The study protocol was approved by the Central South University Institutional Review Board. All methods used in this study were carried out in accordance with the approved Ethical Guidelines of Central South University. Informed consent was obtained from all subjects prior to the study.

Cord blood (CB) was obtained from 10 normal full-term deliveries in the Women and Child Health Hospital of Hunan Province. UCB-EPCs were isolated and cultured as previously described (34). Briefly, CB was diluted 1:1 with Dulbecco's phosphate-buffered saline (DPBS; Gibco, Grand Island, NY, USA), and then overlaid onto 1.077 g/ml Ficoll-Paque™ Premium (GE Healthcare, Logan, UT, USA). The liquid was centrifuged for 30 min at 400 × g. Monocytes were collected and washed with DPBS. The cells were seeded on tissue culture plates coated with fibronectin (Millipore, Billerica, MA, USA) in EGM-2 (Lonza, Rockland, ME, USA) at 37°C, 5% CO₂ humidified incubator. The culture medium was changed every other day until the EPC colonies appeared. The cells were harvested for expansion and freezing after they reached 80–90% confluence.

Human adipose tissues were obtained from Xiangya Hospital of Central South University (Changsha, China) and digested with 2 mg/ml collagenase I, 2 U/ml dispase and 2 mg/ml hyaluronidase (all purchased from Sigma-Aldrich, St. Louis, MO, USA) for 90 min at 37°C. The digested tissues were centrifuged (1,000 rpm for 10 min) and the stromal vascular fraction (SVF) was washed with DPBS. SVF was then cultured in Dulbecco's modified Eagle's medium-F12 (DMEM/F-12) containing 10 ng/ml basic fibroblast growth factor (bFGF; Gibco) and 10% fetal bovine serum (FBS). The medium was changed every 2 days. The cells were harvested for expansion and freezing when the cells reach 80–90% confluence. The cells at passage 4 were used for the following experiments.

The EPC single-cell suspension was generated into the concentration of 1×10⁷ cells/ml. The cells were then incubated respectively with anti-human CD31-FITC (eBioscience, San Diego, CA, USA), vascular endothelial growth factor receptor (VEGFR2)/KDR-PE (R&D Systems, Minneapolis, MN, USA), CD144-FITC (Abcam, Cambridge, UK), CD34-PE, CD45-FITC (both from Biolegend, San Diego, CA, USA), CD14-FITC (eBioscience), CD29-PE, CD90-PE and SSEA4-PE (all from Biolegend). Briefly, 100 µl cell suspension was incubated with 5 µl antibody solution at 4°C for 30 min in the dark. After washing twice with phosphate buffer saline (PBS), cells were resuspended in 400 µl PBS and analyzed with a FACSAria I (Becton-Dickinson, San Jose, CA, USA) and Becton-Dickinson CellQuest software.

The Alexa Fluor 488 Annexin V and prop-idium iodide kit (Invitrogen, Carlsbad, CA, USA) was used for the analysis of apoptosis. Briefly, 1×10⁵ cells were harvested and washed twice with cold PBS, then resuspended in 100 µl binding buffer. Subsequently, 5 µl Annexin V-FITC and 1 µl propidium iodide were added to the solution. Following 15 min of incubation, 400 µl binding buffer were added to the solution, and the cells were analyzed using a flow cytometer (BD Accuri™ C6 Flow Cytometer; BD Biosciences San Jose, CA, USA).

The TUNEL apoptosis detection kit (Beyotime, Shanghai, China) was also used for the analysis of cell apoptosis. Briefly, the EPCs were fixed with 4% paraform/PBS, followed by permeabilization with 0.1% Triton X-100 for 2 min on ice. The cells then underwent TUNEL staining in the dark for 1 h at 37°C. After washing twice with PBS, the suspension was analyzed by flow cytometry (BD Accuri™ C6 Flow Cytometer; BD Biosciences).

EPCs at passage 2, 3, 4, 5, 6, 7 and 8 were seeded at 1×10⁵ cells/well for 4 wells in 6-well plates, respectively. Following 3 days of incubation in EGM-2, the cells were digested with TrypLE Express (Gibco) and resuspended into single-cell suspension, followed by counting under a light microscope (IX71; Olympus, Tokyo, Japan). The proliferation index was calculated as follows: proliferation index = total number at day 3/1×10⁵.

