Exogenous Pro-Angiogenic Stimuli Cannot Prevent Physiologic Vessel Regression

Journal of Surgical Research 135, 218 –225 (2006) doi:10.1016/j.jss.2006.04.006

Exogenous Pro-Angiogenic Stimuli Cannot Prevent Physiologic Vessel Regression 1,2 Ankush Gosain, M.D., Ph.D.,*,†,3 Annette M. Matthies, Ph.D.,‡ Julia V. Dovi, Ph.D.,‡ Adrian Barbul, M.D.,§ Richard L. Gamelli, M.D.,† and Luisa A. DiPietro, D.D.S., Ph.D.¶,3 *Department of Cell Biology, Neurobiology and Anatomy, †Department of Surgery, and ‡Department of Microbiology and Immunology, Loyola University Medical Center, Burn and Shock Trauma Institute, Maywood, Illinois; §Department of Surgery, Sinai Hospital of Baltimore and the Johns Hopkins Medical Institutions, Baltimore, Maryland ¶Center for Wound Healing and Tissue Regeneration, College of Dentistry, University of Illinois at Chicago, Chicago, Illinois Submitted for publication January 9, 2006

Background. In healing wounds, rising levels of vascular endothelial growth factor (VEGF) induce a period of robust angiogenesis. The levels of proangiogenic factors in the wound begin to decline just before a period of vascular regression, suggesting that these mediators are necessary to sustain vessel density. The purpose of this study was to determine if the maintenance of pro-angiogenic stimuli in the wound would prevent physiological vessel regression. Materials and methods. A standard subcutaneous sponge wound model was modified by the addition of a mini-osmotic pump, allowing manipulation of the wound milieu by the addition of exogenous growth factors. After initial characterization of this model, exogenous VEGF (10 ␮g/mL), FGF (10 ␮g/mL), PDGF (10 ␮g/mL), or VEGF (10 ␮g/mL) plus FGF (10 ␮g/mL) were delivered to wounds and blood vessel density analyzed by immunohistochemistry. Results. VEGF administration resulted in a transient increase in wound vessel density (P < 0.05). None of the pro-angiogenic growth factors (VEGF, FGF, PDGF, VEGF/FGF) were able to prevent vascular regression (P ⴝ NS). Conclusions. These findings suggest that the antiangiogenic signals that mediate physiological vascular regression in wounds are strongly dominant over pro-angiogenic stimuli during the later phases of wound 1 Presented at the 1st Annual Academic Surgical Congress (Association for Academic Surgery), San Diego, CA, February 2006. 2 Ankush Gosain and Annette M. Matthies contributed equally to this work. 3 To whom correspondence and reprint requests should be addressed at Center for Wound Healing and Tissue Regeneration, College of Dentistry, University of Illinois at Chicago, Chicago, Illinois 60612-7211. E-mail: [email protected].

0022-4804/06 $32.00 © 2006 Elsevier Inc. All rights reserved.

healing. Clinical manipulation of anti-angiogenic signals in addition to the currently used pro-angiogenic targets may be needed to achieve therapeutic modulation of blood vessel density. © 2006 Elsevier Inc. All rights reserved.

Key Words: angiogenesis; regression; wound; VEGF, FGF, PDGF.

INTRODUCTION

Physiological angiogenesis, the growth of new capillaries from pre-existing vasculature, is highly regulated, allowing for the controlled formation and regression of vessels [1]. Vascular endothelial growth factor (VEGF), a potent pro-angiogenic factor, is critical to angiogenic processes [2] and the blockade of VEGF leads to a significant decrease in angiogenic activity in wounds, tumors and the corpus luteum [3, 4]. The pro-angiogenic phase of tissue repair is mediated primarily by VEGF [2, 5]. In healing wounds, a robust angiogenic response is observed just after the peak level of VEGF is reached in the wound bed. New vessels sprout and grow to a maximum vessel density three times that observed in uninjured skin. After maximal capillary density is attained, a distinct phase of vessel regression begins. In this phase, newly formed vessels are pruned until the vessel density is comparable to that found in normal skin [6]. In contrast, many pathological conditions (e.g., tumors, arthritis) lack a phase of vessel regression. The maintenance of an angiogenic phenotype is a key element in the pathogenesis of these diseases [7]. The events contributing to vessel regression at sites of physiological angiogenesis are not well understood.

