Topical Sonic Hedgehog Gene Therapy Accelerates Wound Healing in Diabetes by Enhancing Endothelial Progenitor Cell–Mediated Microvascular Remodeling
Background— Sonic hedgehog (Shh) is a prototypical morphogen known to regulate epithelial-mesenchymal interaction during embryonic development. Recent observations indicate that exogenous administration of Shh can induce angiogenesis and may accelerate repair of ischemic myocardium and skeletal muscle. Because angiogenesis plays a pivotal role in wound repair, we hypothesized that activation of the hedgehog pathway may promote a favorable effect on microvascular remodeling during cutaneous wound healing and thereby accelerate wound closure. Because diabetes is associated with impaired wound healing, we tested this hypothesis in a diabetic model of cutaneous wound repair.
Methods and Results— In Ptc1-LacZ mice, cutaneous injury resulted in LacZ expression, indicating that expression of the Shh receptor Patched was induced and therefore that the Shh signaling pathway was intact postnatally and upregulated in the process of wound repair. In diabetic mice, topical gene therapy with the use of naked DNA encoding for Shh resulted in significant local gene expression and acceleration of wound recovery. The acceleration in wound healing was notable for increased wound vascularity. In bone marrow transplantation models, the enhanced vascularity of the wound was shown to be mediated, at least in part, by enhanced recruitment of bone marrow–derived endothelial progenitor cells. In vitro, Shh promoted production of angiogenic cytokines from fibroblasts as well as proliferation of dermal fibroblasts. Furthermore, Shh directly promoted endothelial progenitor cell proliferation, migration, adhesion, and tube formation.
Conclusions— These findings suggest that a simple strategy of topically applied Shh gene therapy may have significant therapeutic potential for enhanced wound healing in patients with impaired microcirculation such as occurs in diabetes. (Circulation. 2006;113:2413-2424.)
Received June 23, 2005; de novo received November 21, 2005; revision received March 7, 2006; accepted March 10, 2006.
Nonhealing skin ulcers are one of the most serious consequences of diabetes, resulting in more hospitalizations than any other diabetic complication and representing a major contributing factor to amputation in diabetics.1 A number of factors have been implicated in the predisposition to nonhealing wounds observed in diabetes, including macrovascular and microvascular disease, peripheral neuropathy, and impaired angiogenesis of wounds. Of these factors, angiogenesis is considered to play a pivotal pathophysiological role by virtue of its requirement for successful wound repair.
Clinical Perspective p 2424
Our previous studies have shown that sonic hedgehog (Shh) induces neovascularization in part by upregulation of multiple families of angiogenic growth factors, such as vascular endothelial growth factor (VEGF) and angiopoietins.2 Very recently, we also demonstrated that the innate hedgehog (Hh) pathway is activated after myocardial ischemia and that Shh gene therapy may have therapeutic potential in this setting.3
The aim of the present study was to determine whether the endogenous Shh pathway is physiologically involved in cutaneous wound healing in adults and to investigate the therapeutic potential of augmenting Shh signaling with the use of a simple, topical gene therapy to enhance diabetic wound vascularization and therefore accelerate wound repair. We evaluated topical Shh gene therapy in the leptin receptor–deficient diabetic mouse, which is an established model of deficient wound healing associated with diabetes.4 Furthermore, we also evaluated the role of local Shh gene therapy on the contribution of bone marrow (BM)–derived endothelial progenitor cells (EPCs) using bone marrow transplanted (BMT) mice.
These studies reveal that topical Shh gene therapy can accelerate wound healing, promoting neovascularization by enhanced recruitment of BM-derived progenitors, thereby suggesting the potential for a practical, simple gene therapy to mitigate one of the most serious complications of diabetes.
Human Shh Plasmid
For construction of the human Shh plasmid (phShh), the amino-terminal domain of human Shh coding sequence was selected with the use of pCMV-ScriptPCR mammalian expression vector (Stratagene, La Jolla, Calif). PhShh is a 4878-bp plasmid that contains the 600-bp amino-terminal domain coding sequence of human Shh. Expression of Shh gene is modulated by the presence of promoter sequences from cytomegalovirus to drive Shh expression. Downstream from the Shh cDNA is an SV40 polyadenylation sequence. The plasmid contains a gene that confers neomycin/kanamycin resistance to the host cells.
C57BLKS/J-m+/+Leprdb mice (db/db mice), C57BLKS/J (wild-type of db/db mice), green fluorescent protein (GFP) transgenic (Tg) mice, and C57BL/6 (wild-type of GFP Tg mice) were obtained from The Jackson Laboratory (Bar Harbor, Me). Nuclear localization signal (NLS)-Ptc1-LacZ mice or their wild-type littermates were kindly provided by Dr M.P. Scott (Stanford University, Stanford, Calif). We also used BMT mice created by transplantation of BM from GFP Tg mice. BMT mice were prepared as previously described with minor modification.5,6 BM cells were collected from femurs and tibias of donor GFP Tg or wild-type B6 mice by aspiration and flushing. Recipient mice were lethally irradiated with 12.0 Gy, and BMT from the Tg mice was performed. At 4 weeks after BMT, by which time the BM of recipient mice was reconstituted with the transplanted BM, skin wounds were made on the recipient mice, as described below. All procedures were performed in accordance with the Institutional Animal Care and Use Committee of St Elizabeth’s Medical Center.
