SEARCH

SEARCH BY CITATION

Keywords:

  • nonunion;
  • vascular endothelial growth factor;
  • growth factors;
  • gene-activated matrix;
  • smart biomaterials

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Healing of fractures is dependent on vascularization of bone, which is in turn promoted by VEGF. It was shown that 0.1 and 1 mg of pVEGF165-GAM led to a significant increase in vascularization and bone regeneration in defects that would otherwise have led to atrophic nonunions.

Introduction: One reason for lack of bone healing in nonunions is the absence of vascularization. In skeletogenesis, which is tightly linked to angiogenesis, vascular endothelial growth factor (VEGF) promotes the vascularization of the growth plate and transformation of cartilage to bone. We postulate that a gene-activated matrix (GAM), created with a plasmid coding for human VEGF165, coated on a collagen sponge could efficiently accelerate bone healing in large segmental defects.

Materials and Methods: Sixty New Zealand white rabbits received a 15-mm critical size defect on one radius, which was filled with either 0.1 or 1 mg plasmid-DNA as GAM. Radiographs were obtained every 3 weeks. After 6 or 12 weeks, animals were killed. New bone was measured by μCT scans. Vascularity was measured using anti-CD31 staining of endothelial cells in 18 regions of interest per implant.

Results: Scaffold and control plasmid showed no defect healing, whereas most of the animals in the VEGF groups showed partial or total bone regeneration. Significantly more bone was found in the VEGF groups, with no significant differences between the 0.1- and 1-mg groups. Immunohistochemical staining of endothelial cells revealed that the VEGF groups showed two to three times the number of vessels and a significantly larger endothelial area after 6 weeks. Twelve weeks after surgery, the amount of vascularization decreased, whereas more new bone was detectable.

Conclusions: The rabbit critical size defect was appropriate in size to produce atrophic nonunions. We showed that angiogenesis and osteogenesis can be promoted by a VEGF165-GAM that is an appropriate tool to induce bone healing in atrophic nonunions.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

AN EARLY VASCULAR response is essential for the normal progress of fracture healing. Fracture hematoma contains the angiogenetic growth factor vascular endothelial growth factor (VEGF), which has the capability to induce angiogenesis.(1) During endochondral bone formation, VEGF couples blood vessel formation, chondrocyte apoptosis, cartilage remodeling, and endochondral growth plate ossification.(1,2) Inhibition of VEGF decreases angiogenesis, callus mineralization and bone healing.(3) VEGF is expressed within the first 11 days of bone healing by angioblasts, osteoprogenitor cells, chondrocytes, and osteoblasts.(4) During bone formation and fracture healing, there is a cross-talk between endothelial cells and osteoblasts in which VEGF plays a key role: osteoblast-like cells produce VEGF while VEGF enhances osteoblast differentiation.(5) This confirms that re-establishment of the circulation is an early event in fracture healing and osteogenesis may be amplified by endogenous production of angiogenic mediators.(6)

VEGF is known to stimulate therapeutic vascular growth. Direct application of VEGF plasmids has been used in clinical trials in leg ischemia and coronary artery disease with positive results.(7) Induction of angiogenesis by VEGF was also achieved after perivascular application of minimal amounts of plasmid.(8)

Combining the success of direct gene transfer of pVEGF in therapeutic angiogenesis and the important role of angiogenesis and especially VEGF in bone healing, we hypothesized that phVEGF165 could promote bone healing in a critical size defect and prevent nonunions. Street et al.(3) showed that delivery of VEGF protein over a period of 7 days stimulates bone repair in rodents, whereas Peng et al.(9) found no such effect.

Our major issue was the application of VEGF over the time required because VEGF is a very unstable, short-lived protein in vivo and very costly to produce. For promotion of angiogenesis, it is preferable to deliver lower doses over a period of several days from an actively expressing transgene rather than single or multiple boluses of recombinant protein.(10,11) Gene therapy has the potential to maintain an optimal dose and local concentration over a period of time, thus using its advantages of minimal side effects and long-term efficiency.(7)

Bonadio et al.(12) described a type of local plasmid gene transfer technology known as the gene-activated matrix (GAM). GAM carriers serve as a scaffold that holds DNA in situ until endogenous repair cells arrive. It has previously been used to encode for human PTH (hPTH) implanted in canines where a time- and dose-dependent secretion of hPTH could be found. Likewise, Fang et al.(13) used GAM to deliver BMP to rat bones. Both yielded a positive effect of their GAM on bone healing.

