Therapeutic potential of vasculogenesis and osteogenesis promoted by peripheral blood CD34-positive cells for functional bone healing

Abstract

Failures in fracture healing are mainly caused by a lack of vascularization. Adult human circulating CD34+ cells, an endothelial/hematopoietic progenitor-enriched cell population, have been reported to differentiate into osteoblasts in vitro; however, the therapeutic potential of CD34+ cells for fracture healing is still unclear. Therefore, we performed a series of experiments to test our hypothesis that functional fracture healing is supported by vasculogenesis and osteogenesis via regenerative plasticity of CD34+ cells. Peripheral blood CD34+ cells, isolated from total mononuclear cells of adult human volunteers, showed gene expression of osteocalcin in 4 of 20 freshly isolated cells by single cell reverse transcriptase-polymerase chain reaction analysis. Phosphate-buffered saline, mononuclear cells, or CD34+ cells were intravenously transplanted after producing nonhealing femoral fractures in nude rats. Reverse transcriptase-polymerase chain reaction and immunohistochemical staining at the peri-fracture site demonstrated molecular and histological expression of human-specific markers for endothelial cells and osteoblasts at week 2. Functional bone healing assessed by biomechanical as well as radiological and histological examinations was significantly enhanced by CD34+ cell transplantation compared with the other groups. Our data suggest circulating human CD34+ cells have therapeutic potential to promote an environment conducive to neovascularization and osteogenesis in damaged skeletal tissue, allowing the complete healing of fractures.

Figures

FIGURE 1

Mobilization of mouse ScaI+Lin− cells by fracture stress. A: Percentage of peripheral blood ScaI+Lin− cells to Lin− MNCs shows an increase with the peak at day 1 after fracture. Black line represents isotype control (negative control), and blue line indicates sample data. x axis shows fluorescent intensity of cells, and y axis demonstrates the number of cells. B: Number of ScaI+Lin− cells significantly peaked at day 1 after fracture and then gradually decreased. *P < 0.05.

FIGURE 2

Phenotypic characterization of human peripheral blood MNCs and CD34+ cells by FACS and RT-PCR analysis. A: FACS analysis to evaluate positivity for CD34 in MNCs (left) and CD34+ cells (right). Numbers are percentages of the cells positive for CD34. B: FACS analysis to characterize CD34+ cells by positivity for various cell surface markers. Human CD34+ cells were positive for CD133, c-Kit, CD45, CD31, and CD105 and negative for VE-cad, KDR, and Tie2. Numbers are percentages of double-positive cells for CD34 (y axis) and each antibody (x axis). C and D: RT-PCR analysis for human-specific genes of endothelial (C) and osteoblastic (D) lineages was performed in freshly isolated CD34+ cells and MNCs. The analysis of CD34+ cells revealed weak expression of hCD31 and hOC but not of the other endothelial marker (hVE-cad) or bone-related marker (hCol1A1). Cultured human umbilical vein endothelial cells and OBs were used for positive control for human-specific endothelial and bone-related genes. E: RT-PCR analysis at the single cell level of the CD34+ cells. Four of 20 CD34+ cells expressed both CD34 and hOC.

FIGURE 3

Recruitment of fluorescent-labeled CD34+ cells into fracture site. A–C: Histochemical staining for isolectin B4 (green), a rat-specific EC marker, using tissue samples of fracture sites obtained from rats receiving Qtracker-labeled human cells (red) 1 week after fracture in the granulation area shown as zone a in M. More efficient recruitment of human cells with neovasculature-like structure was accompanied with enhancement of neovascularization by recipient rat cells in animals treated with CD34+ cells (A) compared with those receiving MNCs (B) or PBS alone (C). D–F: Double-fluorescent immunostaining for human leukocyte antigen (HLA)-ABC (red) and smooth muscle actin (green) using the tissue sample of granulation area. The massive recruitment of the human cells in the granulation area with relatively rare incorporation along the inner layer of SMA-positive smooth muscle cells was observed in CD34+ group. The human cells lining along the inner layer were morphologically compatible with ECs (D). In contrast, less recruitment of human cells in the granulation area and no human cells along the inner layer of arterioles were observed in animals receiving MNCs (E). Human cells were not found in the PBS group (F). G–I: Pursuit of Qtracker-labeled human cells 1 week after fracture in newly formed bone area shown as zone b in M. Human cell (red) recruitment into the newly formed bone area (surrounded by broken lines) was more abundant in animals treated with CD34+ cells (G) compared with those receiving MNCs (H) or PBS alone (I). J–L: Fluorescent immunohistochemistry for HLA-ABC demonstrated massive recruitment of human cells (red: arrows) in CD34+ group (J). In contrast, less recruitment of the HLA-ABC-positive cells was observed in animals receiving MNCs (K) and no human cells in PBS group (L). M: Representative photomicrograph of immunostaining for isolectin B4 (brown) and toluidine blue counter staining for anatomical indication of following areas: zone a, granulation zone; zone b, newly formed bone zone; cb, cortical bone; ca, cartilage; wb, woven bone. N: The number of HLA-ABC-positive cells observed in zone a was significantly greater in CD34+ group compared with other groups. **P < 0.01. O: The number of HLA-ABC-positive cells in zone b was significantly greater in CD34+ group compared with other groups. *P < 0.05; **P < 0.01. Blue fluorescence indicates DAPI for nuclear staining. Scale bars = 100 μm (A–C, G–F); 50 μm (D–F, J–L). Original magnifications: ×100 (A–C, G–I); ×200 (D–F, J–L); ×40 (M).