Preparation of DNA/Methylcellulose Pellets
DNA/methylcellulose pellets were prepared, as described previously.7 Briefly, 100 μg of phShh or LacZ plasmid was diluted in double-distilled H2O (20 μL). Plasmid containing double-distilled H2O or double-distilled H2O alone was mixed with an equal volume of 1% methylcellulose prepared in double-distilled H2O. Then this solution was spotted onto bacterial plates and allowed to dry at room temperature for 2 hours. The dehydrated pellets were removed intact from the plates with forceps. Immediately after wounding, the dehydrated pellet containing plasmid was applied to the wound.
Creation of Wounds and Topical Application of PhShh
All mice were between 8 and 12 weeks of age at time of wounding. Mice were placed in individual cages and subjected to wounding. Wounding was performed as described previously.8 After induction of deep anesthesia by injection of sodium pentobarbital (160 mg/kg IP), full-thickness excisional skin wounds with the use of 8-mm skin biopsy punches were made on the backs of mice. Immediately after wounding, the methylcellulose pellet containing phShh or LacZ plasmid was applied to the wound. Then the wound was covered with the semipermeable polyurethane dressing OpSite (Smith & Nephew, London, United Kingdom). Opsite, skin, and muscle surrounding the wound were sutured together with 6-0 Prolene to prevent the mouse from taking off the OpSite. For inhibitory study, 10 μg of anti-VEGF antibody (R&D Systems, Minneapolis, Minn) was topically injected to the wound through Opsite twice a week to neutralize VEGF bioactivity.
Analysis of Wound Closure
A total of 5 db/db mice were used at each time point. Wound closure was documented with a digital camera (Nikon Coolpix 995, Nikon, Tokyo, Japan) on days 0, 5, 10, and 14. Images were analyzed with the NIH Image J analyzer by tracing the wound margin with a fine-resolution computer mouse and calculating pixel area. Wound closure was reported as the percentage closed and calculated as Percentage Closed=[(Area on Day 0−Open Area on Final Day)/Area on Day 0]×100, as described previously.7 The areas of the wounds were compared with a 1-way ANOVA test to determine statistical significance.
Wounds were harvested 10 days after application of either control (LacZ) or phShh. Histological scores were assigned in a blinded manner according to the method previously described.8
Fluorescence Microscopic Evaluation of Wound Vascularity
Fourteen days after creation of wounds and application of either PhShh or LacZ plasmid, animals were prepared for vascular labeling with rhodamine-conjugated BS1 lectin (Sigma-Aldrich, St Louis, Mo). Before euthanasia, 75 μL of BS1 lectin was injected into the left ventricle to visualize functional vasculatures in the healing wound. This was allowed to perfuse for 10 minutes in the animal. After 10 minutes, the chest was entered, the left ventricle was cannulated, and the right ventricle was incised. The animal was perfused with phosphate-buffered saline and fixed with 4% paraformaldehyde. The wounds were then harvested from the dorsum of the animals. Vascularity was analyzed as described previously.9
Immunofluorescence and Immunohistochemistry
Immunohistochemistry for Shh was performed on 4% paraformaldehyde-fixed paraffin-embedded tissue sections (5 μm thick). Tissues were blocked with 3% hydrogen peroxide and 5% goat serum and treated with a 1:200 dilution of rabbit polyclonal anti-Shh antibody (Santa Cruz Biotechnology, Inc, Santa Cruz, Calif) followed by a biotinylated goat anti-rabbit antibody and Vectastain ABC reagent (Vector Laboratories, Burlingame, Calif). Multicolor immunofluorescence was performed on frozen sections (6 μm thick). Labeling of functional vessels was performed by injection of rhodamine-conjugated BS1 lectin before euthanasia, as described above. For immunofluorescence double staining of CD31 and LacZ expression, rat anti-CD31 antibody (BD Biosciences, San Jose, Calif) and rabbit anti–β-galactosidase antibody (Cortex Biochem, Inc, San Leandro, Calif) were used. For immunofluorescence double staining of Ptc1 or Smoothened (Smo) and VEGF, rabbit anti-Ptc1 antibody (Research Diagnostics Inc, Concord, Mass), rabbit anti-Smo antibody (Santa Cruz), and goat anti-VEGF antibody (Santa Cruz) were used. Normal rabbit IgG (Santa Cruz) and normal goat IgG (Santa Cruz) were used as isotype controls. Green fluorescence was generated with fluorescein isothiocianate (FITC) streptavidin (Vector Laboratories) and biotinylated anti-rat antibody (Vector Laboratories) or biotinylated anti-goat antibody (Vector Laboratories). Red fluorescence was generated with Cy3-conjugated anti-rabbit antibody (Vector Laboratories).
Cultured EPCs were costained with acetylated low0density lipoprotein (acLDL)-DiI (Biomedical Technologies, Stoughton, Mass) and FITC-conjugated isolectin B4 (Vector Laboratories), both of which are features characteristic of endothelial lineage.10,11 Rabbit anti–β-galactosidase antibody was used to detect LacZ expression in EPCs derived from NLS-Ptc1-LacZ mice. Blue fluorescence was generated with AMCA streptavidin (Vector Laboratories) and biotinylated anti-rabbit antibody (Vector Laboratories).
For X-gal staining, skin tissues from NLS-Ptc1-LacZ mice were harvested and processed as described previously.12 For the negative control, skin from NLS-Ptc1-LacZ mice harvested immediately after wounding (day 0 wound) and wild-type wounded skin were used. Histological sections were counterstained with nuclear fast red.
Fibroblasts were isolated from wild-type C57BL by placing explants (1×1 mm) on plastic culture dishes that were subsequently cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum (FBS) after the tissue had dried to the plate. Tissue explants were grown to confluence and passages as needed.