218

GOSAIN ET AL.: CONTROL OF BLOOD VESSEL REGRESSION

Studies in adult mice have shown that VEGF is required to sustain immature blood vessels and downregulation of VEGF expression results in vessel regression [8]. In the resolving wound, the level of VEGF declines just before the onset of vessel regression and the loss of the VEGF survival signal has been hypothesized to be causative in vessel regression [9]. This hypothesis is supported by the observation that VEGF withdrawal has been shown to decrease vessel density in tumors [10]. While a large body of evidence suggests that VEGF is the single most important pro-angiogenic growth factor in the healing wound [5], other growth factors have been implicated in promotion of angiogenesis [11]. Fibroblast growth factor-2 (FGF-2) induces both mitogenic and chemotactic responses in endothelial cells and plays a central role in embryonic vasculogenesis [11–13]. Platelet-derived growth factor (PDGF) promotes proliferation and migration of mesenchymal cells via the receptor tyrosine kinases PDGF-R␣ and –␤ [14]. In the present study, we hypothesized that sustained levels of pro-angiogenic growth factors in the wound would prevent physiological vascular regression. To test this hypothesis, we characterized a model system in which the process of physiological angiogenesis mirrors that seen in cutaneous wounds, while allowing us to manipulate the wound milieu in clearly defined ways. We investigated the effect of the addition of exogenous recombinant VEGF 164 protein to the wound site at three different stages in wound angiogenesis: robust vessel growth, maximal vessel density, and vessel regression. Additionally, the ability of other known pro-angiogenic growth factors to sustain vessel density in healing wounds was investigated. We demonstrate that physiological blood vessel regression takes place despite the delivery of exogenous pro-angiogenic growth factors to the wound. MATERIALS AND METHODS Sponge-Wound Constructs A standard sponge wound model modified to allow constant VEGF administration to the wound site was used (Fig. 1) [15]. In this model, a subcutaneously implanted polyvinyl alcohol sponge serves as the wound site. This model is modified by the addition of a mini-osmotic pump joined to the sponge by tubing, allowing the continuous administration of growth factors directly to the sponge wound. The sponge-tubing constructs were created using Silastic laboratory tubing (0.76 mm ⫻ 1.65 mm, Dow Corning Corporation, Midland, MI). Two holes were cut in the tubing, one on each side, approximately 2 cm from the distal end. Polyvinyl alcohol sponge material (M-PACT Worldwide Management Corporation, Eudora, KS) was cut into 8 mm disks and channels through which the tubing was fed were created using a spinal needle. The tubing was fed through the sponge and the holes in the tubing were centered in the middle of the sponge (Fig. 1A).

219

FIG. 1. Modified sponge wound model. See text for a complete description. A, The enlarged sponge depicts the holes in the tubing over which the sponge is positioned. B, Dorsal view of the animal and placement of the sponge and tubing after insertion. Note the suture placed at the distal end of the tubing to ensure that all of the infusate exits the tube into the sponge. C, Ventral view of the animal and placement of the pump after attachment to the tubing is shown. Incisions are represented by dashed lines.

Implantation of Sponge-Wound Constructs All animal procedures were reviewed and approved by the Loyola University Institutional Animal Use and Care Committee. Female BALB/c mice (Jackson Laboratory, Bar Harbor, ME) aged 8 to 9 weeks, were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). The dorsal and ventral sides of the animals were shaved and scrubbed with betadine. A transverse incision was then made on the dorsal interscapular region. The sponge-tubing construct was placed subcutaneously caudal to the dorsal incision (Fig. 1B). A second incision was then made transversely overlying the ventral thorax. A subcutaneous tunnel was created over the shoulder of the animal connecting the dorsal incision to the ventral incision. The tubing portion of the construct was

220

JOURNAL OF SURGICAL RESEARCH: VOL. 135, NO. 2, OCTOBER 2006

passed through this tunnel and inserted into a subcutaneous pocket caudal to the ventral incision. The incisions were closed using two surgical wound clips dorsally and one surgical wound clip ventrally. The day of initial surgery was considered day 0.