In order to measure the migration of the EPCs, 1.5×10⁵ cells at passage 4, 6 and 8 with or without pre-treatment with tyrphostin AG1295 (Sigma-Aldrich) at 20 µM for 1 h were seeded in the upper Transwell chamber (BD Biosciences) in serum-free medium, with 500 µl DMEM with 10% FBS in the lower chamber. After 24 h, cells that did not migrate through the pores were carefully wiped out with a cotton-tipped swab. The filters were fixed in 90% alcohol, followed by staining with 0.1% crystal violet (Meryer, Shanghai, China). After washing with PBS 3 times, the filters were observed under an inverted microscope (Olympus).

To examine protein expression in PDGF-BB-stimulated cells, the EPCs were harvested and lysed. Proteins were subjected to sodium dodecyl sulfate-poly-acrylamide gel electrophoresis (SDS-PAGE) and transferred onto PVDF membranes. The membranes were incubated at 4°C with primary antibodies overnight [anti-platelet-derived growth factor receptor-β (PDGFR-β; ab32570), anti-phospho-PDGFR-β (ab16868), anti-PI3K (ab86714), anti-phospho-PI3K (ab182651), anti-Akt (ab8805), anti-phospho-Akt (ab38449), or anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH; ab9485); all from Abcam], and then stained with horseradish peroxidase-coupled secondary antibodies (ab131366; Abcam). Finally, the bands were visualized by chemiluminescence (Amersham Pharmacia Biotech, Amersham, UK).

To examine the effects of PDGFR-β, EPCs at passage 4, 6 and 8 were pre-treated with PDGF-BB (PeproTech, Rocky Hill, NJ, USA) at 40 ng/ml for 24 h. The cells were then used in the subsequent experiments. To examine whether the PDGFR-β/PI3K signaling pathway is involved in the PDGF-BB-induced biological function changes of EPCs, EPCs were treated with 20 mM tyrphostin AG1295 (Sigma-Aldrich) for 1 h, followed by PDGF-BB stimulation as mentioned above. The cells were then used in the subsequent experiments.

Chromosomal DNA was extracted using Qiagen DNeasy Blood and Tissue kit according to the manufacturer's instructions. DNA from human embryonic stem cells was used as control or reference DNA. DNA was used as templates in SYBR-Green qPCR with specific primers. The primer sequences for telomere (tel) and 36B4 (single copy gene) genes were as follows: tel (tel1b, CGG TTT GTT TGG GTT TGG GTT TGG GTT TGG GTT TGG GTT; tel2b, GGC TTG CCT TAC CCT TAC CCT TAC CCT TAC CCT TAC CCT); 36B4 (36B4u, CAG CAA GTG GGA AGG TGT AAT CC; 36B4d, CCC ATT CTA TCA TCA ACG GGT ACA A). Two PCR runs were performed for each sample: one to determine the cycle threshold (Ct) value for telomere; the other to determine the Ct value for the amplification of 36B4. PCR was performed in a total volume of 20 µl, including 10 µl of SYBR-Green qPCR mix, 1 µl of each forward and reverse primer (final concentration: 400 nM for telomere; 300 nM for 36B4), 1 µl each DNA sample and 7 µl H₂O. Amplifications were carried out in triplicate in 96-well microtiter plates. The thermal cycling conditions for telomere PCR were as follows: 95°C for 10 min (stage 1), followed by 35 cycles of 95°C for 5 sec, 56°C for 10 sec, and 72°C for 1 min (stage 2), and finally followed by 95°C for 5 sec, and 60°C for 10 sec. For 36B4: 95°C for 10 min (stage 1), followed by 40 cycles of 95°C for 5 sec, 58°C for 10 sec, 72°C for 40 sec.

A 96-well plate was covered with Matrigel (BD Biosciences). The EPCs (4×10³) were suspended in 50 µl EGM-2 and seeded on Matrigel. The plate was incubated at 37°C. Images of tubules were captured after 2 h using a Camera Nikon TE2000-U (Nikon, Tokyo, Japan).

