Transplanted cord blood-derived endothelial precursor cells augment postnatal neovascularization

Abstract

Endothelial precursor cells (EPCs) have been identified in adult peripheral blood. We examined whether EPCs could be isolated from umbilical cord blood, a rich source for hematopoietic progenitors, and whether in vivo transplantation of EPCs could modulate postnatal neovascularization. Numerous cell clusters, spindle-shaped and attaching (AT) cells, and cord-like structures developed from culture of cord blood mononuclear cells (MNCs). Fluorescence-trace experiments revealed that cell clusters, AT cells, and cord-like structures predominantly were derived from CD34-positive MNCs (MNC(CD34+)). AT cells and cell clusters could be generated more efficiently from cord blood MNCs than from adult peripheral blood MNCs. AT cells incorporated acetylated-LDL, released nitric oxide, and expressed KDR, VE-cadherin, CD31, and von Willebrand factor but not CD45. Locally transplanted AT cells survived and participated in capillary networks in the ischemic tissues of immunodeficient nude rats in vivo. AT cells thus had multiple endothelial phenotypes and were defined as a major population of EPCs. Furthermore, laser Doppler and immunohistochemical analyses revealed that EPC transplantation quantitatively augmented neovascularization and blood flow in the ischemic hindlimb. In conclusion, umbilical cord blood is a valuable source of EPCs, and transplantation of cord blood-derived EPCs represents a promising strategy for modulating postnatal neovascularization.

Figures

Figure 1

Enumeration of MNCCD34+ in umbilical cord blood and adult human peripheral blood. Umbilical cord blood contained 10-fold excess of MNCCD34+ compared with adult human peripheral blood. AP < 0.001.

Figure 2

Comparison of numbers of AT cells and cell clusters between umbilical cord blood and adult human peripheral blood. Greater numbers of AT cells (a) and cell clusters (b) developed from similar numbers of MNCs from cord blood (n = 10) than from adult peripheral blood (n = 7) examined at both days 7 and 14 of culture. AP < 0.001. Representative photomicrographs of AT cells and a cell cluster developed from cord blood MNCs are shown at the top of each panel.

Figure 3

Development of spindle-shaped and AT cells, and cord-like structures from MNCCD34+. (a) When labeled MNCCD34– alone were plated on fibronectin, most of the cells did not attach to fibronectin and had a round shape. (b) When labeled MNCCD34+ alone were plated on fibronectin, only limited numbers of cell clusters and AT cells appeared within 7 days. (c) When green fluorescence-labeled MNCCD34– were coplated with unlabeled MNCCD34+, fluorescence-positive MNCCD34– participated in cell cluster formation, but they were located only at the center of the clusters. AT cells and cord-like structures were mostly fluorescence negative. (d) Conversely, when fluorescence-labeled MNCCD34+ were coplated with unlabeled MNCCD34–, the AT cells, cell clusters, and cord-like structures were mostly positive for the fluorescence. (e, f) From coculture of fluorescence-labeled MNCCD34+ and unlabeled MNCCD34–, numerous fluorescence-positive cell clusters developed, and some of the cell clusters connected together to form linear cord-like structures. Bars: 20 μm in a–d, 100 μm in e and f. (g) Bar graphs showing the numbers of AT cells and cell clusters developed from four different conditions of cell culture (1,000 cells/mm2). A limited number of AT cells and cell clusters appeared from MNCCD34+ alone. The numbers of developed AT cells and cell clusters were greatly enhanced by coculture of MNCCD34+ with MNCCD34– when compared with culture of MNCCD34+ alone. The numbers of AT cells and cell clusters derived from coculture of MNCCD34+ and MNCCD34– were comparable to those from total MNCs.

Figure 4

AT cells had EC-like functions and expressed cell surface markers in vitro. At day 7 of culture (a) AT cells took up DiI-ac-LDL. (b) AT cells were positive for endothelial constitutive NO synthase (ecNOS) immunostaining. (c and d) Incubating AT cells with DAF-2 DA, a NO-specific fluorescence indicator, showed that hVEGF165 (50 ng/mL) (d), but not saline (c), induced NO release from AT cells. Bars, 50 μm in c and d. (e) Flow cytometric analyses of AT cells at day 7 of culture (n = 8). Horizontal bars indicate positive antigen gates as determined by an isotype-matched IgG control study. AT cells were positive for KDR (52 ± 5%), VE-cadherin (62 ± 2%), CD34 (62 ± 7%), and CD31 (57 ± 9%), but not for CD45 (7 ± 3%). Data are percentage of positive cells.

Figure 5

Participation of cord blood–derived AT cells in neovascularization in vitro and in vivo. (a) The phase-contrast (left) and fluorescent (right) photomicrographs at the same microscopic field are shown. Green fluorescence-labeled AT cells contacted unlabeled HUVECs on matrix gel within 24 hours of coculture. (b) The labeled AT cells were incorporated into the EC network formed on the basement matrix gel within 3 days after coculture. (c–g) Frozen sections of tissues harvested from the ischemic hindlimb of nude rats 2 weeks after transplantation of fluorescence-labeled cord blood–derived AT cells. The labeled AT cells were sprouting from a place near the cell injection site indicated by * (c). AT cells further migrated into the interstitial regions among preserved skeletal myocytes (indicated by M). Numerous labeled AT cells were incorporated into and formed capillary-like networks (arrows in e and g), as well as tubular structures with a round lumen (arrowheads d–g). (f) Green fluorescence signals indicate localization of transplanted AT cells forming numerous capillary-like networks (arrows). (g) The phase-contrast photomicrograph of an adjacent section histochemically stained with alkaline phosphatase (AP) shows viable capillary ECs (dark blue colors) at identical locations of the incorporated AT cells seen in f. Bars, 50 μm.

Figure 6

Augmented neovascularization by local transplantation of either umbilical cord blood–derived (UCB-derived) or adult peripheral blood–derived (APB-derived) EPCs in the ischemic hindlimb of immunodeficient nude rats. (a) Unilateral (left) hindlimb ischemia (arrowheads) was surgically induced in nude rats. Either APB-derived EPCs (3 × 105 cells/animal, n = 6), UCB-derived EPCs (3 × 105 cells/animal, n = 8), or saline (control, n = 7) were locally transplanted into the ischemic hindlimb at postoperative day 3. Representative serial LDPI analyses showed progressive increases in blood flow in the ischemic hindlimb (red to white color) of rats injected with EPCs obtained from APB or UCB. In contrast, recovery of blood flow was retarded in the ischemic limb (blue to green color) in a control animal. (b) Computer-assisted quantitative analyses of hindlimb blood flow demonstrated significantly enhanced ischemic/normal limb blood perfusion ratios in rats injected with EPCs of either origin (APB or UCB) compared with controls. AP < 0.001. (c) Capillary ECs were identified in the sections of ischemic tissues under a light microscope after immunohistochemical staining for vWF (brown reaction products) and alkaline phosphatase (AP) staining (blue reaction products). Light microscopy revealed increased capillary densities in rats receiving EPCs of either origin (APB or UCB) compared with controls. Bars, 50 μm. (d) Quantitative analyses showed increased capillary densities in the ischemic hindlimb tissues of the two EPC-transplanted groups as compared with control animals examined 14 days after induction of limb ischemia. AP < 0.001.