Treatment Groups and Implantation of Mini-Osmotic Pumps For initial characterization, sponges were harvested at various time points from day 8 to day 56 for analysis (n ⫽ 4 – 6 per group, per time point). For immunohistochemical analysis, sponges were embedded in Tissue Freezing Medium (Triangle Biomedical Sciences, Durham, NC) and stored at ⫺80°C. For ELISA analysis, wound fluid was eluted from harvested sponges into 1.0 mL of 1X PBS containing Complete Protease Inhibitor Cocktail (Roche Molecular Biochemicals, Indianapolis, IN). After the temporal peak in endogenous VEGF had been identified as day 17 (see Results), a separate experiment was initiated in which either VEGF 164 (10 ␮g/mL, R&D Systems, Minneapolis, MN) or 1X PBS (control) was administered via the mini-osmotic pumps (n ⫽ 6 –12 per group, per time point). Animals were re-anesthetized with ketamine and xylazine intraperitoneal injection. A second ventral incision was made approximately 5 mm caudal to the first incision and a distal subcutaneous pocket created. A mini-osmotic pump (DURECT Corporation, Cupertino, CA), dispensing 0.5 ␮L/hr for 14 days, was attached to the construct tubing and implanted in the subcutaneous pocket (Fig. 1C). On day 28, animals were sacrificed and sponges retrieved from their subcutaneous positions. Finally, the ability of other known pro-angiogenic growth factors to sustain vessel density was investigated by the administration of PDGF alone (10 ␮g/mL, R&D Systems), FGF-2 alone (10 ␮g/mL, R&D Systems), and FGF-2 (10 ␮g/mL) in combination with VEGF 164 (10 ␮g/mL). These groups were compared to PBS-treated controls (n ⫽ 3–7 per group, per time point).

Quantification of VEGF Levels in the Wound Sponge wound extracts were centrifuged to remove debris, and filtered through a 1.2 ␮M pore syringe filter. VEGF levels were determined with a commercially available ELISA kit (Quantikine murine VEGF, R&D Systems).

Immunohistochemistry There were 700 ␮m sections (thickness required for collection of an intact sponge section) from frozen embedded sponges prepared for immunohistochemical analysis of PECAM-1, an endothelial cell marker. All incubations and washes were carried out at room temperature; 1X PBS, pH 7.4, was used for all washes. Sections were fixed in acetone overnight. After three 3-min washes, sections were treated with 0.3% H 2O 2 in methanol for 30 min to quench endogenous peroxidase activity. The slides were washed in PBS and then blocked with a 1:10 dilution of normal mouse serum (Sigma Chemical Company, St. Louis, MO) in 1X PBS for 30 min. Sections were incubated in 1.0 ␮g/mL of MEC13.3 rat anti-mouse PECAM-1 antibody (anti-CD31, PharMingen International, San Diego, CA) in 1X PBS for 30 min. After incubation with the primary antibody, the slides were washed and then incubated for 30 min with 13.0 ␮g/mL of biotinylated mouse anti-rat IgG antibody (Jackson ImmunoReseach Laboratories, West Grove, PA). After three 3-min washes, slides were incubated with avidin-biotin-horseradish peroxidase complex (ABC-HRP, Vector Laboratories, Burlingame, CA) for 30 min. After another set of three washes, slides were incubated in a HRP substrate, 3,3=-diaminobenzidine (Kirkegaard and Perry Laboratories, Gaithersburg, MD), for 10 min. Coverslips were mounted with Cytoseal 280 (Stephens Scientific, Kalamazoo, MI).

Analysis of Vessel Density Vessel density was determined by blinded observers as previously described [6]. Briefly, images of PECAM-stained sponge sections were captured using Scion Image Software (version 4.0.2, Scion Corporation, Frederick, MD). For each sponge section the total area of the section was divided into measurable fields of view and digitally imaged at 5⫻ magnification. The PECAM positive area within each field of view was measured and the vessel density was calculated. For each sponge section, the vessel density for each field of view was determined and an average of the fields was calculated to yield the percent vessel density for the sponge section.

