Ang-1 (angiopoietin-1) improves the ineffective angiogenesis and impaired wound healing in diabetes; however, the mechanism underlying this positive effect is still far from being completely understood. In the present study, we investigated whether rAAV (recombinant adeno-associated virus)–Ang-1 gene transfer could improve wound repair in genetically diabetic mice (db+/db+) and the mechanism(s) by which it causes new vessel formation. An incisional skin-wound model in diabetic and normoglycaemic mice was used. After the incision, animals received rAAV–LacZ or rAAV–Ang-1 in the wound edge. After 7 and 14 days, wounds were used to (i) confirm Ang-1 gene transfer, (ii) assess histologically the healing process, (iii) evaluate wound-breaking strength, and (iv) study new vessel formation by PECAM-1 (platelet/endothelial cell adhesion molecule-1) immunostaining. Finally, we investigated VEGF (vascular endothelial growth factor) mRNA and protein levels, eNOS (endothelial NO synthase) expression and VEGFR-1 and VEGFR-2 (VEGF receptor-1 and -2 respectively) immunostaining. The efficiency of Ang-1 gene transfer was confirmed by increased mRNA and protein expression of the protein. rAAV–Ang-1 significantly improved the healing process, stimulating re-epithelization and collagen maturation, increasing breaking strength and significantly augmenting the number of new vessels, as indicated by PECAM-1 immunostaining. However, Ang-1 gene transfer did not modify the decrease in VEGF mRNA and protein expression in diabetic mice; in contrast, Ang-1 increased eNOS expression and augmented nitrate wound content and VEGFR-2 immunostaining and protein expression. Ang-1 gene transfer did not change vascular permeability. Similar results were obtained in normoglycaemic animals. In conclusion, our results provide strong evidence that Ang-1 gene transfer improves the delayed wound repair in diabetes by inducing angiogenesis in a VEGF-independent manner.

INTRODUCTION

Angiogenesis is a multistep process that involves EC (endothelial cell) sprouting from the parent vessel, followed by migration, proliferation, alignment, tube formation and anastomosis to other vessels [1]. Active angiogenesis occurs in adult life only under certain physiological and pathological situations. This process is associated with the expression of cytokines, as well as angiogenic factors such as VEGF (vascular endothelial growth factor) [2]. Members of the VEGF family, including VEGF (also known as VEGF-A), VEGF-B and PlGF (placental growth factor), play important roles in angiogenesis. The effects of VEGF are mediated via receptors termed Flt-1 or VEGFR-1 (VEGF receptor-1) and KDR/Flk-1 or VEGFR-2 (VEGF receptor-2) [3]. VEGF levels are increased during hypoxic conditions, making VEGF-driven angiogenesis a central response to low oxygen tension.

Insufficient angiogenesis is involved in the impaired healing of diabetic skin ulcers [4,5]. Indeed, healing impairment in diabetes is characterized by delayed cellular infiltration and granulation tissue formation, decreased collagen organization and, of course, reduced angiogenesis [4,5]. As far as angiogenesis is concerned, a defect in VEGF regulation, characterized by an altered expression pattern of VEGF mRNA during skin repair, has been shown in db/db mice [6]. In addition, to confirm the pivotal role of VEGF in skin repair abnormalities in diabetic animals, it has been reported that AAV (adeno-associated virus) vector-mediated human VEGF165 gene transfer stimulates angiogenesis and wound healing in genetically diabetic mice [7].

Besides angiogenesis, vasculogenesis may also play a crucial role in the healing process. Vasculogenesis, the in situ differentiation of the primitive endothelial progenitors, known as angioblasts, into ECs that aggregate into a primary capillary plexus, has been shown to be responsible for the development of the vascular system during embryogenesis [8]; however, vasculogenesis is also present in adults and occurs through the action of circulating or resident bone-marrow-derived cells called EPCs (endothelial progenitor cells) [9]. Further cell lineages not bone-marrow-derived may be found at different sites and have been demonstrated to differentiate into ECs under hypoxic conditions or during physiological replenishment of the skin and gut [10]. Moreover, vasculogenesis appears to be more present and effective when angiogenesis is failing: this is the case in the healing of diabetic ulcers in which there is an impairment of haemostasis, inflammation, matrix deposition and, most of all, angiogenesis [11].

Angs (angiopoietins) are known to work in concert with VEGF in vascular growth, development and maintenance as well as in the regulation of microvascular permeability [12]. Ang-1 has been also proposed to stimulate angiogenesis via the activation of the Akt signalling pathway and finally stimulates eNOS (endothelial NO synthase) [13]. This mechanism could be extremely important in the repair of wounds, especially during impaired wound healing, which may be a consequence of metabolic derangements such as diabetes [14].

Recently, Cho et al. [15] reported the beneficial effects of a soluble form of Ang-1, COMP (cartilage oligomeric matrix protein)-Ang1, on the promotion of healing in cutaneous wounds in diabetic mice by using an adenovirus coding for COMP-Ang1; however, the combination of both COMP and Ang-1 did not allow the investigators to dissect out the exact efficacy of Ang-1 in improving the impaired healing of wounds and to identify the underlying specific mechanism of action. In addition, that study [15] did not address the issue of a possible interplay between Ang-1 and VEGF.

Therefore the aim of the present study was to investigate the mechanism(s) underlying the beneficial effects of Ang-1 in the diabetes-induced impairment of wound healing.