The objective of our study was to evaluate whether the application of a VEGF165 plasmid as a GAM could produce a sufficient local concentration of active VEGF protein to promote angiogenesis and bone healing in a critical size defect in the rabbit radius. Osteoneogenesis was accessed radiographically and quantified by μCT, whereas angiogenesis was measured by histomorphometry of CD31+ vessels.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Plasmid preparation

For in vivo experiments, pVEGF165 was used, which was graciously donated by S Nikol.(7,8) In brief, human VEGF (GenBank no. XM 004512) was synthesized and amplified by PCR using the primer pair (5′-GAACCATGAACTTTCTGCT-3′, sense and 5′-CCCGGCTCACCGCCTCTTCT-3′, antisense). The amplified VEGF fragment was integrated into the pCR3.1 plasmid (Invitrogen, Karlsruhe, Germany) into the EcoRI restriction site, resulting in a 5.6-kb vector containing the human VEGF165 cDNA fragment, creating the plasmid pVEGF165. For large scale endotoxin-free plasmid DNA, production of pVEGF165 and pCR3.1 EndoFree Plasmid Giga Kit (Qiagen, Hilden, Germany) was used according manufacturer's protocol.

Transfection of rabbit osteoblasts and generation of conditioned media

Primary rabbit osteoblasts (rOBs) were grown from trabecular bone chips in culture medium (DMEM, 4.5 g/liter glucose; Sigma-Aldrich, Taufkirchen, Germany) containing 10% FCS (Biochrom, Berlin, Germany), 50 units penicillin/ml, 50 units streptomycin/ml, 12.5 mM HEPES, 0.4 mM L-Prolin, 50 mg/liter ascorbic acid, and 0.1 μM dexamethasone. Cells were cultured in 6-well plates under standard conditions at 37°C with 5% CO2. Cells were subcultured and grown to 80% confluence in a 6-well plate, and transfection was performed using metafectene (Biontex, Munich, Germany). For transfection, 1 μg of plasmid DNA of pVEGF165 or empty pCR3.1 was combined with 6 μl of metafectene reagent in 100 μl of serum-free culture medium and incubated with 200,000 cells for 6 h. To estimate the transfection efficiency, green fluorescent protein plasmid (pEGFP; Invitrogen, Karlsruhe, Germany) was used in a parallel experiment. To generate conditioned media, recombinant cells were grown to confluence, washed with PBS, and shifted to culture medium containing 2% FCS in a 6-well plate. Over 4 days, each day, 1 ml of media sample was collected and replaced with fresh medium, and the collected medium was stored at −80°C until analysis.

VEGF ELISA

VEGF protein was determined by a commercially available ELISA kit (R&D Systems, Wiesbaden, Germany) that recognizes the 165 amino acid splice variant of human VEGF (hVEGF165). The assay was performed according to the manufacturer's instructions. Briefly, standards or conditioned media samples (50 μl) were pipetted into an antibody-coated 96-well plate containing 50 μl of assay diluent and incubated for 2 h at 20°C on a shaker. The wells were washed five times with wash buffer, 100 μl of VEGF detection conjugate was added, and the samples were again incubated for 2 h at room temperature. Samples were washed five times, 100 μl substrate buffer was added, the samples were incubated for 30 minutes at 20°C, the reaction was stopped, and the absorption was measured with an ELISA reader (1420 VICTOR2 multilabel reader; Wallac, Turku, Finland) at 450 nm with λ correction at 570 nm. All measurements were performed in duplicate. The detection limit of the ELISA was 3.0 pg/ml.

VEGF bioassay

To determine the biological activity of secreted rhVEGF165, a human umbilical vein endothelial cell (HUVEC) proliferation assay was carried out. The VEGF-responsive HUVECs (Clonetics CC-2519; Cambrex, Verviers, Belgium) were expanded in endothelial cell growth medium (Clonetics CC-3024; Cambrex) and 2% FCS. Cells were plated in 96-well flat bottom culture plates (5000 cells/well) and incubated at 37°C in a humidified atmosphere (5% CO2) for 24 h. The medium was replaced with fresh endothelial cell growth medium containing either conditioned medium of pVEGF165 recombinant rabbit osteoblasts, conditioned medium of control-plasmid recombinant osteoblasts, or fresh culture medium with 2% FCS, respectively. After 4 days, cell proliferation was assayed with the WST-1 cell proliferation kit (1644807; Roche, Mannheim, Germany) according to the manufacturer's instructions.

