Subcutaneous adipose tissue contains a lot of stem cells [adipose-derived stem cells (ASCs)] that can differentiate into a variety of cell lineages. In this study, we isolated ASCs from Wistar rats and examined whether ASCs would efficiently differentiate into vascular endothelial cells (ECs) in vitro. We also administered ASCs in a wire injury model of rat femoral artery and examined their effects. ASCs expressed CD29 and CD90, but not CD34, suggesting that ASCs resemble bone marrow-derived mesenchymal stem cells. When induced to differentiate into ECs with endothelial growth medium (EGM), ASCs expressed Flt-1, but not Flk-1 or mature EC markers such as CD31 and vascular endothelial cadherin. ASCs produced angiopoietin-1 when they were cultured in EGM. ASCs stimulated the migration of EC, as assessed by chemotaxis assay. When ASCs that were cultured in EGM were injected in the femoral artery, the ASCs potently and significantly inhibited neointimal formation without being integrated in the endothelial layer. EGM-treated ASCs significantly suppressed neointimal formation even when they were administered from the adventitial side. ASC administration significantly promoted endothelial repair. These results suggested that although ASCs appear to have little capacity to differentiate into mature ECs, ASCs have the potential to secrete paracrine factors that stimulate endothelial repair. Our results also suggested that ASCs inhibited neointimal formation via their paracrine effect of stimulation of EC migration in situ rather than the direct integration into the endothelial layer.
- vascular endothelial cells
- endothelial repair
cell-based therapy has been recently applied to the field of cardiovascular medicine. Among a variety of stem or progenitor cells that can be used for regeneration of heart and blood vessels, endothelial progenitor cells (EPCs) and bone marrow-derived mesenchymal stem cells (BMMSCs) are the ones that are most popularly used in this field. Although EPCs were originally isolated from human peripheral blood using the hematopoietic stem cell marker CD34 for positive selection, they are believed to reside mainly in the bone marrow (1, 19). EPCs can differentiate into vascular endothelial cells (ECs) in vitro and stimulate angiogenesis in vivo through integration in the endothelial layer of new forming capillaries (1). EPCs that are induced to differentiate into ECs are engrafted in the endothelial layer and inhibit neointimal formation via stimulation of EC regeneration (5). EPCs that are differentiated into ECs also form the endothelial layer on the surface of prosthetic grafts (5). BMMSCs have the potential to differentiate into mesenchymal tissues such as bone, cartilage, fat, and muscle (14). In addition to their capacity to differentiate into mesenchymal tissues, CD34-negative BMMSCs have been shown to differentiate into ECs (13). Furthermore, BMMSCs reportedly secrete paracrine factors that potentially stimulate angiogenesis (10). Bone marrow-derived cells have been used clinically to treat cardiovascular diseases and have turned out to be useful in some reports (22, 24). The problem in the clinical application of EPCs and BMMSCs is that bone marrow aspiration is usually necessary to prepare EPCs and BMMSCs. This procedure is somewhat painful for patients.
Recently, subcutaneous adipose tissue has been drawing much attention, because it contains a lot of mesenchymal stem cells that potentially differentiate into a variety of cell lineages including adipocytes, chondrocytes, osteocytes, and skeletal muscle (4). If these mesenchymal stem cells are useful for the regeneration of heart and blood vessels, adipose tissue will be a promising source of stem cells in the field of cardiovascular medicine, because it is easy to collect by local anesthesia. In fact, it has been reported that adipose-derived stem cells (ASCs) stimulate angiogenesis in the mouse hindlimb ischemia model (9, 11, 15, 16). However, the mechanism by which ASCs stimulate angiogenesis remains to be debated. ASCs promoted angiogenesis by being engrafted in the endothelial layer and stimulating neovascular formation (9, 15) or by producing angiogenesis-stimulating factors without integration into the endothelial layer (11). It also remains controversial whether ASCs can efficiently differentiate into ECs in vitro. Furthermore, it remains unclear whether ASCs can be efficiently integrated in the endothelial layer and inhibit neointimal formation.
In this study, we isolated ASCs from Wistar rats and examined whether ASCs would efficiently differentiate into ECs in vitro. We also examined whether ASCs would be engrafted in the endothelial layer and inhibit neointimal formation using a wire injury model of the rat femoral artery.