The MSCs were seeded onto 24-well plates at 3×10⁴ cells/ml and incubated until 80% confluency. The EPCs at passage 4, 6 and 8 were then seeded on the MSC monolayer at 2×10⁴ cells/ml and incubated in EGM-2. After 6 days of co-culture, UEA-I (Vector Laboratories, Burlingame, CA, USA) was used to perform the staining of EPCs. Images were acquired using a fluorescence microscope (ELX800; BioTek Instruments, Inc., Winooski, VT, USA) and a Nikon photographic system (Nikon Eclipse Ti-S; Nikon). Quantification analysis was carried out using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Total RNA was extracted from the cells using TRIzol reagent (Life Technologies, Shanghai, China) according to the manufacturer's instructions. cDNA was synthesized using the Transcriptor First Strand cDNA synthesis kit (Roche, Basel, Switzerland). qPCR was performed using a Lightcycler 480 SYBR-Green I Master system (Roche) according to the manufacturer's instructions. GAPDH were used as an internal control. The sequences of the human primers were as follows: VEGF-A sense, AGG GCA GAA TCA TCA CGA AGT and antisense, AGG GTC TCG ATT GGA TGG CA; transforming growth factor-β1 (TGF-β1) sense, CTA ATG GTG GAA ACC CAC AAC G and antisense, TAT CGC CAG GAA TTG TTG CTG; PDGF-B sense, CTC GAT CCG CTC CTT TGA TGA and antisense, CGT TGG TGC GGT CTA TGA G; ANG-1 sense, GCC TGA TCT TAC ACG GTG CTG and antisense, GCA TCA AAC CAC CAT CCT CC; PDGFR-β sense, GGA GAG GGC AGT AAG GAG GA and antisense, ATG GTG TCC TTG CTG CTG AT; TIE-2 sense, TGT GCT GTT CCT TCT TGC CT and antisense, GCA CCT TCC ACA GTT CCA GA; VEGFR2 sense, GCA GAA CAG TAA GCG AAA GAG and antisense, TGA GGC AAG AAC CAT ACC ACT; interferon gamma receptor (IFNGR)1 sense, TAA ATG GAG ACG AGC AGG AAG and antisense, TGA ATA CCA GGC TAA GCA CTA; IFNGR2 sense, TTT AGA GTC GGG CAT TTA AGC A and antisense, TCA GGA CCA GGA AGA AAC AGG; fibro-nectin 1 (FN1) sense, ACA AAC ACT AAT GTT AAT TGC CCA and antisense, AAC TCC CAG GGT GAT GCT TG; laminin subunit alpha 2 (LAMA2) sense, CTG TTG CTG ATA ACC TCC TCT T and antisense, AGT TCT TGA TGC TAC GAT ACG G; integrin subunit beta 1 (ITGB1) sense, CCT ACT TCT GCA CGA TGT GAT G and antisense, CCT TTG CTA CGG TTG GTT ACA TT; integrin subunit alpha 1 (ITGA1) sense, GTG CTT ATT GGT TCT CCG TTA GT and antisense, CAC AAG CCA GAA ATC CTC CAT; collagen type IV alpha 1 chain (COL4A1) sense, CCA GGG GTC GGA GAG AAA G and antisense: GGT CCT GTG CCT ATA ACA ATT CC; GAPDH sense, AGA AGC CCA GCC AGT CGC CAT CA and antisense, AGC AAA GCC CGC CTT ACA GAG CC. PCR was performed in a total volume of 20 µl, including 10 µl of SYBR-Green qPCR Mix, 1 µl of each forward and reverse primer (10 µmol/l), 1 µl each cDNA sample, and 7 µl H₂O. Amplifications were carried out in triplicate in 96-well microtiter plates. Thermal cycling conditions were as follows: 95°C for 5 min, followed by 45 cycles of 95°C for 10 sec, 60°C for 10 sec, and 72°C for 10 sec, and finally followed by 95°C for 5 sec, and 60°C for 10 sec.