GAM

GAMs were prepared the day before surgery. Commercially available collagen sponges (Kollagen-resorb; Resorba Clinicare, Nuernberg, Germany) were cut into small pieces of desired dimensions (15 × 5 × 5 mm) using a sterile blade under a laminar flow hood. Either 0.1 or 1 mg of pVEGF165 or pCR3.1 DNA diluted in 200 μl PBS was pipetted on the sponges and allowed to soak overnight at 4°C.

Animals

Animals were treated in compliance with the guiding principles in the Care and Use of Animals. The Committee on Animal Experimentation of Baden-Wuerttemberg approved the experiment. Six- to 9-month-old female New Zealand white rabbits (NZWR; n = 60) weighing 3.1–5.8 kg (mean, 4.4 kg) were kept in separate cages, fed a standard diet, and allowed free mobilization during the study. Skeletal maturity was verified radiographically by examination of the closure of epiphyseal plates.

The 60 rabbits were randomly assigned to five treatment groups. The collagen sponge was either not loaded (group 1) or loaded with 0.1 or 1 mg of the empty control plasmid pCR3.1 (groups 2 and 3) or the active plasmid containing the VEGF165 gene (groups 4 and 5). Groups 1–3 were used as controls.

Surgical procedure and treatment

Our animal model was adapted from Tuli and Gupta(14) and Wittbjer et al.(15) Unilateral 15-mm-long critical size defects were prepared in the radial diaphysis. The rabbits were anesthetized with an intramuscular injection of ketamine hydrochloride (50 mg /kg body weight, Hostaket; Intervet, Tönisvorst, Germany) and xylazin (5 mg/kg body weight, Rompun; Bayer Vital, Leverkusen, Germany). An antibiotic (netilmicin, 4 mg/kg body weight, Certomycin; Essex Pharma, Munich, Germany) was administered perioperatively.

After superomedial incision of 3 cm, soft tissues were dissected, and the bone exposed by gently retraction of the muscles. A tourniquet was not necessary because the muscles were not injured, and bleeding was not observed except from the skin incision. A Hohmann retractor was placed between the ulna and radius to protect the ulna. A 15-mm segmental defect in the middle of the diaphysis was created with an oscillating saw under irrigation with 0.9% sterile saline solution. Five millimeters of periosteum was stripped from each side of the radius.

The gap was irrigated with sterile physiological saline solution, and the collagen sponge was placed into the gap. Muscles were replaced over the defect, and soft tissue and skin were closed with 4–0 resorbable sutures (Ethilon; Ethicon, Norderstedt, Germany). Fixation of the osteotomized radius was unnecessary because of the fibro-osseous union of ulna and radius proximal and distal to the surgical site.(16) Postoperatively, 4 mg/kg body weight of carprofen was given as needed for pain. There were no differences in carprofen administration between the groups. Water and food were supplied ad libitum.

After 6 or 12 weeks, the animals were killed. Probes were excised en bloc with soft tissues attached to them and immediately placed in 70% ethanol.

Radiographic evaluation

Standardized anterior-posterior and lateral radiographs were taken immediately postoperatively and every 3 weeks thereafter until death. We used an ultra-high definition film, 44 kV and 2.2 mA, with a constant X-ray-to-object-to-film distance of 171 cm. Beneath the limb, a step wedge was placed to allow semiquantitative evaluation of BMD.

μCT scans

A μCT 40 (Scanco Medical, Bassersdorf, Switzerland) was used to evaluate the amount of new bone within the defect area. Approximately 32 slices/mm were taken, and 3-D volumes were reconstructed. A virtual tube was built between the beginning and end of the defect, and the percentage of new bone filling this tube was measured. A complete anatomical restoration of bone structure comparable with the contralateral intact bone was interpreted as 100% healing.

Histology and immunohistochemical staining for blood vessels

Retrievals were fixed in cold 70% ethanol for a minimum of 5 days. After fixation, they were decalcified with 25% EDTA for 4 weeks and embedded in paraffin so that the two bones lay parallel to the upper surface of the embedding block.

Serial longitudinal sections of 5 μm were cut at depths of 0.5, 1, and 1.5 mm and mounted on “superfrost” glasses. Each section included host bone proximal and distal to the defect.

Alternate sections were stained with Alcian blue or anti-CD31. Alcian blue-stained sections allowed an overview of new bone and cartilage.

Immunolocalization of blood vessels was performed with a monoclonal mouse anti-human CD31 antibody (clone: JC70A; Dako Cytomation, Glastrup, Denmark; dilution 1:20) overnight at 4°C. Primary antibody was detected with streptavidin-biotin linked polyclonal goat anti- mouse IgG complex (code:115–065–146; Jackson ImmunoResearch, West Grove, PA, USA) according to the protocol. Counterstaining was avoided because of possible influencing of computer-based image evaluation.