Ex vivo expansion of EPC was performed as recently described.13 In brief, BM cells obtained by flushing the tibias and femurs were plated on rat plasma vitronectin-coated (Sigma-Aldrich) culture dishes and maintained in EC basal medium-2 (Cambrex, Rockland, Me) supplemented with 5% FBS, human VEGF-A, human fibroblast growth factor-2, human epidermal growth factor, insulinlike growth factor-1, and ascorbic acid. After 4 days in culture, nonadherent cells were removed by washing, new medium was applied, and the culture was maintained through day 7.
Quantitative Real-Time Reverse Transcriptase–Polymerase Chain Reaction
Skin samples were harvested 4 days after surgery and homogenized in RNA-Stat (Tel-Test Inc, Friendswood, Tex). RNA was isolated according to the manufacturer’s instructions.
In an in vitro study, wild-type fibroblasts were plated in noncoating 35-mm plates at a density of 10 000 cells per well in Dulbecco’s modified Eagle’s medium containing 10% FBS for 24 hours. Wild-type EPCs were plated in noncoating 35-mm plates at a density of 4×105 cells per well in EC basal medium-2 containing 5% FBS for 24 hours. Oct-Shh protein (hydrophobic modified protein designed to increase its activity)14 was supplemented at the appropriate concentration (0, 0.5, 1, and 5 μg/mL) in serum-free culture medium. Cells were harvested after 24 hours, and RNA was extracted with the use of RNA-Stat according to the manufacturer’s instructions. Total RNA was reverse-transcribed with the use of Taqman Mutiscribe RT Kit (Biosystems), and amplification was performed on the Lightcycler (Roche) with the following primers and probes: human Shh: forward 5′-AAGGACAAGTTGAACGCTTTGG-3′, reverse 5′-TCGGTCACCCGCAGTTTC-3′ and FAM-CTCCTGGCCACTGGTTCATCACCG-TAMRA; mouse Shh: forward 5′-GCAGCAAGTACGGCATGCT-3′, reverse 5′-GGATGTGAGCTTTGGATTCATAGTAG-3′ and FAM-CTGGCTGTGGAAGCAGGTTTCGACTG-BHQ; Gli1: forward 5′-CACCACCCTACCTCTGTCTATTCG-3′, reverse 5′-TCCTGTAGCCCCCTAGTATCCA-3′ and FAM-CCCAGCATCACCGAAAATGTTGCC-BHQ; PTC1: forward 5′-CTCTGGAGCAGATTTCCAAGG-3′, reverse 5′-TGCCGCAGTTCTTTTGAATG-3′ and FAM-AAGGCTACTGGCCGGAAAGCGC-TAMRA; VEGF: forward 5′-CATCTTCAAGCCGTCCTGTGT-3′, reverse 5′-CAGGGCTTCATCGTTACAGCA-3′ and FAM-CCGCTGATGCGCTGTGCAGG-BHQ; angiopoietin-1: forward 5′-GGGACAGCAGGCAAACAGA-3′, reverse 5′-TGTCGTTATCAGCATCCTTCGT-3′ and FAM-TTGATCTTACACGGTGCCGATT-BHQ; angiopoietin-2: forward 5′-TCAGCCAACCAGGAAGTGATT-3′, reverse 5′-AGCATCTGGGAACACTTGCAG-3′ and FAM-CACAAAGGATTCGGACAATGACAAATGCA-BHQ; stromal cell–derived factor 1α (SDF-1α): forward 5′-CCTCCAAACGCATGCTTCA-3′, reverse 5′-CCTTCCATTGCAGCATTGGT-3′ and FAM-CTGACTTCCGCTTCTCACCTCTGTAGCCT-TAMRA; 18S: forward 5′-CGGGTCGGGAGTGGGT-3′, reverse 5′-GAAACGGCTACCACATCCAAG-3′ and FAM-TTTGCGCGCCTGCTGCCTT-BHQ. The relative level of expression of the target gene mRNAs was calculated by the comparative threshold cycle (CT) method, with normalization for the control gene, 18s. Differences of CT values were calculated for each mRNA by taking the mean value from duplicate reactions and subtracting the mean value of 18s RNA from duplicate reactions. The fold change in the expression of each target gene by treated cells relative to control cells was calculated as follows: Relative Expression=2DCT.
Western Blotting for Shh
Skin samples were harvested 4 days after surgery and homogenized in lysis buffer. Protein extracts were used for Western blotting analysis of Shh. Proteins were detected with the use of primary antibody, rabbit polyclonal against Shh (Santa Cruz Biotechnology).
The proliferative activity of cells treated with Shh was examined with the use of CellTiter 96 nonradioactive cell proliferation assay (Promega, Madison, Wis) according to the manufacturer’s instructions. Briefly, subconfluent cells (fibroblasts: 5000 cells per well; EPCs: 10 000 cells per well) were reseeded on 96-well flat-bottomed plates with 100 μL of the growth medium. Then cells were treated by Shh (0, 0.5, 1, 5, and 10 μg/mL) and incubated for 48 hours at 37°C. The absorbance at 570-nm wavelength was recorded with the use of a 96-well ELISA plate reader (Bionetics Laboratory, Kensington, Md).
EPC migrations were evaluated with a modified Boyden’s chamber assay as described previously.15 Briefly, the polycarbonate filter (5-μm pore size) (GE Infrastructure, Fairfield, Conn) was placed between upper and lower chambers. Cell suspensions (5×104 cells per well) were placed in the upper chamber, and the lower chamber was filled with medium containing human recombinant VEGF (50 ng/mL) (R&D Systems) or Shh protein (0, 0.5, 1.0, 5.0, 10.0 μg/mL). The chamber was incubated for 16 hours at 37°C and 5% CO2. Migration activity was evaluated as the mean number of migrated cells in 5 high-power fields (×40) per chamber.