MATERIALS AND METHODS

rAAV (recombinant AAV) vector preparation and characterization

Two rAAV vectors were used in the present study, one expressing the LacZ reporter gene (control gene) and the other cDNA for mouse Ang-1 under the control of the strong and constitutive CMV (cytomegalovirus) immediate early promoter. Both constructs were based on the plasmid pFU-5, kindly provided by Dr Nicholas Muzyczka (Department of Molecular Biology and Genetics, University of Florida, Gainesville, FL, U.S.A.) [16]. pAAV–LacZ was obtained by substituting the GFP (green fluorescent protein) open reading frame with the LacZ gene from the plasmid pCH110 (Amersham Biosciences). The cDNA for Ang-1 was obtained by RT-PCR (real-time PCR) amplification of total RNA with appropriate primer pairs and, again, cloned to replace GFP in the pFU-5 vector digested with BamHI and EcoRI. Infectious vector stocks were generated in HEK-293 cells (human embryonic kidney cells), cultured in 150-mmdiameter Petri dishes, by co-transfecting each plate with 15 μg of each vector plasmid, together with 45 μg of the packaging/helper plasmid pDG (kindly provided by Dr Jürgen A. Kleinschmidt, Program of Infection and Cancer, German Cancer Research Center, Heidelberg, Germany) expressing AAV and adenovirus helper functions [17]. At 12 h after transfection, the medium was replaced with fresh medium and, 3 days later, the medium was collected and the cells harvested by scraping. After three freeze–thaw cycles in a dry ice/ethanol bath and a 37 °C water bath, cell lysates were fractionated using ammonium sulfate precipitation. rAAV particles were then purified by CsCl gradient centrifugation in a SW41Ti rotor at 288000 g for 36 h. Between 12 and 16 fractions of ten drops each were collected by inserting a 21-gauge needle below the rAAV band, and their refractive index was determined. The six fractions with an index closest to 1.3715 (corresponding to a density of 1.40 g/cm3) were dialysed against PBS at 4 °C overnight and stored at −80 °C. rAAV titres were determined by measuring the copy number of the viral genomes in pooled dialysed gradient fractions. This was achieved by a competitive PCR procedure [18] using primers and competitors mapping in the CMV promoter region common to all of the vectors. The purified viral preparations used in the present study had particle titres of approx. 1×1012 viral genomes/ml. The infectious titre in this preparation is usually 100 to 1000 times less, depending on the type of viral preparation and the target tissue. Each animal received the same dose of rAAV as stated in the experimental protocol.

Animals

All procedures complied with the standards for care and use of animal subjects as stated in the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, National Academy of Sciences, Bethesda, MD, U.S.A.). Female genetically diabetic C57BL/KsJ Lepdb mice (referred to as db+/db+ mice) and their normoglycaemic C57BL littermates (referred to as db+/+m mice) were produced by the Jackson Laboratories and obtained from Charles River Laboratories. Animals were 14 weeks old at the start of the experiments. Diabetic mice were obese and severely insulin resistant, as shown by hyperglycaemia (plasma glucose, 560±70 mg/dl in diabetic mice compared with 143±8 mg/dl in normoglycaemic mice; P<0.05) and hyperinsulinaemia (plasma insulin, 14.2±0.48 ng/ml in diabetic mice compared with 1.43±0.05 ng/ml in normoglycaemic mice; P<0.05). Diabetic mice weighed 30–32 g, whereas non-diabetic littermates weighed 22–24 g. The animals were housed individually, maintained under controlled environmental conditions (12-h light/dark cycle at approx. 23 °C), and provided with standard laboratory food and water ad libitum.

Experimental protocol

After general anaesthesia with sodium thiopental (80 mg/kg of body weight, intraperitoneally), the hair on the back was shaved and the skin was washed with povidone/iodine solution and wiped with sterile water. Two full-thickness longitudinal incisions (4 cm) were made on the dorsum of the mice, and the wound edges were closed with 4-0 silk surgical suture placed at 1 cm intervals. A total of ten animals from each group were killed after 7 and 14 days respectively, and the wounds were divided into three segments (0.8 cm wide and 1.2 cm long). The caudal and cranial strips were used for molecular analysis and wound-breaking strength, whereas the central one was used for histological and immunohistochemical evaluations. The animals were randomized to receive a single dose of either rAAV–Ang-1 or rAAV–LacZ (approx. 100 μl; 1×1011 viral particles). Both rAAVs were administered in the wound edges immediately after incision through eight intra-dermal injections with a total volume of approx. 100 μl, thus infecting 100% of the wound area.

RT-PCR for Ang-1 and VEGF mRNA

Total RNA was extracted from skin using a commercial kit (TRIzol® reagent; Invitrogen), according to the manufacturer's instruction, and was quantified spectrophotometrically (Biophotometer; Eppendorf). Total RNA was then treated with DNaseI to digest residual DNA contamination and, subsequently, 5 μg of total RNA was reverse-transcribed using a High Capacity cDNA Archive kit (Applied Biosystems) and random primers, according to manufacturer's instructions. A portion (10 ng) of total cDNA was used to quantify the amount of Ang-1 and VEGF cDNA by RT-PCR, as well as β-actin cDNA as an endogenous control. Both reactions were carried out in the same microwell (biplex) with TaqMan Universal PCR master mix and Assays-on-Demand™ ready-to-use primers and probes with different reporter dyes (Applied Biosystems).

The progression of PCRs was monitored on a 7300 Real Time-PCR System (Applied Biosystems), and the relative quantification was determined using a standard curve method for both target and endogenous reference. After normalization, the median Ang-1 and VEGF levels in normal skin from control mice was considered as the calibrator (1× sample), and the results were expressed as an n-fold difference relative to normal controls (relative expression levels).

The skin from diabetic and non-diabetic mice not subjected to rAAV injection were also tested.

Determination of Ang-1, VEGFR-1, VEGFR-2 and eNOS protein levels by Western blot analysis

Briefly, skin samples were homogenized in 1 ml of lysis buffer A [20 mmol/l Hepes (pH 7.6), 50 mmol/l β-glycerophosphate, 1.0 mmol/l dithiothreitol, 2.0 mmol/l EGTA, 1% Triton and 10% (w/v) glycerol] containing protease and phosphatase inhibitors (2 mmol/l PMSF, 1 mg/ml aprotinin, 2 mg/ml pepstatin A, 5 mg/ml leupeptin and 100 mmol/l sodium vanadate; Sigma) using an Ultra Turrax (IKA) homogenizer. The homogenate was centrifuged at 15000 g for 15 min at 4 °C. The supernatant was collected and used for protein determination using the Bio-Rad Laboratories protein assay kit.

