tube formation by EPCs transfected with orwithout Ad-GFP, Ad-CCN1, negative control siRNA orCCN1-siRNA, in the presence or absence of anti-CCN1 antibody,were performed as described in Materials and methods,attached to Matrigel-coated 24-well culture plates andincubated at 37 °C for 18 h. EPCs images were captured andanalyzed by Leica Qwin system as depicted in the legends toFig. 2. Values are presented as the mean±S.D. of total length perfield (⁎p<0.05 versus control; #p<0.05 versus Ad-CCN1).
3206 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
cular cells and especially secreted factors, such as VEGF, SDF-1,FGF, PDGF and many more, play important roles in the regulationof EPCs and thus in neovascularization. Although the applicationof a single factor has entered the clinical arena, i.e., VEGF, largescale trials have not yielded expected beneficial results [21–23].Thus, other potential factors and interactions between thesefactors remain to be further elucidated.
In the present study, we explored the role of novel matrixprotein CCN1 in the regulation of EPCs proliferation and function.CCN1, belonging to the emerging CCN family, is a secreted, extra-cellular matrix associated proangiogenic factor. CCN1 is expressedby all types of vascular cells in response to a variety of physical andchemical stimuli such as growth factors, proteases, ischemia,hypoxia, and shear stress [24–27]. In addition, aberrant expressionof CCN1 is associated with several diseases such as wound healing[28], atherosclerosis [29] and restenosis following percutaneoustransluminal coronary angioplasty (PTCA) [30], supporting thatCCN1 may play an important role in the self-renewal program,specifically EPCs-involved vascular regeneration. Here, we showedthat CCN1 was present at a fairly low or undetectable level inquiescent EPCs, but was rapidly up-regulated upon stimulationwith serum, VEGF, ECs-cocultivation and EC-conditioned medium.In accordance with other observations, we also identified theinducible expression of CCN1 in ECs. It is tempting to speculate thatmatrix CCN1 may act on EPCs in autocrine and paracrine manners,contributing to neovascularization under physiological and patho-logical conditions.
Our previous work demonstrated that mesenchymal stem cellsco-cultured with mature ECs underwent milieu-dependent differ-entiation towards ECs [19]. In addition, observations from othersshowed that co-culture with ECs efficiently supported hemato-poietic cells proliferation and trafficking [31] and was able topromote proliferation, migration and differentiation of neuralprogenitor cells [32]. In this study, we showed that mature ECs, theimportant cellular component of vascular microenvironment, alsoimpacted on the biological properties of EPCs. The proliferation,migration and angiogenesis activities of EPCs were all significantlyenhanced by either coculture with ECs or ECs-conditionedmedium. More importantly, no obvious differences were foundbetween treatments of ECs and ECs supernatant that emphasizedthe critical paracrine role played by mature ECs. Herein, we soughtto investigate the role of CCN1 thereafter. First, we blocked CCN1 inthe co-culture system with an anti-CCN1 antibody and found thatECs-augmented EPCs migration and tube formation were signifi-cantly attenuated by the antibody, suggesting a CCN1 dependentmechanism of ECs effects. Next, we constructed an adenoviralvector expressing rat CCN1 and transduced it into EPCs. Exogenousoverexpression of CCN1 directly enhanced the migration and tubeformation of EPCs as expected, whereas no effect was observed oncell proliferation. Then, we knocked down endogenous CCN1 bysmall interfering RNA, and further confirmed that RNAi-mediatedknockdown of endogenous CCN1 inhibited the serum and VEGF-induced proliferation, migration and tube formation of EPCs. Thiswas consistent with the above observations.
However, the effect of CCN1 on cell proliferation and migrationis controversial [32–36]. CCN1 enhanced both migration andproliferation in vascular ECs and fibroblasts, and suppressed apop-tosis in breast cancer cells, while CCN1 expression was associatedwith cell death in neuronal progenitor cells and CCN1-inducedapoptosis in fibroblasts and endometrial cancer cells. Moreover,
recent findings showed that CCN1 inhibited mesangial cellsmigration induced by PDGF-BB, but did not affect cell proliferation[37].What impact CCN1may play on EPCs is currently unknown. Inour study, we demonstrated that CCN1 promoted the proliferationandmigration of EPCs, using both loss and gain function assays. Ourresults raised a possibility that, under stimulation of vascularinjury, CCN1 secreted by local ECs and other vascular cells mightrecruit circulating EPCs to sites of injury and facilitate regenerationand neovascularization. The seeming inefficacy of Ad-CCN1 trans-fection on the proliferation of EPCs could possibly be attributed tothe cytotoxicity of adenoviral vectors. However, combined withprevious observations by others, our results support the cell type-specific activities of CCN1. Well, the underlying mechanisminvolved in these properties is not entirely elucidated and is stillunder investigation.
