- Top of page
- Defining human endothelial progenitor cells
- Disclosure of Conflict of Interests
Summary. There is no specific marker to identify an endothelial progenitor cell (EPC) and this deficiency is restricting the ability of an entire field of research in defining these cells. We will review current methods to define EPC in the human system and suggest approaches to define better the cell populations involved in neoangiogenesis. PubMed was used to identify articles via the search term ‘endothelial progenitor cell’ and those articles focused on defining the term were evaluated. The only human cells expressing the characteristics of an EPC, as originally proposed, are endothelial colony forming cells. A variety of hematopoietic cells including stem and progenitors, participate in initiating and modulating neoangiogenesis. Future studies must focus on defining the specific hematopoietic subsets that are involved in activating, recruiting, and remodeling the vascular networks formed by the endothelial colony forming cells.
Defining human endothelial progenitor cells
- Top of page
- Defining human endothelial progenitor cells
- Disclosure of Conflict of Interests
The term endothelial progenitor cell (EPC) remains difficult to define due to the lack of a specific and unique cell surface or molecular determinant that permits prospective isolation of this cell. In this overview, we will discuss the most common methods for EPC identification and highlight some pitfalls that may be encountered with each approach. We will discuss some of the limitations to the field that will need to be overcome for the development of a consensus definition of a human circulating EPC. It is important that the field continues to strive to identify accurately which cells form the EPC population in order to assure future clinical trials of using the most approriate cell therapy.
In the human system, EPC are defined using three general approaches. First, peripheral blood mononuclear cells are isolated and plated on fibronectin coated tissue culture plates with a variety of endothelial growth factors. After several days, the non-adherent cells are removed and the adherent cell population remaining, which express the ability to ingest acetylated low density lipoprotein and to bind certain plant lectins, is considered to represent EPC. An inverse correlation between the circulating concentration of these putative EPC and an increased risk for developing coronary arterial disease has been reported [1,2]. However, this definition fails to recognize that fibronectin coated tissue culture wells have been used to isolate human blood monocytes . Human peripheral blood monocytes can be cultured in serum containing medium with a variety of growth factors that promote expression of numerous proteins normally expressed by primary endothelial cells [4–6] and such modulating culture conditions may lead to pathways of alternative macrophage activation that support enhanced wound healing and angiogenic activities [7,8]. Thus, use of the short term adhesion and culture method for putative EPC isolation is fraught with complexities in identifying which cell may be playing a functional role attributable to this cultured population.
A second approach to putative EPC identification utilizes monoclonal antibodies and fluorescence activated cell sorting (FACS) analysis to enumerate specific cell populations. Asahara et al.  reasoned that putative EPC may express cell surface markers shared by hematopoietic stem cells (HSC) since endothelial and blood cells share a similar mesodermal origin during embryonic development. Thus, Asahara et al. cultured CD34+ cells (15.7% enriched) on fibronectin coated dishes and observed emergence of spindle shaped cells that expressed a variety of proteins generally expressed by primary endothelial cells. Furthermore, the CD34+ cells or cells expressing vascular endothelial growth factor receptor 2 (KDR) localized, upon intravenous injection, to ischemic regions in immunodeficient mice following instrumentation to produce hindlimb injury. In sum, this paper suggested that CD34+ cells in human adult peripheral blood may function as EPC in postnatal vascular repair.
As some endothelial cells may circulate in the bloodstream in healthy and diseased human subjects [10,11], Peichev et al.  attempted to devise a separation protocol to identify EPC from circulating mature endothelial cells (CECs). In addition to searching for cells expressing CD34 and KDR, these authors included CD133 expression as a discriminating marker. Indeed cells expressing CD34, KDR, and CD133 were identified from mobilized adult peripheral blood, umbilical cord blood, and human fetal liver samples. This expression pattern was also observed on some cells coating the luminal surface of implanted left ventricular assist devices in human subjects suggesting that one could use CD34, CD133, KDR as markers for circulating EPC in human subjects . Of interest, this triple positive population was never directly injected into human subjects and demonstrated to home to the implanted device.
Subsequently, numerous papers have been published where an EPC population has been identified using CD34 or CD133 or KDR or any combination thereof (reviewed in Ref. ) and the concentration of the EPC has been used to correlate with the severity of a disease state [14–17]. However, recent studies have clarified that the actual cell population enriched in the CD34, CD133, KDR fraction is a hematopoietic progenitor, not an EPC. Whether examining umbilical cord blood or mobilized adult peripheral blood, cells expressing CD34, CD133, and KDR are enriched for hematopoietic colony forming cells and they do not form endothelial cell lined vessels in vivo [18,19]. Thus, at present, the three antigen combination of Peichev et al.  has failed to identify EPC specifically.