MATERIALS AND METHODS
Anti-CD34, anti-vascular endothelial (VE)-cadherin, anti-Flk-1, anti-Flt-1, and anti-proliferating cell nuclear antigen (PCNA) antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-CD31 antibody was obtained from ABR Affinity BioReagents (Golden, CO). Anti-CD29-FITC antibody was purchased from BioLegend (San Diego, CA). Anti-CD90 antibody was obtained from AbD Serotec (Oxford, UK). Anti-angiopoietin-1 (Ang-1) antibody was purchased from Abcam (Cambridge, MA), and human Ang-1 was obtained from R&D Systems (Minneapolis, MN).
ASCs were cultured from male Wistar rats as previously reported with slight modification (15). In brief, inguinal subcutaneous adipose tissue was excised and minced in phosphate-buffered saline (PBS) on ice. The minced tissue was then digested at 37°C for 1 h in PBS containing 2% bovine serum albumin and 2 mg/ml collagenase (Sigma, St. Louis, MO). The digested tissue was filtered through a 100-μm nylon mesh and centrifuged at 600 g for 10 min. After lysis of red blood cells in 1× lysis buffer containing (in mM) 154 NH4Cl, 14 NaHCO3, and 0.1 EDTA (pH 7.3), the pellets were plated in 100-mm dishes at a density of 30,000 cells/cm2 in a 1:1 mixture of Dulbecco's modified Eagle's medium (DMEM) and F12 medium containing 10% fetal bovine serum (FBS). Six hours after the cells were plated, the medium was changed to remove nonadherent cells. The adherent cells were cultured in DMEM-F12-10% FBS and split several times to expand the cells. Passages 2 to 3 were used for the experiments. To induce differentiation into ECs, ASCs were cultured in endothelial growth medium-2MV (EGM; Lonza Walkersville, Walkersville, MD) on fibronectin-coated dishes. EGM consists of endothelial basal medium-2 (Lonza Walkersville) containing 5% FBS plus growth factors such as epidermal growth factor, hydrocortisone, vascular endothelial growth factor (VEGF)-A, basic fibroblast growth factor (bFGF), and insulin-like growth factor (IGF)-1. ASCs were also cultured on fibronectin-coated dishes in endothelial basal medium-2 containing 5% FBS (EBM) as the negative control. Human umbilical vein ECs (HUVECs) were purchased from Sanko-Junyaku (Tokyo, Japan) and cultured using HuMedia-EG (Kurabo, Osaka, Japan). Rat vascular smooth muscle cells (VSMCs) were cultured from rat thoracic aortas following the explant method, as previously described (17), and maintained in DMEM containing 10% FBS. NRK-52E cells, a cell line derived from rat renal tubular cells, were cultured in DMEM containing 5% FBS.
Cultured ASCs were trypsinized and incubated in a blocking buffer (PBS-containing 3% FBS) for 30 min on ice. Approximately 5 × 105 cells were incubated with primary antibodies reactive to CD34, VE-cadherin, CD29, CD90, or isotype-matched control IgGs. After being washed, the cells were incubated with secondary antibodies coupled with FITC, when the primary antibodies were unlabeled. Following the wash, samples were analyzed with an EPICS XL flow cytometer (Beckman Coulter, Fullerton, CA).
RNA extraction and real-time PCR.
Total RNA was extracted using TRIzol reagent (Gibco, Rockville, MD) according to the instructions provided by the manufacturer. Total RNA was subjected to reverse transcription using an Omniscript RT kit (Qiagen, Tokyo, Japan). The expression of a variety of genes including Flk-1, Flt-1, CD31, VE-cadherin, VEGF-A, bFGF, Ang-1, hepatocyte growth factor (HGF), IGF-1, and glyceraldehyde-3-phosphate dehydrogenase was examined by real-time PCR using an SYBR green dye. Primers used are listed in Table 1. Real-time PCR was performed using an ABI PRISM 7000 sequence detection system (Applied Biosystems, Foster City, CA).
Western blot analysis.
Western blot analysis was performed as previously described (21). To measure the contents of Ang-1 in culture medium, the culture medium was concentrated using Amicon Ultra-4 centrifugal filter devices (Millipore, Billerica, MA).
ASCs were cultured in EBM or EGM for 7 days, and the medium was changed to DMEM-0.2% FBS. The medium was collected after 12 h and used for chemotaxis assay. Chemotaxis assay was performed using a chemotaxis assay chamber according to the instructions provided by the manufacturer (Neuro Probe, Gaithersburg, MD). In brief, the medium collected from ASCs that contains chemoattractants was placed under the filter. HUVECs were suspended in DMEM-0.2% FBS at a density of 1 × 106 cells/ml, and 25 μl each of the cell suspension was placed on the filter. After 24 h, the upper side of the filter was scrubbed with a cotton swab and washed with PBS to scrape off cells attaching to the side. The lower side of the filter was fixed with 100% methanol, and the cells on the lower side were stained with hematoxylin. The cell number on the filter was counted in three random high-power fields (×100) in each well, and the average of the cell number was used for statistical analysis.