Protein samples (30 μg) were denatured in Laemmli sample buffer (Bio-Rad Laboratories) containing 5% (v/v) 2-mercaptoethanol and were separated by SDS/PAGE on a 12% (w/v) polyacrylamide gel. The separated proteins were transferred on to a nitrocellulose membrane in transfer buffer [39 mmol/l glycine, 48 mmol/l Tris (pH 8.3) and 20% (v/v) methanol] at 200 mA for 1 h. The membranes were stained with Ponceau S (0.005% in 1% acetic acid) to confirm equal amounts of protein and were blocked with 5% (w/v) non-fat dried milk in TBS (Tris-buffered saline)/0.1% Tween for 1 h at room temperature (25±2 °C). After incubation, the membranes were washed three times for 10 min each in TBS/0.1% Tween and incubated with a primary antibody against Ang-1, VEGFR-1, VEGFR-2 or the constitutive and phosphorylated (Ser1177) forms of eNOS (Chemicon and Cell Signaling) in TBS/0.1% Tween overnight at 4 °C. After washing three times for 10 min each in TBS/0.1% Tween, the membranes were incubated with a specific HRP (horseradish peroxidase)-conjugated secondary antibody (Pierce) for 1 h at room temperature. After washing, the antibody complexes were determined by ECL® (Amersham Biosciences), according to the manufacturer's protocol. The protein signal was quantified by scanning densitometry using a bioimage analysis system (Bio-Profil; Celbio). The results from each experimental group are expressed as relative integrated intensity compared with normal skin measured within the same batch. Equal loading of protein was assessed on stripped blots by immunodetection of β-actin with a rabbit monoclonal antibody (Cell Signaling) diluted 1:500 and HRP-conjugated goat anti-(rabbit IgG) (Pierce) diluted 1:15000. All antibodies were purified by Protein A and peptide affinity chromatography.

Determination of VEGF protein in wounds

The amount of VEGF in wounds was determined by ELISA. Briefly, tissues were homogenized in 1.0 ml of PBS containing complete protease inhibitor cocktail (Boehring Mannhein). Homogenates were centrifuged to remove debris and were filtered through a 1.2-μm pore syringe filter. Analysis was performed with a commercially available VEGF-specific ELISA kit (R&D Systems). The amount of VEGF was expressed as pg/wound.

Determination of NO2/NO3 (nitrate/nitrite)

Wound samples were frozen in liquid nitrogen until use. NO products were determined in wound lysates using the Griess reaction. Skin wounds samples were homogenized in 2× lysis buffer B [1× lysis buffer B: 1% (w/v) Triton X-100, 20 mmol/l Tris/HCl (pH 8.0), 137 mmol/l NaCl, 10% (w/v) glycerol, 5 mmol/l EDTA, 1 mM PMSF and 15 μg/ml leupeptin]. The tissue extract was cleared by centrifugation, and the supernatant was diluted 1:1 with water. Lysate (1 ml) was cleared by centrifugation at 10000 g for 30 min. Cleared wound lysates (200 μl) were incubated at 4 °C and mixed with 20 μl of sulfanilamide (dissolved in 1.2 mol/l HCl) and 20 μl N-naphtylethylendiamine dihydrochloride (Sigma). After 5 min at room temperature, A540 was measured with a reference wavelength at A690. Combined NO2/NO3 levels were determined by reducing NO2 to NO3 with nitrate reductase in the presence of NADPH. For this, 80 μl of sample was incubated with 10 μl of Aspergillus nitrate reductase (1 unit/ml; Sigma) and 10 μl of 1 mmol/l NADPH for 1 h at 27 °C before the addition of the Griess reagent. All samples and standards were assayed in triplicate. Results are expressed as means±S.D. of NO2/NO3.

Histology

All tissue specimens were fixed in 10% neutral buffered formalin for at least 24 h at room temperature. After fixation, perpendicular sections to the anterior posterior axis of the wounds were dehydrated in graded ethanol, cleared in xylene and embedded in paraffin. Section (5 μm thick) were mounted on glass slides, dewaxed, rehydrated with distilled water and stained with haematoxylin/eosin or with Masson's trichrome, according to routine procedures for light microscopy. For the amount of granulation tissue, the areas proximal to the incision were evaluated in all of the sections on each slide by three pathologists blinded to the experimental protocol. As part of the histological evaluation, all slides were examined by means of an eyepiece grid under the microscope from 5× to 100× magnification. The following parameters were evaluated and scored: (i) re-epithelialization, (ii) dermal matrix deposition and regeneration, (iii) granulation tissue formation and remodelling, and (iv) angiogenesis. The edges of the wound in each of the sections, as well as normal control wounds, were used as comparisons for scoring.

Immunohistochemistry

Paraffin-embedded tissues were sectioned (5 μm), and antigen retrieval was performed using 0.05 mol/l sodium citrate buffer (pH 6.0). Tissues were treated with a primary antibody against PECAM-1 [platelet/EC adhesion molecule-1; Santa Cruz Laboratories], VEGFR-2 (Cell Signaling) and VEGFR-1 (Abcam). The secondary antibody was provided by Pierce, and the location of the reaction was visualized in brown with 3,3′-diaminobenzidine tetrahydrochloride (Sigma). Nuclear Fast Red counterstain was used. Slides were then mounted with coverslips.

To assess the angiogenic response, MVD (microvessel density) was estimated after PECAM-1 staining. Briefly, five areas/section were randomly selected in the dermis just proximal to the wound site and the positive ECs were counted under high-power field (×40 magnification) by three pathologists blinded to the treatment of the samples. To assess positive VEGFR-1 and VEGFR-2 staining, five areas/section were randomly selected in the dermis and the positive ECs were counted under high-power field (×40 magnification) by three pathologists blinded to the treatment of the samples.