The third approach to defining EPC relies on in vitro colony forming cell assays. Two assays have been evaluated extensively and include the colony forming unit-Hill (CFU-Hill)  and endothelial colony forming cell (ECFC) assays [21–23]. In a direct comparison of these assays , adult peripheral blood or cord blood plated in the CFU-Hill assay gives rise to cells that do express many proteins similar to primary endothelial cells, but the CFU-Hill also express numerous myeloid progenitor cell markers and mature into macrophages that express non-specific esterase and ingest bacteria . Finally, CFU-Hill and their progeny fail to spontaneously form human blood vessels when implanted into immunodeficient mice. In contrast, ECFC express cell surface antigens like primary endothelium, clonally propagate and replate into secondary and tertiary ECFC, form capillary-like structures in vitro, but most remarkably, form human blood vessels in vivo (in immunodeficient mice) and inoculate with murine vasculature to become part of the murine systemic circulation . Thus, ECFC display all of the properties of an EPC while the CFU-Hill assay identifies hematopoietic cells. Other investigators have subsequently confirmed these dramatic differences in putative EPC populations using a co-culture assay for testing in vitro angiogenic activity  or other measures of cell identity [26,27].
A potential role of hematopoietic cells in driving the angiogenic process is being increasingly appreciated  and may be an important correlate of patient disease state . In an attempt to develop a robust clinically useful protocol to examine both circulating progenitor cells (CPC) and CEC, Duda et al.  utilized monoclonal antibodies to four cell surface markers to discriminate between CPC and CEC. Duda et al. reported that CEC can be defined as a discrete population of cells expressing CD31brightCD45−CD34dimCD133− whereas CPC can be defined as CD31+CD45dim CD34brightCD133+ cells. Use of this protocol permitted identification of CEC at a frequency of 0.1%–6.0% of blood mononuclear cells and has been used as a marker for the predicting the outcome of women with breast cancer .
The ability to identify CEC using a standard FACS approach drew correspondence from Dignat-George et al. . Previous work in the field had made it clear that the ability to identify primary human endothelial cells spiked into a blood sample was limited to a minimum of 180 endothelial cells mL−1 of blood . More recently, an immunomagnetic separation protocol utilizing CD146 had been validated as a robust assay for measuring CEC . In normal human subjects, the number of CEC typically ranges from 0 to 10 cells mL−1 of blood. Thus, a comparison of the number of CEC identified in the protocol by Duda et al.  with the data from Woywodt et al.  has demonstrated nearly a one log difference.
Duda et al.  responded to this correspondence by pointing out that the differences in the number of CEC measured in the blood of human subjects could be due to the immunomagnetic separation technique and the FACS protocol identifying CD31brightCD45−CD34dimCD133− cells may represent two different populations. Indeed the CD31brightCD45−CD34dimCD133− events did not show evidence of CD146 expression. While Duda et al.  have not yet sorted cells from the CD31brightCD45−CD34dimCD133− events to examine the cell size and morphology for evidence that these events are indeed CEC, the fact that this population diminishes in the blood of cancer patients receiving an antiangiogenic agent, suggests that these cells are playing an important role in the disease pathogenesis .
The differences in methods to isolate CEC are relevant to the broader discussion of defining an EPC since ECFC may be derived from the endothelial intima of blood vessels . The site of origin of the circulating ECFC is presently unknown, although Lin et al.  have suggested that the human circulating endothelial cells with the greatest proliferative potential may be transferrable via a bone marrow transplant. The specific lineage of origin of the human circulating endothelium with high proliferative potential from the bone marrow has not been delineated though numerous reports suggest an origin distinct from the HSC .
Our working model of the cellular elements that participate in neoangiogenesis includes a variety of hematopoietic cells, rare circulating ECFC, and the ECFC resident in the vascular endothelium (Fig. 1). We hypothesize that vascular or tissue injury, ischemia, or the presence of a tumor may stimulate the recruitment of hematopoietic cells which attach to the endothelium present in the site of interest, disrupt the endothelial barrier, migrate into and through the vessel and into the matrix of the tissue, release chemotactic and growth promoting factors for the ECFC resident in nearby vascular endothelium, and stimulate the ingrowth of the ECFC to form the vessels that will restore normal blood flow to the injured site or be sequestered by the tumor for the angiogenic switch and unregulated growth. This working model fits more closely with traditional views of angiogenesis rather than evoking postnatal vasculogenesis (and large numbers of EPC) as a major component of the vascular ingrowth.
Figure 1. Circulating and resident cells involved in neoangiogenesis. This figure depicts the rare circulating low proliferative potential (LPP) and high proliferative potential endothelial colony forming cells (HPP-ECFC) that become the new vessels at neoangiogenesis sites. More abundant circulating hematopoietic cell subsets modulate the initiation, recruitment, and formation of the new vessels via stimulation of the circulating and resident ECFC.
Download figure to PowerPoint
Challenges in moving forward in the field of neoangiogenesis will be to develop methods to specifically identify the hematopoietic cells that are implicated as playing important roles in angiogenesis. It is quite possible that the type of injury, site of injury, and age of the patient may all influence the number, type, and functional competence of bone marrow-derived hematopoietic cells that are recruited to initiate the process of angiogenesis. It is also probable that the proliferative potential of the ECFC within the vasculature may vary with age and disease process and thus, may be more or less amenable to recruitment by the hematopoietic cells to form the new vessels via migration and sprouting angiogenesis. While a search for novel antigens that permit prospective isolation of the circulating and resident ECFC is warranted, improved methods to measure these rare cells in normal and diseases subjects must be developed. The recent enrichment of CEC using immunomagnetic beads and subsequent FACS analysis reported by Widemann et al.  may be a fruitful beginning.