Histomorphometric evaluation

On each slide, six standardized regions of interest (ROIs) were photographed at a ×10 magnification with the Axioplan 2 (Carl Zeiss, Oberkochen, Germany). Two respective ROIs were located at the proximal, middle, and distal part of the defect. One of them was located at the ulnar cortex and one at the radial cortex. Slices were randomly selected before staining by a blinded technician, and blood vessels were marked by two independent examiners blinded for groups and treatment. They were counted with help of the Axiovision 3.1 program (Carl Zeiss). After this preparation, the percentage of immunohistochemically stained area in each slide was evaluated by means of Analysis software (Soft Imaging Systems, Münster, Germany) by a third blinded examiner.

Statistic evaluation

The results are expressed as means ± SD. Statistic analyses were performed with SPSS 10.01 (SPSS, Chicago, IL, USA). Because of the small group sizes, nonparametric tests were used. Separate analyses were done on each of the two different time-points for all five groups. Differences between the five treatment groups were analyzed with the Kruskal-Wallis test for independent samples. Differences between two groups (e.g., the two concentrations or periods) were determined with the Mann-Whitney test for independent samples. The significance level was set at p ≤ 0.05. In the case of multiple analyses, the significance level was adjusted.

The box-and-whisker plots display the first and third quartiles as the ends of the box, the maximum and minimum as the whiskers, and the median as a vertical bar in the interior of each box.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

pVEGF165-encoded rVEGF165 is bioactive

Rabbit osteoblasts were transfected with pVEGF165 with an estimated efficiency of about 30%. Recombinant cells secreted increasing amounts of rVEGF165 post-transfection that reached a plateau phase around day 3 (Fig. 1A). When HUVECs were incubated with the day 4 conditioned medium of pVEGF165 recombinant rOBs, cell proliferation was stimulated, whereas conditioned medium of control transfected rOBs failed to induce a proliferative response in HUVECs (Fig. 1B). This control served to exclude the possibility that endogenous rabbit VEGF and/or other factors secreted by normal rOBs were responsible for growth stimulation and indicated that the gene product of pVEGF165 was biologically active.

thumbnail image

Figure FIG. 1.. pVEGF165-encoded rhVEGF165 is biologically active. (A) Time-course of rhVEGF165 production by transfected rOBs. Cells were transfected with 1 μg/200,000 cells of each plasmid DNA or PBS as a negative control. Conditioned medium was analyzed over 4 days. Media samples were collected at time-points indicated and subjected to a human VEGF ELISA to determine the amount of secreted rhVEGF165-protein. (B) Proliferative response of HUVECs. Day 4 conditioned media of pVEGF165- or pCR3.1-transfected rOBs was collected. At the start of the experiment and 4 days after incubation with day 4 conditioned medium, cells were subjected to the WST-1 assay to determine the proliferative response of HUVECs.

Download figure to PowerPoint

pVEGF165-GAM enhanced bone formation

Radiographic evaluation:

Serial radiographs taken in two standard projections every 3 weeks after surgery (Fig. 2) were judged by two independent examiners blinded for groups. A semiquantitative evaluation scale adapted from Yasko et al.(17) was used. Bony bridging of the gap of up to 25% was defined as “minor bone formation”; between 25% and 75%, as “major bone formation”; and a bridging from one side to the other, as “total healing.” The number of animals with no, minor, major, and total healing is listed in Table 1. None of the critical size defects of the control groups bridged with new bone. Furthermore, most of the bony ends of the radii became thinner and atrophic (Fig. 2). A higher range of partial and total bone healing was found after VEGF165 application. No significant differences in bone healing were found between the two dosage groups.

Table Table 1.. Number of Animals in Each Group Showing Bone Formation
Thumbnail image of
thumbnail image

Figure FIG. 2.. Radiologic course of bone formation in one representative animal of the (A) control and (B) VEGF group. The GAM without the VEGF gene led to an atrophic nonunion (A), whereas VEGF165 led to a bridging of the bone gap after 6 weeks (B).

Download figure to PowerPoint

Volume of newly formed bone measured with μCT:

In multiplanar reconstructions of μCT scans (Fig. 3), the volume of newly formed bone was determined. Only new bone that was found in a virtual tube between the two ends of the radius was measured. Bone outside of this tube was regarded as overwhelming bone formation and ignored.

thumbnail image

Figure FIG. 3.. 3-D rendering of radius defects reconstructed from μCT images of (A) control or (B) VEGF-treated rabbits. In B, new bone formation is shown, whereas A shows an atrophy of the adjunct parts of the radius.