In each reaction, a positive control of human angiosarcoma was used. For negative controls, the primary antibodies were replaced by sodium citrate buffer (pH 6.0).

Breaking strength

The maximum load (breaking strength) tolerated by wounds was measured in a blinded manner on coded samples using a calibrated tensometer (Instron), as described previously [19]. The ends of the skin strip were pulled at a constant speed (20 cm/min), and breaking strength was expressed as the mean maximum level of tensile strength (g/mm) before separation of the wounds.

In vivo vascular permeability

In vivo vascular permeability was evaluated in an additional set of animals. Briefly, female diabetic mice and their normoglycaemic littermates (normoglycaemic) were treated with either rAAV coding for Ang-1 or rAAV coding for LacZ (approx. 100 μl; 1×1011 viral particles). On the day of the experiment (day 14), the mice were injected intradermally with 5 μg of compound 48/80 (Sigma) dissolved in 50 μl of physiological saline into the wound edges immediately after an intravenous injection of a 0.5% Evans Blue solution (5 ml/kg of body weight). Compound 48/80 administration is followed by increased leucocyte rolling, increased leucocyte adhesion and increased vascular permeability in postcapillary venules. Animals were killed 30 min after injection of compound 48/80 and the skin of the reaction locus was removed for a quantitative determination of the extravasated dye. Evans Blue was extracted from each piece of dissected skin with 1 ml of 6 mol/l KOH at 45 °C for 6 h. This extract was neutralized by 1 ml of 6 mol/l HCl, and then 2 ml of acetone was added. The extract was clarified by filtration, A595 was measured, and the extracted Evans Blue content was then calculated from the calibration curve.

Statistical analysis

All results are expressed as means±S.D. Before analysis, all end points were tested for normality by using the Kolmogorov–Smirnov test. The test suggested that data were normally distributed and, therefore, analysis was performed using a two-way ANOVA. In all cases, P value <0.05 was selected as the criterion for statistical significance.

RESULTS

Efficacy of Ang-1 gene transfer in wounds

Ang-1 is normally expressed in the skin during adult life and is not affected by diabetes. The efficacy of rAAV-mediated gene transfer was assessed by RT-PCR and Western blot analysis at days 7 and 14 after wounding. The unwounded skin from diabetic and non-diabetic mice not subjected to rAAV injection had levels of Ang-1 mRNA comparable with those obtained from rAAV–LacZ-injected animals (diabetic mice, 5.5±1-fold compared with β-actin; normoglycaemic mice, 5.1±0.8-fold compared with β-actin). At day 7 after injury, the expression of Ang-1 mRNA was increased slightly in the wounds of diabetic mice administered rAAV–LacZ (5.7±0.9-fold compared with β-actin), whereas Ang-1 gene transfer significantly augmented Ang-1 mRNA expression at each time point (P<0.001 compared with rAAV–LacZ) (Figure 1). Similar results were obtained in normoglycaemic mice (Figure 1).

Ang-1 mRNA levels in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 1
Ang-1 mRNA levels in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. mRNA levels were determined by RT-PCR. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

Figure 1
Ang-1 mRNA levels in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. mRNA levels were determined by RT-PCR. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

The increase in the mRNA expression of Ang-1 resulted in a significant increase (P<0.001) in protein expression at 7 and 14 days after wounding in both normoglycaemic and diabetic mice administered with rAAV–Ang-1 compared with animals treated with rAAV–LacZ, as confirmed by Western blot analysis (Figure 2).

Ang-1 protein expression in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 2
Ang-1 protein expression in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Upper panel, representative Western blot analysis. Lower panel, quantification of Ang-1 protein expression. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

Figure 2
Ang-1 protein expression in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Upper panel, representative Western blot analysis. Lower panel, quantification of Ang-1 protein expression. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

Expression of eNOS in wounds

Uninjured skin had a very low content of phospho-eNOS (1.1±0.9 and 0.5±0.08 integrated intensity in normoglycaemic and diabetic mice respectively). Wounding caused an enhanced expression of phosho-eNOS in both strains of animals at day 7 (Figure 3A). In wounds treated with rAAV–Ang-1, a greater increase in phospho-eNOS was detected in both normoglycaemic and diabetic mice compared with rAAV–LacZ-treated animals (Figure 3A).

eNOS expression (A) and NO products (B) in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 3
eNOS expression (A) and NO products (B) in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. (A) Upper panel, representative Western blot analysis of constitutive and phosphorylated (p-eNOS) forms of eNOS. Lower panel, quantification of phospho-eNOS protein expression. (B) The NO products NO2/NO3 were measured as described in the Materials and methods section. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice; §P<0.05 compared with normoglycaemic mice.

Figure 3
eNOS expression (A) and NO products (B) in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. (A) Upper panel, representative Western blot analysis of constitutive and phosphorylated (p-eNOS) forms of eNOS. Lower panel, quantification of phospho-eNOS protein expression. (B) The NO products NO2/NO3 were measured as described in the Materials and methods section. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice; §P<0.05 compared with normoglycaemic mice.

NO products in wounds

The concentration of NO2/NO3 in wound lysates from rAAV–LacZ- or rAAV–Ang-1-injected normoglycaemic and diabetic mice was determined. Wounding caused a marked increase in NO2/NO3 content compared with control uninjured skin at day 7 (3.3±0.6 and 1.1±0.08 nmol/g of tissue in normoglycaemic and diabetic mice respectively). A significantly greater increase in wound NO products was observed in rAAV–Ang-1-injected normoglycaemic and diabetic mice compared with rAAV–LacZ-treated animals (Figure 3B).