In vitro Cord Formation Assay There was 1 mL of collagen gel containing 4 mg/mL of rat tail collagen (Upstate Biotechnology, Lake Placid, NY) in PBS, pH 7.4, plated onto each 35 mm dish [5, 16]. Murine endothelial cells, SVEC4-10 (American Type Culture Collection, Rockville, MD), at a concentration of 5 ⫻ 105 cells/mL, were first incubated for 30 min at room temperature in DMEM containing 10% FBS and 100 ng/mL of control murine VEGF 164 (R&D Systems). Some cell suspensions also received 100 ng/mL of VEGF retrieved from mini-osmotic pumps on day 28. After this incubation, 1 mL of the cell suspension was then plated on the prepared collagen gel. After 4 h of incubation on collagen gels, the cells were photographed at five randomly chosen fields on each blinded culture dish. Cord-like structures were counted per field and the average number of cord-like structures per field calculated for experimental and control groups. For each independent experiment, the numbers of endothelial cords formed in the presence of control murine VEGF 164 were considered as maximal (100%), and experimental values were calculated as a percentage of maximal cord formation.

Statistical Analysis The mean and standard error of mean were calculated for each experimental group. Statistical analysis was performed using GraphPad Prism (Version 4.0, GraphPad Software, San Diego, CA). Data that were described over time were analyzed by two-way ANOVA followed by Bonferonni post-comparison testing. Data described at single time points were analyzed by unpaired t-test. Results with a P value less than 0.05 were considered statistically significant.

RESULTS Characterization of Angiogenesis in the Sponge-Wound Model

In the course of healing, blood vessels originating from pre-existing vessels in the overlying skin infiltrate the sponge, reach a maximal vessel density, and finally enter a period of vascular regression. Newly formed capillaries were first observed at the lateral edges of sponges between days 8 to 12 (Fig. 2A) and a robust angiogenic response was observed from day 12 to 24. During this time period, new vessels migrated from the lateral edge toward the center of the sponge to form a dense vascular bed throughout the sponge (Fig. 2A). Vessel density, measured by image analysis of CD31⫹ area, reached a maximal level (8.6 ⫾ 1.0%) by day 24 (Fig. 2B). The maximal vessel density was comparable to the maximal value for vessel density observed in other models of wound angiogenesis [6, 17]. A sharp decrease

GOSAIN ET AL.: CONTROL OF BLOOD VESSEL REGRESSION

221

FIG. 2. Characteristics of angiogenesis in the sponge wound model. A, Images of gross sponges and representative histological sections from days 8, 12, 24, 40, and 56. The increase vessel density from days 8 to 24 and the subsequent regression from days 24 to 56 are seen on both the gross and microscopic scale. B, VEGF concentration in untreated sponges was measured by ELISA and reached a peak level at day 16. Vessel density was determined by PECAM staining and found to peak at day 24. These patterns are similar to those seen in other wound models.

in vessel density was observed immediately after day 24, and a distinct phase of vessel regression was observed from days 24 to 56 (Fig. 2B). Measurement of endogenous VEGF levels demonstrated that the level of VEGF peaked (327.3 ⫾ 35.6 pg/ml) on day 16 (Fig. 2B). The level of VEGF then gradually declined, reaching 111.4 ⫾ 37.8 pg/ml at day 26. Together, the pattern of VEGF levels and vessel density determined for the sponge wound model are similar to the patterns we have observed in another established model of wound healing, with the peak VEGF level occurring several day before maximum vessel density [6]. Supraphysiologic Levels of Biologically Active VEGF Are Delivered to the Sponge Wound

To sustain the level of VEGF at the wound site throughout the time of vascular regression, delivery of exogenous VEGF 164 by mini-osmotic pump was initiated at day 17. Assessment of the VEGF levels in the

sponges confirmed that the osmotic pump infusion yielded significantly elevated VEGF levels throughout the time course studied (Fig. 3, P ⬍ 0.05). An in vitro experiment was performed to confirm the biological activity of exogenously administered VEGF. Mini-osmotic pumps were attached to implanted sponge-tubing constructs at day 17 and the pump opening was closed immediately after attachment. At day 28, the remaining recombinant VEGF 164 within the pump was retrieved. The activity of the retrieved VEGF 164 was examined with an in vitro endothelial cell cord formation assay. Endothelial cells plated on collagen and cultured in the presence of VEGF isolated from the pump formed significantly more cord-like structures than endothelial cells incubated with media alone (Fig. 4, P ⬍ 0.05). The activity of VEGF isolated from the pump was comparable to fresh recombinant VEGF 164, confirming that the recombinant growth factors present in the mini-osmotic pump retain robust biological activity.