After 36 hours of incubation with Shh, EPCs were washed with PBS and gently detached with 0.25% trypsin. After centrifugation and resuspension in EC basal medium-2 containing 5% FBS, identical cell numbers were replated onto vitronectin- or laminin-coated culture dishes and incubated for 30 minutes at 37°C. Adherent cells were counted by independent blinded investigators.16
Tube Formation Assay
Endothelial tube formation was assessed with the use of Matrigel assay (BD Biosciences). After 24 hours of incubation with Shh, EPCs were washed with PBS and gently detached with 0.25% trypsin. Cells were seeded with a density of 3×104 cells per well on a 4-well chamber coated with 250 μL Matrigel and incubated with EGM-2 containing 5% FBS for 48 hours at 37°C. Tube formation was examined by phase-contrast microscopy.
All results are presented as mean±SEM. Statistical comparisons between 2 groups were performed by Student t test. Multiple groups were analyzed by 1-way ANOVA test followed by appropriate post hoc tests to determine statistical significance. Probability values <0.05 were considered statistically significant. All in vitro experiments were repeated at least in triplicate and analyzed.
The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree to the manuscript as written.
Shh Signaling Pathway Is Activated Postnatally in Wound Healing
We first examined the expression of the Hh receptor, which is a marker for activation/inactivation of the Hh/Ptc/Gli pathway, during wound healing. We visualized Ptc1 expression by X-gal staining after skin wounds in mice that carry a mutation of 1 allele of the Ptc1 gene consisting of insertion of a LacZ reporter gene upstream of the ptch coding region (NLS-Ptc1-LacZ mice).17 We prepared 2 types of negative controls (Figure 1a, 1b). Figure 1a shows X-gal staining of the wound of a wild-type mouse. No X-gal–positive area was observed. Figure 1b shows X-gal staining of the skin of Ptc1-LacZ mice harvested immediately after wounding. X-gal–positive areas were infrequently observed, not at the wound edge but within hair follicles within the uninjured skin. Three days after wounding, whole-mount X-gal staining shows upregulation of Ptc1 at the wound edge and in the area surrounding vessels (Figure 1c, 1d). Microscopically, X-gal–positive cells were localized mainly to hair follicles in adult skin just after injury (day 0 wound) (Figure 1e). In the dermal mesenchymal area, only a few cells adjacent to the hair follicles were positive. The number of Ptc1-positive cells was increased in wounds at day 1 and day 3 compared with newly created wounds (Figure 1e, 1f, 1g). Morphologically, Ptc1 expression was observed in 3 types of cells: spindle-shaped mesenchymal cells, round-shaped infiltrating cells, and microvascular endothelial cells (Figure 1h). Double staining of CD31 (green) and β-gal (red) confirmed that several Ptc1-positive cells were endothelial cells, and CD31-negative Ptc1-positive cells, which may be mesenchymal cells or infiltrating cells, were observed near part of the vessels (Figure 1g).
Transgene Expression of Human Shh Plasmid
The expression of Shh after gene therapy in the wound was detected by Western analysis. Figure 2a shows an immunoblot for Shh protein 4 days after topical application of phShh to the wound, revealing expression of Shh protein. In the control group, only slight expression of Shh protein, corresponding to endogenous mouse Shh, was detected. To confirm that the increased Shh proteins in the phShh-treated group detected by Western blotting were not endogenous but derived from phShh, reverse transcriptase–polymerase chain reaction (RT-PCR) was done. RT-PCR for human Shh mRNA expression shows that the phShh results in a significant increase in human Shh mRNA levels in the wound tissue (Figure 2b). No human Shh mRNA was detected in normal skin and LacZ plasmid–treated wound. There are no significant differences in murine Shh expression between normal skin, LacZ plasmid–treated wound, and phShh-treated wound. This result also confirmed that our human Shh primer does not detect murine Shh. Immunoperoxidase staining for Shh from wounds 4 days after treatment with phShh shows a large number of positive-staining cells at the edge of the wound (Figure 2f). In particular, Shh was expressed by a variety of proliferating cells resembling fibroblasts and keratinocytes (Figure 2g and 2h). In contrast, the wounds in the control group showed no specific staining for Shh (Figure 2c to 2e).
Shh Accelerates Wound Closure and Increases Angiogenesis in Diabetic Wounds
Figure 3a shows wound areas of control (LacZ plasmid–treated), phShh-treated, and phShh plus VEGF blocking antibody–treated diabetic mice. The Shh-treated wound shows granulation of more than half of the wound area compared with scant epithelialization in the controls. As a consequence, Shh treatment resulted in a significantly smaller wound within 5 days of treatment (Figure 3a and 3b). By day 14, this effect was most significant (percent wound closure, 25.0±2.0% versus 65.3±7.6%; P<0.0001, control versus phShh; Figure 3b). To evaluate the possible negative effects of LacZ plasmid, methylcellulose pellet alone was applied on the wound. There was no significant difference between LacZ plasmid and this control. Consistent with these findings, histology revealed increased cellular infiltration, collagen deposition, and thick granulation tissue in phShh-treated wounds (Figure 3c). The histological score of wounds from mice treated with phShh was significantly higher (2.3±1.0 versus 9.5±2.1; P<0.001, control versus phShh; Figure 3d). Treatment with VEGF blocking antibody significantly inhibited the acceleration of wound healing in phShh-treated wounds (percent wound closure at day 14, 65.3±7.6% versus 34.7±4.1%; P<0.001, phShh versus phShh+VEGF antibody; Figure 3b).
