Objective: The low number of postnatal endothelial progenitor cells (EPC) in the circulation limits their therapeutic application in cardiovascular medicine. Processed lipoaspirate (PLA) cells differentiate into osteoid, adipose, muscle, and cartilaginous cells. This study examines the potential of PLA cells as a source of EPCs.
Methods: PLA cells obtained from human lipoaspirates were cultured for 1 week in serum-depleted medium to form three-dimensional cell clusters (3DCC). The phenotype of 3DCC-derived cells was assessed by immunofluorescense staining and FACS analysis.
Results: Flow cytometry showed that 45 ± 5% of cells derived from the 3DCC expressed Flk-1, a marker of early EPC, whilst only 4 ± 0.5% of freshly isolated PLA were Flk-1+. The proportion of Flk-1+ cells increased to 98 ± 2% during culture in hematopoietic stem cell medium. When cultured in an endothelial cell (EC)-specific medium, Flk-1+ cells also expressed Ve-cadherin, von Willebrand's factor (vW), and a lectin receptor, and took up low-density lipoprotein. Incorporation into an endothelial cell tubular network confirmed their functional activity.
Conclusion: This report describes the first isolation and culture of Flk-1+ cells from human adipose tissue. The feasibility of the extraction and culture of these cells in increased numbers suggests that such autologous cells will be useful for applications ranging from basic research to cell-based therapies.
Recent findings indicate that hemangioblasts or more mature endothelial progenitor cells (EPC) persist into adult life, at which time they may circulate, differentiate, and contribute to the formation of new blood vessels [1,2].
Rapid revascularization of injured, ischemic, and regenerating organs is essential for the restoration of their physiological function. The angiogenic switch initiates the revascularization process by recruiting EPC that assemble into neovessels. Because tissue injury disrupts the permissive environment necessary for the recruitment of EPC, introduction of exogenous progenitors may facilitate revascularization. To date, umbilical cord blood and bone marrow have represented the main sources of EPC in postnatal life .
Like bone marrow, adipose tissue contains an easily isolated stroma. This cell population, processed lipoaspirate (PLA) cells, can be isolated from human adipose tissue aspirates, and like mesenchymal stem cells, differentiates into bone, fat, muscle, and cartilage cells . Several studies suggest that mesoderm-inducing factors affect the development of embryoid bodies (EB) and hematopoietic differentiation [5,6]. The in vitro differentiation of embryonic stem cells illustrates the importance of mesodermal cells in the development of endothelial cells (EC) .
Flk-1, a receptor for vascular endothelial growth factor (VEGF), is one of the earliest markers of EPC . The results presented here identify a novel source of Flk-1+ cells in adult life, namely, human adipose tissue, and show that these cells provide a source of mature EC. This raises the possibility of infusing autologous Flk-1+ cells to repair injured blood vessel walls, initiate neovascularization, or facilitate the regeneration of ischemic tissue, among other potential applications.
2. Materials and methods
2.1. PLA cell isolation
Fig. 1 summarizes the main steps in cell culture and differentiation. All tissue culture reagents were obtained from CAMBREX (Life Technologies), and Falcon™ plastic cell culture ware was from Becton-Dickinson Labware. Human adipose tissue aspirates were collected conforming to the principles outlined in the Declaration of Helsinki.
PLA from fresh human lipoaspirates were cultured as described elsewhere . Briefly, contaminating debris and red blood cells were removed by extensive washing of the aspirates in phosphate-buffered saline (PBS). The cells were then dispersed by gentle agitation of the washed aspirate for 30 min at 37 °C in 0.015% collagenase (Type II-B, Sigma-Aldrich). The collagenase was inactivated by dilution with DMEM containing 10% fetal calf serum (FCS), 2 mM glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin (DMEM-FCS). The cell suspension was then centrifuged for 10 min at 1200 × g and the pellet was resuspended in DMEM-FCS and filtered though a 100-μm mesh. The cells were plated in conventional culture plates and grown to confluence in DMEM-FCS.
Confluent cells were split and seeded in DMEM-FCS at a density of 40,000 cells/cm2. Four days after seeding, cell differentiation was initiated by replacing the culture medium with DMEM lacking FCS. Within 1 week of culture in the absence of FCS, PLA formed three-dimensional cell clusters (3DCC). These were scraped off and 2–3 clusters were seeded per well in a 6-well plate in DMEM-FCS. Endothelium-like cells began to grow within 24–48 h. These cells were subcultured in hematopoietic stem cell medium [RPMI-1640: DMEM: F12, 0.1% bovine serum albumin (BSA) and 50 μM 2-mercaptoethanol] . This Flk-1+ population was seeded into collagen IV-coated flasks and grown in endothelial cell medium, EGM-2 containing 2% fetal bovine serum, supplemented with single aliquots of vascular endothelial growth factor (VEGF), fibroblast growth factor-2, epidermal growth factor, insulin-like growth factor-1, and ascorbic acid (Clonetics).
2.3. Flow cytometry
Trypsinized cells were washed in PBS and incubated for 30 min at 37 °C in flow cytometry buffer (PBS containing 1% BSA and 0.1% azide) containing conjugated antibodies to CD133 (Miltenyi Biotech), CD34, CD45 (Becton-Dickinson), CD31 (Dako), Ve cadherin (Bender MedSystems), Flk-1 (Sigma-Aldrich), or, to assess background fluorescence, conjugated nonspecific IgG.
2.4. LDL uptake and lectin binding
LDL uptake was assessed by incubating cells for 4 h at 37 °C with 2.5 μg /mL acetylated LDL labeled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine (acLDL-DiI, Molecular Probes). Cells to be analyzed by fluorescence microscopy were then fixed for 15 min in 1% paraformaldehyde and incubated for 2 h in 10 μg/mL FITC-labeled Ulex europaeus agglutinin-1 (UEA-1).