Furthermore, our study observed that transfection of Ad-CCN1into ECs increased the release of growth factors and chemokinessuch as VEGF and MCP-1, and the supernatant from transfectedECs also resulted in an increase of VEGF and MCP-1 levels in EPCs.It was intriguing to observe such effects of CCN1, as both VEGF andMCP-1 were thought to be important mediators of EPCs recruit-ment, proliferation and eventually regeneration and neovascular-ization. However, as an immediate early gene, CCN1 was inducedby growth factors and cytokines such as VEGF, which has beenproven by our study and studies from others. A previousexperiment using microarray analysis identified that in humanumbilical vein endothelial cells (HUVECs), VEGF induced morethan two-fold increase of 139 cDNAs, and CCN1 was among them[14]. Kuiper and colleagues [38] reported that VEGF induced CCN1expression in retina in vivo, as well as in cultured retinal endo-thelial cells and pericytes in vitro. An additional study reportedthat stimulation of osteoblasts with VEGF resulted in up-regula-tion of CCN1 mRNA and protein [39]. In fact, on the contrary, CCN1has been known to regulate VEGF expression in human skinfibroblasts [28] and bladder smooth muscle cells [40]. Addition-ally, CCN1-deficient mice exhibited a substantial down-regulationof VEGF, which contributed to placental vascular insufficiency andreduced vessel integrity [11]. In present study, our results showedthat VEGF (and serum) up-regulated CCN1 in ECs and EPCs, andvice versa, strengthening the hypothesis that a paracrine loop mayexist between CCN1 and VEGF that regulates EPCs-mediatedneovascularization and re-endothelialization. After vascularinjury, CCN1 is induced by VEGF and other stimuli in localvascular milieu, especially in ECs. Then, matrix-associated CCN1 inturn up-regulates synthesis of VEGF, MCP-1 and some otherfactors, directly and/or indirectly promotes proliferation, recruit-ment, and angiogenesis activities of circulating EPCs. EPCsrecruited into local vessels may equally synthesize and secretCCN1 and VEGF in a circuit, thereafter stimulate migration andproliferation of adjacent ECs or support more circulating EPCs,further enhancing the vascular repair process in an amplifying-cytokine network. Furthermore, limited induction of CCN1, albeitto various extents, may be shown during pathological process as aresult of vascular cells dysfunction, suggesting that inadequateCCN1 impacts on the vascular repair process and that targetingCCN1 may be a valuable approach in future therapy.
However, we cannot exclude other factors besides VEGF andMCP-1 are induced in EPCs and ECs by CCN1, and the signalingintermediates responsible for the functional interplay betweenthose factors were not addressed in this study. To date, several
3207E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
possibilities have emerged concerning the mechanism for CCN1effects on EPCs [41–45]. First, as an extracellular matrix protein,CCN1 might exert its actions on EPCs via mediating the matrixmicroenvironment and release the matrix-binding cytokines andchemokines. Second, CCN1 behaved as a ligand for heparin sulfateproteoglycan and various integrins, and thus it might directlyinduce factors such as VEGF and MCP-1 through “outside-insignaling” transduction in EPCs and ECs. Third, CCN1 could func-tionally synergize with other factors such as VEGF and MCP-1 toenhance EPCs proliferation, recruitment and angiogenesis activ-ities in an autocrine and paracrine fashion. Therefore, theinterplay between the actions of CCN1 and other paracrine factorsseems complex in microenvironment that affects on EPCs proper-ties and remains to be more extensively investigated in furtherstudies.