Figure 4a shows hematoxylin-eosin staining of whole wounds revealing improved wound appearance and increased granulation tissue formation in a phShh-treated wound. Wound angiogenesis was analyzed with the use of fluorescent BS1 lectin to visualize neovascularization in the resected wounds on day 14 (Figure 4b). Ten-micrometer frozen sections were used for analysis. Figure 4b shows a comparison of neovascularization at the wound margin in control (LacZ plasmid)–treated, phShh-treated, and phShh plus VEGF blocking antibody–treated diabetic mice after 14 days. Compared with the control group, Shh-treated wounds had significantly enhanced vascularity, with sprouting toward the center part of the wound. Shh significantly enhanced wound vascularity as assessed by the percentage of the pixels in each image that were fluorescent (4.79±0.32% versus 14.63±0.901% versus 5.57±1.82%, control versus phShh versus phShh+VEGF antibody; P<0.001; Figure 4c).
Ptc1 and Smo Expression in Wounds
Immunofluorescence staining for Ptc1 and Smo in the area surrounding the wound was performed with costaining for VEGF (Figure 5). The number of Ptc1-positive cells was increased in phShh-treated wound compared with control, whereas the number of Smo-positive cells was similar in the 2 groups. The Ptc1- or Smo-positive cells were colocalized with immunoreactivity for VEGF. These data suggest that these cells have the ability to produce VEGF after activation of the Ptc1/Smo pathway, stimulated by Shh, consistent with prior data.2,18
Increased BM-Derived EPCs in Shh-Treated Wounds
To determine whether topical phShh augments recruitment of BM-derived EPCs into sites of wound repair, we chose a chimeric mouse model. GFP/BMT mice are the recipients of BM from donors in which cells express GFP. This enabled identification of cells that are derived from a BM progenitor. After recovery from the BMT procedure, wounds were created on their dorsum and then randomly assigned to treatment with phShh or control plasmid. Fourteen days after wounding, these mice were euthanized, and wounds were harvested. To visualize the functional neovasculature in the healing wound, 5 animals of each group were injected with rhodamine-conjugated BS1 lectin via the left ventricle before euthanasia, as described above. Figure 6a shows the contribution of BM-derived EPCs into the neovasculature of the healing wound edge at day 14. Red identifies BS1 lectin binding cells (endothelial cells) in functional vessels, and green fluorescence indicates GFP+BM-derived cells. Double-positive cells (white arrow) indicate BM-derived EPCs incorporated into the neovasculature. Notably, several double-positive cells are incorporated in microvasculature structures. BM-derived EPC incorporation into the neovasculature in the phShh-treated group was significantly increased compared with the control group (P<0.001 versus control; Figure 6b).
Shh Upregulates Production of Angiogenic Cytokines and Proliferation of Fibroblast
To identify the potential mechanisms responsible for the therapeutic effect of Shh on wound healing, we evaluated mRNA expression of a panel of candidate genes in primary cultured adult dermal fibroblasts (Figure 7a). There was upregulation of the Hh-related transcription factor Gli1, meaning that the Hh pathway is intact in these adult cells. As expected, the expressions of VEGF and angiopoietin-1 were upregulated by Shh; however, the expression of angiopoietin-2 was not significantly altered. In addition, the expression of SDF-1α, a trafficking chemokine for hematopoietic stem cells, was also increased.
The effects of Shh on dermal fibroblast proliferation were examined with the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt assay (Figure 7b). Shh increased proliferative activity of fibroblasts in a dose-dependent manner (0 μg/mL, 0.148±0.004; 0.5 μg/mL, 0.187±0.007; 1 μg/mL, 0.232±0.004; 5 μg/mL, 0.243±0.007; 10 μg/mL, 0.191±0.006; *P<0.0001 versus 0 μg/mL).
Shh Upregulates EPC Proliferation, Migration, Adhesion, and Tube Formation
Prior studies have indicated that mature endothelial cells express the Shh receptor Ptc1 and that Hh signaling is essential for endothelial tube formation.19,20 We hypothesized that EPCs might also exhibit functional Ptc1 expression.
Cultured EPCs derived from NLS-Ptc1-LacZ mice were used for immunofluorescence analysis (Figure 8a). EPCs were identified by costaining with acLDL-DiI and FITC-conjugated isolectin B4, both of which are markers of endothelial lineage. Ptc1 expression in EPCs from NLS-Ptc1-LacZ mice was visualized as blue fluorescence with the use of anti–β-galactosidase antibody. Accordingly, triple labeling for acLDL-DiI, FITC-conjugated isolectin B4, and β-gal identifies EPCs that express Ptc1.
To examine the direct effects of Shh on EPCs, we performed a series of in vitro assays. First, we used RT-PCR (Figure 8b) to assess gene expression in cultured EPCs. Shh (1 μg/mL) increased Ptc1 gene expression in EPCs exhibiting a biphasic response (0 μg/mL, 0.19±0.046; 1 μg/mL, 0.329±0.061; 10 μg/mL, 0.178±0.006; *P<0.05 versus 0 μg/mL). Interestingly, VEGF gene expression was not increased in EPCs by Shh.