A recombinant adenovirus that expresses green fluorescence protein (AdGFP) was obtained from Quantum Biotechnologies (Montreal, Canada). ASCs were infected with AdGFP at a multiplicity of infection of 40 and used for in vivo experiments.
Wire injury model.
All procedures involving experimental animals were approved by the Institutional Committee for Animal Research of the Tokyo University. Transluminal mechanical injury to rat femoral artery was performed as previously described (18). Male Wistar rats (8 to 10 wk old) were anesthetized with pentobarbital sodium injected intraperitoneally, and a groin incision was made under a surgical microscope. A guide wire (0.46 mm diameter) was introduced through a small muscular branch of the femoral artery proximally to the aortic bifurcation and withdrawn. ASCs cultured in EBM or EGM (106 cells) for 7 days were injected into the femoral artery and incubated for 30 min with the proximal and distal sides of the artery clamped. In some experiments, ASC suspension was dropped around the femoral artery from the adventitial side after wire injury.
The femoral arteries were fixed by perfusing them with 4% paraformaldehyde and processed for paraffin embedding. Cross sections (2 μm) were cut, deparaffinized, rehydrated, and stained with hematoxylin and eosin. For immunohistochemistry, the sections were incubated with primary antibodies reactive to PCNA and CD31. The sections were then incubated with biotinylated secondary antibody and finally horseradish peroxidase-labeled streptavidin according to the instructions provided by the manufacturer (DAKO). The sections were counterstained with hematoxylin.
Values are means ± SE. Statistical analyses were performed using analysis of variance followed by the Student-Newman-Keuls test. Differences with a P value of <0.05 were considered statistically significant.
Characterization of ASCs.
Cell surface markers were first analyzed using flowcytometry analysis (Fig. 1). In contrast to previous reports showing that ASCs isolated from humans expressed CD34 to some extent (9, 15, 16), CD34 expression was negative in ASCs derived from Wistar rats. The expression of VE-cadherin was also negative, suggesting that the contamination of ECs was negligible. The expression of CD29 and CD90 was positive. These cell surface markers were reportedly positive in BMMSCs (14). Therefore, our results suggested that ASCs isolated from Wistar rats resembled mesenchymal stem cells rather than hematopoietic stem cells.
Expression patterns of mRNA and protein in ASCs.
We cultured ASCs on fibronectin-coated dishes and examined mRNA expression in ASCs cultured in either EBM or EGM (Fig. 2A). The expression of Flk-1 was not significantly induced in ASCs cultured in EGM compared with those cultured in EBM. The expression of mature EC markers such as CD31 and VE-cadherin was significantly suppressed when ASCs were cultured in EGM (data not shown). In contrast, the expression of Flt-1 significantly increased in ASCs cultured in EGM compared with those cultured in EBM. We also examined the mRNA expression of proangiogenic factors and antiapoptotic factors that ASCs might secrete. The expression of VEGF-A and IGF-1 significantly decreased in ASCs cultured in EGM compared with those cultured in EBM, probably because EGM contains VEGF-A and IGF-1 to induce the differentiation into ECs. The expression of bFGF was not significantly changed between EBM-cultured ASCs and EGM-cultured ASCs. The expression of HGF was significantly suppressed in ASCs cultured in EGM. In contrast, the expression of Ang-1 significantly increased in ASCs cultured in EGM compared with those cultured in EBM. We next examined the expression level of some of these genes at the protein level (Fig. 2B). The expression of Flk-1, CD31, or VE-cadherin was not detected in ASCs cultured in EGM until up to 14 days after incubation with EGM. In contrast, the expression of Flt-1 was detected 3 days after incubation with EGM and peaked 7 days after culture in EGM. The ASCs cultured in EGM also secreted a higher amount of Ang-1 into the culture medium than EBM-cultured ASCs. Since Flt-1 is expressed in monocytes/macrophages as well as ECs (20), the expression of CD14 was examined by Western blot analysis and immunostaining, but its expression was not detected in ASCs cultured in EGM (data not shown), suggesting that the contamination of monocytes/macrophages was negligible. Collectively, our data suggested that ASCs do not appear to have the potential to differentiate into mature ECs in vitro, although ASCs express Flt-1. Our results also suggested that ASCs might have the capacity to promote neovessel formation via a stimulation of the recruitment of ECs in situ, because ASCs, especially cultured in EGM, produced a significant amount of Ang-1.
ASCs stimulate migration of HUVECs.