Procedures involving animals and their care were conducted in conformity with NIH guidelines (NIH Publication no. 85-23, revised 1996) and was approved by the Animal Care and Use Committee of the Central South University. A total of 30 male BALB/C nude mice, weighing 20–25 g were anesthetized with 4% chloral hydrate by intraperitoneal injection. The right femoral artery and vein were coagulated and then cut out to induce critical ischemia at day 0. Twenty-four hours later (day 1), the mice were randomly divided into 3 groups and received cell transplantations by tail vein injection: EPCs at passage 4, EPCs at passage 6 and EPCs at passage 8 (n=10).

The mice were anesthetized using 4% chloral hydrate, and then examined using a Laser Doppler Perfusion Imager (LDPI; Moor Instruments, Devon, UK) on days 0, 7, 14, 21 and 28. The animal was placed on a 37°C heating pad for 2–5 min to allow acclimation to the ambient conditions before measurements were taken. The results are reported as the perfusion ratio of the ischemic limb relative to the contralateral untreated hind limb.

Mice were euthanized at day 28. The quadriceps femoris muscles of the mouse hindlimbs were isolated and fixed in 10% buffered formalin, dehydrated in 30% sucrose solution and embedded in paraffin. Sections (7 µm-thick) were cut and mounted on slides. The samples were used for hematoxylin and eosin (H&E) staining (Beyotime) and Masson's trichrome staining (Baso Diagnostics Inc., Zhuhai, China). The sections were observed and captured using a microscope (Sunny CX40; Sunny, Guangdong, China) and the corresponding SmartV350Dc software.

All experiments were repeated at least 3 times independently. Data are expressed as the means ± standard deviation (SD). Comparisons between groups were performed by one-way ANOVA using SPSS 17.0 (SPSS, Inc., Chicago, IL, USA). For animal exterior recovery study, Kruskal-Wallis was used. A value of P<0.05 was considered to indicate a statistically significant difference.

EPCs were isolated from human UCB. In serial subculture, the EPCs exhibited a uniform cobblestone morphology, with no observed differences among the cells at different passages (Fig. 1A). In order to measure the proliferative ability of the EPCs, cells at different passages were seeded quantitatively in culture plates, and the total cell number after 3 days of culture was calculated to reflect the proliferation index. The proliferative ability of the EPCs was significantly decreased with the increase number of passages in culture (Fig. 1B). In subsequent experiments, we used EPCs at passage (P)2, P4, P6 and P8 for the examination of cellular properties. The cell apoptotic rate was then analyzed by TUNEL assay (Fig. 1C). No significant difference was observed in the apoptotic rate of the EPCs at different passages, and the average apoptotic rate was 4.45±2.75% at P2, 5.85±1.35% at P4, 6.3±0.6% at P6 and 5.5±0.8% at P8 (Fig. 1D). This trend was then confirmed by Annexin V/PI staining (Fig. 1E). In addition, the quantification of telomere length was measured by qPCR in order to determine the senescence of EPCs at different passages. The results revealed that although no significant changes were observed in telomere length among the EPCs at P2, P4, P6 and P8, telomere length exhibited a decreasing trend as the passage number increased in culture (Fig. 1F).

The expression of surface markers was analyzed (Fig. 2A). In the expansion process from passage 2 to 8, all EPCs homogeneously exhibited positive expression for the endothelial marker, CD31 (>90%), and the mesenchymal marker, CD29 (>95%), and expressed low levels of monocyte differentiation antigen CD14 (<7%), hematopoietic-related antigen CD45 (<6%) and SSEA4 (<4%) (Fig. 2B). The expression of CD34 was maintained at approximately 40%. of note, the expression of CD90 (from 2.79±2.12% at P2 to 1.3±0.44% at P8, P<0.05) and that of the endothelial markers, CD144 (VE-Cadherin) (from 50.18±23.75% at P2 to 15.86±8.77% at P8, P<0.01) and KDR (VEGFR2) (from 53.32±14.63% at P2 to 30.28±18.48% at P8, P<0.01), was downregulated with increasing number of passages (Fig. 2B). This indicated that the EPC phenotype was partly altered during the process of in vitro expansion.