In conclusion, the present study provides evidence that CCN1, anovel extracellular matrix-associated angiogenic factor, is inducedby serum and VEGF in both ECs and EPCs, and mediates themilieu-dependent activities of EPCs, such as proliferation, migra-tion, and tube formation. In addition, CCN1 up-regulates VEGF andMCP-1 levels in both ECs and EPCs, suggesting a paracrine loopmediated by cytokines and chemokines induced by CCN1 mayamplify the proangiogenic effects of CCN1. These findingsdemonstrate an important role of CCN1 in neovascularizationand re-endothelialization by mediating EPCs proliferation, recruit-ment, and angiogenesis. Thus, future studies are warranted toelucidate the complex mechanism and therapeutic potential ofCCN1 in the angiogenesis and tissue regeneration processmediated by EPCs.
Acknowledgments
Appreciation goes to Dr. Lei Hao for her encouragement and kindhelp in this work. These studies were supported in part by theNational Natural Science Foundation of China (grant 30470729and 30770852).
Appendix A Supplementary data
Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.yexcr.2008.08.001.
R E F E R E N C E S
[1] W. Risau, Mechanisms of angiogenesis, Nature 386 (1997)671–674.
[2] T. Asahara, H. Masuda, T. Takahashi, C. Kalka, C. Pastore, M. Silver,M. Kearne, M. Magner, J.M. Isner, Bone marrow origin ofendothelial progenitor cells responsible for postnatalvasculogenesis in physiological and pathologicalneovascularization, Circ. Res. 85 (1999) 221–228.
[3] C. Urbich, S. Dimmeler, Endothelial progenitor cells:characterization and role in vascular biology, Circ. Res. 95 (2004)343–353.
[4] T. He, L.A. Smith, S. Harrington, K.A. Nath, N.M. Caplice, Z.S.Katusic, Transplantation of circulating endothelial progenitorcells restores endothelial function of denuded rabbit carotidarteries, Stroke 35 (2004) 2378–2384.
[6] M. Koyanagi, C. Urbich, E. Chavakis, J. Hoffmann, S. Rupp, C.Badorff, A.M. Zeiher, P.A. Starzinski, J. Haendeler, S. Dimmeler,Differentiation of circulating endothelial progenitor cells to acardiomyogenic phenotype depends on E-cadherin, FEBS Lett.579 (2005) 6060–6066.
[7] Y.Wu, J.E. Ip, J. Huang, L. Zhang, K. Matsushita, C.C. Liew, R.E. Pratt,V.J. Dzau, Essential role of ICAM-1/CD18 in mediating epcrecruitment, angiogenesis, and repair to the infarctedmyocardium, Circ. Res. 99 (2006) 315–322.
[8] D.R. Brigstock, R. Goldschmeding, K.I. Katsube, S.C. Lam, L.F. Lau, K.Lyons, C. Naus, B. Perbal, B. Riser, M. Takigawa, H. Yeger, Proposalfor a unified CCN nomenclature, Mol. Pathol. 56 (2003) 127–128.
[9] B. Perbal, NOV (nephroblastoma overexpressed) and the CCNfamily of genes: structural and functional issues, Mol. Pathol. 54(2001) 57–79.
[10] T.M. Grzeszkiewicz, V. Lindner, N. Chen, S.C. Lam, L.F. Lau, Theangiogenic factor cysteine-rich 61 (CYR61, CCN1) supportsvascular smooth muscle cell adhesion and stimulates chemotaxisthrough integrin alpha(6)beta(1) and cell surface heparan sulfateproteoglycans, Endocrinology 143 (2002) 1441–1450.
[11] F.E. Mo, A.G. Muntean, C.C. Chen, D.B. Stolz, S.C. Watkins, L.F. Lau,CYR61 (CCN1) Is essential for placental development and vascularintegrity, Mol. Cell Biol. 22 (2002) 8709–8720.
[12] A.M. Babic, M.L. Kireeva, T.V. Kolesnikova, L.F. Lau, CYR61, productof a growth factor-inducible immediate-early gene, promotesangiogenesis and tumor growth, Proc. Natl. Acad. Sci. U.S.A. 95(1998) 6355–6360.
[13] V. Fataccioli, V. Abergel, L. Wingertsmann, P. Neuville, E. Spitz, S.Adnot, V. Calenda, E. Teiger, Stimulation of angiogenesis by Cyr61gene: a new therapeutic candidate, Hum. Gene. Ther. 13 (2002)1461–1470.
[14] M. Abe, Y. Sato, cDNA microarray analysis of the gene expressionprofile of VEGF-activated human umbilical vein endothelial cells,Angiogenesis 4 (2001) 289–298.