We, therefore, examined whether ASCs would stimulate the migration of ECs by chemotaxis assay (Fig. 3). Ang-1 was used as the positive control for the chemotaxis assay. Ang-1 (10 ng/ml) significantly stimulated the migration of HUVECs, and this effect was significantly suppressed when Ang-1 was preincubated with anti-Ang-1 antibody. Culture medium harvested from EGM-cultured ASCs significantly stimulated the migration of HUVECs compared with that harvested from EBM-cultured ASCs. This stimulatory effect was partially but significantly blocked by the preincubation of the culture medium with anti-Ang-1 antibody, suggesting that Ang-1 was, at least partly, responsible for the migration-stimulating effect.
ASCs inhibit neointimal formation via stimulation of endothelial repair in a paracrine fashion.
We next examined the function of ASCs in vivo using the wire injury model of the rat femoral artery. When injected in the femoral artery, EBM-cultured ASCs slightly but significantly inhibited neointimal formation compared with wire injury without cell administration. EGM-cultured ASCs potently and more significantly inhibited neointimal formation compared with EBM-cultured ASCs (Fig. 4A). In accordance with these results, the number of PCNA-positive cells in the neointima was significantly suppressed in the group that was administered EBM-cultured ASCs compared with the group that received no cells. The number of PCNA-positive cells in the neointima was more significantly reduced in the group that was administered EGM-cultured ASCs compared with the group that was administered EBM-cultured ASCs (Fig. 4B). Because ASCs cultured in EGM potently inhibited neointimal formation, we studied the mechanism whereby these cells inhibited neointimal formation. We examined whether EGM-cultured ASCs were engrafted into the endothelial layer and contributed to the repair of the endothelial layer after the wire injury. EGM-cultured ASCs were infected with AdGFP before injection into the artery. One day after injection, EGM-cultured ASCs were detected in the endothelial layer. However, EGM-cultured ASCs were barely detected in the endothelial layer 3 and 14 days after injection (Fig. 5A), suggesting that EGM-cultured ASCs inhibited neointimal formation without integrating into the endothelial layer. To confirm the specificity of the green fluorescence detected in the endothelial layer, we also injected ASCs without AdGFP infection and examined autofluorescence of the femoral artery (Fig. 5B). Green fluorescence was not detected in the endothelial layer in this case, suggesting that the green fluorescence detected in the endothelial layer was derived from AdGFP-infected ASCs that stayed in the endothelial layer. We, therefore, hypothesized that EGM-cultured ASCs potently inhibited neointimal formation by secreting paracrine factors that stimulate the repair of the endothelial layer, because we found that these cells produce a significant amount of Ang-1. To examine this possibility, we next administered EBM- and EGM-cultured ASCs from the adventitial side of the femoral artery after wire injury (Fig. 5C). Interestingly, EGM-cultured ASCs more significantly inhibited neointimal formation compared with EBM-cultured ASCs even when these cells were administered from the adventitial side. To further examine the role of paracrine factors secreted by ASCs, we originally tried to knock down endogenous Ang-1 production using small interfering RNA technology. However, the transfection efficiency of small interfering RNA by lipofection into rat ASCs was <10%, making it very difficult to examine the effect of gene knockdown in ASCs. We instead used rat VSMCs that also produce Ang-1 and NRK-52E cells that barely produce Ang-1. Rat VSMCs produced ∼50% of Ang-1 mRNA compared with ASCs cultured in EBM, and NRK-52E cells produced <1% of Ang-1 mRNA compared with ASCs cultured in EBM, as assessed by real-time PCR (data not shown). When rat VSMCs were administered from the adventitial side, they slightly but significantly inhibited neointimal formation compared with the NRK-52E cells administration (Fig. 6). These results also suggested that Ang-1 produced by ASCs might be, at least in part, implicated in the suppression of neointimal formation. We finally examined whether ASC administration from the adventitial side would stimulate the repair of the endothelial layer (Fig. 7). We examined the ratio of the endothelial layer positively stained with CD31. Endothelial repair was significantly enhanced by the administration of EGM-cultured ASCs compared with administration of EBM-cultured ASCs.
In this study, we isolated ASCs from Wistar rats and examined their characteristics in vitro and in vivo. ASCs obtained from Wistar rats expressed CD29 and CD90 but not CD34, suggesting that the ASCs we used resembled BMMSCs rather than hematopoietic stem cells. Although several studies demonstrated that human ASCs contain a large population of cells that express CD34 and that these CD34-positive cells differentiate into endothelial-like cells in vitro and in vivo (9, 15), ASCs do not always express CD34 in mice (6, 11). Although we do not know the reason for this discrepancy, cell surface markers of ASCs may differ among species.