222

JOURNAL OF SURGICAL RESEARCH: VOL. 135, NO. 2, OCTOBER 2006

FIG. 3. Mini-osmotic pumps deliver supraphysiologic levels of VEGF to the wound. The VEGF levels determined by ELISA in sponges infused with recombinant VEGF 164 were elevated throughout the time course studied as compared to PBS-infused control sponges (P ⬍ 0.0001).

Exogenous VEGF Transiently Increases Wound Vessel Density But Fails to Prevent Regression

Implanted sponge wounds infused with VEGF164 exhibited patches of dense vascularity and a significant increase in the vessel density during the time period of

active capillary growth in this model (12.2 ⫾ 1.7% versus 5.5 ⫾ 0.8% at day 20, P ⬍ 0.001, Fig. 5). This indicates that exogenous VEGF enhances the pro-angiogenic phase of healing in the sponge model. The dense vascular patches and increased vascularity that were observed in the VEGF-treated sponges were transient, and by day 24 VEGF-treated sponge wounds exhibited similar patterns of vascularity to controls (Fig. 5). Sponge wounds infused with VEGF 164 were also examined during the phase of vessel regression (days 28, 32, and 36). In contrast to wounds from the proangiogenic phase, no enhancement of vascularity or dense vascular patches were observed in VEGF 164 infused sponges collected during the period of vascular regression. At all time points during the regression phase, implanted sponges infused with recombinant VEGF 164 showed no significant difference in the vessel density when compared to control sponges infused with PBS (Fig. 5). The results demonstrate that wound vessel regression proceeds despite the presence of sustained, elevated VEGF levels. Exogenous PDGF, FGF, and FGF/VEGF Fail to Prevent Vascular Regression

To evaluate the ability of other, known proangiogenic mediators to sustain vessel density in heal-

FIG. 4. Exogenously administered VEGF is biologically activate. A, For each individual experiment, cord formation in cultures with control recombinant VEGF164 was set as 100% cord formation. The percentage of cord formation in cultures incubated with VEGF isolated from osmotic pumps was not significantly different from the percent cord formation in cultures incubated with the control recombinant VEGF164. B, Representative photographs of cell cultures treated with media, control VEGF164, and VEGF retrieved from osmotic pumps. The enlarged cord in the upper left corner of each photograph is representative of the boxed cord in the photograph of the cell culture (100⫻ magnification).

GOSAIN ET AL.: CONTROL OF BLOOD VESSEL REGRESSION

FIG. 5. Recombinant VEGF 164 transiently increases wound vessel density but fails to prevent vascular regression. Sponge wounds infused with recombinant VEGF 164 exhibited a significantly higher vessel density compared to PBS-infused control sponges at day 20 (*P ⬍ 0.001). The vessel density of sponge wounds infused with recombinant VEGF 164 was not significantly different from PBSinfused control sponges at days 24, 28, 32, and 36.

ing wounds, PDGF, FGF-2, and VEGF 164 were delivered to wounds starting at day 17. Vessel density analysis of sponges retrieved at day 28 revealed no significant differences between control, PDGF-treated, FGF-treated and VEGF/FGF-treated sponges (Fig. 6, P ⫽ NS). These findings indicate that pro-angiogenic mediators cannot overcome the late anti-angiogenic environment of the healing wound. DISCUSSION

