Umbilical cord blood (UCB) has serve as a source of hematopoietic stem and progenitor cells as similar as bone marrow (BM). Allogeneic UCB transplantation has been used to treat a variety of malignant diseases like leukemia and lymphoma, and non-malignant diseases such as severe hematological diseases and inherited metabolic diseases (1). Recent studies provide convincing evidence that UCB contains not only hematopoietic progenitors, but also several types of stem and progenitor cells (Table 1). They include highly immature cell fractions like embryonic-like cells providing a high proliferative capacity and a variety of differentiation potentials (2–4). UCB-derived stem cells have several advantages over adult stem cells like BM, because UCB can be collected without painful procedure, easily available in virtually unlimited supply, and has not been exposed to immunologic challenge. In addition, cord blood transplantation is now an established field with great potential and no serious ethical issue by establishment of public UCB banks throughout the world. Therefore UCB-derived non-hematopoietic stem cells may provide an attractive cell source for tissue repair and regeneration. It is generally accepted that UCB contains endothelial progenitor cells (EPCs) (5), mesenchymal stromal cells (MSCs) (6), unrestricted somatic stem cells (USSC) (2), very small embryonic-like stem cells (VSELs) (3), multi-lineage progenitor cells (MLPCs) (4), and neuronal progenitor cells (7). This review focuses on the isolation, characterization, and differentiation capacity of the non-hematopoietic stem/progenitor cells derived from human UCB and their potential use in cell based therapies.

Circulating EPCs in adult human peripheral blood were originally identified from CD34⁺ mononuclear cells (MNCs) in 1997 by Asahara et al. (8). CD34⁺ cells from human peripheral blood form clusters of cells comprised of round cells centrally and sprouts of spindle-shaped cells at the periphery within 5 days, when the cells were plated on fibronectin-coated dishes in endothelial growth media. Murohara et al. (9) has shown that UCB is a valuable source of EPCs, and transplantation of UCB-derived cultured EPCs augments neovascularization and blood flow in a rat ischemic hind limb. Cultured EPCs from cord blood have a greater proliferative activity than those from other sources (5). A number of experimental and clinical studies have revealed that ischemic cardiovascular diseases such as peripheral artery disease, acute myocardial infarction, brain infarction, and diabetic neuropathy can effectively be treated by UCB-derived EPC transplantation (5, 10). It has been speculated that the paracrine effect of cultured EPCs are responsible for the modest effects in patients because there is no evidence of long-term engraftment of EPCs into newly formed vessels. The cultured EPCs characterized by endothelial cell-colony forming units (CFU-ECs) express not only endothelial markers CD31, CD105, CD144, CD146, vWF, UEA-1 and KDR, but also monocyte/macrophage markers CD14, CD45, and CD115. In addition, the cultured EPCs possess myeloid progenitor cell activity, differentiate into phagocytic macrophages, and fail to form perfused vessels in vivo (11). These findings suggest that the cultured EPCs are hematopoietic myeloid progenitor cells. Recently, Ingram et al. (12) have identified other EPCs with blood vessel-forming ability, termed endothelial colony-forming cells (ECFCs), which are also referred to as blood outgrowth endothelial cells (13), from human peripheral blood and UCB. ECFC colonies appear on days 5 to 10 for UCB or day 14 to 21 for adult peripheral blood samples when the MNCs are cultured on collagen-coated tissue dishes in endothelial growth media. The number of colonies generated from 10⁶ MNCs was significantly smaller in ECFCs (0.017±0.04) compared to that in CFU-ECs (4.35±2) (11). ECFCs express endothelial markers CD31, CD105, CD144, and CD146, but not hematopoietic cell markers CD45 and CD115. ECFCs are characterized by robust proliferative potential and by their ability to form perfused blood vessels in vivo when transplanted with collagen fibronectin matrix into immunodeficient mice (11, 12). ECFCs are enriched in UCB compared to adult peripheral blood. In addition, UCB-derived ECFCs have greater proliferative activity and enhance vessel forming ability compared to adult peripheral blood-derived ECFCs (12, 14). Thus, UCB-derived ECFCs may more effectively contribute to vascular regeneration.