ASCs used in this study expressed Flt-1 when they were cultured in EGM. However, ASCs did not express Flk-1 or mature EC markers such as VE-cadherin and CD31. Thus ASCs used in this study did not appear to have the capacity to differentiate into mature ECs. ASCs may resemble bone marrow-derived cells that express Flt-1 and are recruited to sites of ischemia (7, 12). Although several studies showed that human ASCs had the potential to differentiate toward ECs in vitro by demonstrating the expression of endothelial markers such as CD31, the efficiency varies so much, probably because the methodology whereby they induced differentiation of ASCs into ECs differs from study to study (3, 8, 9, 15). One study demonstrated that mouse ASCs could differentiate into ECs in vitro by examining the expression of CD31 and VE-cadherin, but its efficiency seemed to be very low (11). Recently, Boquest et al. (2) examined the methylation profiles of EC-specific gene promoters such as CD31 and VE-cadherin and showed that the promoters of CD31 and VE-cadherin were hypermethylated in ASCs and that these promoters seemed to have a relatively small potential to be activated in ASCs. Moreover, whether ASCs can differentiate into ECs in vivo remains debatable. Several studies demonstrated that ASCs were integrated into capillaries in hindlimb ischemia models and improved blood flow via the stimulation of angiogenesis (9, 15). However, ASCs could reportedly stimulate angiogenesis and restore blood flow in hindlimb ischemia models without being engrafted into capillaries, probably because of their paracrine effects (11). Therefore, the efficiency of ASCs to differentiate into ECs in vitro and in vivo appears to differ, depending on the cell culture conditions, animal model used, and animal species. Future studies will be required to elucidate an appropriate strategy to efficiently induce differentiation of ASCs into mature ECs.
We, therefore, examined paracrine factors that ASCs produce. ASCs reportedly produce a variety of proangiogenic and antiapoptotic factors such as VEGF, IGF, HGF, and bFGF. Although the production of these factors did not increase when ASCs were cultured in EGM, the production of Ang-1 was significantly increased. Furthermore, ASCs appeared to secrete functionally active Ang-1, as assessed by chemotaxis assay using HUVECs. These results suggested that ASCs potentially promote repair of the endothelial layer via stimulation of migration of ECs in situ.
To test this hypothesis, we administered ASCs in a wire injury model of rat femoral artery. EGM-treated ASCs significantly inhibited neointimal formation without being engrafted into the endothelial layer. EGM-treated ASCs also significantly suppressed neointimal formation even when they were administered from the adventitial side. Endothelial repair occurred more rapidly in rats administered EGM-cultured ASCs compared with those administered EBM-cultured ASCs. The rapid endothelial repair was accompanied by less cell proliferation in the neointima. Furthermore, the administration of VSMCs that also produce Ang-1 slightly but significantly suppressed neointimal formation, whereas the administration of NRK-52E cells that barely produce Ang-1 did not. These results suggested that ASCs inhibited neointimal formation in a paracrine fashion via the stimulation of endothelial repair. These results also suggested that Ang-1 was, at least in part, implicated in the inhibitory effect of ASCs on neointimal formation. It has been reported that EPCs, when injected in the carotid artery, were engrafted in the endothelial layer in a balloon injury model. However, it remains unclear how long the injected EPCs can survive and proliferate in the endothelial layer. EPCs were not detected in the endothelial layer 30 days after the administration in that study (5). Thus, although EPCs seem to be more effectively integrated in the endothelial layer than ASCs, EPCs may also stimulate endothelial repair via a stimulation of EC migration in situ, because EPCs also reportedly produce paracrine factors such as VEGF (23). These possibilities should be addressed in the future.
In summary, although rat ASCs do not differentiate into mature ECs, they produce paracrine factors such as Ang-1, especially when they were cultured in EGM. These factors seem to stimulate the migration of ECs in situ and the repair of the endothelial layer in vivo. Although the capacity of ASCs to differentiate into mature ECs may be low, ASCs will be useful for cell-based therapy to treat cardiovascular diseases such as hindlimb ischemia, acute myocardial infarction, and the prevention of restenosis after angioplasty via their capacity to produce paracrine factors that stimulate angiogenesis and endothelial repair.
This study was supported in part by research Grants from Japan Science Technology Agency (Core Research for Evolutionary Science and Technology) (to T. Nagano and Y. Hirata).
There exist no conflicts of interest.
↵* M. Takahashi and E. Suzuki contributed equally to this work.
- Copyright © 2010 the American Physiological Society