This study of the regulation of wound angiogenesis resulted in three important observations. First, physiological vessel density can be enhanced when a proangiogenic factor is administered during the period of active vessel formation. Secondly, this maximal vessel density cannot be maintained simply by the administration of pro-angiogenic factors. Finally, it appears that the ultimate vessel density in a system of normal physiological angiogenesis cannot be increased over an inherent set maximum. Angiogenesis plays an essential role in many physiological conditions, including pregnancy, menstruation, and wound healing [18]. For these processes to occur normally, the process of angiogenesis is tightly regulated. After new vessel growth, the pruning of newly formed vessels occurs during a period of vessel regression. In multiple disease states, including inflammatory diseases and malignant tumors, the absence of vessel regression allows for the persistence of vessel growth, and such dysregulated angiogenesis contributes to the pathogenesis of these diseases [7]. In healing wounds, VEGF is the primary mediator of

223

the pro-angiogenic phase of tissue repair [2, 5]. Although numerous studies have demonstrated a role for VEGF in the initiating events of wound angiogenesis, its role in the process of wound vessel regression has not been studied. The current study demonstrates that exogenous VEGF 164, administered to a wound site during a period of robust angiogenesis, yields a sharp and statistically significant increase in vessel density. This increased vascularity includes the appearance of dense patches of blood vessels characteristic of VEGF-treated tissue [19]. Our observations are consistent with previous reports showing that physiological angiogenesis in hair follicles and skin substitutes can be drastically enhanced by VEGF treatment during periods of active vessel growth [20 –22]. Interestingly, we observed that although the administration of VEGF led to patches of dense vasculature and increased vessel density during active angiogenesis, these features resolved during the phase of vessel regression in the sponge wound. These findings suggest that wound vessel density is controlled both spatially and temporally to meet the physiological requirement of the tissue. The most dramatic finding of the current study is that vascular regression proceeds despite the administration of supra-physiological concentrations of biologically active pro-angiogenic mediators to the wound site. This observation provides a new paradigm for the understanding of the regulation of physiological angiogenesis. The concept that net angiogenic effect depends upon a balance of pro-angiogenic and anti-angiogenic mediators is well accepted [7, 23]. However, the description of such a dominant anti-angiogenic environment within the resolving wound provides new information about how the level of vascularity is maintained in normal tissues. Indeed, the environment of the resolving wound does not appear to simply halt the growth of

FIG. 6. Exogenous PDGF, FGF-2, and FGF-2/VEGF 164 fail to prevent vascular regression. Sponge wounds infused with either recombinant FGF-2 or a combination of recombinant FGF-2 and VEGF 164 exhibited a vessel density similar to PBS-infused control sponges at day 28 (FGF-2 1.07 ⫾ 0.04 ⫻ control, VEGF/FGF-2 1.04 ⫾ 0.04 ⫻ control, P ⫽ NS).

224

JOURNAL OF SURGICAL RESEARCH: VOL. 135, NO. 2, OCTOBER 2006

new vessels, but to also require the regression of existing ones. While scant evidence for this anti-angiogenic environment was previously available, our findings provide novel evidence that during late stages in wound resolution, anti-angiogenic stimulation predominates over the stimuli of pro-angiogenic factors. The precise nature of the anti-angiogenic factors involved in wound vessel regression is not known and multiple possibilities exist. Down-regulation of VEGFreceptor-2 (VEGF-R2) may prevent endothelial cells from responding to pro-angiogenic stimuli late in angiogenesis or there may exist an alternate ligand for VEGF-R2 that prevents VEGF signaling. Evidence for alternate ligands for VEGF-R2 comes from experiments in which knockout mice lacking in either VEGF or VEGF-R2 die early in embryogenesis, with a more severe phenotype seen in those animals lacking the receptor [24, 25]. Extracellular matrix cleavage products generated during tissue remodeling may also inhibit wound neovascularization [26]. In addition, endothelial cells may be influenced by the secretion of growth factors by inflammatory cells at the wound site, resulting in the inhibition of vessel formation and the onset of vessel regression [27, 28]. Although wound angiogenesis is primarily mediated by soluble factors, direct cell-cell contact or cell-basement membrane contact may also influence the endothelial cell response during wound vessel regression. Finally, there is increasing evidence that the method of delivery of growth factors influences the ability to grow durable and functional vessels [29]. Pericyte coverage of newly formed vessels may play a role in vascular survival. Pericytes have been shown to express PDGF receptors and respond to PDGF in vitro [30, 31]. Additionally, PDGF-B deficient mice form numerous capillary microaneurysms that rupture in late gestation [32, 33]. These aneurysmal areas are devoid of pericytes, suggesting that pericytes may contribute to the mechanical stability of the capillary wall. Further evidence for this concept comes from the observation that these microaneurysms are similar in morphology to those seen in diabetic retinopathy [34], a disease state also characterized by the absence of capillary pericytes. Finally, blockade of PDGF-B or PDGFR-␤ signaling results in an immature microvasculature that is highly dependent on VEGF for survival [35, 36]. The regulation of wound angiogenesis and vessel regression is a complex process. Although the proangiogenic phase of wound healing appears to be controlled primarily by one key mediator (VEGF), wound vessel regression may be the under the regulation of multiple factors. Our findings suggest that the antiangiogenic signals that mediate physiological vascular regression in wounds are strongly dominant over proangiogenic stimuli during the later phases of wound healing. Clinical manipulation of anti-angiogenic sig-