Histological assessment

Delivery of Ang-1 in the wound edge produced a significant improvement in the healing process, as shown by haematoxylin/eosin staining and Masson's trichrome staining for the assessment of epidermal regeneration and the thickness of granulation tissue (Figures 4 and 5). Complete skin normalization in untreated mice occurred at day 22±1 after wounding in normoglycaemic mice and at day 35±2 in diabetic mice. No side effect or signs of inflammation were observed in either normoglycaemic or diabetic mice injected with the viral vectors. Diabetic animals usually have a thicker epithelial layer than normal mice; in addition, the different layers are less differentiated and adipose infiltrates are present in the dermis, impairing the normal elasticity of the skin and, as a consequence, it is more prone to a delayed healing.

Representative haematoxylin/eosin staining of skin wound samples at day 14 from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 4
Representative haematoxylin/eosin staining of skin wound samples at day 14 from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. (A) Wound from an rAAV-LacZ-treated normoglycaemic mouse showing that re-epithelialization is almost complete, and granulation tissue is well formed without inflammatory infiltrates. (B) Wound from an rAAV-Ang-1-treated normoglycaemic mouse showing complete re-epithelialization with the presence of hair and the restoration of normal architecture in dermis. (C) Wound from an rAAV–LacZ-treated diabetic mouse showing a thin and immature epithelial layer, with the presence of crusting and poorly formed granulation tissue. (D) Wound from an rAAV–Ang-1-treated diabetic mouse showing that re-epithelialization is almost complete, with the absence of spongiosis or crusting, the presence of well-organized granulation tissue and no evidence of inflammation, oedema or erythrocyte extravasation.

Figure 4
Representative haematoxylin/eosin staining of skin wound samples at day 14 from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. (A) Wound from an rAAV-LacZ-treated normoglycaemic mouse showing that re-epithelialization is almost complete, and granulation tissue is well formed without inflammatory infiltrates. (B) Wound from an rAAV-Ang-1-treated normoglycaemic mouse showing complete re-epithelialization with the presence of hair and the restoration of normal architecture in dermis. (C) Wound from an rAAV–LacZ-treated diabetic mouse showing a thin and immature epithelial layer, with the presence of crusting and poorly formed granulation tissue. (D) Wound from an rAAV–Ang-1-treated diabetic mouse showing that re-epithelialization is almost complete, with the absence of spongiosis or crusting, the presence of well-organized granulation tissue and no evidence of inflammation, oedema or erythrocyte extravasation.

Representative Masson's trichrome staining of skin wound samples at day 14 (upper panel) and quantification of collagen bundles (lower panel) from diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 5
Representative Masson's trichrome staining of skin wound samples at day 14 (upper panel) and quantification of collagen bundles (lower panel) from diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Values are means±S.D. of seven experiments. Double-headed arrows indicate the thickness of the collagen layer. *P<0.001 compared with rAAV–LacZ-treated mice. Magnification, ×10.

Figure 5
Representative Masson's trichrome staining of skin wound samples at day 14 (upper panel) and quantification of collagen bundles (lower panel) from diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Values are means±S.D. of seven experiments. Double-headed arrows indicate the thickness of the collagen layer. *P<0.001 compared with rAAV–LacZ-treated mice. Magnification, ×10.

At day 14, wounds of diabetic mice administered with rAAV–LacZ still had incomplete re-epithelialization, the presence of a crust, very low organized granulation tissue and marked adipose substitution in the dermis. By contrast, at the same time point, wounds from rAAV–Ang-1-treated diabetic mice had complete re-epithelialization, well-organized dermal layers, well-formed granulation tissue, reduced adipose substitution and mature collagen bundles (Figures 4 and 5).

Normoglycaemic mice injected with rAAV–LacZ had almost complete re-epithelialization at day 14; the administration of rAAV–Ang-1 resulted in an improved re-epithelialization process, as demonstrated by the presence of hair follicles in the dermis (Figures 4 and 5).

Evaluation of new blood vessel formation

PECAM-1 immunostaining was investigated at 14 days after surgical procedures to confirm neovessel formation (Figure 6). In fact, this protein represents a highly specific marker for ECs. As shown in Figure 6 (upper panels), positive staining in the granulation tissue of rAAV–Ang-1-injected diabetic mice was mostly expressed in small vessels and capillaries, whereas, in those mice receiving rAAV–LacZ, reactivity was found only in pre-existing vessels, demonstrating an augmented angiogenesis mediated by Ang-1 administration.

Representative immunohistochemical PECAM-1 staining of skin wounds samples (upper panel) and MVD quantification (lower panel) at day 14 from diabetic mice (db+/db+) treated with either rAAV–LacZ or rAAV–Ang-1

Figure 6
Representative immunohistochemical PECAM-1 staining of skin wounds samples (upper panel) and MVD quantification (lower panel) at day 14 from diabetic mice (db+/db+) treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Arrows indicate positive staining in endothelial cells. Magnification, ×40. Values are means±S.D. of seven experiments. *P<0.05 compared with rAAV–LacZ-treated mice.

Figure 6
Representative immunohistochemical PECAM-1 staining of skin wounds samples (upper panel) and MVD quantification (lower panel) at day 14 from diabetic mice (db+/db+) treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Arrows indicate positive staining in endothelial cells. Magnification, ×40. Values are means±S.D. of seven experiments. *P<0.05 compared with rAAV–LacZ-treated mice.

Using the images obtained from PECAM-1 staining, the tissue samples from each group were assessed for microvessel density (Figure 6, lower panel).

Breaking strength

Ang-1 gene transfer improved the tensile strength of the injured tissue at day 14 (Figure 7). The breaking strength of incisional wounds in diabetic mice treated with rAAV–Ang-1 was significantly higher compared with those of diabetic mice treated with rAAV–LacZ. Similar results were observed in wounds obtained from normoglycaemic mice (Figure 7).