Download figure to PowerPoint

All three control groups were similar and showed only minor new bone formation (p > 0.5) (Fig. 4). Defects treated with pVEGF165-GAM revealed a significantly enhanced bone volume in both groups compared with the respective control group (p < 0.01) at 6 and 12 weeks after surgery. No differences were evident between the two VEGF165 plasmid doses (6 weeks: p = 0.5; 12 weeks: p = 0.9; Mann-Whitney test), indicating that 0.1 mg of pVEGF165 was sufficient to enhance bone healing.

thumbnail image

Figure FIG. 4.. Bone volume in μCT scans. The volume of newly formed bone, which was found in an imaginary tube between the ends of the critical size defect, is shown. Significantly more bone was found after pVEGF165 application compared with the respective control groups after 6 and 12 weeks (*p < 0.05).

Download figure to PowerPoint

Bone volume showed a trend to increase between 6 and 12 weeks in all groups carrying plasmids, but did not reach significance (VEGF 0.1 mg: p = 0.055; VEGF 1 mg; p = 0.337; Mann-Whitney test).

pVEGF165-GAM enhanced vessel formation

Angiogenesis was quantified in immunostained sections by determination of the number of new vessels in predefined areas (Fig. 5) and by measuring the area covered with CD31+ endothelial tissue. No significant differences could be found between the three control groups neither after 6 nor 12 weeks for both parameters. When the control groups at 6 weeks were compared with the 12-week control groups, a slight yet insignificant decrease in vascular density was observed.

thumbnail image

Figure FIG. 5.. One ROI in a longitudinal histological section of the radius 12 weeks after surgery. Respective ROIs at the proximal and radial part of the histological sections are shown. On the left side of the images, the proximal part of the radius can be seen. CD31 immunostaining shows more newly formed vessels (dark) in (B) the VEGF group than in (A) the control group.

Download figure to PowerPoint

In both VEGF165 groups, significantly more vessels were counted than in the respective control groups (p < 0.005) at 6 and 12 weeks (Fig. 6). Likewise, the area covered with CD31+ endothelial cells was significantly larger (p < 0.001; Fig. 7). In the 1 mg VEGF165 plasmid group, slightly more vessels were seen than in the 0.1 mg VEGF165 plasmid group after 6 (p = 0.09) and 12 weeks (p = 0.39). Endothelial area also did not show a significant dose-dependent relationship.

thumbnail image

Figure FIG. 6.. Mean number of vessels per ROI. The total amount of vessels were counted in 18 ROIs of each retrieval, and mean values of all ROIs per sample were determined. In the VEGF groups, significantly more vessels were found compared with the respective control groups after 6 and 12 weeks. In the two VEGF groups, significantly fewer vessels were counted after 12 weeks than after 6 weeks (*p < 0.05).

Download figure to PowerPoint

thumbnail image

Figure FIG. 7.. Percentage of the ROIs that was covered with endothelium stained by anti-CD31 The area covered by CD31+ tissue in the respective ROIs was significantly larger after VEGF treatment compared with the respective control groups after 6 and 12 weeks (*p < 0.05).

Download figure to PowerPoint

We found a time-dependent effect of VEGF165 application on the proliferation of vessels. In both groups, significantly fewer vessels were seen after 12 weeks along with new bone formation (VEGF 0.1 mg, p = 0.006; VEGF 1 mg, p = 0.01). Decreasing vessel density along with consistent endothelial area was interpreted as progression of multiple small vessels into larger vessels over time.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Fracture disrupts the vascularity of the bone, leading to necrosis and hypoxia of adjacent bone. In cases where the effect is significant enough, atrophic nonunions may form.(18) Similar situations were simulated in rabbits using a critical size defect model(14–16) and showed that a 15-mm radial defect filled with a collagen sponge alone will lead to atrophic nonunions.

BMPs have been used in several trials to promote bone healing of such critical size defects in animal models and showed that bone healing could be accelerated.(17,19) Moreover, it has been shown in a clinical trial that BMP-2 is superior to autograft in spinal fusions.(20) In cases of nonunion, however, BMP did not lead to a higher rate of bone healing.(19,21,22) Bostrom and Camacho(19) described that, after treatment with BMP-2, 25% of the nonunions required a secondary bone graft procedure. Kujala et al.,(22) who failed to achieve complete unions with a composite of biocoral and BMP in 8 of 10 scaphoid nonunions, contended that BMP was incapable of reaching target cells secondary to inadequately vascularized tissue.