nals in addition to the currently used pro-angiogenic targets (e.g., anti-VEGF antibody) may be needed to achieve therapeutic modulation of blood vessel density. ACKNOWLEDGMENT This work was supported by NIH T32-GM08750 (AG), R01GM50875 (LAD), R01-GM55238 (LAD).

REFERENCES 1.

2.

3.

4.

5.

Ferrara N. Role of vascular endothelial growth factor in regulation of physiological angiogenesis. Am J Physiol Cell Physiol 2001;280:C1358. Nissen NN, Polverini PJ, Koch AE, Volin MV, Gamelli RL, DiPietro LA. Vascular endothelial growth factor mediates angiogenic activity during the proliferative phase of wound healing. Am J Pathol 1998;152:1445. Ferrara N, Chen H, Davis-Smyth T, et al. Vascular endothelial growth factor is essential for corpus luteum angiogenesis. Nat Med 1998;4:336. Howdieshell TR, Callaway D, Webb WL, et al. Antibody neutralization of vascular endothelial growth factor inhibits wound granulation tissue formation. J Surg Res 2001;96:173. Matthies AM, Low QE, Lingen MW, DiPietro LA. Neuropilin-1 participates in wound angiogenesis. Am J Pathol 2002;160:289.

6.

Swift ME, Kleinman HK, DiPietro LA. Impaired wound repair and delayed angiogenesis in aged mice. Lab Invest 1999;79: 1479.

7.

Folkman J. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1995;1:27.

8.

Alon T, Hemo I, Itin A, Pe’er J, Stone J, Keshet E. Vascular endothelial growth factor acts as a survival factor for newly formed retinal vessels and has implications for retinopathy of prematurity. Nat Med 1995;1:1024.

9.

Giavazzi R, Sennino B, Coltrini D, et al. Distinct role of fibroblast growth factor-2 and vascular endothelial growth factor on tumor growth and angiogenesis. Am J Pathol 2003;162:1913.

10.

Benjamin LE, Golijanin D, Itin A, Pode D, Keshet E. Selective ablation of immature blood vessels in established human tumors follows vascular endothelial growth factor withdrawal. J Clin Invest 1999;103:159.

11.

Nissen NN, Polverini PJ, Gamelli RL, DiPietro LA. Basic fibroblast growth factor mediates angiogenic activity in early surgical wounds. Surgery 1996;119:457.

12.

Stavri GT, Zachary IC, Baskerville PA, Martin JF, Erusalimsky JD. Basic fibroblast growth factor upregulates the expression of vascular endothelial growth factor in vascular smooth muscle cells. Synergistic interaction with hypoxia. Circulation 1995; 92:11.

13.

Tomanek RJ, Lotun K, Clark EB, Suvarna PR, Hu N. VEGF and bFGF stimulate myocardial vascularization in embryonic chick. Am J Physiol 1998;274:H1620.

14.

Heldin CH. Structural and functional studies on plateletderived growth factor. Embo J 1992;11:4251.

15.