Tensile strength of wounds obtained at day 14 from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 7
Tensile strength of wounds obtained at day 14 from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

Figure 7
Tensile strength of wounds obtained at day 14 from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

VEGF mRNA and protein expression

To understand the basis of the angiogenic process, the expression of VEGF in wounds was investigated by RT-PCR at days 7 and 14 after injury. In fact, diabetic animals have a distorted VEGF pattern, which is responsible for an inadequate angiogenetic process. In the wounds of diabetic mice administered with rAAV–LacZ, VEGF mRNA was markedly lower than in normoglycaemic mice at each time point (P<0.001 compared with normoglycaemic mice). Treatment with rAAV–Ang-1 did not modify the disrupted VEGF mRNA expression in diabetic mice at day 7 and 14 (P value was not significant compared with rAAV–LacZ) (Figure 8A).

VEGF mRNA levels evaluated by RT-PCR (A) and VEGF protein studied by ELISA (B) in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 8
VEGF mRNA levels evaluated by RT-PCR (A) and VEGF protein studied by ELISA (B) in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice at day 14, or diabetic mice on day 7.

Figure 8
VEGF mRNA levels evaluated by RT-PCR (A) and VEGF protein studied by ELISA (B) in skin wound samples from normoglycaemic (db+/+m) and diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice at day 14, or diabetic mice on day 7.

rAAV–LacZ-treated normoglycaemic mice had a typical marked induction of VEGF mRNA at day 7, which returned to normal values at 14 days after injury, and this expression was not affected by Ang-1 gene transfer (P value was not significant compared with rAAV–LacZ), suggesting that Ang-1 was not able to induce VEGF expression (Figure 8A).

To confirm this result, VEGF protein expression was investigated by ELISA and an essentially identical pattern was obtained (Figure 8B). Administration of rAAV–Ang-1 did not modify the disrupted VEGF protein production in diabetic mice at days 7 and 14 (P value was not significant compared with rAAV–LacZ) (Figure 8B). In addition, rAAV–LacZ-treated normoglycaemic mice had a typical marked increase in VEGF protein expression at day 7, which declined to normal values at day 14 after injury, and this expression was not affected by Ang-1 gene transfer (P value was not significant compared with rAAV–LacZ), confirming that Ang-1 was not able to induce VEGF expression (Figure 8B).

Assessment of VEGFR-2 immunostaining and protein expression

At day 7, immunostaining was performed for VEGFR-2, a marker that identifies endothelial precursor cells [10,20]. These cells are not only circulating and bone-marrow-derived, but they are also present in subcutaneous adipose tissue. rAAV–Ang-1 gene transfer augmented VEGFR-2 expression in ECs in diabetic mice (Figure 9), thus contributing to the healing process. Similar results were obtained in normoglycaemic mice (results not shown). In contrast, a very slight staining for VEGFR-1, which was mostly expressed by mature ECs, was observed in both strains of mice administered with rAAV–Ang-1, suggesting further that VEGF is not involved in Ang-1-driven angiogenesis. However both Ang-1 and VEGF are able to recruit endothelial precursors [10,20].

Representative immunohistochemical VEGFR-2 and VEGFR-1 staining of skin wound samples at day 7 from diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

Figure 9
Representative immunohistochemical VEGFR-2 and VEGFR-1 staining of skin wound samples at day 7 from diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

VEGFR-2 staining of a skin wound sample from diabetic mice treated with (A) rAAV–LacZ or (B) rAAV–Ang-1. (C) VEGFR-1 staining of a skin wound sample from diabetic mice treated with rAAV–Ang-1. Arrows indicate positive staining in endothelial cells. Magnification, ×40. (D) Quantification of VEGFR-positive ECs per field in diabetic mice treated with either rAAV–LacZ or rAAV–Ang-1. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

Figure 9
Representative immunohistochemical VEGFR-2 and VEGFR-1 staining of skin wound samples at day 7 from diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1

VEGFR-2 staining of a skin wound sample from diabetic mice treated with (A) rAAV–LacZ or (B) rAAV–Ang-1. (C) VEGFR-1 staining of a skin wound sample from diabetic mice treated with rAAV–Ang-1. Arrows indicate positive staining in endothelial cells. Magnification, ×40. (D) Quantification of VEGFR-positive ECs per field in diabetic mice treated with either rAAV–LacZ or rAAV–Ang-1. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice.

To confirm this, VEGR protein expression was quantified in skin wounds by Western blot analysis. A low expression of VEGFR-1 protein was present in both diabetic (Figure 10A) and normoglycaemic (results not shown) mice injected with rAAV–Ang-1. By contrast, rAAV–Ang-1 enhanced VEGFR-2 protein expression in both diabetic (Figure 10A) and normoglycaemic (results not shown) mice.

VEGFR expression in diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1, and in vivo vascular permeability

Figure 10
VEGFR expression in diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1, and in vivo vascular permeability

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. (A) VEGFR expression at day 7 in skin wound samples from diabetic mice treated with either rAAV–LacZ or rAAV–Ang-1. Values are means±S.D. of seven experiments. *P<0.05 compared with rAAV–LacZ-treated mice. (B) In vivo permeability in normoglycaemic (db+/+m) and diabetic mice. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice, or diabetic mice.

Figure 10
VEGFR expression in diabetic (db+/db+) mice treated with either rAAV–LacZ or rAAV–Ang-1, and in vivo vascular permeability

Mice were treated with approx. 100 μl of rAAV, equivalent to 1×1011 virus particles. (A) VEGFR expression at day 7 in skin wound samples from diabetic mice treated with either rAAV–LacZ or rAAV–Ang-1. Values are means±S.D. of seven experiments. *P<0.05 compared with rAAV–LacZ-treated mice. (B) In vivo permeability in normoglycaemic (db+/+m) and diabetic mice. Values are means±S.D. of seven experiments. *P<0.001 compared with rAAV–LacZ-treated mice, or diabetic mice.

In vivo vascular permeability

At day 14, in vivo vascular permeability was evaluated by administering compound 48/80, which causes Evans Blue leakage. As shown in Figure 10(B), intradermal stimulation of 5 μg of compound 48/80 caused a significantly leakage of Evans Blue that had been intravenously injected before compound 48/80. No significant differences were observed between the groups treated with either rAAV–LacZ or rAAV–Ang-1.