[15] K. Grote, G. Salguero, M. Ballmaier, M. Dangers, H. Drexler, B.Schieffer, The angiogenic factor CCN1 promotes adhesion andmigration of circulating CD34+ progenitor cells: potential role inangiogenesis and endothelial regeneration, Blood 110 (2007)877–885.
[16] N. Schütze, R. Schenk, J. Fiedler, T. Mattes, F. Jakob, R.E. Brenner,CYR61/CCN1 and WISP3/CCN6 are chemo attractive ligands forhuman multipotent mesenchymal stroma cells, BMC Cell Biol. 8(2007) 45–52.
[17] X.H. Zhao, L. Huang, Y.G. Yin, Y.Q. Fang, J.H. Zhao, J.F. Chen,Estrogen induces endothelial progenitor cells proliferation andmigration by estrogen receptors and PI3K-dependent pathways,Microvasc. Res. 75 (2008) 45–52.
[18] J. Rice, D.L. Courter, C.M. Giachelli, M. Scatena, Molecularmediators of alphavbeta3-induced endothelial cell survival,J. Vasc. Res. 43 (2006) 422–436.
[19] X.J. Wu, L. Huang, Q. Zhou, Y.M. Song, A.M. Li, J. Jin, B. Cui,Mesenchymal stem cells participating in ex vivo endotheliumrepair and its effect on vascular smoothmuscle cells growth, Int. J.Cardiol. 105 (2005) 274–282.
3208 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 4 ( 2 0 0 8 ) 3 1 9 8 – 3 2 0 8
[23] S.B. Freedman, J.M. Isner, Therapeutic angiogenesis for coronaryartery disease, Ann. Intern. Med. 136 (2002) 54–71.
[24] J.S. Han, E. Macarak, J. Rosenbloom, K.C. Chung, B. Chaqour,Regulation of Cyr61/CCN1 gene expression through RhoA GTPaseand p38MAPK signaling pathways, Eur. J. Biochem. 270 (2003)3408–3421.
[25] U.R. Pendurthi, T.T. Tran, M. Post, L.V. Rao, Proteolysis of CCN1 byplasmin: functional implications, Cancer Res. 65 (2005)9705–9711.
[26] M. Kunz, S. Moeller, D. Koczan, P. Lorenz, R.H. Wenger, M.O.Glocker, H.J. Thiesen, G. Gross, S.M. Ibrahim, Mechanisms ofhypoxic gene regulation of angiogenesis factor Cyr61 inmelanoma cells, J. Biol. Chem. 278 (2003) 45651–45660.
[27] K.D. Hilfiker, K. Kaminski, A. Kaminska, M. Fuchs, G. Klein, E.Podewski, K. Grote, I. Kiian, K.C. Wollert, A. Hilfiker, Regulation ofproangiogenic factor CCN1 in cardiac muscle: impact of ischemia,pressure overload, and neurohumoral activation, Circulation 109(2004) 2227–2233.
[28] C.C. Chen, F.E. Mo, L.F. Lau, The angiogenic factor Cyr61 activatesa genetic program for wound healing in human skin fibroblasts,J. Biol. Chem. 276 (2001) 47329–47337.
[29] A. Hilfiker, K.D. Hilfiker, M. Fuchs, K. Kaminski, A. Lichtenberg, H.J.Rothkotter, B. Schieffer, Expression of CYR61, an angiogenicimmediate early gene, in arteriosclerosis and its regulation byangiotensin II, Circulation 106 (2002) 254–260.
[30] K.J. Wu, A. Yee, N.L. Zhu, E.M. Gordon, F.L. Hall, Characterization ofdifferential gene expression in monkey arterial neointimafollowing balloon catheter injury, Int. J. Mol. Med. 6 (2000)433–440.
[31] B. Jazwiec, A. Solanilla, C. Grosset, F.X. Mahon, M. Dupouy, L.V.Pigeonnier, F. Belloc, K. Schweitzer, J. Reiffers, J. Ripoche,Endothelial cell support of hematopoiesis is differentially alteredby IL-1 and glucocorticoids, Leukemia 12 (1998) 1210–1220.