Since there was no difference observed in cellular properties between the EPCs at P2 and the EPCs at P4, including proliferative ability, apoptotic rate, telomere length and surface marker expression, and the total cell number obtained at P4 is greater than that at P2, we selected the EPCs at P4, P6 and P8 for use in further experiments.

To compare the angiogenic ability of the EPCs at different passages in vitro, the EPCs was seeded on Matrigel. As shown in Fig. 3A, the EPCs at P4, P6 and P8 all assembled into tubular-like structures, but the more integrated network formed by the EPCs at P4 was not observed in the EPCs at P6 and at P8 in particular. In addition, the EPCs at P4 and P6 formed more network junctions and vascular rings than those at P8 (P<0.001) (Fig. 3B).

The EPCs were subsequently seeded on the monolayer of MSCs. After 6 days of culture, EPCs without feeders only exhibited a scattered distribution, while the EPCs seeded on MSCs formed capillary-like networks (Fig. 3C). Notably, the EPCs at P4 and P6 assembled into more organized and integrated networks than the EPCs at P8, which was quantified by an increased number of vascular junctions (P4, P<0.01, P6, P<0.01, compared with P8) and longer tubular-like structures (P4, P<0.01, P6, P<0.05) (Fig. 3D). Thus, the EPCs at P4 and P6 exhibited an enhanced angiogenic ability in vitro. These findings collectively incidated that the expanded EPCs at different passages exhibited inequable angiogenic properties in vitro.

The angiogenic properties of the EPCs were further examined in a mouse model of hind limb ischemia. After 24 h of right femoral artery ligation and excision surgery, the EPCs at P4, P6 and P8 were transplanted into mice by tail intravenous injection. Blood perfusion was detected using a LDPI on days 0, 7, 14, 21 and 28 (Fig. 4A). The statistical analysis of the blood perfusion rate in the leg revealed no significances among the different mouse groups transplanted with EPCs at different passages, although the transplation of of EPCs at P4 on day 21 showed a certain advantage (Fig. 4B). However, the perfusion condition in the paw revealed the statistical superiority of EPCs at P4 and P6. The transplantation of EPCs at P6 (0.46±0.25) more efficiently improved blood flow in the paw of the ischemic limb on day 7 compared with the transplantion of EPCs at P8 (0.22±0.11; P=0.017). On day 14, the transplantion of EPCs at P4 (0.56±0.16, P=0.002) and P6 (0.51±0.18, P=0.014) led to a relatively higher perfusion rate than the transplantion of EPCs at P8 (0.32±0.11). No statistically significant difference was observed among the 3 groups on days 21 and 28 (Fig. 4C).

We defined hind limb recovery after 28 days as five progressive levels, including limb salvage, swollen foot, amyotrophy, mild loss of limb and severe loss of limb (Fig. 4D). The proportion of limb salvage in the groups transplanted with EPCs at P4 and P6 was approximately 30%, whereas the injection of EPCs at P8 resulted in almost no final limb salvage (Fig. 4E). The transplantation of EPCs at P8 caused approximately 60% amyotrophy. Notably, the rate of limb loss (including mild and severe loss of limb) in the 3 groups was >20%, and was even >30% in the group injected with EPCs at P4.

We further evaluated the therapeutic effects of EPCs at different passages on muscle degradation of the ischemic hind limb by histological examination. H&E staining revealed that following the transplantation of EPCs at P4, the muscle fibers were arranged neatly in a regular round shape, with small gaps among the muscle bundles (Fig. 5A). The nucleus located on the edge of the muscle fibers. These were very similar to the normal control group. However, in the groups transplanted with EPCs at P6 and P8, the muscle fibers became smaller and wizened, with the nucleus located in the center of the muscle fibers. The gap among the muscle bundles became larger, and was filled with the infiltrated cells and hyperplastic connective tissue. Fibrous morphology was even observed in the group transplanted with EPCs at P8. These features were similar to those of the negative control (NC) group, although to a lesser extent. Furthermore, Masson's trichrome staining revealed that fibrosis was markedly attenuated following the transplantation of EPCs, compared with the negative control (Fig. 5B). The EPCs at P4 and P6 exerted more positive effects.