Next, we evaluated the effect of Shh on EPC migration (Figure 8c). The migratory response of EPCs toward different concentrations of Shh was measured with the modified Boyden chamber migration assay. The effect of Shh on migration peaked at 1 μg/mL; higher concentrations elicited less stimulation (*P<0.001 versus 0 μg/mL Shh). Furthermore, the peak effect of Shh on EPC migration was greater than the migratory response induced by 50 ng/mL of VEGF (†P<0.001 versus 50 ng/mL VEGF).
EPC proliferation was examined with the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt assay (Figure 8d), disclosing that Shh increased proliferative activity in a dose-dependent manner (0 μg/mL, 0.202±0.002; 0.5 μg/mL, 0.203±0.003; 1 μg/mL, 0.212±0.001; 5 μg/mL, 0.223±0.001; 10 μg/mL 0.226±0.002; *P<0.001 versus 0 μg/mL).
To study the EPC adhesion (Figure 8e), EPCs were first incubated with Shh (0, 1, 10 μg/mL) for 36 hours. After replating on vitronectin- or laminin-coated dishes, EPCs preexposed to Shh exhibited a significant, dose-dependent increase in the number of adhesive cells at 30 minutes (vitronectin: 0 μg/mL, 141.8±18.012 cells; 1 μg/mL, 371.4±49.878 cells; 10 μg/mL, 532.8±24.897 cells; laminin: 0 μg/mL, 13.6±1.166 cells; 1 μg/mL, 17.4±1.435 cells; 10 μg/mL, 36.2±5.286 cells; n=8; *P<0.001, **P<0.01 versus 0 μg/mL).
Finally, tube formation assays were performed to evaluate in vitro capillary morphogenesis induced by Shh exposure of EPCs. Figure 8f shows phase-contrast micrographs of EPCs in the presence or absence of 2 μg/mL Shh protein for 48 hours. EPCs formed tubular structures after exposure to Shh, whereas tube formation was rare in the control-treated EPCs.
Hh proteins are morphogens that play critical roles in organ development in embryogenesis and regulate epithelial-mesenchymal interactions that are crucial to morphogenesis of the nervous system, somite, limb, lung, gut, hair follicle, and bone.21–28
Several recent studies have suggested that Hh proteins might be involved in the vascularization of certain embryonic tissues. Hypervascularization of the neuroectoderm is seen after transgenic overexpression of Shh in the dorsal neural tube,29 whereas a knockout of the zebra fish Shh homologue results in disorganization of the endothelial precursor cells and inability to form the dorsal aorta or axial vein.30 In addition, Shh-deficient mice lack proper vascularization of the developing lung.23
We investigated the hypothesis that the endogenous Shh pathway is physiologically involved in the skin wound healing process, in which angiogenesis plays a critical role, and that augmenting the Shh signaling pathway could accelerate wound healing by promoting angiogenesis. Specifically, we were interested in testing this hypothesis in a wound model in which angiogenesis is known to be deficient, ie, diabetes.
We detected upregulation of Ptc1 during wound healing, indicating that this pathway is intact in adult skin and upregulated during wound repair. Upregulation of Ptc1 expression occurs in mesenchymal fibroblasts, endothelial cells, and infiltrating cells. The ability of fibroblasts to respond to Shh stimulation has been reported previously.2 We show that dermal fibroblasts are activated by exogenous Shh stimulation, resulting in the expression of several angiogenic cytokines and induction of proliferative activity.
Because of the very short half-life of Shh protein, we used Shh gene therapy to extend the duration of exposure of the target tissue to the therapeutic agent. The methylcellulose carrier method results in rapid and relatively long-term expression of the transgene.7 This longer duration of expression of the transgene is likely attributable to the methylcellulose carrier that retains DNA at the wound site and slows diffusion of the plasmid into the tissue.
The critical role of angiogenesis in the acceleration of wound healing by Shh was disclosed in our studies by application of VEGF neutralizing antibody, which resulted in abrogation of the therapeutic effect of Shh on wound neovascularization with an attendant delay in wound closure. These data suggest that the effects of Shh on wound healing are mediated to a large extent by the augmentation of neovascularization via VEGF and are not the result of stimulating reepithelialization alone. If the effects of Shh on wound healing were mediated only by growth of keratinocytes, the inhibition of neovascularization by blocking of VEGF signaling would not be expected to have a significant effect on wound closure. Shh may promote reepithelialization via direct effects on keratinocytes; however, our data suggest that the acceleration of neovascularization via VEGF upregulation is a key feature mediating the effects of Shh. It is also known that sufficient granulation tissue is needed for reepithelialization.1 We have shown the development of thick granulation tissue with abundant neovessels in Shh-treated wounds. This finding also suggests that Shh accelerates granulation tissue formation in part via a proangiogenic effect. However, our results do not completely eliminate the possibility that other effects of Shh on wound reepithelialization, for example, through activating keratinocytes, could contribute to the acceleration of wound closure. It is possible that Shh also promotes reepithelialization directly as well as by augmenting neovascularization and granulation tissue formation, and these multifactorial effects of Shh might contribute to the overall effect on wound closure.
The effects of Shh on endothelial cells has been shown previously; Shh dose dependently induced capillary morphogenesis by endothelial cells through the Gli-independent PI3-kinase pathway.19,20 In our wound-healing model, Shh induces large and thick microvasculature sprouting toward the center part of the wound. This morphogenic effect of Shh plus the induction of expression of angiogenic cytokines from fibroblasts may contribute to the formation of mature vessels.
We also identified Ptc1 expression in EPCs. Recent studies suggest that there is no direct effect of Shh on cellular responses, such as proliferation, migration, and serum-deprived survival, in cultured endothelial cells.20 However, EPCs are distinguishable from mature endothelial cells and could be said to share certain characteristics with undifferentiated cells, which are a primary target of Shh during embryonic life. Therefore, we examined the effects of Shh on EPCs and found several direct effects, including induction of proliferation, enhanced adhesion, increased migration, and tube formation.