MSCs are adherent stromal cells, initially isolated from the bone marrow (15). MSCs have been defined by their ability to self-renew and differentiate into the cells that form mesodermal tissue. Recent studies demonstrated that MSCs also show ability to suppress immune responses in vitro and in vivo. In contrast to hematopoietic stem cells, MSCs lack an unique surface antigen for positive selection. MSCs are suggested to be positive for CD13, CD29, CD44, CD73, CD90 (Thy-1), CD105 (endoglin), and CD166 and negative for CD14, CD34, CD38, and CD45. From 2000, several researchers reported that UCB contains multipotent MSCs (6). Subsequently, Lee et al. (16) has shown that clonally expanded MSCs from UCB exhibit multilineage differentiation potential including osteogenic and chondrogenic lineages. It has been shown that UCB MSCs also can differentiate into cardiomyocytes (17), hepatocyte (16, 18), and neuroglial cells (16, 19). Interestingly, UCB MSCs exhibit no or less adipogenic differentiation capacity (20–22). Recent studies demonstrated that frequency of MSC colony in UCB is much lower than that in the other tissue such as BM, adipose tissue, and umbilical cord tissue (Wharton's jelly) (21, 23, 24). The ability to obtain MSCs is from only 30% to 60% of UCB units with a frequency of up to 2.3 MSC clones/108 mononuclear cells (25, 26). Javed et al. (27) reported that 24∼28 week gestational age UCB generate predominantly MSC colonies, although MSCs are rarely identified in 37∼40 week gestational age UCB, suggesting circulating concentration of MSCs is rapidly decreased in full-term UCB. In addition, the isolation efficiency of UCB MSCs depends on time from collection to isolation and volume/cell number of UCB (20). Additional studies on development of the efficient isolation and expansion procedures for UCB MSCs are expected. MSCs possess various properties of clinical interest, including their wide-ranging differentiation potential, their capacity for engraftment (28, 29), their immunosuppressive effects (30) and their supporting effects on in vivo expansion of HSCs (31).

In 2004, Kögler et al. (2) reported a pluripotent cell population within UCB, which they named USSC. USSCs can be isolated and expanded from UCB MNCs based on their capacity to adhere to a plastic surface as same as MSCs. USSCs are adherent, spindle-shaped cells with high proliferative activity, although they have a very low primary frequency in UCB. Cell surface antigens profile of USSCs is similar to MSCs (CD13⁺, CD29⁺, CD44⁺, CD90⁺, CD105⁺, CD14⁻, CD34⁻, and CD45⁻). Differences between MSCs and USSCs include absent expression in USSCs of CD50, CD62L, CD106, and HAS1, all of which are present in MSCs. USSCs can differenciate into cells of all three germ layers including osteoblasts, chondroblasts, adipocytes, hematopoietic cells and neuroglial cells in vitro and in vivo. Furthermore, transplantation of USSCs in non-injured fetal sheep resulted in engraftment of USSC-derived hematopoietic cells, albumin-producing hepatic cells and cardiomyocytes (2). The high adipogenic differentiation activity is likely to be a functional marker to discriminate between USSCs and UCB MSCs (32). Since USSCs express typical MSCs markers and expand in MSC-supporting media, these findings suggest that USSCs are of an earlier cell type than MSCs. USSCs also have hematopoiesis-supporting stromal activity (33) and capacity to enhance the homing of CB CD34-positive cells in vivo (34). The hematopoiesis-supporting activity and related cytokines expression levels of USSCs are significantly higher than those of BM MSCs (33). Thus, USSCs may be a potential cell source for cell-based therapy in facilitating homing and engraftment for cord blood transplant recipients.

VSELs were initially isolated from mouse BM as a population of small Sca-1⁺lin⁻CD45⁻ cells that express CXCR4 receptor with embryonic characteristics (3). The cells are characterized by their very small in size (2∼4 μm) that are smaller than red blood cells. Subsequently, the VSELs were also identified in human UCB by same research group (35). To isolate the VSELs from human UCB, two-step isolation procedure (removal of erythrocytes by hypotonic lysis combined with multiparamtor FACS sorting) has been developed (36). The VSELs in UCB are also very small (3∼5 μm) and enriched in a population of CXCR4⁺AC133⁺CD34⁺lin⁻CD45⁻MNCs. The cells express several markers for embryonic pluripotent stem cells such as SSEA-4, Oct-4, and Nanog (35). Mouse BM-derived VSELs form cellular clusters like embryoid bodies when they plate over C2C12 feeder cells and differentiate into cells from all three germ layers including neural cells, pancreatic cells, and cardiomyocytes (3). Recently, McGukin et al. (37) reported that UCB-derived embryonic-like stem cells could be efficiently differentiated into neuronal cells in their refined culture system. in vivo, VSELs transplantation reconstitutes hematopoiesis in lethally irradiated mice (38) and improves cardiac function and remodeling in the mouse acute myocardial infarction model (39). It is conceivable that VSELs are originated from epiblast on the basis of phenotypical similarities with primordial germ cells. It is still obscure that UCB-derived VSELs also possess the pluripotent capacity as similar as BM-derived VSELs.