DISCUSSION

Blood vessels form through two distinct processes: vasculogenesis and angiogenesis [21]. In vasculogenesis, ECs differentiate de novo from mesodermal precursors, whereas in angiogenesis new vessels are generated from pre-existing ones. Angiogenesis and neovascularization can be unwanted processes in certain diseases, including cancer and diabetic retinopathy. However, formation of the new blood vessels can also help alleviate some disease states, as in the formation of collateral circulation in ischaemic myocardium and limbs as well as during wound healing.

VEGF and Ang-1 have potent effects on angiogenesis and vasculogenesis respectively. VEGF-specific activities include EC survival, proliferation, migration and tube formation, whereas Ang-1 has no proliferative effect, but it is a potent inducer of vasculogenesis, recruiting and differentiating mesodermal resident precursors as well as bone-marrow-derived EPCs. These cells have the surface marker VEGFR-2 (or flk-1) specific for committed as well as young ECs and, driven by growth factors, migrate from the lateral plate mesoderm towards the dorsal aorta to the site of neovasculogenesis [2023].

In the present study, the rAAV-mediated gene transfer of Ang-1 improved the altered healing process, enhanced the impaired angiogenesis and increased wound- breaking strength in diabetic mice; however, Ang-1 gene transfer did not modify the disrupted pattern of VEGF expression. In contrast, Ang-1 stimulated the expression of VEGFR-2 in new vessels close to the wound site in diabetic mice.

In diabetes, the need for neovascularization arises from inadequate VEGF production and release in wounds; thus the classic angiogenetic process is extremely delayed and poor and, as a consequence, therapeutically valuable approaches have been investigated with the aim of stimulating the expression of the impaired angiopoietic factors [24]. However, the results of the present study clearly suggest that active angiogenesis may also occur in diabetes without the involvement and intervention of VEGF. Under these specific circumstances, vasculogenesis may represent an alternative pathway to induce new blood vessel formation. More specifically, our present findings strongly suggest that Ang-1 causes, at least in diabetes, angiogenesis in a VEGF-independent manner, possibly through the EPCs. The enhanced PECAM-1 staining in ECs at the wound site confirms active angiogenesis. It could be hypothesized that the slight positive staining for VEGFR-1 is a consequence of the low, but still present, VEGF protein expression in the skin of diabetic animals after surgical injury. On the other hand, the marked staining for VEGFR-2 after Ang-1 administration could be related to the presence of EPCs that usually present this marker [2,10].

To date, it is well established that NO substantially regulates wound matrix deposition [25], as inhibition of wound NO synthesis was paralleled by decreased wound collagen accumulation and wound-breaking strength. Interestingly, we found in the present study that active eNOS protein expression was low at the wound site in diabetes-impaired healing in db/db mice. This observation clearly suggests that, in the diabetic wound-healing process, there is also a defect in the NO pathway which might also contribute to impairing the healing process. Ang-1 gene transfer succeeded in reverting the impairment in NO. In fact, rAAV–Ang-1 increased both eNOS expression and NO production. Therefore the activation of this NO pathway might represent an additional mechanism of action of Ang-1, with Ang-1 and eNOS acting in concert to sustain an effective improvement in new vessels formation.

In conclusion, the results of the present study are extremely relevant in the management of patients with diabetes who have delayed wound healing. In fact, our present results provide strong evidence that Ang-1 gene transfer improves the delayed wound repair in diabetes by stimulating angiogenesis, apparently without VEGF involvement, that could be more detrimental than useful as it happens during diabetic retinopathy. Furthermore, unpublished observations from our laboratory suggest that VEGF gene transfer does not affect Ang-1 expression and the combined delivery of VEGF and Ang-1 does not cause a greater effect than the delivery of one of the two angiogenic factors (A. Bitto, F. Polito and F. Squadrito, unpublished work). Whether this finding might have important clinical implications remains to be investigated fully and deserves further pre-clinical and clinical investigation.

Abbreviations

     
  • AAV

    adeno-associated virus

  •  
  • Ang

    angiopoietin

  •  
  • CMV

    cytomegalovirus

  •  
  • COMP

    cartilage oligomeric matrix protein

  •  
  • EC

    endothelial cell

  •  
  • eNOS

    endothelial NO synthase

  •  
  • EPC

    endothelial progenitor cell

  •  
  • GFP

    green fluorescent protein

  •  
  • HEK-293 cells

    human embryonic kidney cells

  •  
  • HRP

    horseradish peroxidase

  •  
  • MVD

    microvessel density

  •  
  • NO2/NO3

    nitrate/nitrite

  •  
  • PECAM-1

    platelet/EC adhesion molecule-1

  •  
  • PlGF

    placental growth factor

  •  
  • rAAV

    recombinant AAV

  •  
  • RT-PCR

    real-time PCR

  •  
  • TBS

    Tris-buffered saline

  •  
  • VEGF

    vascular endothelial growth factor

  •  
  • VEGFR

    VEGF receptor

References

References
1
Conway
 
E. M.
Collen
 
D.
Carmeliet
 
P.
 
Molecular mechanisms of blood vessel growth
Cardiovasc. Res.
2001
, vol. 
49
 (pg. 
507
-
521
)
2
Ferrara
 
N.
Gerber
 
H. P.
LeCouter
 
J.
 
The biology of VEGF and its receptors
Nat. Med.
2003
, vol. 
9
 (pg. 
669
-
676
)
3
Tammela
 
T.
Enholm
 
B.
Alitalo
 
K.
Paavonen
 
K.
 