Shh also increases the expression of SDF-1α, a trafficking chemokine for hematopoietic stem cells, providing a potential explanation for the enhanced recruitment of BM-derived progenitor or stem cells documented in our BMT models.
Taken together, these data suggest that fibroblasts are one of the major mediators of Shh activity during wound healing. It is well known that hair follicle keratinocytes (or epidermal stem cells) contribute to reepithelialization during the wound-healing process,31 suggesting that follicle keratinocytes and dermal papilla cells are key targets of Shh signaling.
The Hh gene was identified as a critical regulator of cell-fate determination during embryogenesis in Drosophila 2 decades ago.
There are 3 highly conserved Hh genes in mammals: Shh, desert hedgehog (Dhh), and Indian hedgehog (Ihh). Shh is the most widely expressed during development among the 3 hedgehog family members.27,32 Hh signaling occurs through the interaction of Hh ligand with its receptor, Patched-1 (Ptc1 encoded by Ptch).33 After Hh binds Ptc1, subsequent activation of Smo initiates signaling events that lead to the regulation of transcriptional factors belonging to the Gli family, which modify the expression of downstream target genes of the Hh pathway, including Ptch and Gli themselves.34–37 Thus, Ptc1 and Gli are required components as well as transcriptionally induced targets of the Hh signaling pathway.
We examined Ptc1 and Gli1 expression as markers for activation/inactivation of the Hh/Ptc/Gli pathway. However, Hh can also induce a Gli-independent pathway such as the orphan nuclear receptor COUPTF-II38 or PI3-kinase.20 We recently discovered that Shh acts as an indirect angiogenic agent and may trigger neovascularization through Shh/Ptc1 signaling, specifically in mesenchymal cells,2 and the inhibition of Shh signaling is sufficient to decrease ischemia-induced local angiogenesis and the upregulation of VEGF from skeletal fibroblasts.39 These results indicate that the activation of the Shh signaling pathway is crucial for angiogenic response. Thus, the embryonic morphogen Shh may have therapeutic potential for pathological conditions that need angiogenesis such as lower extremity ischemia, myocardial ischemia, and diabetic foot ulcer.
Upregulation of Shh signaling has been reported in humans and in animal models of basal cell carcinoma.40–42 This might suggest that transfection of Shh could be limited as a clinical strategy for wound healing because of a potential risk for cancer induction. However, it should be noted that reports regarding the carcinogenic effect of Shh involved transgenic or mutated mice and the observation that a mutation in the Shh pathway is found in a percentage of patients with basal cell carcinoma. It is probable that the high dose and long-term exposure of Shh resulting from the genetic models, as well as the dysregulated signaling that occurs in the presence of the human mutation, are factors in the development of cancer. Thus, it is possible that the lower dose and short-term exposure to the unmutated Shh, via naked DNA gene therapy, would not generate a similar degree of risk, if any. Nevertheless, a cautious stance is certainly prudent as one approaches the use of Shh clinically.
Regulation of cell proliferation, migration, adhesion, and capillary differentiation is essential to develop functional neovasculature during wound healing. The findings of the present study suggest the possibility that topically applied Shh gene therapy might play a novel role in postnatal neovascularization, not only indirectly but directly via stimulation of EPCs. Finally, vascular remodeling, induced by modulating the Hh pathway, might have therapeutic potential in patients with microvascular dysfunction such as diabetic foot ulceration, which is a major cause of morbidity for the growing population of diabetics. Future investigation is warranted to determine the potential clinical utility of this approach.
This study was supported in part by National Institutes of Health grants (HL-53354, HL-57516, HL-77428, HL-63414, HL-80137, PO1HL-66957). Oct-Shh protein was a generous gift from Curis Inc. We thank M. Neely for secretarial assistance.
Dr Losordo received research grant support from Curis Inc, Baxter Inc, Corautus Genetics Inc, Cordis, Anormed, and Boston Scientific Corp. The other authors report no conflicts.
Kusano K, Allendoerfer KL, Munger W, Pola R, Bosch-Marce M, Kirchmair R, Yoon YS, Curry C, Silver M, Kearney M, Asahara T, Losordo DW. Sonic hedgehog induces arteriogenesis in diabetic vasa nervorum and restores function in diabetic neuropathy. Arterioscler Thromb Vasc Biol. 2004; 24: 2102–2107.
Goodson WH III, Hunt TK. Wound collagen accumulation in obese hyperglycemic mice. Diabetes. 1986; 35: 491–495.
Ii M, Takenaka H, Asai J, Ibusuki K, Mizukami Y, Maruyama K, Yoon YS, Wecker A, Luedemann C, Eaton E, Silver M, Thorne T, Losordo DW. Endothelial progenitor thrombospondin-1 mediates diabetes-induced delay in reendothelialization following arterial injury. Circ Res. 2006; 98: 697–704.
Qin G, Ii M, Silver M, Wecker A, Bord E, Ma H, Gavin M, Goukassian DA, Yoon YS, Papayannopoulou T, Asahara T, Kearney M, Thorne T, Curry C, Eaton L, Heyd L, Dinesh D, Kishore R, Zhu Y, Losordo DW. Functional disruption of alpha4 integrin mobilizes bone marrow-derived endothelial progenitors and augments ischemic neovascularization. J Exp Med. 2006; 203: 153–163.