MLPCs were identified as a multipotent stem cell population which rarely exist in UCB (40). The cells can be isolated from UCB units by using an antibody-based cell separation medium (PrepaCyte-MLPC; BioE Inc.) followed by plastic adherence. Freshly isolated MLPCs exhibit a leukocyte-like morphology and express CD45, CD34, CD133, CD9, nestin, SSEA-3, SSEA-4 and several MSC markers such as CD13, CD29, CD44, CD73, CD90, and CD105. After expansion by culturing on plastic surface, the cells change their morphology into fibroblast-like and lose hematopoietic stem cell makers CD34 and CD133, and embryonic stem (ES) cell markers SSEA-3 and SSEA-4 but continuously express MSC markers. Interestingly, MLPCs also can be isolated from UCB MSCs by single cell cloning (4). MLPCs distinguish themselves from MSCs by showing more extensive expansion and higher plasticity. A phenotypic feature of MLPCs is higher expression of CD9 compared with UCB MSCs and BM MSCs. The microarray analysis suggests that MLPCs are relatively quiescent primitive cells that are capable of wide plasticity yet are not committed to lineage, compared to BM MSCs (40). It has been reported that MLPCs differentiate into cells from all three germ layers such as endodermal hepatopancreatic precursor cells, mature hepatocytes and type II alveolar cells, mesodermal adipocytes, chondrocytes, osteoblasts, myocytes and endothelial cells, and ectodermal neurons, astrocytes and oligodendrocytes. Despite of their extensive expansion ability and high plasticity, the cells do not have abilities of spontaneous differentiation and teratoma formation like ES cells, suggesting the cells provide an attractive transplantation cell source for regenerative medicine.

In 2002, Buznska et al. (7) isolated UCB-derived cells with ability of neurosphere formation and neural differentiation potential. They used a combination of CD34-immunomagnetic depletion strategy and subfractionation according to cell surface adhesive properties to enrich the cells with neuronal progeny. The procedure efficiently removes the cells expressing CD34 and CD45. Subsequently, culturing the cells at a low cell density in the presence of epidermal growth factor (EGF) results in the neuro-sphere formation. The same group also established a clonogenic non-immortalized UCB neural stem cell line (UCB-NSC) that could be maintained in culture at different stages of neural progenitor development (41). The neurospheres derived from UCB-NSC express neural stem/progenitors markers nestin and glial fibrillary acidic protein (GFAP) and differentiate into neuronal, astrocytic and oligodendroglial lineages. McGuckin et al. (42) has been reported the similar type of neurospheres can be induced by discrete hematopoietic lineage (CD45, CD33, CD7 and glycophorin-A) negative cell population from CB MNCs, followed by culturing in medium containing thrombopoietin, flt-3 ligand, and c-kit ligand (TPOFLK). Recently, our research group found that p75NTR⁺ cell fraction in UCB MNCs efficiently form neurospheres and differentiate into neural cells, astrocytes and oligodendroglial cells. The freshly isolated p75NTR⁺ cells express various neural crest specific marker genes such as Slug, Snail, Twist, Wnt-1, and Sox9, suggesting the cell fraction is originated from neural crest stem cells. Many studies have demonstrated that human UCB transplantation enhance functional recovery in animal models for stroke (43), amyotrophic lateral sclerosis (44, 45), traumatic brain injury (46), and spinal cord injury (47, 48), while it is not clear which of these cells are important for functional recovery. Because very few transplanted cells are found in the brain, even when delivered intracerebrally, it is more feasible that they secrete trophic factors that enhance endogenous mechanisms of brain repair.

The recent studies provide convincing evidence that UCB contain not only hematopoietic progenitors, but also several other types of stem/progenitor cells from very primitive embryonic-like cells to relatively mature neuronal and endothelial progenitor cells. Some of these cells may be the overlapping populations of stem cells that have been described by different investigators and given various names due to their different isolation and expansion strategies. Important questions concerning their origin, differentiation potential, tumorigenicity, and availability still need to be elucidated. In addition, methods to expand specific population in sufficient numbers for transplantation must be established. It is interesting that UCB contains embryonic-like pluripotent stem cells lacking tumorigenesis, while efficient isolation and expansion strategies of these cells are not yet determined. Recently it was reported that somatic cells can be reprogrammed to a pluripotent state by forced expression of four transcriptional factors, including Oct4, Sox2, c-Myc and Klf4 (49). These induced pluripotent stem (iPS) cells are expected to be an attractive cell source for regenerative medicine because of their pluripotent ability, their lack of an immunological rejection reaction, and eliminating of current ethical issues surrounding ES cell research. However, low efficiency of reprogramming and tendency to induce malignant transformation compromise the clinical utility of iPS cells. Because of the lower incidence of tumorigenesis, more efficient preparation, and no use retroviral vectors and gene transmission, UCB-derived stem cell based therapy may be more useful for the clinical application of regenerative medicine compared to iPS cell based therapy. Further studies to evaluate the possibilities for both research and therapeutic applications for UCB-derived stem cells are required.