The biology of vascular endothelial growth factors
Cardiovasc. Res.
2005
, vol. 
65
 (pg. 
550
-
563
)
4
Goodson
 
W. H.
Hunt
 
T. K.
 
Studies of wound healing in experimental diabetes
J. Surg. Res.
1977
, vol. 
22
 (pg. 
221
-
227
)
5
Bohlen
 
H. G.
Niggl
 
B. A.
 
Adult microvascular disturbances as a result of juvenile-onset diabetes in db/db mice
Blood Vessels
1979
, vol. 
16
 (pg. 
269
-
276
)
6
Altavilla
 
D.
Saitta
 
A.
Cucinotta
 
D.
, et al 
Inhibition of lipid peroxidation restores impaired vascular endothelial growth factor expression and stimulates wound healing and angiogenesis in the genetically diabetic mouse
Diabetes
2001
, vol. 
50
 (pg. 
667
-
674
)
7
Galeano
 
M.
Deodato
 
B.
Altavilla
 
D.
, et al 
Adeno-associated viral vector-mediated human vascular endothelial growth factor gene transfer stimulates angiogenesis and wound healing in the genetically diabetic mouse
Diabetologia
2003
, vol. 
46
 (pg. 
546
-
555
)
8
Peichev
 
M.
Naiyer
 
A. J.
Pereira
 
D.
, et al 
Expression of VEGFR-2 and AC133 by circulating human CD34+ cells identifies a population of functional endothelial precursors
Blood
2000
, vol. 
95
 (pg. 
952
-
958
)
9
Asahara
 
T.
Murohara
 
T.
Sullivan
 
A.
, et al 
Isolation of putative progenitor endothelial cells for angiogenesis
Science
1997
, vol. 
275
 (pg. 
964
-
967
)
10
Urbich
 
C.
Dimmeler
 
S.
 
Endothelial progenitor cells functional characterization
Trends Cardiovasc. Med.
2004
, vol. 
14
 (pg. 
318
-
322
)
11
Wetzler
 
C.
Kampfer
 
H.
Stallmeyer
 
B.
Pfeilschifter
 
J.
Frank
 
S.
 
Large and sustained induction of chemokines during impaired wound healing in the genetically diabetic mouse: prolonged persistence of neutrophils and macrophages during the late phase of repair
J. Invest. Dermatol.
2000
, vol. 
115
 (pg. 
245
-
253
)
12
Holash
 
J.
Maisonpierre
 
P. C.
Compton
 
D.
, et al 
Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF
Science
1999
, vol. 
284
 (pg. 
1994
-
1998
)
13
Babaei
 
S.
Teichert-Kuliszewska
 
K.
Zhang
 
Q.
Jones
 
N.
Dumont
 
D. J.
Stewart
 
D. J.
 
Angiogenic actions of angiopoietin-1 require endothelium-derived nitric oxide
Am. J. Pathol.
2003
, vol. 
162
 (pg. 
1927
-
1936
)
14
Stadelmann
 
W. K.
Digenis
 
A. G.
Tobin
 
G. R.
 
Impediments to wound healing
Am. J. Surg.
1998
, vol. 
176
 (pg. 
39S
-
47S
)
15
Cho
 
C. H.
Sung
 
H. K.
Kim
 
K. T.
, et al 
COMP-angiopoietin-1 promotes wound healing through enhanced angiogenesis, lymphangiogenesis, and blood flow in a diabetic mouse model
Proc. Natl. Acad. Sci. U.S.A.
2006
, vol. 
103
 (pg. 
4946
-
4951
)
16
Zolotukhin
 
S.
Putter
 
M.
Hauswirth
 
W. W.
Guy
 
J.
Muzyczka
 
N.
 
A ‘humanized’ green fluorescent protein cDNA adapted for high-level expression in mammalian cells
J. Virol.
1996
, vol. 
70
 (pg. 
4646
-
4654
)
17
Grimm
 
D.
Kern
 
A.
Rittner
 
K.
Kleinschmidt
 
J. A.
 
Novel tools for production and purification of recombinant adenoassociated virus vectors
Hum. Gene Ther.
1998
, vol. 
9
 (pg. 
2745
-
2760
)
18
Diviacco
 
S.
Norio
 
P.
Zentilin
 
L.
, et al 
A novel procedure for quantitative polymerase chain reaction by coamplification of competitive templates
Gene
1992
, vol. 
122
 (pg. 
313
-
320
)
19
Pierce
 
G. F.
Mustoe
 
T. A.
Senior
 
R. M.
, et al 
In vivo incisional wound healing augmented by platelet-derived growth factor and recombinant c-cis gene homodimeric proteins
J. Exp. Med.
1988
, vol. 
167
 (pg. 
974
-
987
)
20
Asahara
 
T.
Masuda
 
H.
Takahashi
 
T.
 
Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization
Circ. Res.
1999
, vol. 
85
 (pg. 
221
-
228
)
21
Patan
 
S.
 
Vasculogenesis and angiogenesis
Cancer Treat. Res.
2004
, vol. 
117
 (pg. 
3
-
32
)
22
Visconti
 
R. P.
Richardson
 
C. D.
Sato
 
T. N.
 
Orchestration of angiogenesis and arteriovenous contribution by angiopoietins and vascular endothelial growth factor (VEGF)
Proc. Natl. Acad. Sci. U.S.A.
2002
, vol. 
99
 (pg. 
8219
-
8224
)
23
Shim
 
W. S.
Li
 
W.
Zhang
 
L.
, et al 
Angiopoietin-1 promotes functional neovascularization that relieves ischemia by improving regional reperfusion in a swine chronic myocardial ischemia model
J. Biomed. Sci.
2006
, vol. 
13
 (pg. 
579
-
591
)
24
Pandya
 
N. M.
Dhalla
 
N. S.
Santani
 
D. D.
 
Angiogenesis – a new target for future therapy
Vasc. Pharmacol.
2006
, vol. 
44
 (pg. 
265
-
274
)
25
Schaffer
 
M. R.
Tantry
 
U.
Efron
 
P. A.
Ahrendt
 
G. M.
Thornton
 
F. J.
Barbul
 
A.
 
Diabetes-impaired healing and reduced wound nitric oxide synthesis: a possible pathophysiologic correlation
Surgery
1997
, vol. 
121
 (pg. 
513
-
519
)