Peng et al.(9) found a synergistic enhancement of bone formation after application of muscle-derived stem cells that expressed either BMP-4, VEGF, or both growth factors. Furthermore, VEGF enhanced the capillary density in combination with BMP. VEGF alone did not improve bone regeneration in their trial but increased cell survival and augmented cartilage formation and recruitment of mesenchymal stem cells.

Other authors(23,24) have reported direct effects of VEGF on bone remodeling and regeneration, but these studies differed in the administration of VEGF. Because the VEGF protein is very fragile, Eckardt et al.(25) and Street et al.(3) used a miniosmotic pump to deliver the protein to the osteotomized bone. Local gene therapy seems to be a more practical way to deliver a protein over a period of time.(11,12) Hiltunen et al.(23) and Tarkka et al.(24) used adenovirus-mediated VEGF gene transfer to induce bone formation in rabbits or rats but did not use a fracture model.

Our aim was to show that a nonviral application of the VEGF gene is feasible and leads to an increase in angiogenesis and osteoid formation and thus ameliorates bone healing. To circumvent the risk of immunogenicity of adenovirus application, we used the GAM technology established by Bonadio et al.(12) They reported a transfection efficacy of 30–50% of available fibroblasts and a protein delivery for 3 weeks. Shea et al.(11) proved that DNA delivery from polymer matrices is advantageous compared with direct protein or plasmid DNA delivery. Especially for VEGF, the GAM is a sensible way of application because only small amounts of the protein are necessary.(26) Because the technology is already established, we did not measure the transfection rate but rather focused on the outcome. In a pilot trial (data not shown), we proved that the naked DNA persisted long enough on the collagen matrix to allow migrating cells to arrive. Histological evaluation of transplants after only 2 days within subcutaneous pockets showed adequate cells invading the GAM. In situ hybridization for the VEGF165 vector DNA confirmed abundant signals on the GAM. No phVEGF165 signal was obtained in the surrounding soft tissue.

The VEGF165 gene significantly enhanced bone and vessel formation at plasmid amounts of 0.1 and 1 mg, indicating that sufficient VEGF was expressed at both DNA doses. Although group sizes may have been too small to verify differences between the two concentrations, we speculate that the optimum VEGF effect was already achieved at or below 0.1 mg of plasmid. According to Tsurumi et al,(26) the paracrine effect of the secreted gene product may be sufficient to achieve a meaningful biological effect even if VEGF gene expression is limited to a relatively small number of cells. Moreover, the VEGF expression could be at a plateau because of saturated DNA uptake capacity of invading cells. Thus, greater amounts of DNA on the GAM may not necessarily lead to a greater expression of protein.

According to Peng et al.,(9) the ratio of VEGF expression to the presence of other growth factors is crucial to ensure synergistic effects in bone healing, and additional VEGF may not accelerate bone healing unless other growth factors are also supplemented. Their results revealed a similar increase in bone formation with a VEGF to BMP-4 ratio of 1:1–1:5, indicating that a ceiling effect for VEGF exists. If VEGF-transfected mesenchymal stem cells (MSCs) exceed the number of BMP-4-transfected cells, this positive effect decreased. In a similar manner, we did not find overwhelming vessel formation in the 1 mg pVEGF165 group that would indicate that higher dosages of VEGF might lead to the formation of hemangioma and thus prevent bone healing.

In literature, contradictory reports exist concerning the influence of VEGF on bone healing. It is well established that VEGF is essential for bone formation and that blocking of the VEGF receptors suppresses bone healing,(2,3) but the effect of additional VEGF did not lead to uniform results. Hiltunen et al.(23) described an increase in osteoblast number, osteoid volume, and bone volume after adenovirus-mediated VEGF gene transfer to osteoporotic femurs of New Zealand white rabbits. Street et al.(3) found an increase in callus formation after administering VEGF protein by an osmotic pump over 7 days to 10-mm critical size defects of rabbit radii. Their findings mirror ours in that superphysiologic levels of VEGF led to a doubling of callus volume in a critical size defect of the rabbit.

In contrast, Peng et al.(9) could not find an improvement in bone formation in VEGF transfected cells, although it enhanced the effect of BMP when they were combined. The different experimental settings may well have been responsible for the diverse findings reported.