Efron DT, Most D, Shi HP, Tantry US, Barbul A. A novel method of studying wound healing. J Surg Res 2001;98:16.

16.

Goto F, Goto K, Weindel K, Folkman J. Synergistic effects of vascular endothelial growth factor and basic fibroblast growth factor on the proliferation and cord formation of bovine capillary endothelial cells within collagen gels. Lab Invest 1993;69: 508.

GOSAIN ET AL.: CONTROL OF BLOOD VESSEL REGRESSION 17.

18. 19.

20.

21.

22.

23. 24.

25.

26.

27.

Jang YC, Arumugam S, Gibran NS, Isik FF. Role of alpha (v) integrins and angiogenesis during wound repair. Wound Repair Regen 1999;7:375. Risau W. Mechanisms of angiogenesis. Nature 1997;386:671. Larcher F, Murillas R, Bolontrade M, Conti CJ, Jorcano JL. VEGF/VPF overexpression in skin of transgenic mice induces angiogenesis, vascular hyperpermeability and accelerated tumor development. Oncogene 1998;17:303. Yano K, Brown LF, Detmar M. Control of hair growth and follicle size by VEGF-mediated angiogenesis. J Clin Invest 2001;107:409. Supp DM, Boyce ST. Overexpression of vascular endothelial growth factor accelerates early vascularization and improves healing of genetically modified cultured skin substitutes. J Burn Care Rehabil 2002;23:10. Vale PR, Isner JM, Rosenfield K. Therapeutic angiogenesis in critical limb and myocardial ischemia. J Interv Cardiol 2001; 14:511. Carmeliet P. Angiogenesis in health and disease. Nat Med 2003;9:653. Carmeliet P, Ferreira V, Breier G, et al. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 1996;380:435. Shalaby F, Rossant J, Yamaguchi TP, et al. Failure of bloodisland formation and vasculogenesis in Flk-1-deficient mice. Nature 1995;376:62. Nor JE, Mitra RS, Sutorik MM, Mooney DJ, Castle VP, Polverini PJ. Thrombospondin-1 induces endothelial cell apoptosis and inhibits angiogenesis by activating the caspase death pathway. J Vasc Res 2000;37:209. Luster AD, Cardiff RD, MacLean JA, Crowe K, Granstein RD.

28.

29.

30.

31. 32.

33.

34.

225

Delayed wound healing and disorganized neovascularization in transgenic mice expressing the IP-10 chemokine. Proc Assoc Am Physicians 1998;110:183. Albini A, Marchisone C, Del Grosso F, et al. Inhibition of angiogenesis and vascular tumor growth by interferon-producing cells: A gene therapy approach. Am J Pathol 2000;156:1381. Ehrbar M, Djonov VG, Schnell C, et al. Cell-demanded liberation of VEGF121 from fibrin implants induces local and controlled blood vessel growth. Circ Res 2004;94:1124. Bernstein LR, Antoniades H, Zetter BR. Migration of cultured vascular cells in response to plasma and platelet-derived factors. J Cell Sci 1982;56:71. D’Amore PA, Smith SR. Growth factor effects on cells of the vascular wall: A survey. Growth Factors 1993;8:61. Leveen P, Pekny M, Gebre-Medhin S, Swolin B, Larsson E, Betsholtz C. Mice deficient for PDGF B show renal, cardiovascular, and hematological abnormalities. Genes Dev 1994;8: 1875. Lindahl P, Johansson BR, Leveen P, Betsholtz C. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science 1997;277:242.

Buzney SM, Frank RN, Varma SD, Tanishima T, Gabbay KH. Aldose reductase in retinal mural cells. Invest Ophthalmol Vis Sci 1977;16:392. 35. Bergers G, Song S, Meyer-Morse N, Bergsland E, Hanahan D. Benefits of targeting both pericytes and endothelial cells in the tumor vasculature with kinase inhibitors. J Clin Invest 2003; 111:1287. 36. Song S, Ewald AJ, Stallcup W, Werb Z, Bergers G. PDGFRbeta⫹ perivascular progenitor cells in tumours regulate pericyte differentiation and vascular survival. Nat Cell Biol 2005;7:870.