For further analysis, we detected the expression of angiogenic-related factors in the EPCs at different passages by qPCR. Although there were no statistically significant differences observed in the expression of angiogenic cytokines (such as VEGF-A, PDGF-B, ANG-1 and TGF-β1) among the EPCs at different passages (Fig. 6A), significant changes were observed in the expression of some extracellular matrix (ECM) components (Fig. 6B). The expression levels of ITGA1 and LAMA2 in the EPCs at P8 were higher when compared with those in the EPCs at P4, and the expression levels of ITGB1 and COL4A1 were significantly decreased in the EPCs at later passages.

In addition, angiogenic-related receptors on EPCs were measured following PDGF-BB stimulation. As shown in Fig. 6C, both PDGFR-β and VEGFR2 were significantly highly expressed in the EPCs at P8. However, when the receptor expression was measured by western blot analysis (Fig. 6D), the level of phosphorylated PDGFR-β was found to be significantly decreased in the EPCs at P8 compared to those at P4 (P<0.05; Fig. 6E). Since the binding of PI3K to PDGFR-β has been shown to be important for cell behavior (32), we further examined whether the PDGFR-β/PI3K/Akt signaling pathway is involved in EPCs by examining the phosphorylation levels of PI3K and Akt by western blot analysis (Fig. 6F). The results of statistical analysis indicated that the phosphorylation levels of PI3K and Akt were both significantly decreased in the EPCs at P8 compared with those at P4 (P<0.05 and P<0.05; Fig. 6G and H).

We further examined whether PDGFR-β plays a role in the changes of angiogenesis and migration ability among the EPCs at different passages. The selective inhibitor of PDGFR, tyrphostin AG1295, was used to inhibit the activation of PDGFR-β. Treatment with tyrphostin AG1295 led to less interconnected vascular network being formed by the EPCs at passage 4, 6 and 8 (Fig. 7A). No significant difference was observed among the groups of EPCs at different passages (Fig. 7B). In addition, when seeded on the monolayer of MSCs, the EPCs at different passages pretreated with tyrphostin AG1295 formed a smaller number of tubular-like structures compared with the cells not treated with tyrphostin AG1295 (Fig. 7C), and there was no significant difference observed among the tyrphostin AG1295-treated groups (Fig. 7D). Furthermore, the cell migration ability was examined by Transwell assay (Fig. 7E). The EPC migration ability decreased with the in vitro expansion process without pre-treatment with tyrphostin AG1295 (P<0.05; Fig. 7F). Treatment with tyrphostin AG1295 led to significant decrease in migration in the EPCs at each passage (P<0.05). However, following treatment with tyrphostin AG1295, no significant difference was observed in migration ability among the EPCs at different passages. These results indicated that following treatment with the PDGFR inhibitor, tyrphostin AG1295, the differences in angiogenesis and migration ability of the EPCs at different passages were no longer observed.

As demonstrated above (Fig. 6D and E), the levels of phosphorylated PDGFR-β were decreased in the EPCs with the increasing passage number. In particular, the levels of phosphorylated PDGFR-β were significantly decreased in the EPCs at passage 8 compared to those at passage 4. Subsequently, in order to confirm the effect of PDGFR inhibitor, the expression of PDGFR-β was measured in the EPCs by western blotting (Fig. 8A) following treatment with the PDGFR inhibitor, tyrphostin AG1295. Following pre-treatment with the inhibitor, the EPCs at different passages exhibited no significant difference in the levels of phosphorylated PDGFR-β when stimulated with PDGF-BB (Fig. 8B). Moreover, we also measured the phosphorylation levels of PI3K/Akt by western blot analysis in the EPCs pre-treated with the inhibitor (Fig. 8C); no significant difference was observed in the phosphorylation level of PI3K/Akt among the EPCs at different passages (Fig. 8D and E). These findings indicated that following pre-treatment with the PDGFR inhibitor, tyrphostin AG1295, no significant differences were observed in the levels of phosphorylated PDGFR-β and PI3K/Akt expression among the EPCs at different passages.