Asahara T, Krasinski K, Chen D, Sullivan A, Kearney M, Silver M, Li T, Isner JM. Circulating endothelial progenitor cells in peripheral blood incorporate into re-endothelialization after vascular injury. Circulation. 1997; 96: I-725.
Asahara T, Iwaguro H, Kalka C, Masuda H, Hayashi S-I, Silver M. Gene therapy of endothelial progenitor cell for vascular development in severe ischemic disease. Circulation. 1999; 100: I-481.
Shin D, Garcia-Cardena G, Hayashi S, Gerety S, Asahara T, Stavrakis G, Isner J, Folkman J, Gimbrone MA Jr, Anderson DJ. Expression of ephrinB2 identifies a stable genetic difference between arterial and venous vascular smooth muscle as well as endothelial cells, and marks subsets of microvessels at sites of adult neovascularization. Dev Biol. 2001; 230: 139–150.
Mallat Z, Silvestre JS, Le Ricousse-Roussanne S, Lecomte-Raclet L, Corbaz A, Clergue M, Duriez M, Barateau V, Akira S, Tedgui A, Tobelem G, Chvatchko Y, Levy BI. Interleukin-18/interleukin-18 binding protein signaling modulates ischemia-induced neovascularization in mice hindlimb. Circ Res. 2002; 91: 441–448.
Taylor FR, Wen D, Garber EA, Carmillo AN, Baker DP, Arduini RM, Williams KP, Weinreb PH, Rayhorn P, Hronowski X, Whitty A, Day ES, Boriack-Sjodin A, Shapiro RI, Galdes A, Pepinsky RB. Enhanced potency of human sonic hedgehog by hydrophobic modification. Biochemistry. 2001; 40: 4359–4371.
Walter DH, Rittig K, Bahlmann FH, Kirchmair R, Silver M, Murayama T, Nishimura H, Losordo DW, Asahara T, Isner JM. Statin therapy accelerates reendothelialization: a novel effect involving mobilization and incorporation of bone marrow–derived endothelial progenitor cells. Circulation. 2002; 105: 3017–3024.
Goodrich LV, Milenkovic L, Higgins KM, Scott MP. Altered neural cell fate and medulloblastoma in mouse patched mutants. Science. 1997; 277: 1109–1113.
Kusano KF, Pola R, Murayama T, Curry C, Kawamoto A, Iwakura A, Shintani S, Ii M, Asai J, Tkebuchava T, Thorne T, Takenaka H, Aikawa R, Goukassian D, von Samson P, Hamada H, Yoon YS, Silver M, Eaton E, Ma H, Heyd L, Kearney M, Munger W, Porter JA, Kishore R, Losordo DW. Sonic hedgehog myocardial gene therapy: tissue repair through transient reconstitution of embryonic signaling. Nat Med. 2005; 11: 1197–1204.
Vokes SA, Yatskievych TA, Heimark RL, McMahon J, McMahon AP, Antin PB, Krieg PA. Hedgehog signaling is essential for endothelial tube formation during vasculogenesis. Development. 2004; 131: 4371–4380.
Kanda S, Mochizuki Y, Suematsu T, Miyata Y, Nomata K, Kanetake H. Sonic hedgehog induces capillary morphogenesis by endothelial cells through phosphoinositide 3-kinase. J Biol Chem. 2003; 278: 8244–8249.
Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development. 2000; 127: 2763–2772.
St-Jacques B, Hammerschmidt M, McMahon AP. Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev. 1999; 13: 2072–2086.
Rowitch DH, S-Jacques B, Lee SM, Flax JD, Snyder EY, McMahon AP. Sonic hedgehog regulates proliferation and inhibits differentiation of CNS precursor cells. J Neurosci. 1999; 19: 8954–8965.
Ingham PW. Transducing hedgehog: the story so far. EMBO J. 1998; 17: 3505–3511.
Marigo V, Tabin C. Regulation of patched by sonic hedgehog in the developing neural tube. Proc Natl Acad Sci U S A. 1996; 93: 9346–9351.
Krishnan V, Pereira FA, Qiu Y, Chen CH, Beachy PA, Tsai SY, Tsai MJ. Mediation of sonic hedgehog-induced expression of COUP-TFII by a protein phosphatase. Science. 1997; 278: 1947–1950.
Pola R, Ling LE, Aprahamian TR, Barban E, Bosch-Marce M, Curry C, Corbley M, Kearney M, Isner JM, Losordo DW. Postnatal recapitulation of embryonic hedgehog pathway in response to skeletal muscle ischemia. Circulation. 2003; 108: 479–485.
Oro AE, Higgins KM, Hu Z, Bonifas JM, Epstein EH Jr, Scott MP. Basal cell carcinomas in mice overexpressing sonic hedgehog. Science. 1997; 276: 817–821.
These studies show that a simple, topical application of naked DNA in a biodegradable carrier can effectively accelerate wound healing in a diabetic animal model. The data suggest that a practical approach to gene therapy for wound healing is feasible. The topical route of administration is attractive because it can be patient administered, can be applied repeatedly in resistant cases, and offers a margin of safety because systemic distribution, although not tested in this study, would presumably be very low. Finally, these data are of interest because of the use of a gene that was originally identified for its role in embryonic development. With a transient gene therapy approach, the recapitulation of this signaling pathway in the adult is shown to be therapeutic. This finding underscores the fine line between tissue repair and “regeneration,” with the latter typically considered to constitute the replacement of damaged tissue. In this case a strategy of imitating the signaling of embryonic cells is used and raises interesting questions regarding the wider application of an evolving knowledge of stem cell biology, ultimately obviating the need to use stem cells themselves.