Cells grown to confluence on glass coverslips coated with human fibronectin (7 μg/mL; Sigma-Aldrich) were fixed for 20 min with 3% paraformaldehyde and permeabilized for 5 min with PBS containing 0.1% Triton X-100 . Monolayers were then processed for indirect immunofluorescence staining with polyclonal antibody against von Willebrand's factor (vW; Dako) incubated with Alexa488-conjugated goat anti-Rabbit (Molecular Probes) and analyzed by confocal microscopy.
2.6. Matrigel tubule assay
Thawed Matrigel (Sigma) was added to 24-well plates and allowed to solidify at room temperature. acLDL-DiI-labeled EPCs and human umbilical vein endothelial cells (HUVECs) were seeded in Matrigel-containing wells at 5 × 104 and 6 × 104, respectively, and cocultured at 37 °C. Following completion of capillary formation (24 h), three representative images were recorded from each well using a Leica inverted microscope equipped with digital image capture software.
3.1. Development of 3DCC
Fig. 2 illustrates the development of 3DCC. Initially, monolayers formed that consisted mainly of single cells (Panel A). After 1 week, small cell clusters were observed (Panel B). When cultures were maintained for up to 1 month, the size of the 3DCC increased (Panels C and D). Assay of the ability of these cells to differentiate into adipocytes confirmed their plasticity (data not shown).
3.2. Phenotype of cells derived from 3DCC
3DCC from 2-week cultures were replated and grown in DMEM-FCS. The morphology of the adherent cells growing out from the margins of the clusters resembled that of endothelial cells (Fig. 3A). In five independent experiments, FACS analysis revealed that Flk-1+ cells comprised 45 ± 5% of the population, whereas in freshly isolated PLA, that fraction was only 4 ± 0.5%. These cells were also negative for CD45, a lymphocyte marker, and for CD34 and CD133, early hematopoietic markers (data not shown). Culture of cells in hematopoietic cell medium increased the proportion of Flk-1+ cells from 45 ± 5% to 98 ± 2% (Fig. 3B). In addition, the ability of these cells to proliferate confirmed their expansion capacity (data not shown).
To further characterize the differentiation of PLA cells into Flk-1+ cells, we sorted Flk-1− cells and subcultured them under the same conditions as described for 3DCC. After 1 week in FCS-free medium, 3DCC clusters began to appear, and after culture in hematopoietic medium, FACS analysis showed that 80 ± 5% were Flk-1+, demonstrating once again the capacity of PLA cells to differentiate (Fig. 4A and B).
Culturing populations of 98% Flk-1+ cells in EC medium for 1 day resulted in expression of Ve cadherin, a marker of mature EC, in 30 ± 0.5% of the cells, and this population expanded to 95 ± 2% by the third day (Fig. 5A). The proportion of Flk-1+ cells decreased over time, evidence of differentiation into mature EC. Cells allowed to differentiate for 3 days displayed a cobblestone monolayer typical of EC (Fig. 5B). In addition, they incorporated acetylated LDL and bound UEA-1 (Fig. 5C). Immunostaining for vW was also consistent with the phenotype of mature EC (Fig. 5D). When cocultured with HUVEC, these cells were incorporated into a tubular network, demonstrating their functional competence (Fig. 5E).
Stem cell therapy is a promising approach in cardiovascular medicine. There are now reports  of clinical studies in which EPC have been used for neovascularization of ischemic organs. However, the paucity of EPCs in the circulation limits their therapeutic application. The use of umbilical cord blood, mobilization of EPCs by growth factors, or the local infusion of suspensions of autologous bone marrow cells have been suggested as ways in which to circumvent this problem. However, traditional bone marrow procurement is painful and yields a low number of mesenchymal stem cells .
Our experiments indicate for the first time that PLAs are a satisfactory source of Flk-1+ cells, with a high potential for use in in vitro models of human vascular development and for therapeutic applications.
Although the number of Flk-1+ cells in PLA cultures was low, culture in serum-free medium resulted in enrichment of this cell type. This culture technique allowed the growth of new cellular aggregates (3DCC) that had a morphology resembling that of the blood islands derived from EB, which produce a high percentage of Flk-1+ cells.
Analysis by immunostaining and FACS demonstrated the ability of Flk-1+ cells to differentiate into mature ECs, whilst incorporation into HUVEC tubular networks confirmed their functional competence.
Several studies suggest that mesoderm-inducing factors influence EB development and hematopoietic differentiation . Furthermore, the secretion of angiogenic and antiapoptotic factors by human adipose stromal cells has been reported . This could explain the growth of 3DCC in the absence of serum, and the presence of Flk-1+ cells in these structures. In addition, there are recent reports indicating that adipose tissue, especially the stromal vascular fraction, might contain progenitor cells that can differentiate into mature ECs [13,14].
The method described here yields a large number of Flk-1+ cells that are able to differentiate into mature ECs. Additionally, liposuction for harvesting adipose tissue from which to obtain PLA is technically straightforward, safe, causes relatively little discomfort, and is free of ethical drawbacks. These results raise the possibility of using PLAs as a source of Flk-1+ cells and mature ECs for therapeutic angiogenesis.
We thank Robin Rycroft for his editorial support. JJ is the recipient of a research contract from the Spanish Ministry of Science and Technology (PTQ2002-0272, Programa Torres Quevedo). This work was supported by grants from the Spanish Ministry of Science and Technology (SAF2001-0480 and SAF2004-05481).