Endothelial Progenitor Cells Improve Directly and Indirectly Early Vascularization of Mesenchymal Stem Cell-Driven Bone Regeneration in a Critical Bone Defect in Rats

Scaffold

A commercially available bone graft substitute was studied: Synthetic β-tricalcium phosphate (Chronos®, Synthes, Oberdorf, Switzerland) with a size of 0.7–1.4 mm, 60% porosity, a pore size of 100–500 μm, low mechanical stability, and moderate biodegradability. These small granules exhibit a high surface area; thus, a high amount of cells is seeded. Furthermore, cells migrate easily in the core of the scaffold (14).

Fibronectin Coating of β-TCP

Sterile bone graft granules for all experimental groups were incubated for 60 min in a fibronectin solution (10 μg/ml, Sigma, Deisenhofen, Germany) in PBS without Mg²⁺ and Ca²⁺ (PBS^-/-). The supernatant was then removed and replaced by PBS^-/-. The granules were immediately placed as a dense single layer in a 24-well plate (Nunc, Wiesbaden, Germany) using sterile forceps.

Cell Culture

EPCs were isolated according to the procedure as described previously (14). Briefly, peripheral blood mononucleated cells (PBMCs) were isolated from the buffy coat by density gradient centrifugation (30 min, 900 x g) with Ficoll (1.077 g/ml, Biochrom, Berlin, Germany). The PBMCs were washed twice with cold PBS^w/o (10 min, 900 x g), and each 4 × 10⁶ cells/cm² were cultivated on a fibronectin-coated (10 μg/ml, Sigma), 24-well culture dish in 1 ml of endothelial basal medium (EBM, Cambrex, Verviers, Belgium) supplemented with endothelial growth medium-2 (EGM-2) singlequots at 37°C, 5% CO₂. After 48 h, nonadherent and weak adherent cells were removed, the medium was exchanged, and the cells were cultivated for additional 72 h. A parallel preparation was performed to evaluate the percentage of endothelial cell-like differentiated cells. EPCs were identified as previously described (13). In brief, cells presenting double-positive fluorescence [1,1 dioctadecyl-3,3,3,3-tetramethyl-indocarbocyanine-labeled acetylated low-density liproprotein (DiLDL), UEA-lectinfluorescein isothiocyanate (FITC); Sigma] were considered to be EPCs. For the experiments, the cells were detached by incubation (10 min) with accutase (PAA Laboratories, Linz, Austria), were washed once with MesenCult + supplements (Cell Systems, St. Katharinen, Germany), and were subsequentely adjusted to a density of 2.5 × 10⁵ cells in 100 μl.

MSCs were obtained from the iliac crest bone marrow aspirate of a voluntary trauma patient undergoing pelvic surgery using Ficoll density gradient centrifugation (30 min, 1,100 x g, d = 1.073 g/ml, Biochrom, Berlin, Germany) as previously described (30). Patients were informed and signed a consent form. The study was confirmed by the local ethics committee (249/07).

The cells in the interphase were collected and washed twice using PBS (10 min, 900 x g) containing 2% fetal bovine serum (FBS), resuspended in 3 ml MesenCult + supplements (Cell Systems), and were counted using a Neubauer chamber.

Each 4 × 10⁶ cells were seeded in a 25-cm² culture flask. The cells were expanded over five passages. Subsequently, the cells were detached by incubation with accutase, washed, and resuspended in a medium consisting of 90% fetal calf serum (FCS; PAA Laboratories) and 10% dimethyl sulfoxide (DMSO; Sigma). Aliquots were stored in liquid nitrogen until use.

For the experiments, a portion of the cryoconserved cells were thawed and expanded over two additional passages. The MSCs were then detached by an incubation of 10 min with accutase. After one wash (10 min, 300 g), the cells were resuspended in MesenCult + supplements. The cell suspension was then divided. One part was adjusted to a density of 2.5 × 10⁵ cells in 100 μl and was used subsequently for the experiments. The other portion was used for the confirmation of surface characteristics ascertained to MSCs as described earlier by our group (28). The cryoconservation of the cells ensured that, in all groups, MSCs from one donor in a comparable quality were used.