Our second goal was to show that VEGF could promote angiogenesis and thus prevent the formation of atrophic nonunions. Specimens that received collagen alone or the GAM without the VEGF gene developed atrophic nonunions that, in accordance with the findings of Reed et al.,(27) were not totally avascular. However, most of the vessels counted in these samples came from the surrounding soft tissue, and almost no vessels were seen at the bone-fibrous tissue junction. The addition of the VEGF165 gene was found to increase vessel formation 2- to 3-fold after 6 weeks. With progressive calcification of the cartilage and osteoid within the callus, vessel quantification was obscured over time. Furthermore, the ratio of stained area to vessel density increased indicating that fewer but larger vessels were found after 12 weeks. Comparison of data among various authors studying the effect of VEGF is difficult because of the different experimental settings, methods of evaluation, animals used, and time periods.(9,24) For us, it was important to prove that angiogenesis can be promoted by local gene transfer of VEGF165 and thereby potentially prevent atrophic nonunions.

Our data shows that the model of a GAM as described by Bonadio et al.(12) is promising for delivering pVEGF165 to bone defects, but it should be emphasized that the effect on bone formation was variable. Nevertheless, the GAM model has several advantages over viral vectors or ex vivo transfected cells. Maximal local protein concentration can be assured because the GAM holds the vector within the defect. GAMs can easily be produced in advance and without autologous cells or in vitro culture. In addition to the cost of ex vivo transfection of cells, the disadvantages of apoptosis and possible migration of the transfected cells also need to be considered. Compared with viral vector systems, GAMs have the advantages of low immunogenicity, lower risk of vector spreading into other organs, simplicity of vector design, and relative ease of large-scale production.(28)

In clinical practice, VEGF will probably not be used as a single growth factor for bone formation. It was, however, important to examine the effect of VEGF alone to obtain a better understanding of its role in bone formation. In the future, its combination with other growth factors might improve the osteogenic capacity. The key role of VEGF will be to promote vessel formation (as shown in our study) and bone turnover and thus to prepare the site for other growth factors and cells. This development could lead to an improvement of the healing ratio and not merely accelerate bone formation as has been found for other growth factors in clinical trials.(19,21)