[32] L. Wang, Z.G. Zhang, R.L. Zhang, S.R. Gregg, S.A. Hozeska, Y.LeTourneau, Y. Wang, M. Chopp, Matrix metalloproteinase 2(MMP2) and MMP9 secreted by erythropoietin-activatedendothelial cells promote neural progenitor cell migration,J. Neurosci. 26 (2006) 5996–6003.
[33] T.M. Grzeszkiewicz, D.J. Kirschling, N. Chen, L.F. Lau, CYR61stimulates human skin fibroblast migration through Integrinalpha vbeta 5 and enhances mitogenesis through integrin alphavbeta 3, independent of its carboxyl-terminal domain, J. Biol.Chem. 276 (2001) 21943–21950.
[34] K.C. Chung, Y.S. Ahn, Expression of immediate early gene cyr61during the differentiation of immortalized embryonichippocampal neuronal cells, Neurosci. Lett. 255 (1998) 155–158.
[35] V. Todorovicc, C.C. Chen, N. Hay, L.F. Lau, The matrix protein CCN1(CYR61) induces apoptosis in fibroblasts, J. Cell Biol. 171 (2005)559–568.
[36] W. Chien, T. Kumagai, C.W. Miller, J.C. Desmond, J.M. Frank, J.W.Said, H.P. Koeffler, Cyr61 suppresses growth of humanendometrial cancer cells, J. Biol. Chem. 279 (2004)53087–53096.
[37] K. Sawai, K. Mori, M. Mukoyama, A. Sugawara, T. Suganami, M.Koshikawa, K. Yahata, H. Makino, T. Nagae, Y. Fujinaga, H. Yokoi,T. Yoshioka, A. Yoshimoto, I. Tanaka, K. Nakao, Angiogenicprotein Cyr61 is expressed by podocytes in anti-Thy-1glomerulonephritis, J. Am. Soc. Nephrol. 14 (2003) 1154–1163.
[38] E.J. Kuiper, J.M. Hughes, R.J. Van Geest, I.M. Vogels, R.Goldschmeding, C.J. Van Noorden, R.O. Schlingemann, I. Klaassen,Effect of VEGF-A on expression of profibrotic growth factor andextracellular matrix genes in the retina, Invest. Ophthalmol. Vis.Sci. 48 (2007) 4267–4276.
[39] A.N. Athanasopoulos, D. Schneider, T. Keiper, V. Alt, U.R.Pendurthi, U.M. Liegibel, U. Sommer, P.P. Nawroth, C. Kasperk, T.Chavakis, Vascular endothelial growth factor (VEGF)-inducedup-regulation of CCN1 in osteoblasts mediates proangiogenicactivities in endothelial cells and promotes fracture healing,J. Biol. Chem. 282 (2007) 26746–26753.
[40] D. Zhou, D.J. Herrick, J. Rosenbloom, B. Chaqour, Cyr61 mediatesthe expression of VEGF, alphav-integrin, and alpha-actin genesthrough cytoskeletally based mechanotransduction mechanismsin bladder smooth muscle cells, J. Appl. Physiol. 98 (2005)2344–2354.
[41] N. Nguyen, A. Kuliopulos, R.A. Graham, L. Covic, Tumor-derivedCyr61 (CCN1) promotes stromal matrix metalloproteinase-1production and protease-activated receptor 1–dependentmigration of breast cancer cells, Cancer Res. 66 (2006)2658–2665.
[42] T.V. Kolesnikova, L.F. Lau, Human CYR61-mediated enhancementof bFGF-induced DNA synthesis in human umbilical veinendothelial cells, Oncogene 16 (1998) 747–775.
[43] D. Xie, D. Yin, X. Tong, J.O. Kelly, A. Mori, C. Miller, K. Black, D. Gui,J.W. Said, H.P. Koeffler, Cyr61 is overexpressed in gliomas andinvolved in integrin-linked kinase-mediated Akt andbeta-catenin-TCF/Lef signaling pathways, Cancer Res. 64 (2004)1987–1996.
[44] Y. Chen, X.Y. Du, Functional properties and intracellular signalingof CCN1/Cyr61, J. Cell. Biochem. 100 (2007) 1337–1345.
[45] J.G. Abreu, N.I. Ketpura, B. Reversade, E.M. De Robertis,Connective-tissue growth factor (CTGF) modulates cell signallingby BMP and TGF-beta, Nat. Cell Biol. 4 (2002) 599–604.