Major Histocompatibility Complex-1 (MHC-1) Expression on MSCs and EPCs In Vitro

Detection for human characteristics of MSCs and EPCs in vitro were performed by using FACS analysis. Detached MSCs, respectively, detached DiLDL prestained EPCs, were resuspended in PBS^-/-, which was supplemented with 2 mM EDTA (Sigma) and 0.5% FCS (PAA Laboratories). Each 100-μl cell suspension was incubated with 7.5 μl of FITC-conjugated monoclonal antibodies against human leukocyte antigen (HLA)-ABC (MSCs, BD Pharmingen, Heidelberg, Germany), MSCs received additionally anti-CD90-phycoerythrin (PE; BD Biosciences, Heidelberg, Germany) for 15 min in the dark. After one wash with above-mentioned buffer, the cells were subjected immediately to flow cytometry. A dual laser four-color FACScalibur (BD Biosciences) was used. At least 10,000 events were recorded. For analysis, the software CellQuest pro (BD Biosciences) was applied.

Composite Scaffold: β-TCP Seeded with EPCs and MSCs

Scaffolds were placed into individual 48-well plate without having direct contact with the surrounding well border. The scaffolds were loaded with 5 × 10⁵ cells of either a mixture composed of 50% MSCs and 50% EPCs or 5 × 10⁵ MSCs alone or 5 × 10⁵ EPCs alone. The cells were dripped in a volume of 200 μl to the bone graft layer and were incubated for 10 min at 37°C. The medium containing the nonadhering cells was then removed and rinsed once again over the bone graft layer, followed by incubation as indicated above. This procedure was repeated three times. Then the granules were gently transferred to another well containing 500 μl MesenCult + supplements. The remaining cells in the supernatant and at the bottom of the initial seeding well were isolated and counted, and the seeding efficiency was calculated [(initial cell number – remaining cell number)/initial cell number x 100%].

Animals and Cell Transplantation

All animal experiments were performed in accordance with the institutional animal care and oversight committee (Project F3/03, Regierungspräsidium, Darmstadt, Germany). For cell transplantation experiments, we used male adult athymic rats (RNUrnu rats; Harlan-Winkelmann, Horst, Netherlands), which were older than 12 weeks. Table 1 shows the number and distribution of animals for the evaluation of vascularization of composite scaffold after cell transplantation.

OPEN IN VIEWER

Table 1.

Experimental Setting

Group	Transplanted Scaffold	Number of Animals
1	β-TCP alone	6
2	β-TCP loaded with human MSCs	6
3	β-TCP loaded with human EPCs	6
4	β-TCP loaded with human MSCs and EPCs	6
5	Rat autologous bone	5

A general anesthesia with 1.0 ml of a mixture of Ketavet (100 mg/ml) and Rompun (20 mg/ml) was given intraperitoneally. In order to fix the monolateral external fixator to the anterolateral aspect of the right femur, the right leg of the rat was shaved and cleaned with antiseptic fluid. A lateral longitudinal incision over the femur was made under aseptic conditions. The fascia was cut, and the muscles were separated between the quadriceps femuris and the hamstrings. The bone cortex was drilled using a 0.8-mm bit, and four self-tapping pins were placed at right angles to the surface of the femoral shaft. Four pins were fixed to the external fixateur. An osteotomy was made midshaft between the second and the third pin using an oscillating saw and a critical bone defect of 5 mm was created (Fig. 1). The composite scaffold was implanted into the osseous defect zone, and the wound was closed with continuous subcutaneous stiches using a 4/0-monofilament nylon suture. OPEN IN VIEWER

The animals were fed a standard laboratory diet ad libidum and were housed double in type III cages (800 cm²) in temperature controlled rooms (22°C) having a 12-h light cycle.

Statistics

Results are presented as median and interquartiles (median, 25% quartile, 75% quartile). Kruskal–Wallis test with Dunn's post hoc test for multiplicity was used for comparisons between the groups. A value of p <0.05 indicates statistical significance.