FIGURE 4

Human CD34+ cell-derived vasculogenesis and osteogenesis. Immunohistochemical staining and RT-PCR for human-specific EC or OB markers was performed using tissue samples harvested at week 2. A–D: Differentiated human ECs were identified in zone a as hCD31-positive cells (red) in animals receiving CD34+ cells compared with MNC (B) or PBS group (C). E: RT-PCR analysis of tissue RNA isolated from the peri-fracture site demonstrated the expression of human-specific EC markers (hCD31, hVE-cad) in animals treated with CD34+ cells but not in control animals. Cultured human umbilical vein endothelial cells were used for positive control, and no RNA served as negative control. F–I: Differentiated human OBs were identified in zone b as hOC-positive cells (red) in animals receiving CD34+ cells compared with MNC (G) or PBS group (H). J: RT-PCR analysis of RNA isolated from the peri-fracture site demonstrates the expression of human-specific bone-related markers (hOC, hCol1A1) in animals treated with CD34+ cells, but not in control animals. Cultured hOBs were used for positive control. Blue fluorescence represents DAPI. Broken lines, newly formed bone surface. Scale bars = 50 μm (A–C, F–H); 20 μm (D, I). Original magnifications: ×200 (A–C, F–H); ×400 (D, I).

FIGURE 5

Enhanced intrinsic vascularization and osteogenesis by recipient’s cells and serial improvement of blood flow and enhancement of callus formation after CD34+ cell transplantation. A: Rat-specific vascular staining with isolectin B4 (brown) at week 2 demonstrated enhanced neovascularization in zone a in animals treated with CD34+ cells compared with those receiving MNCs or PBS alone. Intrinsic angiogenesis assessed by capillary density at week 2 was significantly enhanced after CD34+ cell transplantation compared with other treatments. **P < 0.01. B: Representative LDPI at week 0 (1 hour after fracture) and week 2 in each group. In these digital color-coded images, maximum perfusion values are indicated in white, medium values in green to yellow, and lowest values in dark blue. The skin blood flow within fracture site (red square) and intact contralateral site (black square) were evaluated as mean flux, and ratio of the mean flux in the fractured site with that in the contralateral site (mean flux ratio) was calculated. Severe reduction of the blood flow was similarly observed 1 hour after fracture with the periosteum cauterized in all groups, whereas the mean flux ratio at week 2 was significantly greater in animals treated with CD34+ cells compared with those receiving MNCs or PBS alone. *P < 0.05. C: Rat-specific OC staining (brown) to detect intrinsic OBs at week 2 revealed augmentation of osteogenesis in zone b in animals treated with CD34+ cells compared with those receiving MNCs or PBS alone. Intrinsic osteogenesis assessed by the OB density at week 2 was significantly enhanced after CD34+ cell transplantation compared with other groups. **P < 0.01. D: Relative callus area assessed by radiograph at week 2 was significantly larger in CD34+ group compared with other groups. *P < 0.05. Scale bars = 50 μm. Original magnifications, ×200.

FIGURE 6

Inhibition of intrinsic angiogenesis and osteogenesis by anti-angiogenic agent in animals receiving CD34+ cell transplantation. A: RT-PCR revealed that fractured tissue sample contained higher amount of angiogenic factors (hVEGF, hFGF2, and hHGF) in human CD34+ group compared with MNC group. B: Rat-specific vascular staining with isolectin B4 (brown) at week 2 demonstrated reduced neovascularization in zone a in animals treated with CD34+ cells and sFlt1 compared with those receiving CD34+ cells and PBS. Intrinsic angiogenesis assessed by capillary density at week 2 is significantly reduced after sFlt1 treatment compared with PBS. *P < 0.05. C: Representative LDPI at weeks 0 and 2 in each group (red square, fracture site; black square, intact contralateral site). Mean flux ratio at week 2 was significantly less in animals treated with sFlt1 compared with rats receiving PBS. *P < 0.05. D: Representative rat-specific OC staining (brown) to detect intrinsic OBs at week 2 in zone b in animals treated with CD34+ cells and sFlt1 and those receiving CD34+ cells and PBS. Intrinsic osteogenesis assessed by the OB density at week 2 was significantly inhibited after sFlt1 treatment compared with PBS. *P < 0.05. E: Relative callus area assessed by radiograph at week 2 was significantly smaller in sFlt1-treated group compared with PBS group. *P < 0.05. Original magnifications, ×200.

FIGURE 7

Radiographical, histological, and biomechanical evidence of fracture healing after CD34+ cell transplantation. A: Fracture healing was serially assessed by radiographs. By week 8, fracture was healed with bridging callus in all animals receiving CD34+ cells (arrow), but in no rats treated with MNCs or PBS alone. B: Serial assessment of fracture healing by histological examination with toluidine blue staining. Histological evaluation demonstrated the enhanced endochondral ossification consisting of numerous chondrocytes and newly formed trabecular bone at week 2, bridging callus formation at week 4, and complete union at week 8 (arrow) in animals receiving CD34+ cell transplantation. Although thick callus formation at week 2 was observed, the healing process had stopped by week 4, and the callus was finally absorbed at week 8 in animals receiving MNCs or PBS. C: The degree of fracture healing was assessed by Allen’s classification. The degree of fracture healing at weeks 4 and 8 was significantly higher after CD34+ cell transplantation compared with other treatments. *P < 0.05; **P < 0.01. D: Functional recovery after fracture is assessed by biomechanical three-point bending test at week 8. The percentage of all parameters (percent ultimate stress, percent extrinsic stiffness, percent failure energy) showing the ratio of each value in fractured side with contralateral side in animals receiving CD34+ cells was significantly superior to those in animals receiving MNCs or PBS. *P < 0.05.