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We thank K Goetzke and R Foehr of the histology laboratory for great support and effort in performing the histology, T Hennig for preparing the plasmids, and C Lill and S Breit, who helped prepare the setting of this experiment. This project was supported by a grant of the Research Fund of the Orthopedic Hospital, University of Heidelberg.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Street J, Winter D, Wang JH, Wakai A, McGuinness A, Redmond HP 2000 Is human fracture hematoma inherently angiogenic? Clin Orthop 378: 224237.
  • 2
    Gerber HP, Vu TH, Ryan AM, Kowalski J, Werb Z, Ferrara N 1999 VEGF couples hypertrophic cartilage remodeling, ossification and angiogenesis during endochondral bone formation. Nat Med 5: 623628.
  • 3
    Street J, Bao M, deGuzman L, Bunting S, Peale FV Jr, Ferrara N, Steinmetz H, Hoeffel J, Cleland JL, Daugherty A, van Bruggen N, Redmond HP, Carano RA, Filvaroff EH 2002 Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. Proc Natl Acad Sci USA 99: 96569661.
  • 4
    Uchida S, Sakai A, Kudo H, Otomo H, Watanuki M, Tanaka M, Nagashima M, Nakamura T 2003 Vascular endothelial growth factor is expressed along with its receptors during the healing process of bone and bone marrow after drill-hole injury in rats. Bone 32: 491501.
  • 5
    Deckers MM, Karperien M, van der Bent C, Yamashita T, Papapoulos SE, Lowik CW 2000 Expression of vascular endothelial growth factors and their receptors during osteoblast differentiation. Endocrinology 141: 16671674.
  • 6
    Glowacki J 1998 Angiogenesis in fracture repair. Clin Orthop 355 (Suppl): 8289.
  • 7
    Nikol S, Huehns TY 2001 Preclinical and clinical experience in vascular gene therapy: Advantages over conservative/standard therapy. J Invasive Cardiol 13: 333338.
  • 8
    Nikol S, Engelmann MG, Pelisek J, Fuchs A, Golda A, Shimizu M, Mekkaoui C, Rolland PH 2002 Local perivascular application of low amounts of a plasmid encoding for vascular endothelial growth factor (VEGF165) is efficient for therapeutic angiogenesis in pigs. Acta Physiol Scand 176: 151159.
  • 9
    Peng H, Wright V, Usas A, Gearhart B, Shen HC, Cummins J, Huard J 2002 Synergistic enhancement of bone formation and healing by stem cell-expressed VEGF and bone morphogenetic protein-4. J Clin Invest 110: 751759.
  • 10
    Ferrara N 1999 Molecular and biological properties of vascular endothelial growth factor. J Mol Med 77: 527543.
  • 11
    Shea L, Smiley E, Bonadio J, Mooney D 1999 DNA delivery from polymer matrices for tissue engineering. Nat Biotechnol 17: 551554.
  • 12
    Bonadio J, Smiley E, Patil P, Goldstein S 1999 Localized, direct plasmid gene delivery in vivo: Prolonged therapy results in reproducible tissue regeneration. Nat Med 5: 753759.
  • 13
    Fang J, Zhu Y-Y, Smiley E, Bonadio J, Rouleau JA, Goldstein SA, McCauley LK, Davidson B, Roessler B 1996 Stimulation of new bone formation by direct transfer of osteoinductive plasmid genes. Proc Natl Acad Sci USA 93: 57535758.
  • 14
    Tuli S, Gupta K 1981 Bridging of large chronic osteoperiosteal gaps by allogeneic decalcified bone matrix implants in rabbits. J Trauma 21: 894898.
  • 15
    Wittbjer J, Palmer B, Thorngren KG 1982 Osteogenetic prop-erties of reimplanted decalcified and undecalcified autologous bone in the rabbit radius. Scand J Plast Reconstr Surg 16: 239244.
  • 16
    Mackenzie DJ, Sipe R, Buck D, Burgess W, Hollinger J 2001 Recombinant human acidic fibroblast growth factor and fibrin carrier regenerates bone. Plast Reconstr Surg 107: 989996.
  • 17
    Yasko A, Lane J, Fellinger E, Rosen V, Wozney J, Wang E 1992 The healing of segmental bone defects, induced by recombinant human bone morphogenetic protein (rhBMP-2). J Bone Joint Surg Am 74: 659670.
  • 18
    Probst A, Spiegel HU 1997 Cellular mechanisms of bone repair. J Invest Surg 10: 7786.
  • 19
    Bostrom MP, Camacho NP 1998 Potential role of bone morphogenetic proteins in fracture healing. Clin Orthop 355: 244282.
  • 20
    Boden SD, Kang J, Sandhu H, Heller JG 2002 Use of recombinant human bone morphogenetic protein-2 to achieve posterolateral lumbar spine fusion in humans: A prospective, randomized clinical pilot trial. 2002 Volvo Award in clinical studies. Spine 27: 26622673.
  • 21
    Friedlaender GE, Perry CR, Cole JD, Cook SD, Cierny G, Muschler GF, Zych GA, Calhoun JH, LaForte AJ, Yin S 2001 Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J Bone Joint Surg Am 83 (Suppl): 151158.
  • 22
    Kujala S, Raatikainen T, Ryhanen J, Kaarela O, Jalovaara P 2002 Composite implant of native bovine bone morphogenetic protein (BMP) and biocoral in the treatment of scaphoid nonunions-a preliminary study. Scand J Surg 91: 186190.
  • 23
    Hiltunen MO, Ruuskanen M, Huuskonen J, Mahonen AJ, Ahonen M, Rutanen J, Kosma VM, Mahonen A, Kroger H, Yla-Herttuala S 2003 Adenovirus-mediated VEGF-A gene transfer induces bone formation in vivo. FASEB J 17: 11471149.
  • 24
    Tarkka T, Sipola A, Jämsä T, Soini Y, Ylä-Herttuala S, Tuukkanen J, Hautala T 2003 Adenoviral VEGF-A gene transfer induces angiogenesis and promotes bone formation in healing osseous tissues. J Gene Med 5: 560566.
  • 25
    Eckardt H, Bundgaard KG, Christensen KS, Lind M, Hansen ES, Hvid I 2003 Effects of locally applied vascular endothelial growth factor (VEGF) and VEGF-inhibitor to the rabbit tibia during distraction osteogenesis. J Orthop Res 21: 335340.
  • 26
    Tsurumi Y, Takeshita S, Chen D, Kearney M, Rossow ST, Passeri J, Horowitz JR, Symes JF, Isner JM 1996 Direct intramuscular gene transfer of naked DNA encoding vascular endothelial growth factor augments collateral development and tissue perfusion. Circulation 94: 32813290.
  • 27
    Reed AA, Joyner CJ, Brownlow HC, Simpson AH 2002 Human atrophic fracture non-unions are not avascular. J Orthop Res 20: 593599.
  • 28
    Capito RM, Spector M 2003 Scaffold-based articular cartilage repair. IEEE Eng Med Biol Mag 22: 4250.