Human EPCs and MSCs Express MHC-1 Using FACS Analysis

Nearly all MSCs express MHC-1 on their surface (Fig. 2A, B). A concomittant staining with anti-CD90 identifies the MSCs. EPCs express also MHC-1 (Fig. 2C, D). A concomittant DiLDL staining identifies the EPCs. OPEN IN VIEWER

EPCs and EPCs/MSCs Loaded β-TCP Improve Early Vascularization

After 1 week, areas of neovascularization in the critical bone defect of male athymic rats treated with β-TCP, β-TCP + MSCs, β-TCP + EPCs, β-TCP + EPCs + MSCs, or autologous rat bone (Fig. 3A–E) were detected by CD34 histomorphochemistry. OPEN IN VIEWER

Histomorphometric analysis was performed, resulting in a comparison of area fractions in a standardized picture frame (Fig. 4). Values denote median ± quartiles, Kruskal–Wallis test, p < 0.05. OPEN IN VIEWER

Compared to the β-TCP group [2.2, 1.3, 3.5 (% of total pixel)] and β-TCP + MSCs group [3.01, 1.61, 6.31 (% of total pixel)], respectively (Fig. 4), a significant higher amount of vascularization (+#) was observed in samples derived from animals that received β-TCP loaded with either EPCs [8.12, 3.7, 11.62 (% of total pixel), p < 0.05] or coculture of EPCs and MSCs [11.76, 8.2, 15.6 (% of total pixel), p < 0.05].

Also a significant higher amount of early vascularization (*) was observed in the coculture group and EPC group compared to the autologous group [2.57, 1.5, 4.65 (% of total pixel)]. Moreover, we found significant more vessel formation in the coculture group compared to the β-TCP + EPC group (§). No difference between the β-TCP and the β-TCP + MSC group was detectable.

Contribution of Transplanted Human Cells

Using MHC-1 staining, we detected human cells (MSCs, EPCs) after transplantation in the bone defect. Neither in the TCP nor in the autologous group, MHC-1 staining was positive (Fig. 5A–E). Interestingly, MHC-1 staining was easily detectable if MSCs alone or in coculture with EPCs were transplanted. MHC-1 staining was also obvious when EPCs were transplanted, but to a lower degree. One week after cell transplantation, we observed human cells in the area around TCP gaps (decalcified histological section), but also within the center of the bone defect (Fig. 5B–D). OPEN IN VIEWER

Early Vascularization Is Effected Directly and Indirectly by Human EPCs

Based on the results that transplanted EPCs increased following early vascularization, we performed experiments to identify transplanted human EPC incorporation around the vessel formation. Immunohistochemical staining for vWF, an endothelial marker, and MHC-1, a human cell marker, was performed using tissue samples obtained 1 week after creation of a critical size defect (Fig. 6A–E). No human cells were identified after transplantation of TCP granules (Fig. 6A). After transplantation of human MSCs bone formation took place, but early vascularization was rare (Fig. 6B). vWF marked endothelial cells were more abundantly detected as the rarely double-positive cells for vWF and MHC-1 in bone regeneration area (Fig. 6C) after EPC transplantation. Interestingly, more human MSCs and double-positive EPCs (Fig. 6D) were identified when EPCs were coimplanted with MSCs, which led to an improved vascularization in bone regeneration at week 1. Moreover, transplanted human EPCs directly formed new vessels in the critical size defect (Fig. 6D, white arrows). DAPI staining showed a high amount of cells after transplantation of cocultured EPCs/MSCs. Transplantation of autologous bone (Fig. 6E) did not show improved early vascularization. OPEN IN VIEWER

Releasing of VEGF

One week after implantation of β-TCP granules into critical size defects, the degree of VEGF releasing was evaluated (Fig. 7A–E). We observed in animals that received β-TCP loaded with EPCs (Fig. 7C) or both cell types (EPCs/MSCs, Fig. 7D) have a higher amount of VEGF in comparison to animals that received empty β-TCP or β-TCP-loaded with MSCs. A low degree of VEGF release was detected in animals that received autologous bone (Fig. 7E). OPEN IN VIEWER